Hacker News new | past | comments | ask | show | jobs | submit login
Not Long Ago, Lenses Were Much Lighter (photographylife.com)
58 points by midef 42 days ago | hide | past | web | favorite | 74 comments

There are a few things that the article doesn't mention that have impacted the weight of mirrorless lenses over time.

Initially, micro 4/3 systems (and to some extent the other mirrorless systems) were mostly kit lenses and standard primes. Both of which tend to be largely plastic construction. And primes tend to weight less for a given focal length and quality than equivalent zooms.

More recently, "pro" level lenses have been made available in micro-4/3. These tend to made of metal instead of plastic, driving up the weight. And, to keep fixed aperture values, the lenses are physically larger. So, weight goes up.

Then, an interesting side effect of this... a heavy lens on a light body is awkward to handle, so bodies have gotten a bit larger over time as well (plus feature creep). See Olympus' new EM1X, which as the vertical grip and 2nd battery compartment build into the body (vs the EM1, which has both as add-on accessories).

Personally, I like the variety. I have two micro-4/3 bodies (smaller rangefinder-style EP5 and larger SLR-style EM5 mkii) with assorted lenses - primes, pro-level standard zoom, and a mid-level telephoto (physically longer than the standard zoom, but lighter because plastic). I can mix and match for any occasion - if I know I'll be inside in low light, I just take the EP5 and the 17mm prime. Vacation - I'll pack both bodies and 3-4 lenses, but on any given day, only carry a subset. Etc.

I have the E-M5 II also, with the optional grip attachment. It's a great compromise between small and light for vacationing without the grip attachment (which I'll leave off for any lens up through the Olympus 12-40mm f/2.8), but then when I want to use my big Olympus 40-150mm f/2.8 for more serious wildlife photography, I put the grip on.

I definitely appreciate the ability to go small and light with this camera, especially when using the Panasonic 20mm f/1.7 lens. That thing is so tiny! That whole setup can easily fit in a jacket pocket, and the weight isn't too noticeable around your neck after a whole day (unlike the 12-40mm f/2.8 lens).

My normal kit is the E-M5, 12-40mm f/2.8 lens, and a Peak Design Leash worn cross-body. This takes the weight of my neck a bit. And the lens is fantastic.

Though my favorite walking-around kit is the E-P5 with either the 17mm f/1.8 or 25mm f/1.8 with a wrist strap. Small enough that I don't feel like a total wanna-be-pro carrying it, but still plenty camera for good shots.

Yeah, I need to get a cross-body strap. That'd help for carrying it around all day.

A modern full-frame DSLR sensor has greater resolving power than normal medium-format film. You need a really, really good lens to get the full performance out of a >40MPix sensor. You still want to shoot handheld, so you want a really fast lens to minimise the impact of camera shake. Sharp/fast/light - pick any two. Zeiss, Sigma and Sony have broken the old Canon/Nikon duopoly with a range of phenomenally good but relatively heavy lenses that can exploit the full capabilities of modern sensors.

Casual customers who need a "good enough" camera are perfectly satisfied with their smartphone or a cheap kit lens. Aftermarket lenses are inevitably going to target the pixel-peeping crowd who care about every lp/mm, because they spend the big bucks on lenses. There's still a market for lightweight aftermarket lenses, but it's nowhere near as big or profitable as the market for monster glass.

> A modern full-frame DSLR sensor has greater resolving power than normal medium-format film.

This really depends on the film and MF lens. I've shot 6x6 Provia on a Mamiya 6 and had it scanned at 40MP and 80MP final output sizes; the latter was clearly better. Resolution can be even higher with modern fine-grained black and white film.

> You still want to shoot handheld, so you want a really fast lens to minimise the impact of camera shake.

Or image stabilization, which also adds weight and size.

>This really depends on the film and MF lens. I've shot 6x6 Provia on a Mamiya 6 and had it scanned at 40MP and 80MP final output sizes; the latter was clearly better. Resolution can be even higher with modern fine-grained black and white film.

Film scanning is a bit deceptive, because film doesn't have pixels. You'll get more resolution from an 80MPix film scan, but that doesn't mean that the film actually has 80Mpix of useful resolution. A lot of information is lost in the conversion between silver particles and a scanner sensor, so you really need to substantially oversample. In the ideal case for color MF film (a ~200MPix drum scan of 6x7 Velvia 50) you might just about match the resolving power of a A7R III, at the cost of a substantial amount of dynamic range - film has a lot of exposure latitude, but it has substantial innate dynamic range compression.

Technical monochrome films can significantly exceed the resolving power of modern 35mm sensors, but at the cost of ~6 stops of sensitivity. Unless you're shooting landscapes or test charts from a very substantial tripod, you won't benefit from the extra spatial resolution. There's still a place for medium- and large-format film, but the inconvenience outweighs the benefits in the overwhelming majority of cases.

Here's a great article that goes in-depth on comparing the resolution and image quality of MF and LF film vs. images from a D800e. A D850 or A7R III has about 10% more linear resolution than a D800e.


I love this article because it not only looks at lines per mm, it considers other image quality factors like local contrast and color resolution.

The article shows that the Velvia 50 isn't resolving significantly more detail than the D800e's sensor. It's not surprising that the CMS 20 shows more detail. CMS 20 is a technical film that's not usually practical for normal photography.

I'd suggest taking a closer look at the first set of test chart comparison shots. The D800e is not resolving any lines on the chart at all--the "lines" you see in the D800e shot are moire.

You can see the same thing in the comparison shots of the camera. In the D800e shot, the Nikon logo and fine print on the lens are smeared, with yellow and blue artifacts. In the Mamiya Velvia shot, the Nikon logo is sharply defined, with no color artifacts. However, there is grain in Velvia shot.

This is a good example of how there is more to resolution and image quality than just counting megapixels. If you hate grain, the D800e shot is better, despite the moire. If you'll tolerate grain for more resolution, the Mamiya Velvia shot is better.

I've seen the comparison shots - that's what I'm basing my comment on. The medium format film shots don't resolve the lines either - except the CMS 20, which is not a practical film for normal photography. These black and white line pair charts make film look more impressive than it really is, since film has a long "very fine detail at very low contrast" tail. A more instructive comparison would be medium contrast line pairs.

As I said in my other comment, the easy way to prove that a medium format negative can resolve 80MP worth of detail is to provide an 80MP scan which loses detail when downsampled to, say, 40MP. I've asked a bunch of people for examples of such a scan over the years and never received one.

For some reason, the comparison site you link to doesn't provide any of the original files, so it's difficult to draw any conclusions.

>I've shot 6x6 Provia on a Mamiya 6 and had it scanned at 40MP and 80MP final output sizes; the latter was clearly better.

The real test is whether you can find any details visible in the 80MP scan that are lost if it's downsampled to 40MP.

The 80MP scan passed that test for me.

I don't think you understand. The wrong comparison is between the 40MP scan and the 80MP scan. The right comparison is between the 80MP scan and the 80MP scan downsampled to 40MP.

I do understand.

Can you give an example of an 80MP scan that loses detail when downsampled to 40MP?

>Or image stabilization, which also adds weight and size.

In-body stabilization is a great solution for this. f/1.4 primes and in-body stabilization are a great combination for available light photography.

> A modern full-frame DSLR sensor has greater resolving power than normal medium-format film.

Scanned at 5,300dpi (which seems to be the higher end offered by negative scanning services), my medium format 6cm square negative yields a 156 megapixel image. [1]

I'm not aware of any crop sensors offering comparable resolution, in fact this round-up of current cameras seems to suggest <50mpx is more likely and I'm sure some of those are "full frame".

[1] https://www.pixelcalculator.com/?round=&FORM=1&DP=1&FA=&lang...

[2] https://www.practicalphotography.com/group-tests-articles/20...

Scanning at 5300 dpi or 156 megapixels doesn't mean you have that much resolving power (I can scan a negative taken with a cheap lens at any resolution I want but that doesn't mean I magically get massive resolving power out of my crappy camera and lens).

When scanning traditional film you generally want to scan at significantly above the resolving power to capture all the grain structure and other analog characteristics of the film medium.

The easiest way to measure your resolving power is to print a black and white pinwheel - a large number of alternating black and white pie slices around a circle (more pie wedges is better). Pull your camera back and take a photo. Zoom into the pixels of the photo looking for the point near the center where the black and white regions stop being black and white and turn into a gray. Measure the diameter of the gray circle and use the diameter to compute the circumference (measure in pixels, then use the info about your sensor or scanner to convert from pixels to mm). Count the number of black pie slices and divide by the circumference to get your resolution. For extra fun try opening up one of your lenses to its maximum aperture and then compare the results to the same lens stopped down to f/16 or something like that - it's much easier to make a lens that performs well at a small apertures than large ones, and zoom lenses are harder to design well than fixed focal length prime lenses.

To get really trippy, just look at the pinwheel with your eye, and you can use the same technique to measure the angular resolving power of your eye (using your distance from the chart to measure the diameter in angular degrees rather than in mm).

Depends on the scanner! The "listed" DPI targets for scanners often aren't realized in practice. Take a look at the "image quality" section of this plustek optifilm review [0]: this one is actually very good. It's common for the "best" quality to come from scanning at 6,000+ DPI and downscaling to get ~3,000 effective DPI.

In practice, it's a _lot_ of work to get more than 3,200 DPI out of a 120 negative. You can certainly do it if you wet mount and use a flextight X5 scanner [1] (list price $23,000).

There's also the question of how fine the grains are in the film emulsion you're using. If you use ISO 50 film, you're likely to get more than 3,200 DPI of detail in the negative, but you don't have unlimited detail! Faster films have bigger grains - that's why they're faster!

So just keep in mind that the amount of detail you can extract is a combination of the following: how much detail the lens transmits, how much detail the photographic medium records and how much detail the digitization process retains. Digital cameras are a handy hack in that the photographic medium and digitization process are one step! Film is more complex.

[0] https://www.filmscanner.info/en/PlustekOpticFilm135.html [1] https://www.hasselblad.com/flextight-x5/

The easiest way is to print optically rather than digitally. A midrange enlarger lens is easily capable of outperforming a high-end scanner, and high-end apo enlarger lenses are still top dog. Avoiding the quantization inherent to pixel-based formats makes it much easier to deliver resolution to paper.

I agree that there are good aesthetic reasons to avoid digitization all together (or even just digitize prints). Color correction with scanned negatives can be a true nightmare if you want to preserve the look you would get with a traditional print.

Personally, I've always shared pictures primarily through the internet and so traditional printmaking has been less attractive.

Really, you just have to accept that a print made from a negative is a picture-of-a-picture. There is no inherently "right" way to convert a negative to a positive, any more than there is an inherently "right" color balance for a film to make the first capture with. The color balance and contrast levels you select are just an interpretation of the image, just like the initial exposure.

It's not too bad to get color balance right when scanning, as long as you didn't mess up the exposure too badly. Set the white and black points according to the histogram, adjust the gray point until the contrast looks right. Then, if there's color shifts in the highlights or shadows, play with the white/black points on one of the color channels until it looks right. Then pull it into Lightroom and do your final adjustments. It's intimidating at first but mechanically it's not that difficult, after you do a couple rolls you'll have the hang of it.

(background: film is basically three separate emulsions with their own color response curves, that happen to look normal when shot at box ISO. The farther you get away from box ISO, the more color shift you tend to get, eg clouds will start going yellow in an overexposed image)

Also worth noting that a flat scan is preferred to a high-contrast scan in most cases... you can always increase contrast in Lightroom but you can't pull back data that's clipped to zero or doesn't have sufficient bit-depth for good tonality. To phrase it more specifically, you want to try and allocate a good chunk of your bit depth to whatever tonal ranges are significant for a particular image, while trying not to clip anything to pure black or pure white.

Also, you should definitely give wet printing a try, it's a lot of fun for techie types who don't work with their hands for a living. You do pretty much have to be shooting B+W these days, not a whole lot of people still wet-printing color anymore, but as with everything B+W it's quite simple. With variable-contrast paper, you just make a couple exposures on test strips to find your highlight exposure and shadow exposure, then add these to produce a final print (split-filter printing/split-grade printing). You can then scan the finished print, if you like.


It's more of a fun thing than a technical necessity for 8x10 prints of 120 film, there is plenty of detail there for a 8x10 even on a basic flatbed scanner. But if you are doing 8x10s from 35mm film or larger prints from 120, then it's usually noticeably superior to do wet prints. If you are really going balls-to-the-wall on your hybrid digital workflow you can get a decent 8x10 or maybe 11x14 from a 35mm negative, but you'll need something like a Flextight or a drum scanner, which start getting pretty expensive to buy and run. Or, you send it out to a pro shop and pay $100 to have them scan it for you.

I could scan a blank piece of card at 156 megapixels, but that card wouldn't contain 156 megapixels worth of information. You're scanning discrete particles of silver metal, which are effectively opaque and have a finite size. Shoot a test chart and you'll see what film is actually capable of resolving.

Is there a 'Nyquist theorem' that applies to photography?

Basically, to make a film 'faster' you capture phonons from a larger area. Or the other way around, film that can resolve very fine details needs a lot of light.

Film has a built-in low pass filter. Typically 'sampling' in film is random enough that you can capture very fine detail if the contrast is high enough. It behaves quite differently from digital sampling techiques.

For this reason, you want to scan film at a very high resolution. But that doesn't mean that the image actually contains a huge of informaion.

He's comparing back to 2000's, but lenses used to be much much heavier. I have an all-metal Pentax zoom from the 80's that weights over 600g [1].

I believe the U curve he's seeing reflects the process of DSLRs transitioning from being an equipment restricted to professionals, getting popular during the 90's/2000's (pressuring manufacturers to produce cheaper, more lightweight lenses), and finally transitioning back to professional equipment, as cellphones and smaller cameras picked up on image quality.

[1] https://www.pentaxforums.com/lensreviews/SMC-Pentax-A-35-105...

> I believe the U curve he's seeing reflects the process of DSLRs transitioning from being an equipment restricted to professionals, getting popular during the 90's/2000's (pressuring manufacturers to produce cheaper, more lightweight lenses), and finally transitioning back to professional equipment, as cellphones and smaller cameras picked up on image quality.

IDK about that analysis, amateur photography wasn't invented by the digital age. Nikon and Pentax both have a lot of small and light primes and zooms.

I don't think the article is taking into account focal length. A 600mm lens will weigh more than a 100mm lens. Maybe the mirrorless MFGs began with 50, 120, 200mm models and finally getting some quality 400 and 600mm lenses out there which are going to affect weight.

And aperture. I'd love to see weight graphed against aperture and focal length (I guess it'd be a 3D plot), and then see what the correlations are. Does weight increase linearly with focal length? (I'd guess so.) Does it increase super-linearly with aperture? (I'd guess there's a square in that relationship.)

Certainly non-linear (in weight as well as price). Where you have roughly comparable lenses with, say, f4 max aperture (or a zoom with variable f4.5/5.6 or something like that) vs. an f2.8 fixed or zoom.

A good example is Canon 70-200mm.

EF 70-200mm f/4L USM is ~$600 and weighs 705 grams. (This is still a very good, fairly high-end lens.)

Compare to:

EF 70-200mm f2.8L IS III is ~$2100 and weighs over twice as much.

And certainly there are cheaper/lighter (and slower) third-party 70-200mm full-frame zooms and even smaller ones for APS-C sensors.

Along those lines, I use an old 1980s Nikon 70-200mm f/4 as my "long" lens. It weighs about a quarter as much as the modern f/2.8 version. I'll gladly give up a stop I don't really use anyway for something that doesn't feel like I'm lugging a howitzer around.

While the new ones benefit from lighter materials construction, the Nikons at least have vibration reduction built into their lenses rather than sensor also autofocus [in the lens] adds weight vs mechanical. So, on the one hand materials are getting lighter, on the other hand they've added tech which has added weight.

Nikon transitioned from AF motors in the camera to AF motors in the lens, which added weight to the lens (and removed weight from certain low-end consumer bodies). The lens I use has AF, but uses the body's motor.

I don't really care about vibration reduction, stabilization, or any of the other modern sharpness-obsession gear. I'm still shooting a D7000 I've had for close to a decade. For what I'm trying to express, more resolution and sharpness has never really helped.

Maybe the mirrorless MFGs began with 50, 120, 200mm models and finally getting some quality 400 and 600mm lenses out there which are going to affect weight.

This is absolutely true. Check the release history for Olympus lenses - the 40-150mm Pro and 300mm Pro were two of the most recent releases. Both are relatively massive by micro-4/3 standards.

And the most recent releases were the prime f/1.2 pro lenses, which are seriously hefty and massive. The 25mm (50mm FFE), for example, is 410g. Contrast with the pancake Panasonic 20mm f/1.7, which only weighs 100g -- less than a fourth as much!

Yes, this could be just a change in focal length and aperture* product statistics, not any kind of length getting heavier.

*: Observe e.g. most astronomical telescopes have similar magnification, but widely varying apertures.

I don't really know anything about photography, but I'd speculate that after smartphones became good enough for most amateur shutterbugs, lens makers were freed up to focus on people who were willing to accept extra mass for higher performance.

I doubt it. I think it mainly has to do with number of glass elements used which is driven by rising requirements because new sensors. Film was more forgiving.

I, too, don't know much about photography beyond lugging my wife's lenses ;)

Film also accepts rays that strike the surface at oblique angles, while this results in color shift or resolution problems on digital. It was actually technically easier to design high-quality lenses for film because optics did not need to be telecentric (nodal point of the lens placed at infinity). So yes, film was very much "more forgiving" than a digital sensor in this respect.

For example, a lot of early wide-angle designs (eg Biogon) were symmetric, and actually went into the throat of the camera, because if you had a 28mm lens, the nodal point was actually 28mm away from the film. That type of design went away when SLRs needed to put the mirror there, and we switched to lenses that are essentially "reverse telephotos" (the nodal point is farther away than the focal length). Modern lenses for digital cameras take this another step further and put the the nodal point at infinity, so all the rays of light are coming straight into the sensor at a 90 degree angle of incidence. However, since the lens must be a very strong reverse telephoto to accomplish this, it constrains the rest of the design and requires more elements to correct properly.




Thats a very interesting thought. The article is about DSLR and mirrorless system camera lenses though and I guess thats a market that always accepted extra mass for higher performance, in contrast to the market for compact-cameras.

I wonder wether the weight gain is based on the glas or the body. Are there more lenses with metal bodies to give them a luxury touch.

I wonder wether the weight gain is based on the glas or the body.


In the micro-4/3 system, there have been several "pro" level lenses released in the last 5-10 years. These tend to be both physically larger (to let in more light and allow faster fixed apertures) and all-metal construction. More actual glass, metal body that is physically larger = lens weighs substantially more than an equivalent focal length "kit" lens.

But to be fair they are still lighter and smaller than their full-frame equivalents. I recently got rid of my APS-C Nikon gear in favor of Olympus m43 system mainly because of this. Those 24-79 2.8 and 70-200 2.8 were beasts.

Absolutely. I was only attempting to address the upward trend line in the mirrorless (micro-4/3 specifically) graphs in the linked article.

I too sold my Canon DSLR kit for Olympus. I rarely carried the Canon stuff because it was bulky. But, I still wanted a bit more flexibility than the compact 1" sensor offerings.

Lens review are driven by specs, and it is easy to zoom in to 1000% in Photoshop and check whether a lens is sharp or not, and sharpness needs more glass, which is heavy. It is easy to test and easy to understand.

To me it is like the difference between LPs and poorly mastered CDs. It's easy to measure that the CD has more definition but sounds digital and lifeless. The scratches and pops of the LP give it personality.

What makes the CDs sound worse? I have heard this claim repeated A LOT

CDs don't sound like anything - you can't hear the difference between a modern AD/DA system and a length of wire. When recorded and played back using good equipment, they're effectively perfect to the human ear.

Some people subjectively prefer the distortion imparted by vinyl, which is fine. Some people enjoy the theater of pulling a record from a big colorful sleeve and gently placing the needle in the groove, which is also fine. Vinyl isn't better - it's objectively worse as a means of accurately reproducing sound - but some people prefer the experience.

As parent alludes to, I don't think it's a matter of CDs sounding worse in an objective sense, but lacking an affectation of the medium that some people like in vinyl. Think about movies: ordinarily they're shown at 24fps. Higher framerates are objectively higher quality, but don't feel "cinematic" because they aren't what audiences are used to, and in some cases the additional detail makes special effects and makeup more obvious, diminishing the sense of reality.

From an information-theoretic perspective, a vinyl LP is garbage compared to a well-mastered CD. From an aesthetic perspective... people like what they like?

Multiple things.

If the needle is good and the record is new the LP has a better theoretical dynamic range - new record is important here as each time you play the record you physically wear it, so after a few playing it will not be as good. This is the only one that is inherently better.

LPs are sold to "audiophiles", and often have a different mix. The CD will often be compressed until there is very little dynamic range. Even when the record is worn it can still be much better just because the mix on the CD wasn't good in the first place. This is the biggest advantage to LPs, but it should not be overlooked as it applies often.

So you're saying (correctly I think) that it doesn't have anything to do with the CD itself but everything to do with how the album was engineered. CD's don't magically introduce compression, it's about the production trends in mainstream music (which suck).

I'm confident that a good producer and engineer could make an album on CD sound extremely close to an LP, likely by mastering the album similar to the way you would for an LP.

Exactly. CDs can contain basically any audio you'd like... which is part of the cause of loudness wars, after all - you can put things on a CD that just wouldn't work on vinyl.

You can even add vinyl effects (warmness/clicks) in post-processing. I have a few songs in my library that really do sound like you're playing an LP, even though they're completely synthetic.

In the ripping community, there are some well-known albums that were mastered differently for LPs and are considered superior to their digital mastering. And so there are folks who spend $$$ on setups to play back the LP, record it, and then postprocess it to remove clicks. You end up with a digital audio file that has the better mastering. You could put that same file on a CD, after all.

Very much. Classical music often is mixed the same way as enough listeners are musicians who want the better mixes. I'm sure there are other odd (non-mainstream) music that gets good treatment as well.

There's nothing. It's something that's been made up by so called "audiophiles" to justify heavy investments in equipment whose performance does not help with anything.

Our ears are degenerating with every year, but the older those people get, the more expensive the equipment has to be, otherwise they can't grasp the soul of the music etc. pp.

It's just an excuse, CDs are fine.

The LPs basically required a standard mixing, the people able to change it and do something against the standards that sounded any good were rare and knew very well what they were doing.

CDs can be mastered by anyone with results that aren't horrible. So a lot of people that wouldn't touch LP production do produce CDs.

i think he compared "poorly mastered CDs" to LPs and that's fair

Superteles should have been excluded, those really are coming down in weight thanks to advances in optics as well as higher ISO & image stabilization allowing slower glass. Just look at the outlier "max" cases in the charts, and draw the trendline in your mind.

For the serious full-time pro shooting nature, sports, etc. "the best" may still make sense. But with modern full-frame digital sensors, it gets increasingly hard to justify thousands of dollars and pounds of weight for an f-stop. It made a lot more sense when you were maxed out at ISO 25-50 for the best color film quality and somewhere around ISO 200 tops for most pro purposes.

Even if money is no object, just the weight and bulk isn't worth it for a lot of circumstances even if you can get a tad better depth-of-field isolation.

For superteles, f-stops matter because of shutter speed, and increasingly, autofocus performance. It's a unique portion of the photography world, to be sure, but you don't have to be a pro to see real material benefit from an additional f-stop.

Sneaking through the jungle near dusk, shooting 800mm f/8 handheld on a crop body, there's not much I wouldn't do for an extra stop.

I probably shouldn't have drawn a distinction between full-time pros and very serious amateurs in the ultra-tele niche. And that's absolutely a fair point about autofocus especially at the highest focal lengths.

But obviously 800mm lenses are already pretty niche.

400mm + 2x isn't that niche :)

Improved manufacturing techniques for these lenses is likely another reason (on top of what is already mentioned in the comments: high resolution sensors and a changing market for such lenses).

I have found that I prefer modern Voigtlander lenses with my FF mirrorless body. They're small, but hefty with great build quality and image quality. The CV 15mm f4.5 and 40mm f1.2 are all you need for a fun day of photography!

Maybe I missed it, but isn't this just an artifact of full frames becoming more popular since 2000?

All my heavy lenses are from the film era (they fit on my Pentax just fine). My heaviest lens is 40 years old.

The real shift occurred in the 90s when computer-aided design techniques really took off and exotic lens designs (ultra-low dispersion glasses and aspheric surfaces) became affordable to use in mass production. It suddenly became viable to design lenses with many more elements and more exotic elements as well.

A necessary condition was the rise of multi-coating, as each air-to-glass interface increases the amount of reflection and degrades the image. Pentax developed their "SMC" (super-multi-coating) process in the 60s, and licensed it to Hasselblad as the T* coating (in exchange for a license to the Distagon design produced as the K28/2 and P67 55/4 designs). A few other high-end companies like Nikon independently developed their own multi-coatings, but most of the industry muddled along with single-coating until Pentax's patents expired, then immediately copied their process. So the explosion in lens complexity in the 90s was also partially driven by improvements in coatings that made it possible to minimize reflections and loss of contrast in lenses with high element counts (8-15 elements).

(this is why most older lenses were Triplets or Tessar formulas in the 30s and 40s... These lens formulas minimize the number of air-glass interfaces while providing a sufficient degree of sharpness. Modern designs like the Planar or Plasmat have been known for 100+ years (Planar was developed in 1896), but in those days the large number of air-glass interfaces would significantly degrade the contrast of these lenses. The development of single coatings prior to and during World War II drove an increase in lens elements during the 40s and 50s as well, to around 5-7 elements in high-end lenses, which is when you started seeing designs like the Plasmat really take off, like the Symmar or Sironar, and a shift to Planar in consumer cameras, eg Super Takumar or Nikkor-S 50/1.4 types.)

The other thing worth mentioning is that people got fond of zoom lenses in the 90s and 2000s. It is much harder to design a lens that is corrected across a whole range of focal lengths, meaning you need a larger number of elements (and more weight). Meanwhile, people who shoot prime lenses got fond of super-fast apertures, which also require more elements to make.

If you go back to an old-style f/3.5 or f/2.8 prime lens, there are some very lightweight modern designs available (eg Sigma DN 30/2.8 and 19/2.8). If you want a modern 24-105mm f/4 superzoom or a 35/1.4 superfast lens with 12 elements in it, you'll pay for it in weight. People want the faster apertures and wider zooms so lenses are getting heavier.

Those slower lenses are also much easier to correct, so they tend to be sharper than fast lenses are, unless the fast lens is much better corrected. In other words, you are better off shooting a f/2.8 prime at f/2.8 than an f/1.4 prime at f/2.8, in most cases (unless the f/1.4 lens has ~3x the number of elements). Older superfast lenses (eg Nikkor Ai-S 35/1.4) with spherical designs tend not to be that great.

Oh, yeah, I'm sure modern manufacturing also plays a role. Aspheric elements were first used around the 60s, the first I'm aware of being the Kilfitt Makro-Kilar 90/2.8 design, but were extremely expensive to manufacture. Typically you would produce them via single-point diamond turning, essentially using a diamond point on a lathe to produce a non-spherical curve on the glass. CAD/CAM manufacturing strikes again, these designs probably got a lot easier to produce in the 90s and 2000s, both from single-point turning on machine-assisted tools, as well as molded plastic lenses of useful composition.

So, really a convergence of various design and manufacturing improvements over the last 30 years, that made really high-end lenses viable to offer at prosumer-level price ranges.

I still miss my old Pentax screw drive lenses. They were high quality, robust and very light compared to today’s offerings. I think the new focus motors added a lot of weight but I am not sure.

The focus motors in modern lenses are trivially lightweight, incredibly fast and largely silent. Modern lenses are heavy because they've got a whole bunch of big, heavy glass elements held in precise alignment by a big, heavy metal chassis. Modern sensors have an order of magnitude more effective resolution than moderately fast 35mm color film, which has made lenses the image quality bottleneck. Lenses that were good enough for film or 12MPix sensors are visibly soft on a modern full-frame camera.

So are there innovations to address this? Maybe adaptive shaped sensors to accommodate lens aberrations?

What about the image stabilization sensors and motors, as those also trivially lightweight?

The sensors are a few SMD packages on a flat-flex and the motor is a voice coil with a few dozen winds of fine enamelled wire. The mounting system for the stabilized element is marginally more complex, but you're only moving one small element out of the dozen or so elements in a modern prime or the two dozen in a good modern zoom. It's pretty negligible, and it's becoming irrelevant as the industry moves to body-based stabilization.

The Canon 50mm f/1.4 weighs 290g. It has seven elements and acceptable optical performance. The Sigma 50mm f/1.4 Art weighs 815g. Same focal length, same sensor coverage, same maximum aperture, same focus motor technology, but it has thirteen elements and vastly superior optical performance. More and bigger elements are necessary for a better lens, but big lumps of glass are heavy. There's no getting around that.

Makes sense.

The best lens I probably own is my Canon ltm 35/2 which weighs a whopping 120g.

Of course. Unless you need some giant lens, a smartphone's imager and lens is probably good enough. Smartphones have eaten the low, and now the mid, end of cameras.

Or wide angle. Or low light performance. Or depth of field control. Or the best autofocus and high-speed shooting. Or just the quality that comes from a larger sensor.

I don't disagree with your basic point. If you're mostly shooting casual photos in mid focal lengths, smartphones are indeed increasing great and better than most point & shoot cameras have ever been whether film or digital. I use one a lot of the time even though I have plenty of other camera equipment because they often are good enough.

Why is this site spamming with a message asking to show notifications?

That message is an unblockable popup ad imho.

Applications are open for YC Summer 2019

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