Would love it if someone would make a simple player which was controllable by the user.
If you want a cheap device to run rockbox...
> Either develop a a player from scratch or port a subset of Rockbox (no games, FX, etc.) on a small SBC plus added display, DAC, SD and LiPo circuitry. The SBC could well be the ESP32 which would add wireless functionality; *Pi and other Linux boards would be tremendously overkill for the task. Shouldn't be that hard for skilled developers, and it would surely sell among those of us who don't want to use a cellphone.
...this feels like it's going to end up as being an expensive device. There are a bunch of other MP3 players available from companies like FIIO, Cowon, Shanling or others. Would you be able to build something that competes with these? https://www.advancedmp3players.co.uk/shop/Digital_Audio_Play...
EDIT: Thinking about it, if you had braille on the buttons and a clear simple layout you could push it to both markets: people who want a hackable MP3 player, and people who want an accessible mp3 player for things like audio books. (Which is something my Clip Jam is not great for).
The rockbox-compatible sansas (and earlier cowons) were:
1. cheap (circa 30 bucks)
2. very small (I just measured my Clip Zip as: 5cm x 3cm x 1cm )
3. very light
4. has an SD slot (I don't know its max supported size but 32GB works fine for my audiobooks)
5. is easily flashable with rockbox
Dedicated music players tend to have a much longer battery life, are lighter, cheaper to replace (weather, drops, being left on wet ground...). It allows me to listen to things without sacrificing my safety net.
Pretty much all of this, plus not risking a device that I have to rely on for communications. Each of my Sansa players fell like a dozen times without getting a single scratch because they were very light and the screen was small and sturdy; had I used a cellphone I would probably have destroyed at least its screen multiple times or ruined it from prolonged exposition to salty air at the sea.
Also I don't use smartphones due to the above points, and I still have to see a traditional phone with a decent player (and a decent camera, but that's another story).
The PinePhone might change this, however. Silently waiting for it to hit the market.
Am definitely interested in the PinePhone, but suspect that its dimensions will mean that I shall be restricted in not using it while running/cycling and (more importantly for me) as a cheap audiobook player that I fall asleep to in bed and occasionally drop out of the bed onto the floor or sleep on top of.
high battery use, bad DACs, and clumsy interfaces.
There are exceptions, but not many.
Single purpose devices that are designed to do a specific thing just work better -- go figure.
Thanks! I have one in original packaging somewhere. Bought it to give away but never did for various reasons.
It works with Rockbox, has a monochrome OLED screen, can take 2 MicroSD cards, and has two 3.5mm sockets: one for line out, and one for headphones.
The only thing I don't like about it is the headphone socket, which is prone to fail. It feels tight on insertion/removal, and the extra force can eventually cause it to lose connection with the mainboard.
They cost about $100 new, or half that used.
it's quite interesting with more modern rockbox supported pmps, rockbox is run as an SDL program stuck into the modified linux firmware. this device is moderately powerful with a 1GHz MIPS ingenic x1000 CPU. fun to tinker with.
fiio has some nice devices (some with RB support, e.g. the M3K), too. but i wish they had bluetooth. i guess it's not cool with audiophiles so intentionally excluded.
All it would need is an ESP32, IMU, heart rate sensor, micro SD card reader, audio jack, LiPo and touch display. Make it work with the Arduino IDE and things get very exciting for hack-ability.
Honestly, the watch makes me more excited than the phone.
The F/OSS part is the most important. Rockbox allowed an incredible degree of customization. For something like a music/audiobook player personal tastes in interaction with the a small device mean that configurability is key. That, I suspect is why there are only larger devices with loads more buttons and touchscreens available.
The ability to set timeouts and restart from last position on power-on, led light-level, sort-order etc made devices supported by them something that I could use. I have tried various Alibaba sourced devices that were in the US$20-50 range and none of them had the right form-factor and customizability.
I found an Android app called MortPlayer which is very close to Rockbox in terms of functions and configurability. It hasn't been updated in a years, but still works well.
Here's hoping someone tries to fill the market hole.
I don't have hard numbers, but if you need to update as frequently as every second, I don't think you'd gain much over a low-brightness color display with a polarized window for sunlight viewing.
And battery life seems like the only major draw of e-ink, as long as they are limited to one or two colors and a full-screen refresh rate that your eyes can easily register.
People have tried to develop tricks to update faster and minimize how much of the display gets refreshed, but I'm not aware of many successes since a lot of it seems to require reverse engineering; the companies which produce the displays jealously guard any information about the refresh parameters and functionality as highly confidential IP, and the process appears to be heavily dependent on temperature.
Both of these issues can be gotten around with a different refresh algorithm, but of course the datasheet doesn't include any of them. The vendors regard their refresh logic as proprietary and so you have to spend months trying to figure out what will work, like this: https://www.youtube.com/watch?v=MsbiO8EAsGw
As for a watch, I could see getting notifications once a minute being fine.
When developing this, it obviously refreshed much more than once a minute. Nothing ever led me to believe anything bad was happening. No burn in, no artifacts, no weird behavior at all.
I think the datasheet is being super ultra mega conservative, designed to cover the worst possible unit that they would let past QC being put in a product that would cost billions of dollars to recall. If you got an average unit and you're making one of these things.... you can probably refresh it in a loop for decades and never have a problem. I wouldn't bet billions on it, but I will bet $20 on it.
I think I'd be fine with having to tap my wristwatch to update the display. It would be cool if it could just show seconds without tapping, but if having to tap means the difference between a few days of battery life and months of battery life (I'm on a computer most of the day so it would wake only a few times a day), then sure.
I also find e-ink easier on the eyes than any backlit display. At least for reading. For a watch it probably doesn't matter too much.
And the more sunlight the more readable it is.
(Pebble also used transflective LCD and also called it "e-paper".)
I can’t understand why people want to have color, fancy screens on a watch. Do we need colors to see the time? A calendar notification? Heart rate ?
A watch should serve information without distracting, and most importantly, I should be able to go out without thinking about battery life of my watch. It should be reliable enough to carry it in the arm.
I agree with the fancy screen, IMO this is not necessary but that's what users want apparently.
Some argue that the daily habit of charging makes it easier to remember to do it, vs a less frequent schedule.
And people do use the Apple Watch for sleep tracking, charging it during showers and other preparations in the morning or at night, so the battery life is good enough for effectively 24 hours of use.
And suddenly even if your device had some headroom when new it still won't last a day.
An Amazfit Bip equivalent, with an open system would be a great product at least for a niche (Amazfit's interface and functionality is arguably poor).
edit: has this post been downvoted with information at hand? The statement above is technically correct: the Bip has an LCD (see specs), and it lasts very long; therefore, E-ink is not the only option for a display of a long-lasting device.
> The watch has a color LCD display which shows info with a resolution of 176 x 176 pixels. It is a reflective panel which will allow you to save battery while it is in stand by, as it will still be able to show you what time is it.
I'm still using my Pebble Time, I love it. I adore only having to charge it once a week and still being able to see the time on it.
Just the fact that anyone can look at my wrist and see the time without shaking it or anything is worth much more than refresh rate and apps for me.
All I need a smartwatch for is to see messages and notifications from my phone.
When Pebble went under I switched to a Samsung watch with an AMOLED screen and the difference in view-ability is stunning.
I agree with you with regard to the Pebble Time, but the Pebble 2's screen is fantastic. I can see the time even in an abnormally dim room, without the built-in light. (The trade-off, of course, is that the Pebble 2's screen is black and white, but I think that's fine on a watch).
I’m just surprised. I’ve admittedly never owned a Classic (only a Time and P2), but I thought the Classic was supposed to be almost as good as a Pebble 2. And I find my P2 as easy to read as a normal watch.
I expect both the Steel and the P2 are much easier to read than the classic.
I really miss Pebble, which this sort of evokes. Their 'Time' series of watches were incredible, with awesome battery life and a great C SDK.
How has Pine's recent hardware and software support panned out? I remember people saying that the original PineBooks bordered on unusable, but I'm sure they've made a lot of progress since then and I keep meaning to try one of these libre hardware platforms.
Seem like a big opportunity.
What's different in the Chinese business culture - that nobody tries this opportunity ?
When you're selling hardware at-cost and copying OSS software to run on it, it's not really profitable to add additional software maintenance costs like producing an SDK or offering English-language software support. I'm 100% sure there are Chinese companies who offer software and English-language support, but I am also 100% sure those Chinese companies charge similar prices to Western companies, negating the benefits of Alibaba products.
You certainly do have Chinese companies who attempt this: Huawei and Xiaomi come to mind. But their products sell at prices very comparable to Western companies. For the true "Chinese prices" companies who only sell on Alibaba etc... if they're offering Western support and Western innovation, there's no way they're selling at Chinese prices.
The issue is that they have their own ecosystem for software that doesn't really play nice with western IP, so you sort of get stone walled if you come in talking English. https://www.bunniestudios.com/blog/?p=4297
>very small teams of engineers can obtain complete design packages for working phones
They can obtain complete design packages for working phones because those design packages were stolen according to Western laws and customs.
>we also feel that shying away from reverse engineering simply because it’s controversial is a slippery slope
It might be controversial in China, but it's illegal in many Western countries.
>Unfortunately, queries into getting a Western-licensed EDK for the chips used in the Chinese phones were met with a cold shoulder
So instead they decided that stealing the work of Western companies would not and could not be punished in China, so they just stole it.
"Doesn't play nice with Western IP" sounds an awful lot like theft with a fancier name.
Edit: according to Pine64, they will use NRF52832 https://twitter.com/thepine64/status/1173208712669999105 which is the exact SoC the linked watch use.
I'm developing a smart watch product. I'm working with a Chinese company that does the firmware dev.
If there's someone out experienced with SWD etc and might be interested in collaborating on this device please drop me a line. As far as I'm concerned the more open the hardware the better.
There seems to be no way to contact you other than something called "twitter" as per your profile here?
I only know it's #pinetime:matrix.org
What the Amazfit is really missing is an open SDK or way of running apps on it, then it would be pretty much perfect.
It just seems like a no-brainer. A Linux Watch. Please won't someone just make it. (Bonus points if it has probe inputs...)
I need the recording to start in sub-second time, so there is no barrier to start, which makes my use-case unsuitable for Android-level "hackable."
(shamelessly stolen from the other comment)