There are a number of Commell Mini-ITx Motherboards out now spouting Fibre Optic output.

Models LV-677, 676, 675. Would the SPDIF output be unprocessed eg Direct & not coloured by the onboard sound device? If this is the case I can then connect it to my H701 for processing duties?

good question and i'd like to know the answer to this too . If it is and it's just binary data passed straight from the file being played it'd be a great way around bothering with sound cards and any effects they might have.

Most likely the SPDIF output is 'bitperfect'. That is (in most cases) the point. IE: The pc does NOTHING to the sound at all, and the outboard DAC (which receives the SPDIF output from the PC) does ALL the work. Infact, if its outputting correct bitperfect output, the volume control will do nothing to the output level of the spdif. The spdif will contain only the audio data, and will not dither/truncate the output level in order to provide a reduced volume output level. That would need to be done on the PXA side. This occurs after the PXA receives the bitperfect signal, then applies its own dithering methods in order to reduce the volume and output it to the analogue side of the system.

This is my understanding of the process - hope it helps - but it might not be 100% correct. Its just things Ive read in my research. I dont have an SPDIF receiver/unit to test the theories with.

And one of the members I was hoping would have some input. Thanx Ice.

So i'd assume individual program audio output levels eg GPS vs Media player etc would be set on the PC depending on which you want louder than the other.

SPDIF or Optical doesn't allways equal quality, SPDIF normally use a dolby digital, or PCM audio encoding to output the sound, it doesn't merly pass the MP3 through, it does re-encode the file into a format that a DD or PCM (or DTS for that matter) decoder can understand.

either way, at one point or another, the digital signal has to be turned into an analouge signal, for amplification. your better of using the sound processor on your computer, these would almost certainly be better then an external signal processor, because it reduces the number of encodes, and decodes. you can get 24-bit sound cards like the Creative X-fi, that will do a better job, and be cheaper than a dedicated external car signal processor.

I already have the h701 which has far better useability and tuning than a CarPC could hope to perform without additional resource hungry proggies or hardware. I don't want to use the PC to store MP3 either, Ipod can do that much better anyway. I'd prefer large capacity external Hard drives and store my music in CDA or high Q formats. Disk Storage is cheaper than an Ipod

mosoto - i've read on mp3car just recently of someone who has done a back to back test between onboard SPDIF and using soundcard SPDIF and they've said there is a notable difference (better) quality when using the soundcard as opposed to the onboard SPDIF.



spicypeanut - as mosoto has said, as a convenience thing the alpine RUX controller is a much better option, plus adds no overhead as far as load on the computer with extra programs/hardware goes. It's also an excellent option when it comes to interfacing the computer sound form optical to amplifiers analog RCA runs. I won't argue whether it's better to do tuning in hardware or software, because i think that's probably a very complex argument with relevant points on both sides, but the alpine PXA-H701 processor certainly is a nice piece of equipment . I also believe the tuning options may be slightly more advanced than those available through the creative software that comes packaged with their cards, but admittedly i haven't used their top of the line cards. Out of curiousity, do you know if any of the creative cards allow you to perform time alignment?

cheers m8. i'll look that up. It will be interesting to see if they were comparing onboard & outboard sound processing rather than using the SPDIF as a direct out (unprocessed). There would definately be a difference then, no doubt.






don't believe so. Ice uses his carPC for all processing, he'd be able to give better feedback on that, but i've never seen TA with creative software/drivers.

I think what spicypeanut was getting at is that in order to send it out using the SPDIF protocol there needs to be some processing done regardless. I'm guessing that perhaps an onboard SPDIF may not do as gooder job (perhaps a lower sampling rate/bit rate) than a dedicated sound card? Also, there's a PCI card that alot of the people at mp3car have reported impressive results from that supports bitperfect (look up bitperfect on mp3car if you don't know what it is) and sells for around $30 USD. Another is a usb sound card, that unlike the creative does not require external power, and once again has given some impressive results, that goes for about $80 USD (This would rival the Soundblaster 2NX that i'm using at present which retails around the $200 AUD mark). The exchange rate is good at the moment, so it might be worth looking into. If you're interested i can dig up what the two card models were sometime soon and post them up for your reference.

I believe that some creative cards (possibly by using Kx Project drivers) can do per channel time delay. This was more of a home theatre function, so the delay is in feet rather than in centimetres/millimetres like you would need for car audio.

The issue I believe Brycestro has mentioned is related to "Jitter". Bit perfect should equal bit perfect. No ifs/buts. However, if the clocking mechanism is inferior on one of the devices tested, its gonna make the sound different. Clocking is very important,as the bitstream is asynchronous - and if sync is lost the sound Q is reduced.

Once again, I havent had the chance to use SPDIF/digital.. so this all goes towards theory - not practice - sorry guys.

And yes, I currently do ALL processing via my PC (Crossovers, EQ, Time alignment, Phase, etc).

well I couldn't find that comparo, but I found this snipit from the owner of the site:





ok here he's talking about onboard sound being inferior not specifically optical out, which as Ice mentioned should be Bitperfect and not processed. Do you have that link Bry.?

mosoto - windows itself is the main issue with bitperfect. Windows XP uses something called the "kmixer" which modifies the sound (a great thread on it can be found here: http://www.mp3car.com/vbulletin/faq-empori...it-perfect.html). Basically, if the optical out on the mobo doesn't support ASIO or kernel streaming (which at a guess it probably wouldn't), then you won't be able to achieve bitperfect with it. However, all things in proportion, the Kmixer might not make much of a difference on a setup that isn't brilliant to begin with. The equipment/install is probably a much bigger concern for starters. As is shown in the link i put just above, the problem with bitperfect is that if you have a scratched up cd you'll hear it exactly as it's read from the disc.

Thanx Bry. you've answered some of my Q's with that link. Although he is refering to peripheral Soundcards (PCI) the process of getting Bitperfect audio thru' the Optical out will be the same. The Commell boards use the ALC880 codecs which is good news for achieving this. So to counter Spicypeanuts post, Bitperfect is output from the SPDIF and at the same sample rate as the source (CD)

b.t.w. I read some of the article on Vista audio - interesting and although vista does away with Bitperfect it only seems necessary if utilising the processing technology of Windows Vista (incl. Volume control)

If the output volume levels (Wave, Main, etc) in windows are set to maximum, then Kmixer does not need to dither/truncate the 16/24-bit signal in order to provide a reduced volume level. Another point to not is that Kmixer resamples every piece of audio in your system up/down to 48Khz. The resampler algorithm is actually quite poor. This is one reason that Kmixer is also called 'Kmangler', as it is often the reason SQ is lowered, even before you get the sound out of your PC. ASIO/Kernal streaming are ways to bypass Kmixer and allow bitperfect, un'mangled' audio output. I have read, that if you resample your audio (which for Mp3s/CDs is 44.1Khz) up to 48Khz using a resampler plugin inside Winamp (per say), that Kmixer sees that the signal is already 48Khz and leaves it alone. This sort of circumvents the two anti-bitperfect issues we had earlier.

1. Full volume on all sliders - prevents dithering/truncation of signal (1st issue avoided)
2. Resample the audio using a high quality resampler (AudioBurst PowerFX IIRC - theres 2 version, use the free one) inside winamp to satisfy Kmixer that the sample rate is already 48Khz and it doesnt have to resample. (Prevents poor Kmixer resample algorithm)

[Digital volume reduction is done by changing the actual bits in the source signal - therefore a reduced volume level signal cannot be 'bitperfect', as the bits have been changed slightly. Analogue volume control is done by attenuating the actual output voltage rather than changing the bits of the audio data.]

Do a little research into the above...I believe it has worked for some people.

Chaintek AV-710 can be had for about $30. It can be flashed with Audiotrak Prodigy 7.1 drivers, turning it into a bitperfect source, but corrupting the analogue output stages (unlash flashed back to chaintek firmware). I believe the audiotrak drivers are written to avoid Kmixer entirely, hence giving you bitperfect sound for $30. I realise it is a PCI card and not onboard, but its just another option

Thanx m8