Developer Takes 'Retro' Concept to New Level by Creating Physical Winamp Player | Hacker News Reader

Developer Takes 'Retro' Concept to New Level by Creating Physical Winamp Player

xatakaon.com

125 pointspoopcat104 comments

It's proof positive that there exists a generation of devs who have never seen a physical user interface made of buttons, knobs, or sliders. E for effort.

I literally did a double take reading this. The very possibility that an adult in the world has never interacted with analog controls on anything more complex than a toaster trips me out a bit. Like, how in the hell has software driven UI (with all of the problems that entails) managed to take over? Why must literally everything suck?

It's just vastly cheaper unfortunately. Manufacturing buttons requires investing in moulds, etc. and then iterating on a design is very expensive.

Is it really? Touchscreens and the tech stack to drive them aren't exactly simplistic. A handful of simple switches and a few rheostats seems like it'd be cheaper if for no other reason they don't invoke the avalanche of dumb shit that kicks off whenever software is brought into a project.

Labor also isn't getting cheaper, and someone has to assemble and test all those switches. You can automate... but LCD touchscreen is simpler.

I strongly doubt mass produced commodity components like switches are hand assembled or individually tested. It's typical for random samples, say one in a thousand, to be pulled from a batch for testing.

If you are asking which of these things is more likely:

A) The developer chose a touch screen to make the build easier and cheaper

B) The developer has never seen physical buttons

The answer is A)

You think it's faster to get a bunch of switches and rheostats, CAD components so that they fit together while looking like a Winamp skin that looks like a physical device, get them manufactured, test them, do the above five more times to get everything perfect, sand/paint the components to get the specific look, and wire them all up than to throw some CSS lines and HTML elements on a screen?

Honestly since I don't do UI, yes. I also think it's a little weird to think that is particularly difficult. It's really not.

I didn't say "difficult", I said "faster". I do both, and CSS is an order of magnitude faster, maybe more.

Ok but most of the things you listed don't need to be done. You don't need CAD or custom manufactured parts. Just some model paint. If you want to get nitty gritty about it you left out language support, accessibility, unit tests, version control, and documentation out of the UI portion.

How am I going to make the little buttons, the frame, the rails for the pots, etc? If I have to 3D print them, I need to model them.

Some metal and a drill and a tape measure. What ever did people do before 3d printers?

You can buy all of these things. None of that needs to be custom made.

For a one-off project probably not, for the larger ecosystem time's going to get spent on case design and debugging manufacturing regardless so any gains there are a rounding error. That said I've seen talented designers spend weeks chasing cross-browser bugs and I knew a guy who could lay out and solder an entire mixing board over a long weekend so I'm not convinced the time savings are real.

1) Economies of scale. One touchscreen factory replaces arbitrary number of switches and sliders and buttons of all shape and size.

2) Who says anything about touchscreens? Most touch interfaces your interact with aren't displays - they're capacitive buttons. Think your induction stove, or a powerbank, or any other random device you have to touch just right to turn on. Capacitive buttons are dumb pieces of solid-state electronics. Zero moving parts. Zero things to break (except maybe status LED that shines through them so you know where to touch). Cheap as dirt.

The future sucks because it wasn't created for improved quality or functionality - it was created because it was cheaper than the old-school way before it.

Because slapping a touchscreen or an OLED gets you out of the inconvenience of making stuff repairable. "Just chuck it out" because it's impossible to actually repair a touch screen is way more convenient and cheaper for the manufacturer to do than making their products repairable. Look at the modern disposable camera lenses. Even the most expensive ones have no manual controls, but are full of motors and displays making them way less future proof and actually unrepairable.

Tempus fugit. I still get taken aback when someone reminds me 1990s weren't 10 years ago. There's a whole generation of adults now, who grew up with touchscreens. Original iPhone release date isn't perhaps the best milestone here, but arguably the most well known one, and it happened - I can't really believe what I'm writing - 16 years ago.

So yes, it's totally possible for someone doing tech today to have never interacted with complex analog interfaces since being a little kid. Still unlikely for the next couple years, but possible.

How is it not blindingly obvious yet? It's money. Everything sucks because you aren't an executive or a shareholder. Everything sucks because everything isn't for you, it's to extract money from you. And places like this are the Mos Eisley cantina where everyone talks about the best way to do that.

I was so disappointed as well when I saw he just grabbed a screen and called it a day.

I was really expecting some mechanical tricks around the play/pause buttons liked I had on my old cassette player to make them xor-like .. so disappointed. And frustrated I just don't have the skills or time to make one but would love to see someone actually do it.

Since this is like ten of the top ten comments here, complaining that it's not physical, I'd like to point out that physical controls are expensive. Slide pots are about a dollar: https://www.digikey.co.uk/en/products/detail/bourns-inc/PTA2... ; if you want a motorized one, that's way more.

Pushbuttons are similar. The XOR system is literally "radio buttons" or "ganged switches", but I can no longer find that exact item on catalogues. It works by having a bar hooked around all the switches so that when one is halfway-pressed it triggers the unlock across all switches, so the unused one pops out.

Then you need to hook all that up to your microcontroller as inputs, do panel cutouts, etc. The switches tend to be through-hole, complicating manufacturing. Then you discover that you end up needing two boards, one full of inputs and display parts which are through-hole so there isn't space on the back for the actual MCU.

>I'd like to point out that physical controls are expensive.

Do you guys not have money?

Not enough, no.

I can provide you with a PayPal username of you wanna get rid of some 8-)

This will cost $15 in materials for seven buttons, two sliders, and two screens. The "they're expensive!" argument makes sense when you're making a mass-produced item, but for a one-off art piece, $15 won't break the bank.

> I'd like to point out that physical controls are expensive

More expensive than a touchscreen? Especially if we also calculate the energy-costs for usage..

I agree there's a lot that would go into it. Which is why I was so disappointed with the implementation.

As someone who watches content from This Old Tony, Inheritance Machining, GreatScott!, Marius Hornberger and so on I would watch a series of someone building this. That's how I see you can subsidize a project like this these days.

The complaints stem from the fact that the title says it was physical. If the title was "winamp made of diamond" but it was made of plastic instead, it wouldn't help to tell us diamond is expensive. We know physical is more expensive/difficult, that's why we got excited when the title promised it, and then felt let down.

> xor-like

Radio buttons?

I mean .. brain fart?

yes, I meant radio buttons. of course!

I'm laughing so hard at myself now.

I quite like 'XOR buttons'. I wasn't intending to correct you; just to point out that those buttons have a name that originates precisely from the application we're talking about (well... Winamp's not a radio per se, but close enough).

"... and I'll take three floating point controls too"

"You mean sliders?"

What html calls range, like in deranged.

There’s a reason xor interface buttons are called “radio buttons“.

I definitely remember reading some folk etymology that claimed they were so named because the buttons affected each other despite being physically separate, "as if by radio".

"Like buttons on a car radio" seems more plausible to me however.

A massive overstatement. Even ignoring everything else, cars may have started down the touchscreen route (which is insane for anything the driver is using), but that is recent and not universal still. I can't believe there is a professional dev today that hasn't seen a few cars with physical user interfaces.

Someone has a (3d render?) better concept indeed: https://www.youtube.com/watch?v=2nGCihgndng

(the original page is now 404)

I was thinking it would be this one!

That looks amazing!

I wonder what happens if you press the EQ or PL buttons. :)

The whole front facade needs to rotate like James Bond's license plate: https://www.youtube.com/watch?v=2QXCWW--DlY&t=45

Naw a whole nother panel that has to pop out on a swing boom like transformers - complete with an array of luminous self motivated sliders and a dropdown for presets that pops out like an actual oled panel for the list of presets that pops back in again once you make your selection!

What hardware do you need to make the slider? Other than that and finding displays of the correct dimensions, with a 3D printer it should be pretty simple to make.

Motorised faders have been available for decades https://www.youtube.com/watch?v=RPXeULcDa5w

Indeed, and some are relatively inexpensive: https://www.aliexpress.us/item/1005004808214900.html

My favorite demo of motorized faders, also in the context of audio (M4 Motor fader controller for eurorack):

https://youtu.be/VOcJ3yufxtU?t=357

That is so cool.

Off the shelf slide potentiometer would do it

Slider for the song position needs to go both ways: You can move it to change current time or it moves automatically along with the song

Motorized slider. Can be used to do just that.

You only need one display still. Or maybe two smaller ones? Anyway, the frame can go over the display and mask the rest of it.

I love this concept, I'm going to make this.

Please do and show us! Would love to see one in real life.

I will! Apparently it's not as easy as I thought (because of the progress bar) but I'll try regardless!

Archive of the original page:

https://web.archive.org/web/20230522141639/https://www.rickg...

The message about stolen artwork is most unfortunate.

Perhaps the developer wanted to support skins this way. /s

The title implies, the device to have physical dials, sliders and buttons. However, it's a Rpi4 with a touchscreen enclosed in a rectangular box.

There should also be a physical attachment for llama whipping.

This will be in v2. The player in fact will not work unless said attachment is within suitable lashing distance of a llama. CRM (camelid rights management).

Yeah I thought someone had taken the 3D render that's made the rounds of the classic Winamp 3x player and made it into hardware. Not as impressive as that would be.

I think this device is rather meant to be shown to old colleagues than be actually used. If the author wanted to give some late 90s vibes I call it a success, at least for me. I always had it in a screen corner playing some prog rock song in my phones while banging code back then.

Disappointed for the same reason. A device with physical buttons and even led based panels would be the real deal.

Yeah, that was pretty disappointing, I was imagining the same thing as you.

Well.... sort of. It's WinAmp running on a touchscreen inside a box that looks like the Winamp window frame.

I was hoping for physical buttons and sliders.

I'm working on my own version that's going to be all buttons and sliders, but I'm hesitant to mention this anywhere because I'm only working on this sporadically as a learning exercise, and the first video already got me more attention than I thought I'd get.

Since then I've changed to a Pi Pico, and using PIO (programmable pins) makes the touch-delay problems I had magically go away. PIO really is awesome.

I've also replaced the DC motor with a tiny stepper motor, and now everything works peachy. Just waiting for some screens and stuff from AliExpress.

Anyhoo, shooting myself in the foot: https://youtu.be/GjmFbgqFeMQ

Looks great, thanks for sharing.

That capacitive touch slider is really cool.

That physically moving progress slider that can also be operated is absolutely bonkers. It took me a few moments to completely let it sink in.

Those physical concepts at 0:38 look very good.

Agree it's an awesome build.

FYI there is prior art for this sort of control - google 'motorised fader'. Used for audio tech mixing desks.

It used to be like that, back in the days. It was... normal. If you get a chance, play with some old school hifi at a thrift shop. Many have at least a powered volume knob, for control via the remote.

To this day, sound and light control in the entertainment industry and recording studios have 100s of motorized sliders.

Modern consumer devices are the worst.

> old school hifi at a thrift shop

I don't think I've ever seen a physical seeking slider. Fast forward, rewind, sure. Maybe even variable fast forward and rewind, but nothing like this, not even on modern analog devices. The main reason would be that it is hard to make seeking accurate and to make the mechanism as well wearing as a button. What I have seen are remote controllable physical volume dials where pressing the remote will actually move the dial, and something like that might be acceptably robust in a consumer product.

> entertainment industry and recording studios have 100s of motorized sliders

Definitely a thing in professional settings, but haven't seen this on consumer level devices.

It's a motorized slide pot, they're usually used as faders in mixing consoles.

You guys say this like it's an obvious and ubiquitous piece of technology, but then whenever I see someone ask why can't we have smart light switches that work like physical ones (two states) and work both directions (i.e. toggleable remotely), suddenly all I see on the Internet is that it's a weird and stupid idea and why would anyone want that.

It dawns on me that a two-state motorized light switch is a strictly simpler device than a motorized analog slider.

Well, it is obvious, and it used to be ubiquitous. It's also complicated, which makes it expensive and easy to break.

You are right, the seek slider was never common. My initial comment was more general.

I too wish for a motorized light switch. One day I want to make one (I feel that I am slowly reaching the required skills and tools to build it). Of course, the house insurance might not like it.

edit: To be clear, the motorized volume knob was common on consumer devices. It is merely a potentiometer with a motor attached. As far as I knoe it had no sense of position, just turn it until it reaches a hard stop, and the motor is too weak and simply stall. Motorized sliders were and are still common on mixing tables, for sound and light control, and those seem to have positional accuracy. It is quite fun to watch a mixing table move by itself during a performance. I have never seen them used for seek control.

> why can't we have smart light switches that work like physical ones (two states) and work both directions (i.e. toggleable remotely)

Separate command from power!

Like this? (command + relay(s), inside switch box)

https://kb.shelly.cloud/knowledge-base/shelly-1-mini-gen3#Sh...

https://kb.shelly.cloud/knowledge-base/shelly-plus-2pm#Shell...

or this? (command only, inside switch box, either use a separate smart relay or send on/off commands straight to an always-powered smart bulb)

https://kb.shelly.cloud/knowledge-base/shelly-i4-gen3#Shelly...

or this? (4 command + 4 relays, for breaker box)

https://kb.shelly.cloud/knowledge-base/shelly-pro-4pm-v2#She...

Inputs can be set up as button (so, open/close circuit are an input event) or switch (so, open/close circuit are an input state)

Of course it won't physically change the switch position, but the switch works both ways because you can turn it into a toggle, whatever the current position is.

> Of course it won't physically change the switch position, ...

But that is what we are asking for. That the switch physically moves to follow the state of the system.

I want the switches on either side of a hallway to move in sync.

Exactly. If we can have (or at least used to be able to) physical sliders that work as both inputs and outputs, then why no wall switches? I want my wall switches be in sync with the state of smart devices - like, if I turn off a smart light remotely via an app, I want the corresponding wall switch to physically flip.

Yep, but it’s not in old analog hifis. A good reason is that tapes and CDs can’t instantly seek. By the time we got mp3 players, it was all about miniaturization, and eventually flat devices and screens. Slide pots for seeking never happened

I haven't watched the video (intend to, because it looks really cool) but my first thought is to repurpose a motorized fader from a mixing desk.

They're great, old technology but good fun. Only problem is the pricing: https://www.digikey.co.uk/en/products/detail/bourns-inc/PSM0... that's over twenty of your US dollars for one of them.

Professional mixing desks have hundreds of them, and can memorize the positions of the set to save a lot of sound engineer time.

To be fair on the maker, it's the article title that is misleading. Stop crapping on him.

I also couldn't help but notice that the account that posted it is 3 days old, has made 5 submissions, 3 of which are from the same xatakaon(dot)com domain.

I'm confused. Is this website about furtive self-promotion to land a job or get VC money? Or is it an innocent platform for sharing projects?

A link to the actual hackaday blog post: https://hackaday.com/2024/06/06/linamp-the-irl-winamp/

ugh - I really need to get my butt in gear and finish my winamp car stereo thing. The gimmick in mine is the use of a linear servo for the progress bar. I stmbled on a super easy way to control it that i never seen anybody use.

My one holdup is that its hard to find a TFT of exactly the right rize for the upper part. Also the volume/balance sliders get in the way a bit s the mechanical design is harder than anticipated. Also I feel that FDM printing doesn't do it justice.

Update: forgot to include the trick - cheap servos are crap because of the cheap potiometers. The electronics have to work and are the cheapest way to compensate for sub-standard parts elsewhere so they are somewhat ok. So if you combine the board from a 1EUR servo with an ALPS linear motorized slider that has a super high quality potiometer since its far larger and you would hear it being crap - the gears and motor are also good. This combined with the board from the servo gets you super smooth control.

Hah! I'm working on something very similar!

The screens are by far the most tricky part to get right, size and positioning. At first I thought about using one of the super thin Sony mobile screens for the entire thing, then using an overlay to separate them, annnd then to make it tactile with knobs, the knobs would touch the screen, and the phone would send the co-ords to the playback device (RPi in my case).

But that's because I wanted a very "thin" end product. Making it deeper allows for some more creative layout.

I've settled on a linear potentiometer driven by a stepper motor, and it's working out for me so far!

Good luck with yours! Got any links to follow it on?

Here's mine FWIW: https://youtu.be/GjmFbgqFeMQ

I want to make this as well! Maybe we should all collaborate.

I'm currently doing tiny bits of way too many projects, so I'm as reliable as Trumpet Winsock sadly. But we could encourage each other to stop procrastinating!

Yes, thin would be super hard to do - I expect at least a good few CM in depth. The Alps fader alone is almost the biggest factor. This is a link to a German electronics store that offers the same I have been using: https://www.reichelt.com/de/en/alps-rsa0n-studio-fader-motor... I will see if i can find the box with my prototype setup because its way smoother than your setup is. its unbelievable smooth when using a cheap analog servo board.

From your video you came to almost exact the same solution I have. :) I really can suggest trying to run it from a servo board - it has all the stuff for smooth movement already in it.

I'm using a Pi Pico now with a stepper motor (no video of it yet), and it's super smooth, indeed!

I was thinking about using a "SoftPot Membrane Potentiometer" to bring the chonkiness down a little, where the knob on the front would have a little nub at the back which would press against the strip and give you the same kind of effect (but then you need a way to drive that nub, of course).

I also thought about using the stepper motor and spiral combo out of a CD-ROM drive, maybe putting an encoder on it, but the spiral isn't long enough, so meh.

I'm still waiting for screens from AliExpress before I make an updated video, otherwise it'd just be, "look, it's smoother now, yay".

Lack of physical buttons aside, I really miss this era of UI design. Part of it is nostalgia, sure, but there is something comforting about skeuomorphism that we just don't see anymore. Perhaps it was necessary back then to enable a smooth transition towards digital interfaces, whereas now we have digital native generations that don't need it. The sad thing is that now these digital interfaces are driving the design of our physical ones, which I think most consumers would agree is a regression from what we've had before. Manufacturers: stop optimizing for cost, and start optimizing for user experience.

Why does the top-left part of the screen look like a sticker?

It does kind of look like a sticker now that you mention it. I went over to look at the video of the device itself and it seems like it was just the angle of the photo.

https://www.youtube.com/watch?v=5nUYHROVOmo

I think Posy would have something to say about this abomination.

Here is the actual good thing, gone but not forgotten: https://www.youtube.com/watch?v=PkPSDOjhxwM

Wow, I haven't seen these things in a while, but I remember lusting after a piece of Hi-Fi equipment with this kind of display in my teens. I mean, the first frame of that video is already pure VFD porn.

Such a missed opportunity though, should have made buttons and sliders physical as well!

That isn't a physical Winamp player, it's a Winamp only touchscreen device.

https://youtu.be/5nUYHROVOmo

He mentions the main issue for a physical interface is that screens aren't necessarily available in the required dimensions.

Where's the hardware Winamp? All I see is Winamp in fullscreen zoomed in.