First, thanks for the feedback you have provided!

I'm using FLAC files and they are gapless.  I can play them gapless on my iRiver and my PC with foobar2k.

The previous poster suggested CPU may be the culprit and I think they may be right.  If I reduce the quality of the upsampling algorithm, or eliminate it entirely, the issue goes away.  I've done a few things to reduce CPU utilization and they seem to be helping too.

I've read a few articles that suggested upsampling in computers could improve audio quality.  I haven't done A-B testing, but some informal listening going from 16-bit/44100 to 24-bit/48000Hz (with foobar2k) seemed to have a nice positive effect, so I will continue to experiment with it.  Some of the theories have to do with jitter reduction and other factors -- I don't claim to understand it yet.  Here is one example article: http://www.positive-feedback.com/Issue22/nugent.htm

Going from 44100 to 48000 isn't a big change adjustment, but I expect to get a high-quality outboard USB DAC in the next year or so, which would allow me to go to 24-bit/96000Hz minimally or 24-bit/192000Hz.  A couple of higher-end ones (for me at least) are below:

• http://www.benchmarkmedia.com/dac1/
• http://www.belcantodesign.com/prod_eOneDAC3.html

edit - split thread from http://amarok.kde.org/forum/index.php/topic,14617.0.html - eean

If you want that kind of quality you are much much better of with sending the raw PCM stream to an external reciever and let that do the job. That way you ensure that you will get the best possible quality, given with your amp/speaker setup

I am a bit shocked by reading the site you linked to http://www.positive-feedback.com/Issue22/nugent.htm

See, he states that magicly one can make an algorithm that recreates details in the music from thin air :S

Upsampling is one of the BIG advantages of using a computer to drive your music. Digital music is recorded at a particular sample-rate, generally 44.1 kilosamples per second on CD's. Upsampling or resampling adds more samples that were not part of the original recording process, but try to approximate the samples that would have been recorded had the music been recorded at the higher sample rate initially. The added samples are computed with various mathematical algorithms that examine the music waveform prior to and after the time-slots where the new samples will be inserted. These new samples not only add more detail to the music, but improve the dynamics as well. Upsampled data can be 24 bits at 88kHz, 96kHz or 192kHz. Most of the currently available converter chips do not support 192kHz.

The first thing he fail is to understand HOW music is recorded. Music IS recorded at 192KHz or more at professional studios, and the downsampling is only done when mastering the CD's. Second thing, no math - no matter how complex can just recreate missing things - it can however, guess, but that's all it is, a guess. And it will be wrong sometimes and make a wrong guess. Do your self a fawor, do ABX blind tests, and do lots of them - just make sure you are not biased when judging.

And if you really want quality, get an expensive, high quality turntable, pickup and amp - the the frequency range can be well above 44.1KHz if you know how to treat your LPs.

I said upsampling was controversial!

I didn't understand the point about raw PCM stream above?  My music files are uncompressed today, so I shouldn't be losing any information (unless I'm missing something).  My $600 amp's only digital input is through a USB port.  I've experimented between a budget outboard DAC (the $200 Headroom MicroDAC) and my amp's internal DAC and so far the outboard solutions seems better with the same input formats from my Linux source.  Is there some other advantage of PCM?

I do feel confident the author of article is well aware of how CDs are actually recorded and mastered. The guy is an experienced patent-filing engineer and now runs an audio equipment company.  He seems to be using the term 'recorded' where you use the more accurate term 'mastered' but I'm pretty sure he means the same thing. It is possible he is selling snake oil of course, but it won't be because he's unaware of how CDs are recorded and mastered.

No, he does not claim to be recreating the original details (e.g., "magically"), but approximating what might have been there in a way that is pleasing to the ear.

There are many, many examples of products that upsample to deliver purported improvements in sound, and many of these are well-reviewed. It is built into software, DACs, amplifiers, CD players, et cetera.  There are some theories about why "guessing" this information back into the data (with algorithms) can improve quality if executed well, and they don't have anything to do with accurately guessing or recreating the original material.

I agree ABX tests are the right course and will do so when I get the time; at the moment setting this system up is what I do in a very small amount of spare time and a non-ideal environment. I'm trying to strike a good balance between sound quality, budget and convenience.  So I don't anticipate a move back to vinyl (expense, lack of convenience). My experimentation with computer audio is for its practical, ergonomic advantages (I can clear out the CD closet for other uses, my wife can find-and-play the music she likes, playlists, et cetera).  My speakers aren't going to be ideally placed three feet from the wall because my boys will knock them over or stand on them and fight each other with sticks or something.  But given these parameters, I am hoping to improve sound quality and I'm interested in what other folks have experienced.

Anyone else have feedback on whether they like upsampling in their environments?

More interesting articles:
http://forum.stereophile.com/forum/show ... Main=29286
http://www.stereophile.com/asweseeit/344/index.html
http://www.dcsltd.co.uk/technical_papers/aes97ny.pdf
http://www.audioasylum.com/audio/digita ... 33507.html