Digital Outputs sound VERY different... puzzled.

[parrot5]

The issue might be solved.... or not. See my updates: POST #7 and #30 and #37

I just A/B-ed the Coax output of my M-Audio Firewire Audiophile (powered by external power supply) and the Coax output of my Toshiba SD-3980, and they sound SO different from each other. (I used the Beresford DAC, which has two coax input and two coax output, so that I can switch between the two devices with the press of a button)

I thought digital outputs are supposed to be the same? OK, we don't live in an ideal world, but there's no way they should sound so much different? The M-Audio Firewire Audiophile sounds obviously brighter than the Toshiba SD-3980. (Or to put it another way, the Toshiba sounds bassier and muddier.)

To make sure it isn't a faulty cable, I switched the cable, and the sound signatures stay the same.

Now I'm puzzled, and I lost my reference point. For example, when people say a certain DAC is superior, now the transport has to be taken in consideration too, and maybe it is the MOST important piece in the chain, as even the external DAC feeds of this...


As I said, I'm just shocked and puzzled. Anyone shed any light into this issue?

The issue might be solved.... or not. See my updates: POST #7 and #30 and #37

 

[memepool]

If I understand correctly you are listening to the computer hardrive via the M-Audio and a DVD player and comparing the outputs via spdif into the same dac.

Why wouldn't they sound different? On the one hand paying music from a harddrive will have no jitter issues compared to a device which is designed to decode in real time like a DVD/ CD transport. Moreover a DVD transport isn't primarily designed to decode audio so it's not optimised for this task.

Secondly the signal is going through your computer / firewire / usb busses and then through the audiophile's signal processing perhaps? I am not sure how the internal architecture of the audiophile is implemented. It's possible there is some resampling or something going on.

SPDIF in both cases is a specific audio transport protocol so there would be some recoding involved compared to say something like USB/Firewire which is a pure data bus.

 

[tot]

Assuming the bitwise copy is the same (are you sure it is?), it is DAC's fault if they sound different.

In addition to bits, the digital signal carries also timing information and that is always affected by the transport and cable jitter. Proper DAC should be immune to this incoming jitter and "average" the incoming clock.

 

[parrot5]

Thanks for the replies. To clarify some issues that you two raised:

1. Source for the M-Audio Firewire Audiophile is EAC-ripped FLAC, from original CD.
2. Source for the Toshiba SD-3980 DVD player is the original CD.

Resampling issue:

1. For both sources, all effects are turned OFF.
2. For the M-Audio, I'm always using ASIO. There should be no resampling going on, as the soundcard supports 44.1kHz natively.
3. For the Toshiba player, its manual just states that only 44.1kHz audio cds are supported. 44.1Khz inputs may be resampled?

Device paths:

1. For the M-Audio Firewire Audiophile, the path looks like this:
   VIA chipset Firewire PCI card -> externally powered M-Audio Firewire Audiophile SPDIF output -> Beresford DAC
2. For the Toshiba SD-3980, the path looks like this:
   Toshiba SD-3980 SPDIF output -> Beresford DAC

I also have the MicroDAC so I can test whether it's the DAC's fault. I just thought having two coax inputs (the Beresford) is more convenient to switch between the two devices. I'm still puzzled as to the VERY different sound signatures. I'll try the MicroDAC next. In the mean time, any more light on this issue?

 

[d-cee]

are you playing the exact same CD on both? I'm assuming you are...

 

[parrot5]

Quote:

Originally Posted by d-cee /img/forum/go_quote.gif are you playing the exact same CD on both? I'm assuming you are...

O yes of course

And same tracks too.

 

[parrot5]

The issue might be solved. Thanks to the replies above about the possibility of RESAMPLING.

I know the M-Audio Firewire Audiophile should not resample Audio CD's 44.1kHz signal, so maybe it's the Toshiba DVD player.
In my computer, I have a couple of sound tracks of live performances that I rip from DVDs (48kHz). They are direct "demux", not "recorded", so it should be an exact copy (think as exact as Copy and Paste)......

Sure enough, when played in both sources, they sound the same now!

Conclusion: Toshiba DVD player resamples the Audio CD signals, degrading quality while doing it. No thanks.

I also wonder now, how some people believe cheap transports doesn't matter as long as you use an external DAC. I guess it should say "cheap transports that do not resample"


Now I need a dedicated CD player or a DVD player that supports 44.1kHz... sigh....

 

[bigshot]

Did you balance the line level using preamps? Minor output differences can make big differences in sound when you're A/Bing.

Resampling is probably the main cause though. That makes a LOT more difference than the insignificant differences between DACs and the inaudible effects of jitter.

see ya
Steve

 

[parrot5]

I tried using preamp and not using preamp. Sound signature stays different. As I said, when there's no resampling messing around, they sound the same again.

 

[tot]

Quote:

Originally Posted by parrot5 /img/forum/go_quote.gif I also wonder now, how some people believe cheap transports doesn't matter as long as you use an external DAC. I guess it should say "cheap transports that do not resample"

But is it really a transport anymore if it resamples?

 

[audioengr]

The difference is primarily due to the jitter magnitude and spectra.

Steve N.
Empirical Audio

 

[bigshot]

Are you saying that the resampling isn't affecting the sound and jitter is?

See ya
Steve

 

[Jigglybootch]

Quote:

Originally Posted by audioengr /img/forum/go_quote.gif The difference is primarily due to the jitter magnitude and spectra. Steve N. Empirical Audio

I don't buy that.

 

[happybob]

Quote:

Originally Posted by parrot5 /img/forum/go_quote.gif The issue might be solved. Thanks to the replies above about the possibility of RESAMPLING. I know the M-Audio Firewire Audiophile should not resample Audio CD's 44.1kHz signal, so maybe it's the Toshiba DVD player. In my computer, I have a couple of sound tracks of live performances that I rip from DVDs (48kHz). They are direct "demux", not "recorded", so it should be an exact copy (think as exact as EAC)...... Sure enough, when played in both sources, they sound the same now! Conclusion: Toshiba DVD player resamples the Audio CD signals, degrading quality while doing it. No thanks. I also wonder now, how some people believe cheap transports doesn't matter as long as you use an external DAC. I guess it should say "cheap transports that do not resample" Now I need a dedicated CD player or a DVD player that supports 44.1kHz... sigh....

hmm... i know with oppo dvd players, you have the choice of outputing "raw" digital signal or re-sampled pcm digital signal up to 96hz from the spdif outputs. did you check your toshiba dvd player's setup menu to make sure that it's outputing raw/unmodified signal? one more thing, have you tried playing the CD from your computer's CD or DVD player? that would rule out any difference between CD and FLAC files.

 

[parrot5]

To happybob:

1. The Toshiba SD-3980 outputs "PCM" only for Audio CDs. (stated in the manual) (The other option being "Bitstream" but only for DVDs, and then my MicroDAC or Beresford does not decode Dolby etc.... so.. yeah, PCM only then)
2. Using the computer, playback through the CD/DVD drive and FLAC sounds the same to me.

Thanks to everyone for responses. If you have any more thoughts on this issue, keep them coming. Or else I'm losing faith in cheap DVD players....