Open Source, Open Architecture!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The new Sony Play Station 3 might be a good platform to experiment with..

It will have a BlueRay true HD DVD player

Several teraflops of compute power

Wireless game controllers

Estimated price $500 in full production


I would like a fully digital HT with a crossover and room equalization in the player that drives digital coax to the digital amps in each speaker.
 
http://en.wikipedia.org/wiki/GP2X

Look at this, the G2PX video game controller, with a display. Open source, open architecture. Anyone can duplicate it.

Say you use that as your hand held controller. So, you pick what kind of tone controls you will have. You decide how to manage the low bass with your various subwoofers and full range channels. You decide how and when to do Dolby and Thx comp. You decide what to use the surround channels for, maybe even going beyond 7.1. Maybe you take input form a synthesized organ. Such often have dozens of audio channels.

Maybe you are running scratch correction software to digitize vinyl.

Maybe you want digital delay to simulate a much bigger room.

The possibilites are endless, but the principle is that you the user decide, not marketeers.

Say this handheld controller plugs into your digital control preamp, which might be a box running say, a TigerShark DSP chip.

Otherwise the box doesn't have much. Maybe just an emergency volume and mute.

Digital lines to powered speakers. Of course the powered speakers can also take analog input.
 
Open source

Hi,

I have posted a few times before but mainly just reading. I have a great interest in the general theme of open source, but agree with the others that you need to have some idea of what you want. I dont really understand what it is you want from this thread, a list of links?

Im not sure why you are focusing on the games console. It seems over the top for a controller. It may have an open source development platform, but it isnt readily available and Id imagine reasoanbly expensive. I understand that you are citing it as something that we could develop into many differnet things (the point of your first post), but there are at least tens of other products that are hacked or allow open source development - not least a carefully specced pc.

I think you underestimate the completeness of what has already been done.

Look at Giantdisc it can use older palm handhelds to control an audio server via bluetooth i.e. no montior or keyboard required. It is open source and all the programmes it requires are as well. For a competent programmer (not me Im afraid) it would be relatively simple to add further formats simply by using pluggins, and the addition of Brutefir would be a nice touch. It would be simple to add control features to the palm. Surely something like this, a complete system, is a better starting point than a device you 'could' turn into a remote control for something you have not yet even defined?!

There are many minimal linux distributions that can run off usb flash drives and could be modified by a competent programmer to include most functionalities you are dreaming of.

You might want to check
Demudi out as well. It is a linux OS that has been carefully otimised for low latency audio - a 'DAW' as you mentioned earlier. In my oppinion its repositories inlcude the best of open source audio software, and if you want to be active in open source audio development it is a good place to start. You can check it for alternatives to VST and you should definetly look at the JACK audio server, the heart of the system. Sony, as a consortium member, based a prototype 'Hifi' unit on Demudi, see here although I doubt they will be willing to share. looks like what you want to do, at least physically.

Ubuntu Studio is also a very interesting site. it focuses on otpimising Ubuntu in the same way Demudi was developed from debian. It also has interesting links to a number of other projects.

Mythtv is also a popular base for an open source audio system.

I think we could go on all day listing interesting projects, but if you are not willing to commit to a spec nothing will be achieved and this will simply be an unorganised wish list. Sorry if Ive missed your point so far.

Cheers,

Ross
 
I'm excited by Borge's new project and thead, and I will check it out.

Rossco_50

What is already building is a list of links. I don't at all underestimate what has already been accomplished. Really I'm trying to apprasie it. Only then is it possible to disserne what is still missing.

I want the links. But I'm also hoping to get more first hand accounts. We've already gotten some good ones. I want to know what people are actually using now.

I'll check out all your links.

Audio Explorations:
http://groups.google.com/group/audioex_amps_atob
http://groups.google.com/group/audioex_amps_dght
 
Giantdisc?

http://www.giantdisc.org/index.php

This looks very interesting. I hope its not only mp3, because that has lossy compression.

But controlling a main unit with a Palm, or other display equiped BlueTooth device, is very attractive.

I see the best as a low cost box with just a couple of controls. Then most of the controls are on a hand held with a display.

If you want big play lists and jukebox like thing, then there could also be a PC with its hard disk.

There are so many software things going on that its very hard for me to make sense of them all. Addressing that was my real reason for starting this thread.
 
Rossco_50, do you know the Giantdisc people? Do you know where they are geographically?

This is all very interesting to me.

In my mind I've broken it down into 5 parts.

The first 3 are for domestic audio. The second two are not.


1. Control preamp functions. This is my first focus. It includes
Eq, Dolby, THx, decoding of surrounds, delay and ambience, compression and decompression ( needed for some feature films nowdays )

2. Active crossovers, for powered multipamped speakers.

3. Source managment, play lists, library cataloging, juke boxing.

And then:

4. Recording, mixing, and editing functions.

5. Music synthesis.


It seems that the same signal processing plug in standards that work for one of these will work for the other.

The real issue here is just to appraise what the current state of affairs really is. This is daunting.

Audio Explorations:
http://groups.google.com/group/audioex_amps_atob
 
Giantdisc

Hi,

I dont know the giantdisc people (seems main responsibility is in switzerland, but the there are a number of contributors listed). Ive been messing about with the system a little and am currently trying to get a bluetooth adaptor for the palm vx to work with it, but I am no expert - still learning. My main interest is 2 cahnnel audio so havnt actively sought solutions to decoding of surround formats. FLAC is supported so not only MP3, Im using FLAC.

Its audio only, so mythtv is definitely the better way to go if you want movie playback functionality. It provides very good creation/management features for music (FLAC included) and movie playlists, also records tv. If you had a tv card on your pc you would be able to use its interface on your tv and control everything from there - its an open source alternative to Windows media centre. If you use the JACK system I mentioned in my last post you can route all audio output from mythtv to any processing plugin you want - that just leaves identification of the plugins required. Brutefir will do all crossovers, eq, delay etc. Check the mythtv forums Im sure someone has set up the exact configuration you are looking for

If you ran all this on Demudi you would have all that is currently available in terms of open source audio editing/synthesis/mixing etc. within the same system. You will not be able to better their efforts in terms of a professional open source audio system, at least not without difficulty.

I think the 'missing' part of it all is the ability to decode closed source standards such as THX dolby etc using open standards. No too much you can do about that. Find a souncard that is supported under linux that can do such decoding. Or study these standards and simulate their effect with the processing ability of something like brutefir. This is probably the point where a lot of people turn back to windows. There are vst plugins that will decode thx etc but they are not open and are expensive.

found this. its a dsp card with audio ins and outs that can be programmed via linux for a number of uses, one being a 'thx pro logic like' decoder (i.e. its not actually an official decoder) - however its not open source, I dont think, and costs about a grand for the card although the software is free to download.


Cheers,

Ross
 
mythTV is talked about in "Linux Smart Homes for Dummies", by Niel Cherry. It uses MySQL. On a quick perusal, I didn't get the idea that the software actually has access to the media stream itself?

I look at their web site, and I am trying to find out.

http://demudi.agnula.org/wiki/LinksAudio

The above looks extremely interesting.

Truth is, I don't see a real time Linux version as being the way to go long term.

I see Linux as more of an executive oversight, or play list managment utility.

The actual engine for processing the stream, in my mind, would be something much simpler, and something that gives an absolute and uninterruptible guarantee of continuous stream processing.

I guess there are several layers.

1. The stream processing, totally uninterruptible.

2. The user interface and control options. Could use Linux, or something else. Still best if diskless and fanless.

3. An executive overview controller which does playlists and juke boxing. This pretty much has to be a PC with Linux.

In my mind it is also desireable if all of this software is compatible with what is use for recording and mixing, for mixing boards for concerts, and for music synthesis.

I'm trying to understand what is currently available, and the task is daunting.

I also think there is a big role for advocacy here. Meaning that there is so much available, and so many people who want to contribute to it.

There needs to be some overview or communications vehicle, to help coordinate this.

I'm glad that this thread has turned out to be so productive.
 
MYTHTV etc

Hi Zenmasterbrian,

Not quite sure what you mean about the 'audio stream' with regards to mythtv. Mythtv and giantdisc both use mysql as a database for the media and interface the databases with an ordinary media player, most likely Mplayer which supports DVD and FLAC amongst others. Mplayer can output to the JACK server which in turn can be connected to another programme, e.g. brutefir, to allow processing of the audio before it reaches the soundcard for output. So you can do what ever you want to the audio stream, this is a common set up and you shouldnt have trouble finding other people in the relevant forums with the same. Knoppmythis a live cd version of mythtv which will boot off a cd, easy to try it out. You can see that these systems cobble together the features they require from work that has already been done, this what you must do, decide your feature set and then identify the most effective way of achieving it e.g. they wanted DVD and FLAC support so chose a player supports these.

You'll notice on the demudi homepage they are developing an 'embedded' version of the OS, this will mean it can run better on lower spec machines and on a smaller storage footprint. Here is a tutorial on how develop a silent (diskless fanless) system based on the demudi live cd and a mini itx system. You could do something similar and include giantdisc and/or mythtv using a distibution like Slax which is the most configurable mini distribution and it is easy to get it to boot off a usb flash drive (it even has JACK, mysql and Mplayer as packages!). Somebody on the giantdisc forum has got it working with bluetooth on slax.

I know you dont want to use a PC, but given that you imagine the interface to be on one and that you will have to use a hard disk to store media, I see the neater solution as everything being done by the same machine. If you are concerned about powersupply noise there are soltuions like this which are fanless and would allow a quality diy external powersupply. Mini itx or even micro atx motherboards ensure reasonably smal size and decent computing power.

AlmusVCU is by the same developer as brutefir and looks to provide a means of developing a dsp 'box' based on brutefir and interfaced via hardware buttons and an LCD display. This seems to be in the same spirit as the system you envisage.

I realise I might not be helping here, too many options. I think any of the features you are imagining are doable, and in most cases have already been done. Gmarsh recommeded that you get the features you want working on linux and then look to strip that down to something that will work on a blackfin - this is good advice. The job to be done is simply gathering together all relevant existing software, once you have done that, and if people are interested enough, there will be scope for collaborative work.

By the way, in terms of ease of use ubuntu is brilliant. Im using it at the moment. In many ways it is easier to use than windows and just about any audio app you can think of (all those included in demudi) is available for easy install. Its a good starting base if you are not linux savvy. Ubuntustudio is simply a wiki focused on optimising ubuntu for audio so that it is something similar to demudi, demudi is not nearly as polished in terms of user friendliness or support.

Ross
 
Ross, The information you have provided is imense and formidable. Let me just give a quick impression.

PC based media players are choppy, disjunctive. They are not the sort of thing I would want hooked up to speakers at high volume.

Even real time Linuxes are still complex and multithreaded affairs.


Look at the standard digital audio alternative. I can go to a department store and buy a boxed 7.1 system. It can process the data coming off a DVD, and drive all its power amps.

Its memory and cpu requirements are minimal.

I'm looking first to an open source version of that.

I say the user should be able to decide how the tone controls work, how the surround channels are used. All the parameters should be out in the open, not just a few simplistic options.


Now, the kinds of things you have presented are very interesting, and may be the way things go.


But just for handling the audio stream, the digital version of a control preamp, not much is needed.

It does not take 1Ghz of processing or 100Mbytes of memory.



The examples you have presented deal more with media or source management. They are far far more complex and involved.

Does the actual audio play back work well? Is it reliable? Glitch and drop out free? Does it never miss a beat?

Does it never get stuck, never leave you with a dead volume control?


I want to understand all that is being done.

But in my mind this will be best when done in layers.

1. The most basic digital audio control preamp, 2ch, 7.1 ch, as desired.

2. Then we can talk about source management.

3. Then also talk about recording functions.

I'm goinging to look at all this. But I would be very interested to see this minimalist control preamp approach.


That was why I was intrigued when someone mentioned a Sony video game as a base. That was why I put the G2PX game controller as a good way to interface to such a box.

ZMB
 
Additional note:

That control preamp module ought to be built so that it cannot mess up. Build it like a cardiac resusitator. It resets the processor several times per second.

When you turn that volume knob, there is nothing, not possible software glitch that could ever prevent it from responding instantly.

When you press the mute button, the outputs go dead.


To me, that is the way to start.


I have never seen PCs run that smooth. The function they serve is very different.

IMHO, they don't do media very well because the operating systems and the hardware were never designed around an absolute rigid real time requirement.

Think of the digital audio core as something like antilock brakes.

You would not want Windows in you antilock brakes.

I really wouldn't even want an embedded Linux in my antilock brakes either.


The kind of hardware I am talking about is something small and cheap.

It is just a little apendage added to a PC.

The PC can be the user interface, or my GP2X game controller could.


Then, after that works, consider using PCs as media file management centers.

ZMB
 
You talk earlier about the use of a Palm with BlueTooth as a controller. That is similar to what I wanted with the G2PX game controller. Bluetooth would be nice. But mostly it is a convenient display and some finger controls, for a DSP box that would have just a few minimal backup controls.


You could control it with a PC, but PC interfaces are not easy, not quick enough!!!


Anything that uses the mouse is more like a sobirety test.


When I listen to an analog system, and the phone rings, as soon as my hand is on the volume knob, it responds.

Likewise, I have a department store all in one digital stereo in front of me. Its volume control is digital.

But when I touch it, the response feels instantaneous.


Very very different trying to control stop and start or volume on the windows media player.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.