Are all digital outputs created the same ?

[gabrielo]

Hi

Are there quality differences between digital out generated by an on-board sound-card (say N-Force) and a discreet sound card (say creative audigy ) ?

My understanding is that discreet cards can do the DSP onbaord offloading CPU which was very helpful in the old days with Pentium-II running in 350Mhz.
Nowadays with dual-cores running in 2Ghz+ and all the added functionality in the SSE, does it still makes a difference if I'm gonna stream digital out to a standalone DAC ?

Another question, when watching Blue-Ray movie using a Stereo system does it make a difference if sound source is 2 channels AC3, 5.1-AC3, DTS or all the new HD audio format ?
I mean after all I will be down mixing all extra channels into 2 channels, do we have extra info in the HD-Audio formats or just extra channels

Are there tools/drivers/players/sound-cards which will do better down-mixing than other or that it is a simple process which everyone know how to make

Last, assuming that DTS-HD is better than 2 channels AC3, can the down-mixing be done on-the-fly or that I can get better quality by doing pre-processing of that info.

Thanks a lot
/gabriel

 

[Slogra]

Differences can be:
- Samplerate: a lot of cards don't even support 44.1khz out, so they are unable to pass through CD sound perfectly (like Audigy)
- Dolby Digital Live/DTS connect. Cards which support this are able to encode to Dolby Digital or DTS in "realtime" (like N-force)
- HDMI sound. I don't know much about this, but some ATI videocards support sound via HDMI. So i guess it is able to pass through lossless Dolby Digital TrueHD/DTS-MA and PCM multichannel. Which are all much better formats than normal Dolby Digital and DTS.

EDIT: if you downmix to 2 channel, then you don't need Dolby Digital Live or DTS Connect.
And yeah, Dolby Digital TrueHD/DTS-MA and PCM (lossless) multichannel are still better than normal Dolby Digital and DTS (lossy) even when you downmix to stereo at a lower samplerate.

I'm not sure what the differences in downmixing could be, so i can't tell you what would be the best way.

 

[gabrielo]

Thanks for the reply
I don't think sample rate should be an issue since many on-board sound cards (G-Force included) support Intel-HDA standard (Azalia) which has 192-kHz 32-bit quality for two channels (which all I need).
{edit}
Just realized Azalia defines maximum sample rate of 32/192Khz, but typical implementation uses only 16/20/24 bits and not 32 (not sure about sample rate)
Do I get anything from 32bits in low/mid-range DAC or is it only used with very high-end DAC (if any) ?
What is the maximum sample-rate I can use in low/mid-range DAC ?
Is it 192khz or 96khz only ?
I checked specs for C-Audio DACMagic and its max input rate is 24/96Khz
Is this typical for DAC at this price range
{/edit}

Question remains, are all HDA(Azalia) implementations the same or that some card will output better/cleaner/more accurate digital-output ?

I don't want to output any fancy format like DTS/AC3, but a simple 2 channel PCM which will be understood by the DAC unit

If DTS-HS/AC3-TRUE-HD has more info per-channel I will then try to use them instead of the basic AC3/DTS
Any idea if down-mixing DTS-HS/AC3-TRUE-HD to 2 channels on-the-fly is good enough or that a better conversion can be done offline ?

Thanks again
/gabriel

 

[FallenAngel]

Do you have any 32-bit content? If no, don't worry.

Most "modern" DACs will accept at least 96kHz and most will do 192kHz through coax S/PDIF, Optical may get limited by the receiver to 96kHz, but some are OK at 192kHz. Do you have any 96kHz or 192kHz content? If no, don't worry.

The main difference between MOST on-board audio implementations and high quality sound cards is that on-board is made to be as cheap and small as possible. Under most circumstances, even if you do manage to get a bit-perfect signal (and most of the time, you will not), you won't get a transformer-coupled, well timed S/PDIF that you can from a good sound card.

You are asking a lot of questions regarding multi-channel audio while looking at 2-channel DACs, why? A stereo DAC obviously won't process multi-channel input, why not simply go with 2-channel input.

 

[bordins]

Even bits are perfect, but the S/PDIF stream can still have jitter (phase noise), resulting in improper audio reproduction in the DAC.

S/PDIF Jitter: Myth or Reality?
A high performance S/PDIF receiver 2006
JITTER PERFORMANCE OF S/PDIF DIGITAL INTERFACE TRANSCEIVERS: IS MEETING STANDARDS ENOUGH? 2005
The Deficiencies of SPDIF as a Digital Transmission Method 1997
Bel Canto Design DAC2 24/192 Up-sampling DAC (claim: Zero-jitter SPDIF interface)
S/PDIF or USB ?

Jitter suppression and PLL design **

You may compare different S/PDIF cables (short v.s. long, brands) to hear different sounds. I also find my M-Audio Audiophile USB or Firewire Solo has the lowest noise floor.

 

[little-endian]

Quote:

Originally Posted by bordins /img/forum/go_quote.gif Even bits are perfect, but the S/PDIF stream can still have jitter (phase noise), resulting in improper audio reproduction in the DAC.

That's true in principle, however the DAC's problem at the end. Trying to reduce jitter before the conversion is the wrong approach.