Digital Out sound quality

D

Dodgey

Guest
Hi all - I use an HTPC with an SB Audigy - feeding my Dennon3802 via the digital coaxial out.

A couple of questions.

1. Will buying a better sound card give better sound - bearing in mind, as far as I undersand, the SB is merely "passing on" the digital stream from the DVDs to the amp.

2. I use a very shoddy, long, home made phono lead to get the signal to the amp. It's digital, so the quality of the cable should make no difference.?

Thanks! :cool:
 
Hi,

You have touched on perhaps two of the most important issues in the "PC FOR HOME CINEMA USE" category.

The soundcard is the most important part of the chain second only to the DVD drive. One of the most common mistakes people make is to think of the whole thing as "its digital so it will be fine". The soundcard is your first interface between the pc and the 3802. There are so many things to consider, for example, jitter which is the one of the biggest problems there is with digital audio transmission. I recommend you get one of the best soundcards you can. Something like the new Audigy 2 Platinum as a starting point. Why spend all that money on a really good receiver when you are feeding it a poor signal???

All i can say for the cable is - Try others, there are huge differences in quality. Get one that is TRUE 758ohms and heavily screened around the 30mhz region.

Ok?

Cheers.
 
Thanks, I suspected as much. I've never really understood why, but always suspected, that just because it is digital it will be OK.

I'll have a look at new sound cards. Does anyone have any other suggestions (it seems Creative get a bit of a slagging off for HTPC).

I can understand why there might be differences. I have heard aweful sound from cheap DVD players and good sound from good ones. If the digital ignorance theory held true then they should all sound the same!

As for the cable, I thought it just carried 0's and 1's , so as long as they get there then all is OK?
 
Hi,
When I first built my HTPC i thought I would go the soundblaster route.....and went for the cheap and cheerfull thinking exactly the same as you, digital just needs a good feed through, so the SB 4.1 for about £24 went straight in, I was releived when it sounded just great through my Sony 930 reciever so I was happy, then quite recently I changed to a Denon 3802, so thought I wonder what would happen? LOL
And it just so happened at that time we had just reciever at work a new video edit computer around £3500 worth, it had the Audigy platinum in it, so of course one weekend it sorta fell out of that machine and slid into mine to see if it would be worth investing in a better card, guess what...it sounded worse!
Very slighty........difficult to describe but sorta........thicker! muddier! not so light and lively, and I am sure the effects didnt steer so well either..........so first thing Monday morning the Audigy went back into the edit machine, I had had a very smug look on my face! <BG>
 
I don't want to be too contentious here but....

Isn't Jitter mainly a CD phenomenon related to the physical reading of the disk and the fact that it doesn't have accurate timing information as part of the red book standard. I though similar effects can be avoided in trasmission by using a master clock on one device. Certainly my DAC seems to pick up clock from my sound card.

Digital by it's very nature is on \ off. So long as you get a signal at all, the timing and error correction should ensure that there is no nuance in the connection, details etc, but audiophiles swear by premium cables for digital connections (but then they do plenty of other stange things too.)

Say you are converting a WAV to a signal via your sound card and outputting it via spdif, what it is actually doing? I can see that if the process is not given sufficient resources, or you disk is slow, or if you're pulling accross a network then you can get the sound to dropout, but do these factors really have a significant effect on quality. I've done a bit of comaprative listening of CD vs computer to stereo (not surround decoder) DAC and it's close, but the CD wins out by a margin.

Has anyone done any research into this?

Thanks.

SB
 
There's a pretty long and quite boisterous thread on this very subject on the avsforum board if anyone wants to read - I'll try to paraphrase...

Yes digital is 1 and 0, but that stream flows from one device to another along a wire, exactly like an analogue electric signal. So unfortunately unlike the ramblings of the local high street electric retailers, digital does not mean always perfect.

Basically just because 44100 samples travel down the cable, that doesn't mean some of the 0's won't get switched to 1's or vice versa. If you use an excessively long cable or lay it across power cables you will get interference, the point is it may or may not be audible in your system.
 
I've read a good chunk of that thread and basically it seems to come down to poor design of the protocols. I made an erroneous assumption that if you are designing a transport protocol you add error detection and correction (like ieee1394 (firewire) for example) and that timing would be included (after the error of excluding it in cd-audio). This would deal with issues like interference.

This appears not to be the case on spdif for PCM audio, but on DTS as it contains timecode information in the file (as does WAV, incidentally, but I'm decoding it as PCM, I'll look at other options) this should be eliminated.

Hmmm, interesting. More research required methinks...
 
Hi Xeonic,

> Basically just because 44100 samples travel down the cable, that doesn't mean some of the 0's won't get switched to 1's or vice versa.

That isn't jitter though, what you're describing is errors.

Jitter is when the samples are not evenly timed.

For a full explanation

best regards,

Paul.
 
Originally posted by PaulP
Hi Xeonic,

> Basically just because 44100 samples travel down the cable, that doesn't mean some of the 0's won't get switched to 1's or vice versa.

That isn't jitter though, what you're describing is errors.

Jitter is when the samples are not evenly timed.

For a full explanation

best regards,

Paul.

Paul,
Thanks I did know that bit swapping interference wasn't jitter, but just simplified the case without getting into sync references, master clock and all that other stuff. Your link is pretty good anyway :blush:

spamboy,
That thread is good for a laugh to see how serious people can get. I think the reason older SB cards get trashed is because the SPDIF input upsamples everything to 48k, rather than leaving the signal unaltered. Cards like the M-Audio and Terratec leave signals alone, and have better quality ADCS/DACS than SB anyway.
 
Dodgey,
There is another thread here about this exact same thing. I have 2 SB Lives and
the new Audigy2 platinum. If you know what high end decoding and delivery sounds like with
a good dedicated DVD player and external amp/receiver or decoder, then you will be
disappointed with any of the SB's outputs. I also have another PC with an onboard soundcard and its
output although it has an optical out, is pathetic. I think its an AC97 which can't even
pass a 5.1 digital stream, but its stereo output is terrible anyway.
Most people here and over on AVS recommend the M-Audio route, so that's personally what I've
opted for now. I am just awaiting delivery of the card from the US.
 
I think we have to realise that CD-audio and digital surround data are very different, although they share the same SP/DIF connection.

Firstly, I agree that CD audio over SP/DIF is fragile. There's definately differences between digital outputs on diferent soundcards. Apparently this is because such data has no timing or error correction data in it and so the cpu or soundcard clock (if present) must be used to control send speed. If this is out of true, then the receiving DAC's performance suffer because of signal drift.

I read an interesting article that stated that playback digital jitter doesn't matter if you are recording it or capturing it, but does at the end of the chain when you need to convert it back through a DAC.

I also think that Dolby Digital (2.0, 5.1, etc.) is much less susceptible to jitter, because it's data is in packets with CRC error correction information. I think the decoder will have to buffer this to some extent, so IMHO if it was good enough to decode, then it's probably been re-clocked before it gets to the 5 or so DACs.

I haven't heard much difference between the 5 Dolby Digital sources I've owned; Creative Labs DXR2 decoder card, Wharfedale 950 player, Sony 925 player, HTPC SBlive, HTPC Audiophile 2496.

I'd bet these cheap on-board SP/DIF or USB SP/DIF outputs will be just as good as expensive soundcards, but only for Dolby Digital or DTS and won't sound right for CD-audio (or any PCM tracks).

regards,
Rob.
 
Hi Rob,

> I read an interesting article that stated that playback digital jitter doesn't matter if you are recording it or capturing it, but does at the end of the chain when you need to convert it back through a DAC.

Yes.

In the article I linked earlier in this thread, Bob Katz points out that a DAT recording may actually sound better when replayed, than when monitored during recording.

He suggests that this is because during replay, the clock is a local crystal clock, and during record, the clock is derived from the (possibly jittery) source clock. The point being that jitter is only important to the DAC (assuming that it is not so severe that the data cannot be recovered at all), and he has found no evidence that source jitter is in some way recorded.

> I also think that Dolby Digital (2.0, 5.1, etc.) is much less susceptible to jitter, because it's data is in packets with CRC error correction information.

Branxx has posted a link to a lively discussion on this topic over at avsforums.

The gist is that although the data is indeed packetised, and the decoding process buffers the data - that alone is not sufficient to render a jittery incoming clock unimportant.

The question is which clock is used to clock out the data from the buffers to the DACs ?

The avsforums thread suggests that if it is derived from the incoming jittery clock, then the sound may still be degraded.

Best regards,

Paul.
 
Hi,

I've read everything about soundcards.
I thought digital is ok as well.
I have a very cheap Terratec soundcard, but would like to upgrade to a far better one.
Which soundcard is now one of the best for this use?

Thank you,

Frederic
 

The latest video from AVForums

TV Buying Guide - Which TV Is Best For You?
Subscribe to our YouTube channel
Back
Top Bottom