jf -

OT, but do you listen to wbgo through your Squeezeboxes?

I've been playing the station a lot lately, and am impressed by the quality of the stream.

s.

Thanks for the reviews, Jazzfan.

Quote:
As I stated in another post, with a $300 list price the Touch is really going to shake things up in the world of digital audio. JA get one of these marvelous units out to a Stereophile reviewer ASAP, please!

Yes!

And give it a good workout on the test bench, too.

Quote:
The re-buffering was only happening when playing back high resolution files. I stream the files as flac rather than wav.

Curiosity: have you tried streaming FLAC files made with different compression options? (-1, -5, -8, etc.)?

Is there a skip at the beginning of playback or elsewhere due to the rebuffering?

Quote:

I do listen to WBGO and WKCR while I am in my car since I live within range of their on air signals. I find WBGO to be rather conservative in their musical tastes. Rarely, if ever, do they play anything even remotely out of the mainstream, so no John Zorn, no Dave Douglas, no Ken Vandermark and hardly any David Murray or Art Ensemble of Chicago. And then to add insult to injury during week day mornings and afternoons WBGO plays at least 40% to 50% vocals and lots of really bad vocals at that. I sometimes feel that when any new vocal "jazz" (I use the term "jazz" very guardedly since much of the dreck they play sounds alot like Bill Murray's airport lounge act from his SNL days) CD shows up WBGO will play it but you'll never hear the new Henry Threadgill or Anthony Braxton (not even something from one his many standards CDs).

I noticed. I can live with conservative (there are times when Painkiller just doesn't make it) but last night I stumbled into smooth jazz hell.

Disappointing. Thanks for the notes.

s.

edit - I don't know why, but 'bgo *sounds* better on my very modest system than its putative rate.

Quote:
I know that there has been some discussion over on the Slim Devices forum regarding flac files made using the various compression options but I've never really noticed any difference from one flac file to the next.

Nor have I in any context.

My understanding is that the decoding is efficient and the actual work is done on the encoding side.

Thanks for the info re rebuffering.

Quote:
My understanding is that the decoding is efficient and the actual work is done on the encoding side.

That is largely true but decoding the higher compression rates does take more computations (higher order polynomials in the model, higher maxmimum partition order in the residuals). As you can see from this table the variance is approx 15% in decode CPU time between -0 and -8 (compared to a nearly tenfold increase in encode CPU time).

It seems that depending on the sample rate of the file and what else is going on at the time (wired vs wireless networking stack, whether or not VU meters are displayed etc.) that extra 15% can result in the Touches CPU being momentarily maxed out, thereby causing the stuttering.

Nice table, Struts. It makes more of a difference during decoding than I expected.

I am surprised that the Touch has problem with high resolution FLAC with wireless v. Ethernet as two tracks of streaming audio is not that much bandwidth.

Quote:
The Touch does not "stutter" it simply stops playing, as in silence, until the rebuffering is complete and then it starts playing again at the point where it stopped.

You're right jf, "stutter" is probably the wrong word, "sticking" might be a better metaphor here?

Quote:
I believe that the rebuffering i experienced was mainly due to the wireless signal strength (around 70%) and the higher demand put on the network by the high resolution files. As I stated earlier, the Touch has no problem with high resolution when connected via wired Ethernet.

I have been travelling for a couple of days and not yet caught up with the threads over at the Slim forum but when I last checked the bottom line was that the exact cause of the rebuffering problem wasn't yet established. Several folks had reported seeing the problem with wired connections, others not, some were reporting maxed-out CPUs, others not.

One interesting aspect is that iirc among all the sometimes contradictory observations the problem has only manifested itself with FLAC files encoded with higher compression settings. Streaming the equivalent WAV or FLAC -0, even wireless, does not appear to reproduce the problem. That the problem seems to be sensitive to the FLAC compression ratio could point in the direction of a CPU issue, however I admit that at this stage that is pure conjecture.

The Touch represents a quite complex set of interdependent subsystems so it will be fun following the troubleshooting effort and fascinating to find out what is actually causing the problem. It seems there are some ARM optimizations for the FLAC decoder that never made it into the main code branch so even if it is a CPU issue it might still be fixable with a firmware upgrade.

Quite a nice, no-BS review of the Touch over at Computer Audiophile.

Quote:
I am surprised that the Touch has problem with high resolution FLAC with wireless v. Ethernet as two tracks of streaming audio is not that much bandwidth.

I don't think the problem has anything to do with network bandwidth. I still strongly expect this is a CPU issue, possibly exacerbated by the increased latency of wireless vs wired.