[IstariAsuka]

Quote:

Originally Posted by austonia lol... it would take millions and a long development period to make anything equal to what's already on the market, much less something better. There's no way to play catch-up at this point without massive resources at your disposal. And while you're trying to do what's already been done, other companies will continue to advance thier own products....

At least from my point of view you don't have to play catch-up. If all you wanted was a nice DAP, just get whatever is already out--there're several players that have MOST of the features one could want.

If you want a DAP with a good amp, use a line-out and a portable amp.

The problem is that there is no unit with a really good digital out. Except a PC.

Seriously, from my perspective this would be just about the most ideal: HDD based, transportable, gapless, and high quality digital output. Most of the convenience and benefits of a computer hifi setup without the computer.

In fact, it might be easiest to just leave out the analog step completely--or at least the internal amplifier part.

Stuff like physical UI, GUI, all that stuff would only have to be minimally functional. Just enough to get the job done. At least for my vision.

 

[acklee]

It's not going to happen.

You can open-source complex software projects and do reasonably well - there are hundreds of examples. But building a hardware device like a DAP would require all the below:

1) Hardware design, physical / electrical
2) Manufacturing (not the same as HW design)
3) Device firmware
4) PC software

The killer is that all of these would need to be very tightly integrated. You're much better off trying to mod an existing device. Figure out how to boot the device into Linux (it's been done for the iPod), take it apart and solder in component upgrades, etc.

 

[DanT]

Check what I find. I didn't have the time to read through tho. Google is your friend

http://www.ladyada.net/make/minty/index.html

enjoy

 

[acklee]

DanT 1, me 0.

Let me revise my previous statement ... building a DAP which had all the features we wanted would be impossible. Heck, even writing a spec we could all agree on would be impossible.

 

[jsc3]

Quote:

Originally Posted by acklee It's not going to happen. You can open-source complex software projects and do reasonably well - there are hundreds of examples. But building a hardware device like a DAP would require all the below: 1) Hardware design, physical / electrical 2) Manufacturing (not the same as HW design) 3) Device firmware 4) PC software The killer is that all of these would need to be very tightly integrated. You're much better off trying to mod an existing device. Figure out how to boot the device into Linux (it's been done for the iPod), take it apart and solder in component upgrades, etc.

Were we talking about a commercially viable product, or another DIY? For a commercially viable product, I agree - it won't be done. The reason is that anybody who could pull this off would break off from the group and do it themselves, and sell it - start a business.

HW design is not that hard, if somebody is willing to just take charge and do it. As somebody pointed out, it is almost impossible that everybody will agree on features here. I was presuming that mostly off-the-shelf components would be used. Disk, flash memory, cables, LCD display, etc are all readily available.

Manufacturing - I was thinking we were talking DIY here, which means old fashioned hand assembled from mostly off-the-shelf components.

Firmware is the same as software - just in some kind of flash memory. It really isn't magic, but it is not trivial either, as working on a custom platform is initially challenging.

PC software is probably the easiest part, a lot of existing programs could be modified, if we are talking DIY. For commercial product, it could still be done.

The main reason that it won't get done is that everybody here is saying "it will never happen". The only thing that might make it happen is if some *individual* or *small* group just designs the thing and puts the spec out into the world to be freely built/modded.

That is exactly what happened with Linux - Linus Torvalds said one day "here is a linux kernel I have written, what do you think?". What he had was VERY primitive, but it was done. Others started adding to it, and existing Open Source projects (eg GNU) ported to Linux, and it turned into something good.

I make my living doing UNIX computer consulting, mostly Sun Solaris. Solaris used to be the most popular UNIX variant out there, and it is still the most advanced UNIX kernel, but Linux is the most popular, and even Sun sells computers with Linux on them now. Sun has even done some opening of the Solaris source.

I guess I am rambling now. I like Linux, I use it a lot, but I also used Windows (I'm using it now). I use Solaris a lot too, and BSD isn't bad. The key point is that Linux turned into something competitive with existing commercial products that used to be VERY expensive, and this was started by one person's decision to make something *and* keep it freely available.

Linux didn't get to this point by being technically superior (though it is technically competent, sometimes superior in certain areas). Linux got to where it is by one person driving it (Linus) and by being free and gathering a lot of energy behind it. Every Linux distribution has a LOT of other free software (components) added to the kernel (basic design) - it is not just ONE thing, and there is not ONE right way to do it - there are several competing versions.

So much for my preaching - can I get an Amen! I am not volunteering to lead anything, as I am much too lazy, and simultaneously very busy (which ain't right for someone so lazy). But all of you sitting here coming up with reasons why this can't be done make me smile and hope there is one person sitting out there thinking "it really shouldn't be that hard. "

 

[jsc3]

Quote:

Originally Posted by DanT Check what I find. I didn't have the time to read through tho. Google is your friend http://www.ladyada.net/make/minty/index.html enjoy

Excellent. It's primitive, but it's a start. It almost certainly isn't audiophile quality yet, but it probably could be with some mods. The user interface is the biggest drawback but that could be fixed. As somebody else pointed out, modding an existing play is easier than starting from scratch.

The OpenSource BSD license is the most free license there is - you can use this for anything you want, even make mods without showing them to anybody or sharing the source code, and selling it as your own.

 

[jefemeister]

Quote:

Originally Posted by DanT Check what I find. I didn't have the time to read through tho. Google is your friend http://www.ladyada.net/make/minty/index.html enjoy

Nice. I haven't looked through all of this, but it looks like an interesting read if not a good starting point.

 

[jerb]

Hardware has become like legos, building the beasst isnt hard. i can asssemble a computer in under an hour. however software is something entirely different, it often takes me at least 3 hours to install windows and optimize everything for the particular machine.

we would be doing this and a much grander scale. im sure some people on the fourm could assemble a suitable unit in several months (assuming they spent all of their free time working on it) however i dont think that there is any feasable way to do an os. do you know how much programming that is? just boot up windows in safe mode and watch as the pages of scripts start up

. no we would have to write all of those scripts for this little monster.

sorry to be pecimistic but i dont think that this project is do-able

 

[Mr Iriver]

It would be possible if we could get a programmer from Rockbox?

 

[Jmmmmm]

Quote:

Originally Posted by Mr Iriver It would be possible if we could get a programmer from Rockbox?

outsourcing!

what about programmers in india? it worked for dell

 

[plus_c]

Any Head-Fi'ers in EE/CS and need a capstone project for their degree? This sounds like the perfect idea...

 

[IstariAsuka]

Quote:

Originally Posted by plus_c Any Head-Fi'ers in EE/CS and need a capstone project for their degree? This sounds like the perfect idea...

I know a bunch... but I don't think they'd be interested. =(

Linux on the iPod seems to be able to navigate and play music just fine. Thus, it must not be that hard to at least have a simple GUI.

As jsc3 notes, the hardware design would not be that hard, so long as we had some people willing to do it.

In fact, correct me if I'm wrong, the more generic the pieces the better. Why? Well, if we are able to use an x86 compatible processor, then coding would be very easy. Plus we wouldn't need a seperate decoder chip, then.

Think about it: we know that not everyone will agree on what features should be part of the player. If this thing could run an x86 processor, coding for it would be very easy--if it runs on windows/linux it would run on the machine, and vice versa. Thus, you could make a simple software base (telling it how to talk to the DAC, USB controller, controls, etc.), and then people could write whatever plugins for it they wanted. And you could pick and choose which features you wanted, and change them at any point. This would be much like how Foobar is very barebones, but there are plugins for it that allow you to do anything--change the UI, add features, etc.

So, if a VIA processor or somesuch is feasible, it would be great. Fast enough to do what we would need it to do, cool enough to run with a minimal heatsink, and cheap enough to be reasonable.

Or am I completely offbase?

 

[jsc3]

For those who say the programming is too hard, that's because you're not programmers. One motivated person could to most of the integration, at least to get it to the point where there was a running OS that others could modify.

The OS does not need to be written from scratch - that's why I suggested an embedded Linux. If we use generic parts, as the previous poster suggested, there probably wouldn't be many device drivers that need to be written.

As for using an X86 compatible, I would have to check the specs on the VIA C3. It is relatively low power/heat for an X86, but it still might use too much power to be of much use in a small portable powered by AA or 9V batteries, and the processing power is overkill for this project.

Here's an overview of tiny Linux-friendly single board computers available:

http://www.linuxdevices.com/articles/AT8498487406.html

A lot of these are using chips commonly used in PDAs, which are going to have better power/heat characteristics than even the VIA C3. They have clock speeds between 100-500 MHz and have more than adequate processing power to run a DAP.

 

[jefemeister]

Neuros officially released their complete schematics to the public ahwile back. I was having trouble finding the post before.

Blessingx posted about it here:
http://www6.head-fi.org/forums/showthread.php?t=97516

The direct Link:
http://open.neurosaudio.com/archives...uros_is_r.html

 

[IstariAsuka]

Quote:

Originally Posted by jsc3 For those who say the programming is too hard, that's because you're not programmers. One motivated person could to most of the integration, at least to get it to the point where there was a running OS that others could modify.

Awesome, now we just need a motivated programmer.

Quote:

Originally Posted by jsc3 The OS does not need to be written from scratch - that's why I suggested an embedded Linux. If we use generic parts, as the previous poster suggested, there probably wouldn't be many device drivers that need to be written. As for using an X86 compatible, I would have to check the specs on the VIA C3. It is relatively low power/heat for an X86, but it still might use too much power to be of much use in a small portable powered by AA or 9V batteries, and the processing power is overkill for this project.

You're right, it most likely wouldn't work for a battery powered project--I simply wasn't really thinking of battery power at the time. And it is overkill. Sorry.

Quote:

Originally Posted by jsc3 Here's an overview of tiny Linux-friendly single board computers available: http://www.linuxdevices.com/articles/AT8498487406.html A lot of these are using chips commonly used in PDAs, which are going to have better power/heat characteristics than even the VIA C3. They have clock speeds between 100-500 MHz and have more than adequate processing power to run a DAP.

I must say, those are quite nice looking. But how easily available are they, and how much do they cost?

Anyways, the only really good programmer I know is my older brother, but he's busy with his PhD, so definately not enough time, and probably no interest either. I know a bunch of OK programmers, but they for sure wouldn't be interested.

Do you think anyone in the DIY forum here or at DIY audio would be interested and willing in doing design, hardware wise? Because if we can produce a viable hardware design that's relatively easy to assemble and relatively inexpensive, I'm pretty sure people would be more willing to step up to the plate to work on software...