Hacker News new | past | comments | ask | show | jobs | submit login
Build Your Own Professional-Grade Audio Amp on the Sort-Of Cheap (ieee.org)
444 points by chablent 4 months ago | hide | past | web | favorite | 256 comments



Audiophilia is a study in contradiction. The ask is a flat "uncolored" accurate sound. The gear that sells for ridiculous prices is usually tube gear with limited bandwidth and makes ~15 watts at 5-10% (admittedly usually even order harmonic) thd. Hardly uncolored.

That said, and Ive mentioned this before on HN. My (now late) uncle had a bunch of audiophile friends in Dallas. He hung around folks who had spent upwards of 500,000 (yes you read that right) on audio gear. I can't forget the time I rolled up to a modest middle class home, only to find out that inside of this house was a pair of Focal Grand Utopias (selling retail for 250k at the time, though he said he paid around 60k for the pair). He had 2 750 watt tube amps, current era but the brand escapes me. For source material he had a heavily modified Technics SP-10 with 2 tone arm/cartridge combinations and a boutique tube phono preamp. This fed into a similarly heavily modified Dynaco tube preamp from the 1960s. When he really wanted to get serious, though, he had a studer 2" tape machine and a sizable collection of low generation master tape copies. If I hadn't seen this with my own eyes I would say my uncle was full of shit when he told me about it. You know what? it sounded freaking amazing.Im sure it is 90% placebo, and 10% the fact that he was able to play high bandwidth audio to concert levels with no stress or strain from the amplifiers. Was it worth 6 figures of money? not to me. But I am glad I got to hear it...


The three critical paths for high end sound reproduction are: loudspeakers, speaker placement, room acoustics. Of course all parts of sound reproduction matter, but those three are the ones that affect real systems in practice.

Most other factors are rarely an issue and/or inferior parts have a much lower magnitude of impact on sound quality. Sufficiently good source material (256kbps AAC or better) is widely available. Most amplifiers are fine. Almost all wire/cable is perfectly good. High quality DACs are more abundant than you might think (e.g. the digital inputs of any half-decent AV receiver made in the past decade).

Good speakers.

Good placement.

Good acoustics.

Those are the ones to focus on. They're the ones to spend money on. If you wanted to add a fourth to that list, I'd add an honourable mention to in-room calibration systems like Audyssey and DIRAC. Depending on the sorts of imperfections in a given system the opportunity for improvements with DSP might be substantial and dramatic, or marginal to the point of being irrelevant. (For most systems it's somewhere in the middle.)


You’re absolutely right about that. I didn’t even mention how all of the drywall was replaced strategically with acoustic wall board. The entirety of the living room was gutted and refurbished to be the ultimate listening room, complete with stadium seating. Another recollection I just had is that he had a separate electric circuit run to the house(as in a second set of wires coming off the utility pole) so he could properly condition the power feeding his audio gear.

Every aspect of the room was built around the grand utopias and their placement. If you’re not familiar with them check them out here. https://www.focal.com/en/home-audio/high-fidelity-speakers/u...

The subwoofer drivers are unique in that they use an electromagnet instead of the usual strontium or neodymium permanent magnet.

I think what impressed me the most was that this guy was not annoying or in to the hand wavy audiophile hocus pocus. Just super passionate about getting everything done to get the absolute best sound even if the bang per buck was low. He had a wall of vinyl and was happy to let me pick several selections to play. From what I understand, this guy would open his home to anyone interested in a listening session. If anyone is in the DFW area and you can get plugged in to that community it seemed like a great group of folks.


> Another recollection I just had is that he had a separate electric circuit run to the house(as in a second set of wires coming off the utility pole) so he could properly condition the power feeding his audio gear

In my limited experience, "clean power" (I know nil about this so I'm just repeating what I hear) was the most important difference people would "hear". My uncle used to have a Hi-Fi shop and I helped him sometimes. People hearing distortions in their existing setup used to make up for 90% of his customers. In those cases the number one step was to make sure that the power that comes to the system is clean. Being a certified electrician, he used to do his thing and you would almost always hear the "aaah" from the customers the moment you hit play. To me also the difference was generally night and day.


And yet such impressions conflict with how power supply circuits work and how THD measurement works. Wall power is incredibly noisy and there’s nothing much an electrician can do to change that. If this noise had any meaningful effect on the sound—it doesn’t—you would be able to measure it as a variance in THD.

(Now of course that might be less true for some poorly made vinyl/phono level equipment but if you actually cared about audible distortions you wouldn't be listening to the sound of a needle scratching its way across a fragile plastic groove in the first place...)


Maybe then it was because of the isolation that the audio cables got being away from all the chaos created by the random power cables. I don't know. Most of the time, it was a huge improvement. The next step was to fix/remove anything that caused distortions.


Would have been way better off with a bunch of car batteries, the internal resistance and cold cranking amps would have pissed on any utility supply regulation and power supply regardless of the capacitors he had.


I don't know can I judge him. I listen to music all the time. Every imaginable genre that I happen to like. I'm not putting hundreds of thousands to my setup, but a good pair of headphones and a good headphone amp is a must for me in every office I work. Put some cash for Beyerdynamic DT-1770 Pro and been having a handmade headphone amp for the last 15 years, still using it. The sound cards differ, but now I'm ok with a Dragonfly serving the system.

If you spend most of your day wearing headphones, you might invest to them a bit.


Have you checked out the O2 headphone amp by nwavguy? It is mentioned in a toplevel comment in this thread.


I have a Corda HA2 mk3 and it still gives everything I need: enough punch to drive 250ohm headphones and a crossfeed to prevent fatigue when using them hours every day.

Maybe in 50 years when it breaks I'll get a new one :)


The weakest "link" in the "audio-chain" (if you exclude the room) are passive speakers with distortion above 1%

The best speakers i heard to date are from a german company called Backes & Müller. Active crossovers, each speaker has its own poweramp and a sensor attached to the membrane. The signal of the sensor is fed back into a comparator which calculates the difference to the music signal, this is then used to force the speaker to follow the music signal exactly.

>On the basis of physical factors all drivers make mistakes, which one must accept in passive and uncontrolled systems. In contrast, all BM drivers are a part of active error correction systems.

>Active loudspeakers know no conduction loss, no impedance issues; have faultless crossover performance, exact phasing, an optimal amplifier-chassis ratio, more actual output stage power at their disposal, quicker impulse processing and higher dynamics. All of this produces active regulatory intelligence. A sensor system continuously compares and corrects the membrane movements with the actual music signal.

>B&M uses two types of negative feedback: inductive and capacitive. Intelligent electronic fuses ensure that the entire system is reliably shut down in the case of an impending overload. The BM electronic crossover splits the music signal into individual paths of respective frequency ranges and allocates these to their individual output stage. Every individual output stage drives to its respective driver, optimised for its frequency range.

from http://www.backesmueller.de/en/b-m-technology.html

Pricey and not DIY friendly, but you can pick them up used on ebay for "peanuts". W0as thinking about buying an BM8 (3-way) refurbished for 2800 Euros, BM10 (4-way) refurbished costs 10000 Euros, got a BM12 (5-way) used for 1400 Euros.

The company is now owned by http://ksaudio.com


Those look intriguing. I would like to hear them. One of the things I’ve lamented as time and technology march on is how high fi seems to have lost it’s mainstream appeal. It seems like when I was growing up there was a wide continuum of commodity grade to mid grade to prosumer to high end gear available. There was always ridiculous high end stuff but even companies like Sony pioneer and technics had a higher end line. Then there were the smaller companies like adcom, carver, McIntosh, etc that had a reasonably large product line. Now it seems that it’s just commodity or niche boutique gear. Most of those companies still exist, but with a far smaller product assortment or they were absorbed into an audio conglomerate like harmann and badge engineer from some other product line.


Its like cinema for the ears. Once started talking normally and i could not hear the music anymore because the volume was so low yet all the details and bass were still there. When i heard them for the first time there was opera playing, felt like the orchestra was in the same room. It was a small 2-way system.


We had some older B&Ms at our house back in the day (also an audiophile middle class household). As a musician myself, I'd say they're worth the money if you can afford it. (I think they were around $8000 each or something).


Not in that order though, unfortunately: Unless you are using really shitty speakers, room acoustics trumps everything else pretty quickly. For anyone who has never listened to music in a well treated environment I found https://youtu.be/dB8H0HFMylo?t=355 to be pretty eye opening.


Even in well treated environment , speaker placement is critical. I’ve been working in pro audio 30 years and still surprised at how much moving speakers affects their performance.


> Unless you are using really shitty speakers, room acoustics trumps everything else pretty quickly.

I agree completely, but emphasising that point tends to lead to flame-wars in most places on the internet. And it's such an imprecise subject to debate because everyone's room is different. Some rooms can be atrocious, while others can be unintentionally excellent. Some rooms can be easy to improve, some require knowledge and experience in order to improve. Some only require repositioning, some require treatment products that can cost a lot more than many enthusiasts realise.

My favourite video that demonstrates the effectiveness of acoustic treatment:

https://www.youtube.com/watch?v=cp56A6TcL1E


Apparently the iPhone 6 has a surprisingly high-quality DAC: https://www.computeraudiophile.com/forums/topic/22843-how-do...


They all do. The idea of bad quality DACs is mostly FUD. That’s not to say that there aren’t bad ones, but the impact on the sound is rarely anything other than its self-noise floor.

That thread contains little to no evidentiary merit and a boatload of technical ignorance.


It's very easy for an audiophile to start going down the slippery slope into madness.

It usually starts with amplifiers, then speakers, then source equipment... and then they start rearranging furniture, tearing out drywall and putting absorbers everywhere, and then it starts all over again in a cycle.

I find it best to stick with a nice comfy pair of headphones and keep expectations low for everything else. If you're only managing the volume of air between your ear drums and the headphone cups-- that can be done to an obsessive level for vastly less cost (and without divorce) than if we're talking about a whole living room.


I used to sell medium-high end audio visual equipment in Calgary during the height of the oil sands boom. Plenty of money moving around.

I knew customers who built the foundation of their homes around speaker placement, having separate foundations for their towers so that fewer resonant effects from the house would impact the speakers. Haha. Madness is right!

And that $500,000 number is easy when you’re dealing with those circles for sure.


I’ve heard such things. Two points to note:

1. Decent monitor headphones or speakers sound the same. That’s under $500. Even some Sennheiser HD25’s are literally up there.

2. The equipment used to record such music cost less and was lower quality. Adding $500k doesn’t add anything because there isn’t anything there.

Then again I have no problem with people who spend $500k on audio kit. That’s up to them. If it makes them feel good buying something in that range then fair play to them. It’s mostly just art.

What I object to is little bags of stones wrapped around cables, magic boxes which absorb non physical entities in the room and directional cables. Woo and such needs to be taken down.


Not actually true about recording equipment. Top end mixing desks cost > $1m. Top end engineers and producers usually have at least $100k of outboard processing on hand - everything from mic preamps to vintage compressors. Top end mastering studios will work with similar budgets.

Add exotic guitars, vintage synthesizers, and other esoterica, and the total hardware budget for a mainstream commercial hit album is comfortably high six/low seven figures.

Some genres - mostly electronica - are 100% "in the box", generated and mixed with software. But they're usually still mastered in a high end studio with a combination of digital and analog tools on reference-grade monitors.

But all of that equipment is used to produce recordings that "translate" - play well on everything from ear buds to high end systems. That requires processing of its own.

True hifi recordings that aim to be as transparent as possible are very rare.

There's no such thing as "transparent" in pop/rock because it's all artifice. But even in classical recordings the sound has usually been edited and significantly processed, so the abstract concept of fidelity doesn't really apply.


Yes but the point is the signal path in those probably less than $100 end to end. Instruments excepted.

A decent audio grade opamp is $2 for example.


I'm assuming you mean instruments and microphones excepted, because many high-end recording microphones cost $1000 and up. Yes, there are legends of a few albums being recorded with only a few SM57 (a $100 mic), but they're the exception - and most such albums aren't generally regarded as audiophile-quality.

But regardless, even just the input transformer on most mixing consoles costs at least $100. And many extremely high-end consoles use discrete opamps which can get really expensive. There's also a lot of tube-based outboard gear in studios.

Sure, you could build a mixing desk with only $2 opamps and no transformers, no tubes, etc. But that would sound bad. Unlike the audiophile world, recording isn't about doing everything possible to prevent distortion - if it was, all recording would be done with calibrated measurement microphones and other various lab equipment.

So how can distortion be good in recording if it's bad in reproduction? A few reasons. IM distortion, for starters - distorting every channel individually is going to sound much different (and in most cases, better) than distorting the mix of all the channels.

But also, in recording, the distortion is finely tuned to the exact song, by someone who gets paid hundreds of dollars an hour because they are very good at it (and yes, good at getting it to sound even better on a variety of different systems). A recording engineer can use exactly whatever mic preamp they think will sound best for the song. Yes, audiophiles do tend to adjust tiny settings for each song, but not to the degree that recording engineers do - no audiophile is going to own 20 different amps and switch them out for different parts of different songs (or at least not on a regular basis).

Recording engineers can also do so on an individual channel basis - use one preamp for the vocals, a different one for the large-diaphragm condenser over the piano, another different one for a small-diaphragm condenser pointed right at the piano hammers, etc. Unless they're using Dolby Atmos or some other multichannel format, an audiophile doesn't even have that option, no matter how much they enjoy tweaking minor settings.

So no, the signal path is certainly not way less than $100 worth of components - and it wouldn't sound nearly as good if it was.


Yes I exclude instruments and transducers. Those are where the investment should be if anything.

You have no idea what you are talking about regarding signal chain. Discrete opamps are inferior on every possible measurement. This is total nonsense. Utter rubbish.

Do you know what noise figure is?

Do you know what CMRR is?

Do you know what PSRR is?


The original directional cables, sans mythology, are a real thing. Grounding for cable shielding matters a lot, and is plainly audible in many circumstances. "Directional" cables only ground the shield at one end, preventing ground loops and interference from grounding both ends. This is electronics 101.

An interesting aside on cables... I play electric guitar a lot, which is far more sensitive to cables than hi-fi gear, because guitar pickups are high impedance, low output devices - enough that capacitance can have audio-frequency passive filtering, and grounding is a huge problem. With passive pickups, I can easily hear differences in various cables I own. But some of my guitars have active electronics onboard (EMG pickups or other preamps), and produce a low-impedance signal. I can't hear the difference with those guitars.

When it gets down to it, I use the cables that are most reliable and comfortable to handle. Sound quality doesn't even enter into it, whether or not it's audible.


The point that most people miss is it is not the cable, it is the connection, copper oxide is a rectifier (diode) and if you put it in the circuit it erodes the quality pretty quickly.


I don't think that your first paragraph is an honest assessment of the issue. People who buy tube gear are absolutely not looking for "flat, uncolored sound".


> 2750 watt tube amps

WOW! That's ~5% of a Ducati. That's ... insane for a sound system.


A Ducati is a sound system.


What did he listen to ?


The cannons from the 1812 overture, over and over again. Did you hear that? Did you hear that? wasn't that awesome? Hang on I'll play it again, let me just adjust this setting first....


In fairness, though, the 1812 ouverture IS a pretty impressive piece to listen to, if it is properly recorded and reproduced.

While my setup is somewhat (cough) less impressive/over-the-top than the one referenced above, there's no denying listening to a proper 1812 is a physical experience.

I'm partial to Kunzel's recording with the Cincinnati Pops - those cannons are LOUD. (Better be careful not setting the volume too high or you're likely to find yourself a couple of woofers short afterwards.)


Audio is a great way to get kids excited about electronics. When I was in the 6th grade, my father helped me build my own high power amplifier using an LM3876 opamp. I remember clear as day the first time I powered up my amplifier circuit and nothing happened. Hmmm... I brought my face closer to my circuit board and suddenly "BANG!!!!" an electrolytic capacitor blew up right in my face and scared the living daylights out of me. I unknowingly connected the power supply capacitor backwards. I soldered in a new capacitor, and then plugged it in again. Like magic, my home-brewed subwoofer was now cranking out serious bass by an amplifier that "I" built. That moment was magical and surreal - I still get goose bumps thinking about it!


If you just want the amp without building it, you can get a TPA3116 Class D amplifier. It is only $21 USD from aliexpress (with free shipping of course!). It accepts a supply voltage of up to 24V meaning that you can just dig around for your old laptop power supply (19V). It comes with a nice and sleek aluminum case, and has low enough noise floor that I can't detect it at all unless my ear is right up against the speaker. Although slightly less power output compared to the chip from the article, it gets loud enough to drown out someone yelling. This amp from aliexpress has gold plated banana plug terminals, RCA inputs, and a smooth volume potentiometer.

Link: https://www.aliexpress.com/item/Breeze-Audio-BA100-Class-D-P...


I know this isn't what the original article is about but for quite some time I used a Topping TP31 amp, it's the same technology you are talking about.

It's not the best amp I've had, it's not loud enough for parties, or even very loud music, and it doesn't drive my planar headphones well.

What it does do is easily fill my home/office room well above comfortable working & listening level, it drives my Sennheiser headphones pretty well, it's cheap & small and it just works.

For what it is, it's brilliant.


TPA3116 can do about double the power of the TA2024 in the TP31 at about the same distortion.

The TPA3250 from the article, and the TPA3251/TPA3255 that have higher power from the same range are in a different class to those IMO, with much less distortion, and are just ridiculous value for money.


That's interesting, thank you. I got the TP31 as it's got both headphone and speaker out plus it's own DAC. I've had it a few years so will have a look at the newer ones. Thanks.


I guess in technical terms this article is about "making" an amplifier as opposed to "building" an amplifier. "Making" as in "Maker", since the electronics come as pre-made assemblies connected together and put in a case.


"If you wish to make an apple pie from scratch, you must first create the universe." --Carl Sagan


"There are only two colors in the world: black and white." --Ansel Adams


Basically a model kit that inhibits your ability to generalize.


The first step of generalization is copying.


Oh, that one looks neat. I've got another Ali module with integrated Bluetooth for use in my study room. Theoretically the parts have decent specs, but the module has a high noise level and the speaker makes a loud popping sound with a lot of movement on power on :(

Maybe I'll replace it with the amp you linked :)


Yes, you definitely don't want a bluetooth combo. Also, you don't want to buy the 100W version that combines two TPA3116 chips. These models are using a special sync pin/line, but when I bought this model, the manufacturers screwed up the design, it doesn't sync properly, and it's super noisy.


There also seem to be TPA3250 based amps all over Ali Express: https://de.aliexpress.com/item/NFJ-FXAUDIO-FX502S-PRO-HIFI-2...


I don't know much when it comes to audio amp design, but this reminds me greatly of the O2 headphone amp[0] - that anecdotally a few of my friends have soldered for their personal use.

[0] http://nwavguy.blogspot.com/2011/07/o2-headphone-amp.html


I thought of the same thing. Both are inexpensive, straightforward, no-BS projects that leverage the advances made in electronics over the years to deliver quality audio.

If I made one of these, I 'd be tempted to put an ODAC and O2 board in the same chassis to have a complete amplifier.


> If I made one of these, I 'd be tempted to put an ODAC and O2 board in the same chassis to have a complete amplifier.

Such a product is available commercially:

> https://www.jdslabs.com/products/48/objective2-odac-combo-re...


Hearing the ODAC name is quite a throw back! I remember reading NwAvGuy's blog when I first started getting into headphones. I always wondered what happened to him.


Given the way he suddenly went to radio silence, I just assumed he was hired by Apple...


O2 is great and it is quite simple to build. There is even simpler amp called CMoy[1] which is simple enough to be first project for somebody just starting out.

I would start by taking look on some oss version[2] and order PCB from OSH Park[3]. I've built this version few years ago and it works well.

[1] https://tangentsoft.net/audio/cmoy/ [2] https://github.com/akafugu/cmoy [3] https://oshpark.com/shared_projects/dRKK4dgN


I built the cmoy. It was easy. But I know nothing about how it works, and I remember reading warnings on associated web pages like "this is not a guitar amp" etc. I'm curious to know a little more about how these (amps generally) work, why they're seemingly non trivial to design at a reasonable quality level etc.


Guitar amps are very different beasts. If you want to build something simple there is "1 watt guitar amp" project[1] with many different variations.

[1] https://www.electrosmash.com/1wamp


Yeah thanks but it's not that I want to build a guitar amp; I'd just like to read a little about the difference between guitar and headphone amps, different design decisions for headphone amps generally, things like that.


After I started building my own audio gear, I am no longer impressed by what you can buy in the consumer market, no longer how much "audiophile" these systems claim to be. Usually, it's just a simple 2-way boombox sold for $xxx - $xxxx. I started my project after being disappointed with the sound quality of the Apple Homepod. So, I built my own version with inbuilt basic voice command integration. And I couldn't be happier. Funny, that I just finished building mine this week[1][2].

It's also based on Tubes, and the sound is very, very satisfying. Will be happy to share a detailed guide, if anyone's interested.

[1] https://www.instagram.com/p/Bpl71mUHLbm/?taken-by=discovery....

[2] https://www.instagram.com/p/BplhIIqHNhl/?taken-by=discovery....


That looks awesome, post the guide please!


Alright, will post it soon on my Medium1: https://medium.com/build-ideas

Thank you :)


Oh geez they could have left this statement out: "When I wrote my previous article, class-D audio was fairly new, and audiophiles were still arguing about whether class-D amps could sound as good as class-AB or class-A units.

Nobody is having those arguments anymore."

Those arguments are still front and center when it comes to amplifier design, both on the manufacturer and consumer sides.


I'm not convinced audio engineers are having that argument anymore (which is more the audience of IEEE). I get the impression from an engineer I know who's specialized in digital signal processing, and who's done a significant amount of analog work, that anyone who's still having some of these arguments doesn't understand some of the basic maths of the problem space. That basically, you can always digitally encode an analog signal at a higher resolution and fidelity than you can reliably read said signal at any given technology level/pricepoint. Which means that using an ADC -> DAC pipeline will almost always give you better audio output with better signal to noise ratios than strict analog -> analog systems (though possibly at the cost of latency).

Where it gets tricky, is the arguments for how things like tube amps sound. There is nothing that analog components add to a sound that can't be simulated with digital transforms, but the trick is figuring out how various analog components effect the signal. For instance, my friend found that certain tube amps would overdrive components and force them out of spec, and this caused a non linearity in the response of the signal coming from the component that gave the amp it's distinctive distortion effect. Once he figured that out he was able to reproduce the effect digitally. Incidently, he's found that some guitar amps can never be made today, because they were built around a specific batch of a given component, and newer batches didn't have the same out of bounds responses. But this seems to be why so many people are still so big into analog amps, because it takes considerable effort (and apparently engineering skill) to reproduce some of the classic sounds.

Mind you, I'm biased by one of the best engineers I've ever met, so I could be wrong about the general consensus with audio engineers. I'm also mostly using a sample size of 1, though I have corroborated some of his assertions with some of the blogging I've read on the topic. One of the best (I wish I'd bookmarked), was a good overview of encoding by a prolific codec writer, who pointed out that even though a digitized sine wave looks lossy at a low sample rate, it's trivial to rebuild that sine wave perfectly, with just the digitized version (so long as the sample rate was above some multiple of the Nyquist frequency?).


Likely that was xiph on why excessively high sample rates don't do much for the listening experience: https://people.xiph.org/~xiphmont/demo/neil-young.html


Amplifier designers most certainly are. Class D has significant energy efficiency advantages, and significant sonic challenges. It’s taking an analog waveform, whether decoded by a DAC or starting as pure analog, and chopping it into a series of pulses (pulse width modulation). The amp is switching transistors on and off very quickly, rather than using them as a variable linear gain stage. It’s not the same as PCM encoded digital audio, and shouldn’t be confused with it.


> Amplifier designers most certainly are.

Amplifier designers have to sell to a market that is rife with misinformation and a very poor signal-to-noise ratio between technical quality, profit margins and overall market success. I wouldn't expect many of them—especially the ones who have any name recognition—to be concerning themselves with objective technical performance.

Whereas the guy who designed the amplifier circuits for a product where the customer isn't being placeboed by the spec sheet—like powered studio monitors, or consumer devices like the Apple HomePod—is far more likely to be focusing on objective measurement rather than faulty market perceptions.


The energy efficiency advantages in class D are incredible. When my old Technics receiver stopped working because it started overheating, i replaced it with a class D amp on a half-credit card sized circuit board from dx.com and taped it to the back of the TV. 20W per channel, which is plenty for my casual TV watching. The Technics was the size of several stacked pizza boxes, with fans and heat sinks etc.


The thing is that transistors’ behvior is more linear and controlled when fully on or off rather than when used as continuously variable. Class-D can and does sound _amazing_ if built with good theoretical understanding. Check out Hypex amps designed by Dutch engineer Bruno Putzeys on strong control theoretical foundations.


That’s precisely and exactly the opposite of what ‘linear’ means when talking about the operating point of transistors. I think you should rephrase your argument.


I'm going with an understanding similar to what is presented e.g. here: https://www.mwrf.com/components/learn-meaning-amplifier-line...

"The term linearity derives from an amplifier's linear relationship of input power to output power which, in an ideal amplifier, would be precisely related by the gain of the amplifier."

I see it as matching my argument.

Could I ask you to give a definition which clashes? Thanks!


Yes, the definition of linearity is that f(A) + f(B) = f(A+B), which you have contradicted in your GP post. If a transistor is fully on, then doubling the input voltage to it will not double the output, because it's already at maximum. That is very precisely not a linear device. That's why linear amplifiers, like Class A, operate the transistor in a partially switched state, where increasing the input signal (about some bias point, and with feedback to turn a real transistor into a more ideally-behaving one) gives a corresponding increase in output signal.

Class D (switching) amplifiers are _not_ linear, they trade linearity for efficiency. The article you linked to actually says this:

"Over the years, a number of different biasing schemes have been developed for amplifiers, from extremely linear Class A operation, in which the active devices essentially remain powered on through the full sinewave cycle of an input signal, to more power-efficient schemes, such as Class D, in which multiple transistors are switched on and off to conduct different portions of an input waveform."

I think what you are doing is confusing the linearity of the transistor with the linearity of the whole audio amplifier treated as a black box, which is why I suggested you rephrase your argument rather than say you are categorically wrong, however what you said was:

"The thing is that transistors’ behavior is more linear and controlled when fully on or off rather than when used as continuously variable."

... which I'm afraid is really just wrong.


Thanks but no thanks.

When a transistor is off, it’s off. Then it takes a certain input voltage to get it going. Then there’s a varying curve to full output.

In contrast, fully switching on or off “gets rid of the curve” and tranforms the problem into one of timing, and one of a modulation scheme for a series of on or off states. In that respect I contend that it is an useful thought to consider transistors’ behavior (in an amp) as more linear and controlled if used in only the on or the off state. The output quality of Class-D amps in relation to price, size, complexity, and power, and cost suggests to me that the gains are not only in power efficiency.

Class D amps do leave the operation of transistors in their linear domain on the table, and do gain efficiency by that. Overall linearity as a control mechanism is something I stand by saying is increased. You may present an alternate take of course.


I think you are confused. To get a linear amplifier, you operate your output transistors in a state of Saturation. This gives a linear transfer function to the drain based on the Gate to source voltage and you use a method of choice to hold the source constant. This is what a class A amplifier does. An AB amplifier has 2 transistors to extended the range of inputs on the gate for which the output is linear at the cost some distortion when switching between the A and B (hi and low side) transistors.

A class D amplifier uses a smaller amplifier internally (which can be a single transistor or more complicated) that is driven open loop so that it is either hard off or hard on but very fast and efficient. The hard on state is known as full compression because the output power will not change with input power. This is purely non-linear behavior and can be very easily verified by measuring the high IMD components it generates. Class D amplifiers employ significant filtering to suppress these non-linear terms and allow you to recover the signal. Mathematically, they violate the linearity condition of F(ax) = aF(x)

Also you should reread the article you learned from. It directly contradicts you.


It’s not about me; The “you”-ing is incredibly offputting.

We are simply and obviously applying the word “linear” a little differently. I am aware that it is a specific term for certain behavior / a certain operating mode of transistors. I’m referring to it in another wider context on purpose. As another example “linear” in “linear algebra” isn’t about saturation or gates. I’ve gone through this before in this thread and believe my perspective is now pretty complete and clear here and won’t engage further.


It's an interesting contradiction to build a more linear amplifier by intentionally using the transistors themselves in a non-linear fashion.


It is :) But it’s also only a contradiction on the surface of the words. The amplifier is quite linear even though the transistors aren’t used in what’s called their “linear region”, but I’d say they’re still used in a linear fashion.


Actually the opposite is true. Transistors are approximately linear in half-open state, as in Class A, and non-linear at the extremes (when either fully open or fully closed).


Tangentially, the placebo effect is everywhere. If it can affect tasting (wine) and hearing (audiophile gear), I'm surprised not to have noticed its affect in discussing photography.

Will "computational photography" (as touted in the latest iPhones) that alters photos based on criterion of perceived quality end up bringing these kind of inane discussions there too?

Edit: Never mind, I didn't factor in all the arguments about Photoshop altering.


I made my living as a mastering engineer for two years...

To do good work in mixing and mastering, it's important to study the human perceptual mechanism and its propensity for illusion, and it's also important to understand masking effects and how to amplify certain components long enough to tweak them before dialing them back to recede into the gestalt.

It's also important to understand that in sound reproduction the vast majority of the distortion on the analog side comes from transducers (microphones and speakers), not amplifiers. (I'm not even going to bother with the article, because amplifiers aren't an important problem: buy something decently engineered and overpowered, and you're basically all set. Then worry about your speakers and your room, because they're the hard part.)

But it's likewise crucial to grok that the distortion introduced by digital components is disproportionately disturbing because it is enharmonic. And that practically speaking a lot of the damage that gets done in practice happens because people don't handle gain staging right and often don't understand the fundamental principle of preserving high resolution throughout production and only exporting to lower resolution as a final step.

In summary, there are a whole lot of interesting conversations to have on the subject of audio and how perception relates to production. But it's basically impossible to find a community to discuss these subjects with. Either it will be overrun by audiophiles and their pseudoscience, or well-meaning but undereducated skeptics will dominate the discussion with plausible but misguided debunking.

So every once in a while I throw up a piece like this into the middle of a giant discussion, with about as much hope as tossing a message in a bottle into the ocean.


I feel your pain. I produce and mix records - at an amateur level, but enough to get quite good at it. It makes me want to toss all this "objective", "measurable" bullshit from the skeptics right out the window - the subjective experience is all that matters. But it also makes me want to toss the audiophile voodoo out the same window. "Accurate" reproduction? I just want to show them how records are mixed.

My pet example is that I put high-pass filters (12db slope) at 250-300hz on almost everything that isn't a drum or a bass instrument. I'm slaughtering the bottom two octaves of guitars! And if you solo a guitar track, you can hear it. But in the context of a mix, all those bottom two octaves do is make mud and confusion, distracting from the clarity that the kick and bass (guitar || synth) needs. And I need that clarity, because I'm not mixing for $500,000 audiophile systems. I'm mixing for car stereos, boom boxes, Alexas, cheap earbuds, and worse.

So all that matters for me with audiophile systems - and I love audiophile systems - is that they sound good. That doesn't mean they sound accurate.


I agree that your high-pass filters are an application of masking principles. It would be challenging, especially for an untrained person but even for someone who is trained, to perceive enabling/disabling of those filters in the context of the full mix. Instead we solo to isolate the track being adjusted so the change is easily audible -- i.e. no longer masked by the other tracks -- and adjust the filtering anticipating the counter-intuitive but artistically desirable effect when the filtered track is blended back into the mix.

Despite that, your hostility to the skeptics illustrates that you and I still aren't really on the same page (though I don't doubt you produce compelling art, don't question your techniques, and agree 100% about ensuring that aesthetic gestures survive the transition to imperfect end-user environments).

To me, "objective accuracy" is a worthy and desirable intermediate goal for sound artists, a tool which can be leveraged for creative ends the same as a baker understanding food chemistry or an animator understanding Newtonian physics or a modern origami folder understanding programming. However, chasing "objective accuracy" means understanding the wildly non-linear, even bizarre human perceptual mechanism -- which is more of a pattern recognition machine and definitely not a measurement device -- in addition to all the strangeness of acoustics.

The problem with audiophiles is that their approach to "accuracy" is unscientific and doomed from the start. The skeptics on the other hand, have the right idea -- but those who have not done deep study often fail to appreciate just how difficult the problem is and how far off their intuitions are.

I've had some wonderful conversations with individuals whose perspectives are similar to mine, but it still seems impossible to find a hospitable public forum.


When I say "objective", it definitely needs air quotes, because the well-meaning skeptics love to toss it around for some decidedly non-objective opinions. My pet example is quoting "THD" as if it were a useful figure. It's not. The advantage of THD isn't that it is meaningful, but rather that it's easy to measure. Run a 1khz sine wave into the front of the amp, tap a resistive load on the back of the amp, measure the difference. Manufacturers like it because it makes for a competitive-sounding spec. But it has approximately zero to do with how an amplifier behaves in the face of complex musical signals, driving complex speaker loads.

When it comes to recording, my "objective" reality is the sound I hear in the room before it ever hits a microphone. But what comes out of the other end of the microphone is already deeply subjective. The objective ability of hi-fi systems to reproduce that subjective, colored microphone signal (plus whatever processing is involved downstream) is about as objective as they get. But the recording itself? Outside the world of purely electronic music, it's a subjective experience.

I play acoustic guitar every day, listen to unamplified singers/musicians (other than me) at least weekly, play electric guitar at least weekly, gig regularly. These direct, objective experiences of the natural sounds of instruments color my interpretation of both magical-thinking audiophiles and pseudo-scientific skeptics. I know intimately what the natural sound is, I know intimately what the recorded/reproduced sound is, and I know how the entire recording process works, from miking to mixing to mastering. Laws and sausages. The audiophiles and skeptics are both missing that perspective.

So, as a recording/mixing engineer, as a producer, I'm looking not to deliver audiophile accuracy, but rather to deliver the intended intellectual and emotional experience of the music. What feeling is the artist trying to convey? How do I manipulate the sonics to emphasize the musician's intent, as expressed through the lyrics, the composition/arrangement, and the natural tones of their instruments? That's the fun part, for me.

(As an aside, I'm currently working on a particular song of my own, destined for my long-delayed solo album. I wrote the song based on a nightmare I had in great distress. A few years back, I suffered an illness that left me nearly unable to speak, much less sing, and my voice is permanently damaged (I have surgery a few times a year on my vocal cords to keep speaking and not die of suffocation). I recorded the song right after writing it, just a couple of weeks before my first surgery, with a near-useless voice. "My broken voice won't fill the air / I know that you don't really care / You may not need my song but I need to sing it anyway". Fitting with the broken voice is the melody repeated on a piano that hasn't been tuned in over 50 years and has been flooded multiple times, so it's, um, colorful. To both the audiophile and objective mindset, neither the voice nor the piano were worth recording at all. But as an artist trying to communicate an emotional experience, they're vital.)


This is pretty fascinating. I can't add to a discussion, but I would certainly read more about this if you have produced it somewhere.

I'm also curious, what made you leave your job as a mastering engineer?


The profession of audio production is extremely competitive because there is a perpetual oversupply of labor. The total amount of money in the music industry is also deceptively small compared to its high public profile.

I left the field (back in about 2003) because although I felt that I was good enough to compete and survive, I didn't love it enough to grind out a life as a workaholic and just get by.

Most of my writings from that time were forum posts with the perspective of the well-meaning but undereducated skeptic and wouldn't be that interesting to read today. Also, what was probably my best writing happened in forums which don't have public archives which have survived 15-20 years later. (Which amuses me as a former mastering engineer because one of our primary objectives is long-term archival.)

PS: not "enharmonic", "non-harmonic" dammit. The point is that while analog distortion is mostly harmonically related, digital distortion is not and therefore is much more poisonous in small quantities. Mixing especially (though mastering is similar) is all about making changes near or below the threshold of perception which sum to a cumulative perceptible effect.


While there are endless debates about manipulation, when it comes to actual capturing of images, the photographic community (in my experience) has a healthy dose of "the best camera is the one you're carrying" and "a great photographer can take great photographs with any camera but even a great camera won't help a lousy photographer take great photographs" to go with its love of gear (who can help but love gear?)


Camera nerds are the worst. They're worse than audiophiles, if that's possible. Good photography requires nothing more than a bit of reason and observable standard practice, applied to aesthetics. You don't have to look hard to find someone with a $5000 camera/lens combo who doesn't know what the Rule of Thirds is or how to apply it.

Last weekend, I was in DC, and snapping pictures of the Washington Monument on my phone. My sister was shocked at how good they were. It's not hard, really.


Girl at work has 20k camera with 100k megapixels, but then again Huawei just asked to use her pics for their phone screen savers, so...


It's already here, for displays. I've seen lots of internet comments claiming I need the >450ppi screen on the new high-end iPhone, not the miserable old cheapo 326ppi one.


Whenever I see the term audiophile I mentally replace it with "over priced bullshit"

It's just someone being sold snake oil.


I think it might boil down to your philosophy on sound reproduction. Do you want accuracy with some potentials for ugliness if you clip, or do you want coloring and musical distortion? I don't think AM radio sounds good at all, but it is sometimes emotionally comforting to listen to. Fidelity isn't everything.

That said, many high end audio folks are straight up whack jobs.


Google audiophile headphones. Then go to sweetwater and search for studio headphones.

If the price of source gear is a quarter of the price of gear attempting to reproduce source, something is broken.


I like to look on high end headphones as I would any other luxury item - I'm willing to pay for nice materials of a higher quality or a better experience, if the item is right.

That could perhaps explain some of the price - plus I think we're at the point where many of the headphone companies have discovered that there is an ultra-high end market for some of these products.

You can see this in the prices of many top brand flagship headphones - the majority of the well regarded brands have been gradually putting out exceedingly expensive flagships in the past 2-3 years.

10 years ago - then the Sennheiser HD650 was seen as a flagship at $400, now look at the Sennheiser HD820 released earlier this year at $2200.

You aren't going to find too many people who will say that the sound improvements between the two justify an $1800 increase in MSRP.

However if you reframe the perspective to that of a luxury good, given the perception of headphones (regardless of use case) has changed somewhat in the interim and are seen as a fashion accessory and there's a larger enthusiast pool - you can see that manufacturers would be willing to push new boundaries when it comes to price.

However I think it's only the lunatics, or the conspicuous consumers who are ignorant to the law of diminishing returns when it comes to headphones and audio in general.


Some people understand the law of diminishing returns, but have the resources to not care. If a $2000 pair of headphones is no more real financial pain for you than a $200 pair, and they're better despite diminishing returns, then why not get them?


Very correct, that's a perfectly reasonable explanation, especially in terms of the usual consumers of high end luxury goods.


I needed a good set of headphones a few weeks ago, as my work environment has changed and I have more background noise than before. Instead of trying to find something audiophile (I have Grados already) I went to Guitar Center. They had a huge selection and in 30 minutes of trying out different models I had something I was happy with and could leave in an unlocked drawer.


Modern class D amps are terrific, but it's annoying to see THD brought up as a measure of audio quality. THD is terrible measurement - its value is not that it's relevant (except to demonstrate things aren't actually broken), but rather that it's easy to measure.

An amplifier's ability to reproduce a 1khz sine wave into a resistive load tells us nothing about how it behaves in the face of 8-10 octaves of frequency range and the real dynamic range and harmonic complexity of actual musical content, when hooked up to the absurdly complex and nonmusical behavior of real speakers. The 5% THD of single-ended tube amps that people who've never heard one tell us proves how "inaccurate" they are is inaudible, but the grimy, dull sound of cheap amps that have excellent THD specs is plain to hear.

It's frustrating, in part because this stuff is, so far, mostly impossible to measure. It's hard to get a sense of how an amplifier will sound from specs.


I'll never forget my wakeup call on the differences of sound between amplifiers. I had been listening to headphones on my Techniqs receiver, a well regarded unit with a THD of 0.04%. For some reason I plugged the headphones into my Nakamichi cassette recorder instead, and was blown away by how much better they sounded. Until then I had thought that amplifiers were a solved problem and the inaccuracies of transducers like headphones would mask the tiny differences between them. I never expected to be wrong.


The majority of all active speakers produced nowadays are Class-D, it's nothing really new (and the fact they are super cheap).

No idea how this guy wrote this article and didn't come across DIY Audio (or he just didn't mention it in the article). There is a tonne of information here: https://www.diyaudio.com/forums/class-d/


If your listening needs don't involve more than 10-20 watts, inexpensive Class-D amps can sound really good through a pair of good speakers. And because of their efficiency, an external power supply only needs a handful of amps (much lighter). Win-win for me.


TPA3255 can do 75 Watts per channel at 0.003%THD+N, which is pretty damn good IMO.

3e Audio has a nice board using that as well.


I have some questions. My son has been taking guitar lessons using a classic guitar. Using the advice of his instructor we are getting him a begginer's electric guitar for christmas. We also need to get him an amp.

Is the amp in this article the same kind of amp that you could plug an electric guitar into? I know nothing about music or instruments, but I do know about electronics so I was thinking it might be fun to build an amp with my son.

I went to the diyaudio link you suggested, but I didn't immediately see information like in the article (buy this, this, this and this to build the amp). Can you point to where I could find a similar purchase list and assembly instructions on diyaudio? Thanks!


Short answer: No. Guitar amps are weird, not like hi-fi amps at all. In a hi-fi amp, distortion is the enemy. In a guitar amp, distortion is sought after. Driving the pre-amp section of a guitar amp into clipping is extremely common and desirable. Doing the same with a hi-fi amp is not. A hi-fi amp will be typically connected to speakers with a woofer and a tweeter and maybe a mid-range, trying to accurately reproduce all frequencies. A guitar amp typically has nothing but 12 inch speakers. A guitar amp is literally half the instrument. Plugging a guitar into a hi-fi amp is faithfully amplifying 1/2 of an instrument. You'll end up buying a guitar amp anyway. Just get a guitar amp.


> A guitar amp is literally half the instrument.

I'm not a musician and this confused me for some time. I'd see electric guitars with their own amp & speakers then a microphone for the the PA system in front of said speaker. I couldn't understand why the electric guitar wasn't just plugged in, something I read a few years ago was a light-bulb moment.


[Professional Musician / Hobbyist Maker]

Yes, this amp can work — experimental musicians run instruments through much weirder things — but it'll sound strange.

Whether you build an amp or not, you should definitely buy a standard guitar amp, so that he can focus on his guitar playing. A standard guitar amp (solid state or tube, to taste) will let him hear how his playing compares to other musicians (who also use standard guitar amps), helping him gain a critical ear. He'll learn what parts of the tone come from his fingers.

When he's ready to go experimental, then he can start playing with (say) ripping pickups out of guitars and wiring them up to headphones as portable EM detectors. When he gets to that phase, that's the time to start using line amplifiers like the one in this article instead of instrument amplifiers. And drilling hundreds of screws into his guitar.


While you probably could use an amplifier like this for guitar, you usually wouldn't want to unless it's being used for acoustic guitar or vocals. The reason being, an electric guitar amplifier is going to add a lot of character to the guitar tone that you actually would want, where as an amplifier like this is going to strive for a perfect clean reproduction of the original audio. This is why a lot of guitar players prefer to use old tube-based amps, they add a lot of character to the sound through the preamps and power amps that match what you hear on the recordings of your favorite artists and sound great (though, some argue today solid-state modeling amps sound just as good). With that said, there are a lot of projects and schematics on the web which try to recreate some of the popular guitar amps like the Vox AC30 and Fender Twin Reverb, which would be fun to try and build.

Now if you wanted to make a PA system, I don't see why you couldn't use an amplifier like this with some modifications.


You couldn't, not without reengineering the circuit. Guitar output level is much lower than the line level a standard power amp will take.


> Is the amp in this article the same kind of amp that you could plug an electric guitar into?

I'm no expert, but as far as I understand, this is a power amp, which takes an input at line level (~1 volt) up to a level that can power a speaker. Guitar pickups produce a signal around 100 mV, and typically need a "pre-amp" stage before the power amp to amplify this signal to line level.


An amplifier like in this article does just one of the many things a normal combo guitar amp does.

1. Preamp with hi-Z input from guitar to line level

2. Ability to drive into distortion

3. EQ of some sort

4. (Sometimes) reverb

5. Amplify line to speaker level (this article)

6. Midrangey speaker+cabinet that sounds nothing like a normal flat speaker cabinet used for HiFi/PA/keyboard/etc.

However, as others have said, you can use a guitar modeler which performs steps 1-4 outright and simulates step 6 leaving only a need for step 5 and a "normal" speaker.


I used to build guitar amps for fun and repair them for friends. I have parts of a Sound City, Plush, Harmony, Gibson amps in my storage unit. I have some old heathkit thing from the 60s. I have two different Traynors.

Depending on your budget, I'd get a Vox Pathfinder, AC4, or AC10, and then take him to get a pedal or two after Christmas. You can usually find an AC4 used easily. Some of the digital amps are fun, but if he really gets into it, he's more likely to keep one of the amps I mentioned and buy pedals anyway. An AC4 (especially an AC10) will be something which usually sounds really good on it's own.


You can absolutely use stand-alone "hi-fi" amplifiers for guitar; that's what I do. You just can't plug your guitar straight into that; you need some pre-amplifier gear. My gear is old-school in a big rack, described below.

Nowadays, there exist units that provide complete signal processing for guitar: simulating guitar pre-amplifiers, amplifiers and speaker cabinets digitally. You plug your guitar into this, and then it can go straight into a PA system, or your recording console. Something like that can be used with an amplifier like the one being built here.

My rig consists of a 6U portable rack case with a few pieces of gear. The guitar goes into an ADA MP-1 MIDI-programmable tube pre-amp (built in 1987, revamped by me). This goes into a splitter-mixer (called the SMF-1 that I built myself) which bifurcates the signal through a digital effect unit that provides reverb, and then mixes it again. The dry signal remains analog. Then it goes into a 31 band equalizer, and from there to the power amplifier, an Alesis RA-100, which feeds a guitar "4x12" speaker cabinet (four 12" speakers). I could take out the amplifier and drop-in replace it with any rack-mountable amp, provided it doesn't require more than 3U spaces. QSC, Yamaha, Yorkville, whatever.

There also exist dedicated rack-mounted power amplifiers intended for guitar, like the Rocktron Velocity or Marshall Valvestate 8008 (both of these examples are stereo). Many of these kinds of ampls have been produced. These are just power amps; they can't produce any overdrive tones and have only minimal tone control or none at all. The 8008 has a switch on the back for "tube like" behavior, which just means current feedback. That will give it more of a mid-scooped frequency response, induced by the reactive speaker load. I can do something like that with my equalizer.

The amplifier project we see here could be turned into a guitar amp by adding a preamp circuit board into the chassis, fed by a guitar input jack, and exposing some tone and gain controls. The stereo function of the amp could be useful if we add jacks for an external effect loop and make it stereo (mono out, stereo in). Then we need a stereo guitar cabinet: minimally a 2x10 or 2x12 wired for stereo.


For a beginner, I'd get a decent, pre-built guitar amp that "just works" with decent sound and only basic controls (as opposed to a digital amp with millions of mediocre effects). You can always add some DIY effect pedals, or get a whole new amp - but if your son wants to practice and needs to fiddle with the electronics because something is not right again with the DIY amp, this might distract him too much from actually practicing his instrument.

Maybe take him to a store so he can find out which guitar/amp combo he likes most.


No, guitar amps are a totally different beast with many different components.


Amen. Ask a guitarist. Kid needs a real guitar amp. And a couple of decent FX pedals.


Well... I dunno about that. Digital has gotten really really good. I bought a Helix about a year ago and sold my tube amps a few months ago... not that they were BAD, but I could get all the sounds (and way more) out of the Helix.

But obviously a new player isn't going to want to drop $1500 on a rig like that.


I would go with one of the new generation of solid state guitar amps like the Boss Katana. These are Class D amps like the one in the article, but voiced for guitar, and with distortion and reverb effects built in. They're very reasonably priced, and get good reviews from enthusiasts (much better than the solid state amps of 10+ years ago).


Not very cheap and probably outside OPs range. But the Boss Katana Air is an amazing little amplifier, Bluetooth speaker, pedal modelling device etc. And it allows you to connect your guitar to the amp wirelessly. It really works fantastic.

https://www.boss.info/us/products/katana-air/

I'm but an amateur guitarist. I can't detect any latency caused by the radio communication. But I have not been able to find pro muscicians trying out the Katana Air and complaining on the latency. It really works. Over quite long distances too.

The only thing I would have wanted in the Katana Air is rechargable batteries in the Amp itself (the transmitter is charged by the amp). But that is not a major issue.


The Boss Katana is an ideal beginner amp that can handle bedroom practice through small gigs.

It's extremely versatile and sounds great.


Your practical options are:

* Get a standard amp/cab combo.

* DIY something, I'm sure there are guides and such you can follow, but this is only practical in so far as it's a fun project.

* Use your computer. You can get inexpensive cables that plug into your computer USB and there's a variety of software to emulate amps.


Buy a Zoom G3x (or other good multi effects box with amps Sims) you can then chuck that straight into any semi decent guitar amp.


Your first point seems to be reflected in the article, except that, if they are correct, the high end of Class D amplifiers apparently is not "super cheap".

I don't quite get the aggressiveness of your second point: surely just leaving whatever your first association to a topic is out cannot warrant such outrage? It's not like they present this article as groundbreaking innovation. Not only do they cite other DIY efforts. The whole premise of the article is that the author himself already did this ten years ago,.


It comes more down to the design of the board and the quality of the other components on the board. The TPA3250 he mentions is quite a new chip, but if you look on Aliexpress for an older chip like the TPA3116 there is literally thousands of examples: https://www.aliexpress.com/wholesale?catId=0&initiative_id=S...

I imagine this guy must have found the components of a site similar to DIY Audio. It would have been a nice addition to steer people in the direction of more information if they are interested in the topic (which a lot of the comments here seem to be asking about).


> Nobody is having those arguments anymore. There are many class-D amplifiers on the market now, and the best of these are dominating the upper reaches of the high end.

Not quite. Frankly, Class A or AB still has the upper hand from a pure sound quality perspective, and class D struggles to output sufficiently at high levels with lower levels of distortion in my experience.

That said, there have been many small class-D Amps from Lepai and Dayton that are excellent choices for near-field and non-demanding applications. They definitely have their place. The tech is advancing rapidly but it has not supplanted A or AB.

If you want some background in why there is still debate, I highly suggest reading some of Nelson Pass’ interviews and writings:

https://www.google.com/amp/s/parttimeaudiophile.com/2018/02/... https://www.passlabs.com/press/single-ended-class https://www.passlabs.com/press/leaving-class


This may be true of Class D amps commonly used in home equipment, but many high end powered PA speakers use Class D amps and get very loud with minimal distortion.


You'd be surprised how much you can get away with in a PA.

Being powerful and reliable is more important than being all that good, since the room always sounds bad anyway.


It's bizarre how none of these chips seem to have a digital input. All my music is digital, and despite the modern rise of vinyl, I don't think that's unusual. If I want an amplifier like this then I'm forced to have a completely useless DAC/ADC pair degrading the signal quality. Does anybody know of an easily available class-D amplifier with digital input? You'd think that would be an obvious feature for a digital amplifier.


> You'd think that would be an obvious feature for a digital amplifier.

That's because these amplifiers are not digital. The control loop is generally analog. Though there are chips which do in fact use digital inputs and do most of the work via fixed-function DSP, but those are mostly bigger chips targeted at 5.1 or 7.1 systems.


From Wikipedia: "The term "class D" is sometimes misunderstood as meaning a "digital" amplifier. While some class-D amps may indeed be controlled by digital circuits or include digital signal processing devices, the power stage deals with voltage and current as a function of non-quantized time."

So what's it called if it is time quantized? The simple open-loop design depicted on the Wikipedia page could have the analog comparator + triangle wave generator replaced with a digital delta-sigma modulator (1-bit DAC). Then the switching times of the power stage would be quantized to the sample rate multiplied by the oversampling ratio. (Although it seems this open-loop model is an oversimplification, and real class-D amplifiers have feedback, so it wouldn't be that simple in practice.)


All stereo components, except the amp connection to the speaker, should be ethernet. Turntable, CD player, cassette, pre-amp, everything.

Then make it wifi.

Just think, you could eliminate that awful rat's nest of cables. The sound quality would be better. You could locate components where-ever convenient. Each component could have a web browser interface.

It'd be awesome.


Digital, sure. Standardized even, sure. But wifi, going through who knows what network stacks, routers, filtering, and sharing with other traffic? I'd like some guarantees on latency, please; I like my video in sync with my audio.


It would be on your home LAN, so presumably you know what else is using it. I stream video from my computer to a Roku box over ethernet, video consumes far more bandwidth than audio, and it works fine.

Also, if you plug your audio components into its own switch, it should not get interference from the rest of your LAN.


Sonos, d&m heos and Yamaha's multiroom iirc do half of that and yes, latency must be tightly controlled. I imagine with sources also being 802.11 it'd be much harder to keep it in reins.



Linn's exakt system is a little like that - the signal is sent into each speaker in digital form and then decoded and amplified there.


It's not a digital amplifier. A class-D amplifier is basically somewhat like a switched-mode power supply that varies its output voltage in response to swings in the input signal.


They don't have it because they aren't digital at all; a Class D amplifier essentially is to amplifiers what a switching power supply is to power supplies. They're still 100% analog, but they use a different driving method compared to their traditional linear counterparts to maximize performance.


Uh, if you want to "amplify" a digital signal, just multiply all the samples by some number.


Don't they make these ICs with I2S? Then all you need is something that digitally interfaces with your source (eg toslink, usb) and "unpacks" the PCM data and sends it out via i2s.

Edit: look for example at TI TAS5756M, TAS3251 or TAS5825M.


Thanks, those are the kind of chips I was looking for. And S/PDIF to I2S transceivers are also available, so they could be used with common consumer audio equipment.



I kind of want to build one of these amps and connect it to a set of speaker built with these cheap exciters... One of the Youtube channels I follow messed around with them https://youtu.be/zMCyF400l1c but he got that from this other video https://youtu.be/zdkyGDqU7xA where he goes into all sorts of experiments he ran on the exciters to find the right design.


Any recommendations for people that are tempted to build this but should really put their time elsewhere?


Just do some research and get a cheap class D amp off a chinese website. I forgot the one I chose(I'm at work now), but other than some sparks if you jiggle the power connector(haha seriously. can be fixed with a tiny bit of effort), it is an amazing little amp. It also supports optical and bluetooth connections which is really nice.

I put my efforts into building speakers instead. I got a cheap kit (https://www.diysoundgroup.com/overnight-sensations.html) and spent my time learning how to apply a french polish finish. I really enjoyed learning french polishing. It's one of those things that when you finally get it, it becomes immensely satisfying. Yeah, shellac isn't the most durable finish, but it sure is pretty.


> Yeah, shellac isn't the most durable finish, but it sure is pretty.

+1


If you are looking for something already built, the eBay/aliexpress shops of the Shenzhen companies often have the option of selling you a complete amplifier made up of their modules. Which will probably end up cheaper than getting all the parts yourself...


Bang on. A popular amp is a TPA3116 module. You can see thousands of examples on Aliexpress: https://www.aliexpress.com/wholesale?catId=0&initiative_id=S...

For speakers, the cheapest option is to buy some second-hand ones on eBay/Gumtree/Craigslist. Just Google model names and look for ones with good reviews. I like British speakers, so I usually search for Tannoy, B&W, KEF etc.


I'm not sure I understand your question; what do you mean by "but should really put their time elsewhere"?

Are you asking for an amplifier that would deliver similar performance to this, because you don't have the time to build one, or that you suspect you'll come out ahead on build vs buy because of the price you value your time at?


There are these I've been tempted to try out that use the same chip (and a low noise pre amp), but unfortunately I have two perfectly functioning amplifiers already :-) (and a set of powered edifier speakers in the mail - I have a problem). https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2047675.m...

They have a switch mode power supply instead of the toroidal in the IEEE article, so in theory more noise - in real life? probably doesn't matter. THD < 0.01% at 30W, frequency response is +- 1db, the days of having to spend any money for an amp seem gone, < $70


I was planning to build my own system, I wanted a DAC + Tube Preamp + AB Amp. You can get a really good DAC board from https://www.hifiberry.com/shop/boards/hifiberry-dac-dsp/ and for the rest of the system I was going to get it from http://glass-ware.stores.yahoo.net/newhardware.html.


People discard old electronics, some of them of very good quality. Short of that, you can find this stuff on craigslist. The used value of some nice gear is very modest. I'm running one set of my chromecast audio linked speakers on a vintage Pioneer receiver with those funky proprietary speaker jacks I found on the "take it or leave it" table at the dump. I also found a McIntosh with a funky combination of a tube tuner and transistor amp literally on the scrap metal pile. But it needs new power caps.


[for guitars] Depends on what you consider professional. We've made a simple, cheap, modular project that works in hard conditions for a long period of time and has enough quality for a gig.

https://medium.com/@leolorieri/underground-bands-get-more-wa...


Very cool. I was very interested when the IEEE Spectrum original feature was published, and got one of the Tripath amplifiers to play with. It was a commercial product, marketed as inexpensive toy. Then people on Head-Fi Forums started taking these things further. One of the forum members started modding home-theater amplifiers, so I purchased one and it was great.

Vinnie Rossi, Red Wine Audio. It is still amazing. But his stuff is now quite out of my price range.

These amplifiers are great fun, though.


How would one best control the output volume on such a DIY amp? 2-channel logarithmic 10K potentiometer in front of both inputs?

What if you use balanced input?


If your input is balanced, it will be handled by some op-amp circuit or whatever, which converts to single-ended internal signal, like the "balanced line input module" shown in this page:

http://sound.whsites.net/project30a.htm

You can easily insert a volume control right after such a stage.


Yeah, but in the context of this article, I'd have to cut some traces on the PCB to get there. Doesn't sound like the ideal solution to me :)


Perhaps, perhaps not. The schematics don't seem to be available, but let's imagine there is a stage in there which looks roughly like this:

    |\   C        |\
    | >--||-+-----| > 
    |/      |     |/
            <
            < R
            <
            |
            V

If we simply remove/desolder components C and R, we should be able to then install hookup wires for a potentiometer (which includes the C also).

It looks like this particular board has tiny SMT components, which make this sort of thing a pain in the but.


The 3e Audio board has jumpers to choose between single-ended and balanced I believe.


Interesting. I'm tempted to try this out. I have access to a quality soldering iron and can use it decently well (I'm typing this on an ergodox keyboard I built from a kit).

There is something very satisfying about building something yourself that you get use out of every day. Such an amplifier is likely to fall into that category for me.


Any DAC chip and some usb interface for it that you could recommend to go with this?

And slightly OT, but does anybody know any ABX test of balanced input/output signal? I find it hard to believe that there is a single person who can hear the difference.


The big gain for balanced inputs is rejection of signals induced into the wire. Of course optical is the way to go if you are really worried about that. Personally I don't care, my ears are old enough that I no longer hear the difference between a high end and a low end amp.


There are cases where the difference is huge; particularly when you have very long runs between the preamp and the amp. This is on reason why you usually see XLR inputs to monitors. Any noise induced is amplified and a long unbalanced wire is a big antenna.

I've heard of cases where you could receive AM radio stations from such a setup; I lean towards "urban legend" for that though. While it seems theoretically possible (since many amplifier topologies can have resonance when fed signals outside their expected operating range), it seems a bit farfetched to actually end up with intelligible audio.


It's actually pretty common on older guitar amps. I've got a Marshall TMB replica which has this issue when the guitar isn't plugged in. If you turn it up loud enough, you can make out sentences clearly. I thought I was going crazy when I first noticed it.


It's not as far-fetched as you might think. While you aren't likely to see the classic AM demodulator circuit appear in an audio signal path, the nature of AM is that any non-linearity will cause some degree of demodulation to occur, so it can appear in all kinds of places unexpectently (especially if a powerful transmitter is nearby).


It is absolutely not an urban legend. I experienced it myself with my Z5500 speakers! It was a bit soft but on max volume you could clearly make out what was being said


There's thousands of options. Schiit are generally regarded for having high quality products that are relatively cheap.

A good place to find DAC reviews is head-fi: https://www.head-fi.org/forums/dedicated-source-components.7...


Thousands of options is why I asked for a recommendation. I know Schiit has good reviews but I'm just interested in ICs. Something like PCM2707C for USB communication or PCM1794A for DAC or whatever else.


Behringer UCA-202 is the best thing for under $50 IMO. It does not have balanced outputs though.


https://www.audiosciencereview.com has measurements of various DACs. Tl;dr: the topping D50 ($250 MSRP) measures really well, and for a combo amp&dac unit the topping dx7s (currently on massdrop for $360) measures really well. Just a heads up, that site is biased towards objective measurements, the Toppings measure really well, so the site is biased towards them. A lot.

According to that site, Schiit stuff measures really bad.

List of all reviews: https://www.audiosciencereview.com/forum/index.php?threads/m...


A DIY 0.0001% THD Amplifier: The MJR7-Mk5 Mosfet Power Amplifier http://www.renardson-audio.com/mjr7-mk5.html


>When I wrote my previous article, class-D audio was fairly new

Not sure what this means. I had a group undergrad project in 2002 to build and characterise a class-D audio amp. The idea was well-established at the time.


On a similar note, anybody know why you can get a very good A/V receiver for ~$500 but not a pre/pro for under $1k? Is it just the lower manufacturing volume?


It's great that with some research and drive one can make something better and cheaper than the engineered factory version.

Time well invested.


I don’t build audio amplifiers often, but when I do, I always use top-quality components like HPOO.


Hmm, I'd much rather build a Class-D amp from scratch with no chips involved... any tips?




That's not realistic, particularly if you're after a full range amp.


there seems to be too many different thinghs mixed up, "professional amplifier" "high end audio" and "class D" are three different worlds. It's like a Humvee a Maybach and a Prius in the same vehicle.


What's the output power of the amp?


It is refreshing to see these types of articles.

Of all disciplines of human interests and hobbies, Audiophiles are one of the most stubborn, misinformed, anecdotal and delusional bunch. Of course I am generalizing, but every time I engage in a discussion about audio quality, it is a losing battle. Not because of lack of data or supporting evidence, but because of the fundamental dissent for data driven arguments. You cannot argue reason when the first step is to be born with "audiophile ears".

Why is it hard to understand that we have sophisticated metrology and instrumentation that can analyze things better than human ears can? I can understand a separate discussion about personal opinions regarding how "warm" a particular sound is, but that should remain distinctly as - subjective and that's perfectly OK.

It feels like the entire audio industry is promoting pseudoscience barring a very few vendors that actually care. We have a $8500 mp3 player from Sony claiming all kinds of things to the overwhelming abundance of cheap Shenzhen exports. There is just so much noise.

I absolutely abhor audiophiles.


Your generalisation is unfair. Most Audiophiles just want to listen to music the way the artists intended. This is not easy in a industry where colouring the sound is even celebrated, e.g. 'Beats by Dre'.

I recently tried the latest noise cancelling headphones by Sony (1000XM3). The noise cancelling was excellent, but the bass was +5db overblown, boomy and drowned out most of the mids, the treble was dark to hide the NC hiss. It was unlistenable for jazz and classical without EQ. But the mainstream tech press think they are wonderful, even "WhatHifi" gave them 5 stars. An Audiophile is simply someone that cares about sound quality, most people don't.


> Your generalisation is unfair. Most Audiophiles just want to listen to music the way the artists intended.

That is not my experience at all. Audio gear that is 'flat', in that it outputs as close a representation to the recording as possible is largely shunned in audiophile circles. They mostly want to listen to something that 'sounds good', with all the psychoacoustic baggage that comes with such a statement. As an example if the end goal is accurate reproduction then a tube amp is objectively worse than almost any other amp built in the last 30 years, yet they sell!


generalizing what "audiophiles" want is like trying to generalize what the "blacks" want or what "canadians" want. it makes no sense. what sounds 'good' entirely depends on the source and the listener. so immediately you can see there is no right answer. if you want to listen to old elvis records maybe they sound best on old 60s gear. theres supposed to be some noise. all the girls goin crazy listening to it back in the day -- they were listening on the AM radio! the noise was part of it. listening to classical though -- have you ever been? no speaker can reproduce that experience. some come close; they are not noisy speakers. then there is bass, some people love to have lots of it layered thick on top of everything they listen to. some do not. etc, etc, etc


> listening to classical though[]

Yeah.. I vastly prefer a real concert. But some people prefer their home audio setup instead. Even my old aunt.. I heard a guy (self-certified audiophile) complaining about the "bass response" when (apparently for the first time) he went to a classical concert. According to him the orchestra's "bass response" (whatever that means) was sub-standard. I prefer the real concert though (although I mostly go to small avenues). I should add that I have a tiny hearing problem in my left ear - normally I'm not bothered about it at all, but that ear is extremely sensitive to harmonics and quickly turns it into distortion if there's anything not quite right (whatever that is. I have no idea. But I noticed at one time that I can listen at much louder levels with certain high-end speakers. Of the type I could never buy.) Real concerts are bliss though. My ear never complains.


>audiophile circles >They mostly want to listen to something that 'sounds good', with all the psychoacoustic baggage that comes with such a statement

TL;DR the technology is so good and so cheap now that what you get from a $150 set-up today rivals stuff costing much more from a decade ago. I still recommend spending 90+% of your budget over a few hundred dollars on better actual reproduction equipment, be it speakers or headphones.

Not in my audiophile circle (which is mostly head-fi). There are all different price points, and different goals within those. Most reviews I've read in the past decade talk about if the headphones / IEMs are more targeting accuracy or a "fun" sound, but most of the reviews I've ever read on head-fi are very concerned with accuracy and true-to-recording representation. FWIW I'm very much not alone in my interest in things like super low distortion class D, solid-state headphone amps with digital volume control (so you don't introduce a potentiometer / stepped attenuator / LDR into the signal path), ear measuring mics (e.g. https://www.minidsp.com/products/acoustic-measurement/ears-h... ), digital parametric EQs to adjust for not-perfect frequency response, room correction, etc.

The technology in modern decent (decent = what most people would call "high-end", but I'm talking few-hundred-dollar Class D for amps and something along the lines of a properly-engineered ESS 9018-based DAC, maybe $100-200, OR a ~$200-300 headphone DAC+AMP combo, think Oppo HA2-SE level) equipment is so good IMO that headphones / speakers / IEMs make a MUCH MUCH bigger difference than jumping to equipment 10x the cost at the DAC or amp level does. $200 headphones are not bad, but $1000 headphones are damn amazing comparatively. I've owned a dozen pairs of IEMs, and while my gaming 'phones are Shure SE215s, they're not in the same league as my Westone ES60s. Even with that said, most people who have never owned any serious music gear would likely be blown away by just how good the SE215s are, coming out of something like a Fiio K1 or a smartphone / laptop / desktop with a decent DAC+AMP.

Tube amps are still fun to play with, but maybe aside from some 300B snobs, I don't think many people really believe that they're more accurate than modern solid state. Some people also still like vinyl, and will happily spend $10k+ on a record player.

There are also the high-sampling-frequency crowd and the high-bitrate geeks. It's fun to think about, but I personally can't hear a difference between 192kHz/24bit and Spotify "high quality" 44.1/16 with an RME ADI-2 Pro and HD 800S headphones or the ES60s (both balanced, which does sound slightly better to me in both cases). I actually showed up to a headphone meet in Tokyo and pulled out Spotify to demo some flagship IEMs with, and the guy was trying to explain to me how that was like ~eating caviar on a donut.


> ear measuring mics (e.g. https://www.minidsp.com/products/acoustic-measurement/ears-h.... )

That one sounds suspect but the only way to correctly measure headphones is on a binaural head.


Just curious, how do you feel about the Schiit Stack?


Yes the noise surrounding headphones is incredible (pun not intended), even stranger I find the denigration of 'studio monitor' quality very odd. Recently I bought a nice pair of headphones, naturally enough I wanted the ones that colour the music the least, went looking in the headphone section at the nearest consumer electronic shop for the ones that were on their web site and couldn't find them. There were stands of beats, and Sony etc at crazy prices. I asked the guy where they were and they were poked away in the corner in a different section. I laughed - the best headphones that were cheap were hidden away and all the overpriced plastic nonsense were displayed prominently.


Right; and there are real studio monitors, like https://meyersound.com/product/hd-1/ (popular for film) and the Yamaha NS10 https://www.soundonsound.com/reviews/yamaha-ns10-story , popular because it's so bad that if it sounds right on the NS10s, it'll sound right on anything. You don't always want a near-field critical listening device, and most studio headphones must be durable first, leak as little sound as possible second, be cheap third, and sound amazing fourth. Watch some videos of {your favorite artist} recording and look at what they use: it's probably something like a Beyerdynamic DT880 Pro (~$200), Sennheiser HD 280 ($100), or Sony MDR7506 (~$75). Professional musicians on-stage will use something like a Sennheiser ew 300 personal monitor system with custom in-ear monitors (JH, Westone, Sensaphonics, or something cheaper and universal-fit like Senn IE 40 Pro or Shure SE series). Aside from DJs getting paid to wear them at shows, most professional musicians aren't using gear from big-box retailers...they're using stuff from Guitar Center / Sweetwater / zZounds / Musicians Friend / that little music shop around the corner, who sell actual professional equipment.


Well of course, you went to a consumer electronic store. The kit they sell there isn't really meant to be for "professionals".

It's the same for everything they sell at those shops. They are there to upsell shiny shit to people who don't know what they want. Try finding a decent laptop for a decent price at one of those stores.


It wasn't so much that, sure, a consumer electronics shop wouldn't have professional equipment, it was the disconnect between reality and what was being presented as the best.

I've been a bit of an audio enthusiast for about 30 years and back then the expensive stuff was definitely better (mostly). Now its the opposite, the more expensive the stuff is, the worse it is and the dialog surrounding it has become cargo cult mumbo jumbo. I find it disturbing.

In general for laptops performance/cost is still vaguely linear, even in consumer shops, with audio its just a lot of noise even from supposedly professional shops.


The noise cancelling was excellent, but the bass was +5db overblown, boomy and drowned out most of the mids, the treble was dark to hide the NC hiss

Had pretty much the exact same experience with some noise-cancelling in-ear pair of Bose. I was actually amazed at how good the noise cancelling was, but sound reproduction even in absence of any outside noise was very sub-par. Now I can't imagine the engineers didn't measure this thing, so I'm assuming it was done on purpose, maybe aimed for pop music or so?


Yes I'm sure it was done on purpose, just like over-saturated colours on TVs and cameras.


As far a noise sound quality for cancelling wireless closed headphones go, the Sony's are the best along with Sennheiser Momentum 2.0 on-ear.

There are of course much better headphones, and I wouldn't call them audiophile headphones, but I haven't found other wireless ones that do noise-cancelling and isolation as well and still manage to sound acceptable.


If in-ear headphones work for you, you can find both very good isolation and excellent sound quality. I like my Shure SE530's for some peace and quiet on airplanes, and often find myself just enjoying the sound of them (or rather the original recording).


I swear by B&W PX. I prefer them to the more expensive B&O H9i, which are also not bad at all.


Someone I know was recently trying to have a conversation with an audiophile about the possibility of re-purposing some audio equipment to make noise floor measurements for another system that had similar frequency and dynamic range characteristics.

The technique worked just fine, but man, that guy was loony. He was trying to claim, with a straight face, that a GPS disciplined oscillator wasn't stable enough for his audiophile-grade Rube Goldberg machine.


Oh god, jitter... If ever there was a more pseudoscience audio term, I haven't heard it.

I wish there was a separate term between audiophiles who just want to spend a lot of money on hocus pocus, and audiophiles who want the measurably best performance.

Then the audiophile forums would not be polluted with people who spend thousands on cables and other rubbish, when the room response/treatment and speaker selection has a million times more effect on the actual sound.


Of course they’re anecdotal - listening to music is a subjective experience. If someone really thinks their music sounds better after adding $100 cable lifters and taping $50 bags of rocks to their $10,000 system, they are automatically correct. Their ears, their hobby.

It would be like saying that something can’t work because it’s a placebo - even if all they’re getting is a sugar pill, if the symptoms went away, that’s not really something that you can argue with.

I put audiophiles on the same level as sommeliers. The best wine is the one that tastes best to the drinker. You can haul out your GCMS to test the wine and your ABX testing software to test the audio and tell someone their subjective experience is Wrong, but this misses the point. We’re not talking about medical efficacy here, we’re talking about artwork.


Please don't lump all audiophiles in with some sort of distorted (irony...) stereotype.

I don't have a $10k system, or anything close to it, but yes, my system that I put together for about $1000 is way way way way better than what you're listening onm because I care enough to have a good external DAC, and feed that into a good amp paired with either a set of decent used speakers or one of a couple of sets of good headphones.

No "cable lifters" in sight.


It's not so much all audiophiles as all humans. Confirmation bias is very real. Sometimes it matters. Sometimes it doesn't. But we all suffer from it.


Your external dac is indistinguishable to your ears from the internal one in whatever device you are using to supply it with bits. This is precisely what OP is talking about.


No it isn’t, because if it was internal it would be inside my PC... which is pretty electrically noisy. I can run from my PC to the amp directly and here audible clicks every time I move the mouse or the one of the hard drives spins up.


There’s a huge gap in quality between most onboard DACs and externals. It’s not necessarily because the onboard is worse at C-ing the Ds to the As, but because they’re not isolated from the noisy electronics that share the power supply.

My laptop emits different squeals when I scroll a window, for example.


That's just shitty grounding. Try putting one of those 3-to-2 prong adapters on your charging cable and you'll see right away.


Shitty circuit layout maybe, but it's much easier to move the dac outside the noisy box than to fix the cicuit layout. Most computers today can manage a digital audio output, either through HDMI, spdif from the sound card, or via USB. Keeping the analog path away from noise is sensible in a way that 96kHz and oxygen free cables are not.


Nope, occurs even just on battery.


That’s not the DAC.


Right, but it makes the onboard DAC useless. So sidestepping that and shipping the bits out to something else is worth doing.


It’s not even placebo...contextual stimuli actually will change how things sound. If you have decent speakers at your computer, listen to a passage with your monitor on, then the same passage with it off, it will sound quite different.

Another example from audio production, adjusting signal processors while processing is bypassed, and being convinced the sound is changed.

Another one I’ve been noticing is how the resolution of the gain control affects perception of the “power” of the amp. If the resolution of the volume knob is high, meaning it takes more turning motion for the same level increase, the amp “feels” weaker to me, biasing my perception of sound.

The mother of all dilemmas when it comes to comparing audio gear is that the louder source always sounds “better.” This a very powerful perceptual distortion.

So while I think stereotypical audiophile practices are non-sensical, I have no doubt that they really make the systems sound better to the practitioners!


Placebo effects are real.


Ears are easy too fool by visuals. https://www.youtube.com/watch?v=G-lN8vWm3m0


Placebo effects are real profitable.


The issue is I have is that people are being convinced they need to spend $100's and $1000's of dollars to get the placebo effect. If it is indeed just placebo it should be acheivable for free.

Instead what you have is not placebo effect, but perceived value effect.

After working in high-end audio for many years it became apparent that the (high) price tag was in fact the benefit.

Some people just want what there neighbor cannot afford.


I have $2000 speakers and I use a $60 2nd-hand amp from 1999 and a $20 bluetooth dac with them. Audiophiles hate me. Bluetooth is the greatest sin, apparently.

I have AB tested my amp with newer models and while there was a difference, it was very minor and not worth upgrading IMO


If you're listening to uncompressed (losslessly compressed) audio, you're mostly fine with Bluetooth. But if your source is already lossily compressed, they you're multiplying the artifacts of MP3 or AAC by the artifacts of aptX or SBC, and that becomes very audible if you know what to listen for.


>We’re not talking about medical efficacy here, we’re talking about artwork.

if only audiophiles see it this way themselves...


Nothing wrong with vinyl, that’s more about physically handling the media, reading the sleeves, listening to an album the whole way through etc. Than the audio quality. No vinyl enthusiast would say the quality is objectively better, but the experience arguably is


Yea, I removed the Vinyl argument. That's for another time. Nothing wrong with Records but I've had the same issue talking to people about Records wrt to audio quality.

I own Records and like you, I enjoy the experience. There is something amazing about listening to the entire album without distractions and in the sequence that the artist intended. You could do this with mp3/CD, but Records get me in the right mood :-)


There are objective reasons to favor vinyl even today. The 'audio quality' if measured in terms of SNR is objectively much worse than any modern medium, however the mixes used for vinyl tend to not be compressed to death so the dynamic range can offset this shortcoming.


You have a good point and I've heard of this before. The problem is that it depends on the artist/mixing engineer and production process. There is no guarantee.

It is ironic that higher quality mastering is done on a vintage outdated (barring recent resurgence) and objectively lower fidelity medium. It'd be nice to see CD's having great mastering and may be some of them do?


> It is ironic that higher quality mastering is done on a vintage outdated (barring recent resurgence) and objectively lower fidelity medium. It'd be nice to see CD's having great mastering and may be some of them do?

It's because there is a very limited number of people whom care about accurate sound reproduction. The _vast_ majority of listeners use garbage sound reproduction systems (Beats headphones, bluetooth cannon speakers, almost everything Bose makes now, etc). The small segment that cares is also heavily dominated by the audiophile sub-segment, selling vinyl to them is icing on the cake. The segment that _really_ cares about audio quality, those that care but are not 'audiophiles', is _extremely_ small and not worth targeting at the consumer level.


A little OT, but in my little home town the last CD shop disappeared a year or so ago. But you can still buy vinyl in town..


In that case you're comparing mixes, not media.


Yes, but the two are intertwined. You often can not get a flac file from source with the same mix that was put on vinyl, just the unfortunate reality of it. This is exactly what I'm saying, people will say that vinyl "sounds better", and they aren't wrong in a lot of cases. Not because vinyl is a better medium, but because a good mix isn't available in any other medium.


Yes


Also hunting (in stores, second-hand places etc) for particular records is quite an enjoyable experience and the climatic moment of finding something special compared to clicking a couple of buttons, and waiting for the download to finish


What I miss are the album notes. It's easy to include a booklet with massive information in the CD jewelbox, but instead "they" dropped the notes. And these days, you can't even determine who wrote the song, as the track copyright information is missing. It's like ebooks, which often omit the year of publication.


Bingo.

And what I often crave is the less accurate vinyl sound, that coupled with different production values.

Not better, just resonant with me. Easiest way to explain it.


There's also the chance that microphones and signal analysis aren't the best metaphors for how ears and brains process sound. But if the audiophiles you're talking about are endoring $8500 mp3 players, maybe we're talking to different people. That does sound silly.


(EDIT: Parent deleted argument about Vinyl) Vinyl has a few more dimensions than the "quality" aspect. I totally agree that there is no argument for vinyl offering superior quality. However, I have bought a few records simply because they were actually the _easiest_ way for me to buy something I could play back for a few albums/tracks that never saw a proper online release.

It's also easier to display a 12" sleeve as "art" than a CD or a FLAC download ;). Turntables are better decorative living room items than CD players or tape decks, but that's also subjective ;).


This is a Quora answer of mine from (sigh) 2011:

I _like_ vinyl, but I'm happy to admit its almost certainly due to non-audio non-hifi reasons. I like the "ceremony" associated with cleaning the disk, dropping the needle on the lead in track, and getting up to flip it over and do it again every 5 or 6 songs. I like the cover art - there is arguably a loss to the world of a beautiful art form due to the demise of the 12" record and it's large art-work ready packaging. And, as you'll hear on a lot of '90's CDs - I like the 33 1/3 rpm flaws, the clicks, scratches, and wow and flutter - when I hear the opening bars of Tear Drops from Massive Attack's Mezzanine, I get nostalgic, not for the late nineties when it was released, but for the early eighties - that scratched, fuzzy, played-too-much vinyl sound they used - that was the soundtrack to my late teens. Whatever the minor differences in the audio reproduction, those 33 1/3 and 45 rpm noises bring back some wonderful memories... Which is enough reason for me to keep the turntable set up and occasionally spin some discs...


In the late 90's & early 2000's, I bought a ton of vinyl. The majority of it was techno, house, and drum n' bass singles. It was quite literally the only format to buy it in.

Then, the "digital revolution" came to music, and you could buy it online. Which, for nearly every case, was better. Especially for people who weren't in driving distance of a major city.

However, once I started to get a digital collection going, the size was overwhelming. Especially in regards to dj'ing. With vinyl, your entire collection isn't always available. You make little stacks, shift things in and out, and it's much more organic. Digital misses that. Plus, I never really knew track names. I just remembered album or record artwork. It was easier to find things that way.


Much of this is cultural driven by a historical time when we didn't have the tools to measure this stuff, and reproduction was so bad that nearly anything that changed the sound in any way could be argued to be an improvement.

Those days are long gone, but the culture remains.

Getting good sound out of an old victrola--which were very expensive--truly is an art. Getting passable sound out of today's average off-the-shelf electronics doesn't take much at all. But many audiophiles act like are living in the late 1800s.


> Much of this is cultural driven by a historical time when we didn't have the tools to measure this stuff, and reproduction was so bad that nearly anything that changed the sound in any way could be argued to be an improvement.

[citation needed]

State of the art distortion measuring techniques resolved something like 0.002 % in the 70s.


Which is quite a bit later than the timeframe the parent is referring to.


> one of the most stubborn, misinformed, anecdotal and delusional bunch

I've had professional recording artists tell me they won't use MIDI over USB because the volume is lower than the old 9-pin connectors.


Wait, what? The “volume” of MIDI? I’m guessing they’re not referring to cc#7...?

(Also, I’ve never encountered a 9-pin MIDI connector. 5-pin maybe?)


I usually see midi as a 5-pin DIN, but some sound-cards used a DB-9 which required an adapter.


Ah, the joystick port as I called it.


Must be the 5-pin. Been a while since I've used one of the old MIDI connections.


why generalize like this? indeed discussing virtually any subject with virtually anybody (except perhaps within academic circles) usually reveals how "stubborn, misinformed, anecdotal and delusional" their beliefs are

if i had to guess, it sounds like you have poor hearing ;P


{quote} ... I connected it to a pair of 30-year-old, three-way Panasonic speakers, which I frequently use for testing.{/quote} Ummmm. I'm tempted to dismiss the author's entire project based on that. Why not use a pair of decent speakers or some high-end headphones like the Stax electrostatics or Audeze planars? Weakest link in the chain and all that.


I'm probably the only one with this opinion...but I sure miss the old days of tube amps that weren't able to be highly accurate. They would roll-off the on the high end, distort a bit all over, and gave a very warm and relaxing feel to the albums I played obsessively. Now I've owned high-end tube and solid-state gear, and the music hasn't sounded good for decades. Combine that with the very limited bits provided by CDs, preventing any possibility of good playback, and music has been completely destroyed. Old timers like me are dying off, and most of the people who remember good music will be dead. Kind of makes me sad.


>Combine that with the very limited bits provided by CDs, preventing any possibility of good playback

16 bits gives you 96dB SNR (and subjectively more with dithering), which is plenty at non-deafening listening volumes. It's more than even the highest quality vinyl ever had. High bit depth is mostly useful so you've got some headroom during editing.


16 bits give you 96 dB SNR only if your signal is one-bit. In a quiet room with decent headphones set to comfortable level I can easily hear one-bit flips in otherwise silent WAV file.

Not a problem for popular music that's consistently loud, but a real problem for classical music with wide dynamic range.


I tried it and I couldn't hear it. Are you sure you got the endianness right and it was really a least significant bit?


Yes, pretty sure. Had to zoom in a lot in audio editor to even see it. Could hear it nevertheless.

It's important for the room to be quiet, as the sound is very faint and easily masked, but it's there.

It's possible that Bluetooth headphones could lose it because of compression, or cheap DAC in the sound card simply ignores least-significant bits.


That is absurd. Anybody who has read about the history of CD bit depth would know just how absurd that is. And I even personally investigated CD accuracy in burning using an electron microscope to understand how the laser burning process worked on the plastic media. It was absurdly inaccurate. All these claims about how great digital is are just crazy - claims made by people who are relying on old data about hearing accuracy and a religious faith in technology. Anybody who has actually designed instruments using the known laws of physics understands the incredible limitations - and those limitations apply to both characterizing hearing as well as reproducing sound digitally.


>I even personally investigated CD accuracy in burning using an electron microscope to understand how the laser burning process worked on the plastic media. It was absurdly inaccurate.

I can believe that, but the standard was designed to tolerate high bit error rates. 25% of the bits are redundant, as part of a error correction code:

https://en.wikipedia.org/wiki/Cross-interleaved_Reed%E2%80%9...

This scheme works very well, as proved by people consistently ripping bit-for-bit identical copies of CDs pressed or burned with the same audio.


Your statement that CDs have a very limited number of bits is false by every known objective measure (compared to any prior consumer analog medium). CDs afford a bit depth of 16, while tape provides at most 6, and vinyl at most 12. See this great video by signal engineers for more info: https://xiph.org/video/vid2.shtml

I suspect this is a classic case of "you think you want X, but what you actually want is your youth."


Your statements about tape and vinyl are ludicrous and unsourced. What do you think most of your CDs were recorded on in the first place? (Hint: tape)


Yeah, my brother worked in a fairly antiquated recording studio about 10 years ago, and they had 24-bit 96kHz ADCs right next to the microphones, with digital lines carrying the sound data to be recorded onto a computer. All sound processing was performed in this format, with the final mix then being down-sampled to CD format.

This is like when people complain that FM always sounds so much better than the newfangled digital radio, without realising that the station signal was transferred to the transmission stations in digital form, often at lower bit-rate than the newfangled digital radio.

People like the nice familiar sound of low quality audio. Lots of audio equipment deliberately adulterates the sound to "make it sound better".


Not defending the bit depths in parent comment, but the tape a professional recording is made on and a cassette tape (which I assume parent meant by "tape") are a lot different, right?

In any case, I've heard some great sound from Type II and Type IV cassette tapes in the past. The worst part was that tapes I made myself from CD sounded way better than a store-bought cassette of the same album. That's when I learned they really cheaped out on those cassettes (brown-colored Type I tapes)


It’s not that different. Usually 1/2” or 1” wide and run at a faster speed but the chemistry is the same.


Doesn't running faster and having that extra width have the effect of giving more physical space on the tape to the changes in magnetization? And therefore wider frequency range or more stable differences in amplitude?

Sort of like how SP (2 hour) on a VHS tape looks much better than LP (4 hour) or ELP (6 hour).


It does, but plenty of albums have been recorded on 4 track machines to 1/4" tape.

The real difference in quality is probably mostly in having a machine that's properly setup and calibrated, with a good transport to minimize flutter, etc.


The 6 bits worth of SNR (36dB) for tape does sound implausibly low, even for consumer grade compact cassettes. And professional 1/2" reel-to-reel could do substantially better, although still worse than CD quality.


You can get tube preamps that are supposedly good, which might give you the distortion you crave even when paired with a low-distortion Class-D power amplifier.

CDs are still 1000x better than records with all their hiss and distortion in my humble opinion.

Not saying you're one of these people, but I think lots of old-timers who complain about the quality of CD's or anything digital don't really understand how they work - Even though the bits are "limited", they do not actually output a staircase as shown in the explanations, the signal is interpolated between samples by analog filtering on the output. So it's actually outputting a smooth line that goes through the sample points.

Also lots of stuff is available in 96k and above FLAC now if the limited bits are a problem. I get any music I care about in that quality but it's often because the mastering is better on those releases than the actual quality of the reproduction.

Sometimes there's a SACD, Bluray or DVDA release if there's no hi-res digital download available.

Anecdotally I'm happy if I can get anything at 48k 24bit or more.

It's really the loudness war that has ruined music. That's the only thing about records that I consider better than digital - because of the limitations of the medium, they can't brickwall compress everything or the needle will physically jump off the record.

Genuine question: what was so good about music before that is missing in your opinion now?

Also, do you enjoy digital re-releases of records you previously enjoyed as much as before? Why or why not?




Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: