Pc -> Dac, How ?

[spzzzzkt]

Originally posted by Telstar


I was referring to the other Coreworks chip.

Coreworks are selling an IP core not a chip.

[phofman]

francolargo:

I am not sure we understand each other.

The compressed format you have was originally 44.1 or 48kHz?

If you uncompress, i.e. convert back to wav (b.wav), it will certainly sound different than the original a.wav. Both a.wav and b.wav will have the same bitrate of 2 x 16 x 44100 = 1.4Mbps. b.wav will have a limited frequency range since the MPEG-4 coder cuts high frequencies.

Now what I do not understand is why b.wav upsampled to 176.4 should sound worse than b.wav, while upsampling a.wav produces an improvement.

[francolargo]

Hi Phofman,

What I know for sure is that the original compressed files are ripped at 256kb/s and play at 44.1kHz. I assumed that meant that during decoding, fewer samples were repeated in a serial fashion to fill up the bitrate to 1.4mb/s. That's the signal I don't want to upsample to 176.4 kHz.

Frank

[dublin78]

Mid-thread conclusion please: Are we better off using something like an Empirical Audio USB to Spdif converter, or something like a squeezebox duet (assuming both are run into a decent quality DAC)?

http://www.empiricalaudio.com/frComputer_Audio.html

http://www.slimdevices.com/pi_duet.html

[phofman]

Originally posted by francolargo

What I know for sure is that the original compressed files are ripped at 256kb/s and play at 44.1kHz. I assumed that meant that during decoding, fewer samples were repeated in a serial fashion to fill up the bitrate to 1.4mb/s. That's the signal I don't want to upsample to 176.4 kHz.

The MPEG compression works rather differently. Basically, it performs frequency analysis of the incoming wave, breaks the whole band into many sub-bands and records at regular intervals the sub-bands signal levels. Lower bitrate means information only on more significant frequency sub-bands is included in the resultant stream.

The decoder kind of "plays" the frequencies, creating a new waveform (wav) somewhat similar to the original one. The new waveform does not repeat samples in a serial fashion, it just has a different shape from the original wav. Since it is a regular wav (PCM), its upsampling should produce the same improvement (if it is the case) as upsampling the original wav.

[francolargo]

Interesting!

Perhaps with the mpegs I'm noticing an effect similar to one noted over on the ESS Sabre reference thread.

http://www.diyaudio.com/forums/showt...12#post1735012

There it was observed that with a better DAC the poorer quality recordings actually sound worse. We ARE talking about subjective SQ, and in a similar way, perhaps the degradation of mpeg encode/decode is resolution dependent. I was not planning to re-rip most of my 'average'-quality recordings to a higher resolution but now am re-evaluating... It may be necessary to maintain two source resolutions, a lossless one for the living room and a compact one for the iPod...

over-and-out,

Frank in Mpls.

[phofman]

francolargo,

You might want to check out http://mp3fs.sourceforge.net/

[phofman]

Oops I forgot this was not the linux thread Perhaps someone has produced a similar functionality for windows.

[greggp]

Originally posted by phofman
Oops I forgot this was not the linux thread Perhaps someone has produced a similar functionality for windows.

You mean like the basic transcoding feature with most popular media players like Windows Media Player, iTunes, J.R.Media Center, Foobar2000, Media Monkey, etc., etc., etc.

All of these apps will automatically transcode your audio on-the-fly when you transfer tracks to a portable player like the iPod.

[phofman]

Originally posted by greggp
You mean like the basic transcoding feature with most popular media players like Windows Media Player, iTunes, J.R.Media Center, Foobar2000, Media Monkey, etc., etc., etc.

All of these apps will automatically transcode your audio on-the-fly when you transfer tracks to a portable player like the iPod.

Thanks for the insight. The virtual filesystem mp3fs does it slightly differently. It provides a read-only directory where you get a mirror of a master directory structure, and any flac file in the master is automatically converted to mp3 when the mp3 file gets read. It is a standard filesystem directory structure, you can access it with any program or tool.

But in the end, all the alternatives do the job