[KillingTime]

Hi,

Not sure if this is the right sub-forum, but it is to do with source components.

I have a large music collection, and all my CD's were becoming a hassle to manage. I thought I might get around this by ripping the whole lot to HD and use a media player with digital out to play my selections. I'd then use an outboard DAC to get the quality I was used to from the hi (ish) end CD player I was used to.

I could have used a PC as the media player but PC's need lots more attention than dedicated players (need shutting down properly, they need re-installing every so often etc) so I opted for netmedia 101 media player. This is designed to play movies from it's internal HD, but it has SPDIF out and can play mp3's & wma. This is a low end (£50) player but that's really irrelevant here. The point is it has digital audio out, so it's being used as a digital transport of sorts.

I swapped the CD player for the netmedia 101 and kept the rest of my system the same (amp, headphones, DAC, cables etc). There was a difference. It took me while to figure it out, but the 101 sounded less clear, the soundstage was gone too.

Given that all I've changed is the digital transport, how can this be?

The whole point of SPDIF is that 1's and 0's are sent over the cable. If both transports are doing their job properly, then the DAC sees the same 1's and 0's from either, so the sound should be the same(?).

I looked at SPDIF & TOSLINK via the wiki, and it says:

"data is sent using Biphase mark code, which has either one or two transitions for every bit, allowing the original word clock to be extracted from the signal itself."

I assume this means (correct me):

If the DAC uses the timing info in the SPDIF stream to re-assemble the bits, then the quality of the sound will be dependant on the quality of the clock on the transport, because the timing info is embedded into the stream (?).

If the DAC buffers the data and then uses it's own (more accurate) clock, then there's no problem.

I don't know enough about digital audio to know which of the above is correct, or whether they're both correct, the answer being dependant on the the implmentation of the DAC.

More accurate digital clocks are sold as upgrades for CD players, advertising a resultant wider soundstage and cleaner sound, so I'm assuming the stability of the timing signal has a bearing on the quality of the sound that comes out of the same DAC also.

Anyone shed any light on this?

Thanks.

 

[Pars]

Here's a good article, including pointers to some AES stuff, etc.

 

[sejarzo]

Quote:

Originally Posted by KillingTime /img/forum/go_quote.gif I have a large music collection, and all my CD's were becoming a hassle to manage. I thought I might get around this by ripping the whole lot to HD and use a media player with digital out to play my selections. I'd then use an outboard DAC to get the quality I was used to from the hi (ish) end CD player I was used to. .......This is designed to play movies from it's internal HD, but it has SPDIF out and can play mp3's & wma. This is a low end (£50) player but that's really irrelevant here. The point is it has digital audio out, so it's being used as a digital transport of sorts. .....Given that all I've changed is the digital transport, how can this be?

Seems to me that you have changed a whole lot more than just the transport!

Are you not comparing the full resolution, uncompressed datastream from from the analog output of your CD player (or its digital output, fed to an external DAC) with the lossy encoded/compressed then "reconstituted" datastream that is now fed to the external DAC (the end result from ripping to mp3/wma format)?

If your player could handle FLAC or some other lossless compression format, then you would have a valid comparison. I tried to find more detailed info on the "netmedia 101" device but came up empty.

 

[KillingTime]

Pars:
Thanks for the article. It looks like most DACs use the clock info from the SPDIF signal, which in turn comes from the transport. This confirms why the audio sounds different. There are some interesting solutions to this problem, namely clock injection from the DAC to the transport, but this is not found in consumer audio equipment.

sejarzo:

I took a 300kbps mp3 and converted it to WAV, then burned it to CD in CD audio format so it would play on a standard CD player. I used the TOSLINK (optical version of SPDIF) to feed the DAC.

When replacing the CD player, I kept the TOSLINK cable but used the raw mp3 from the hard disk.

The only thing that's changed is the algorithm that converts the compressed mp3 to raw audio. The netmedia 101 uses it's own algorithm from firmware, whereas when I was using the CD player, I was using the software provided by NERO.

[A more accurate comparison would have been to have the 101 play the WAV, but I don't think it supports this.]

I don't think this is the problem though. All the frequencies are there in both playbacks, it's just that the 101 sounds less defined.

The purpose of this post was to find out if the quality of the digital audio transport is important to the quality of the audio if you're using an external DAC, and the article pointed to above (if accurate) answers this (yes, I think).

This would explain why some manufacturers release two versions of the same transport at different price points (Pixelmagic mb100\mb200), the higher end one being more expensive because of the more accurate (read: stable, less jitter) clock.

The point to all this (as I've just found out), is that it's not enough to rely on a good external DAC to realise your audio quality, the quality of the timing signal in your digital transport is important too (unless you've opted for a DAC that reconstitutes the clock signal). So, if you're looking to use your PC as a cheap media player, the design of the sound card is still important, even if your only using the sound card for the digital out.

 

[JaZZ]

KillingTime...

...obviously there's some processing going on in your Net Media 101 player, and be it «just» in the form of MP3 decoding. Which may in fact be the crucial point. MP3s with about 300 kb/s (which is much) can sound indistinguishable from the uncompressed original to average ears, but the decoding algorithm plays an important part. My McCormack UDP-1 does an excellent job with CDs, SACDs and Audio-DVDs, but doesn't sound good with MP3s -- whereas on my computer (with foobar2000) the same MP3s sound virtually as good as their Wave equivalent.

Nevertheless, different digital drives do sound different to my ears, also with uncompressed recordings. I haven't directly compared my computer with my UDP-1 in this respect, though. After all my Bel Canto DAC2 sounds better when fed by the UDP-1 than when fed by the Philips DVD 963 SA. This although the Philips shows better jitter behavior.
.

 

[sejarzo]

Frankly, what I fail to understand is why one would use a "hi (ish) end CD player" to play mp3's. Its value would be in playing what it was intended to play--the original, lossless RedBook CD.

 

[KillingTime]

JaZZ:

Perhaps it's a combination of both, mp3 decoding and the digital transport. I was initially sceptical about problems coming from the decoding because there is very little artistic leway in decompressing mp3's. The same is not true of encoding mp3's and many vendors compete in this arena (Bladenc, lame etc). A great deal is not made public when it comes to how vendors achieve what they publish for their hardware. For example, some of the early soundblaster cards that supported digital audio out were shot down on audiophile forums because they altered the digital audio stream before output (audio enhancement etc) and it was impossible to turn this off. It is quite possible that the digital stream is being altered in some way to conserve processing power on the 101 - in an effort to reduce production costs - or perhaps it's just bad design \ poor quality components - who knows.

It's a cheap HD box with a USB connection and I bought it out of curiosity. Don't use it any more. Gone back to CD's for now.

I would still like to move to a central store for my music at some point. Managing it all is just easier, and the purpose of this post was to find out a bit more about the SPDIF spec and digital transports so I know what to look out for in the future (as well as what to avoid).

sejarzo:

Sometimes compressed audio is all you've got. I preview nearly all my music now on mp3 - and buy it when I find what I like - because of the quality. The rest just gets deleted - wouldn't waste the HD space. I find that I keep 1 in 10 tracks that are auditioned (approx), and when you need to listen to 20 albums, having to burn 20 CD's when the alternative is a simple file transfer is obvious. Getting the most out of an mp3 is a valid persuit in this day and age, and I thought I'd share my failures with others so they can avoid the same.

 

[Edwood]

LOL, no one has brought up I2S in this thread yet.

-Ed