View Full Version : Squeezebox/External DAC - audio quality

Mark Bennett
2004-03-17, 15:45
Tonight we tried again to connect the Squeezebox up to an
external DAC with a high end sound system behind it.

The sound system consisted of:

Audio Synthesis DAX external DAC
Naim Pre-amp (don't know which one)
Naim Power amp (ditto)
Pair of very large KEF Floorstanding speakers (ditto)

We had some technical issues which meant we could only try
with MP3 encoded files (see "Squeezebox/External DAC -
technical issues" thread). However, these were encoded with
Lame at 320kb/s, and also with the quality=0 setting
(supposedly best quality). This was disappointing, since I
was all setup to try Flac streams......

Secondly, we also did not do blind testing. Just simple A/B
testing. I fully accept that this is not a completely reliable
comparison route, and once we have Flac working properly we
will try proper (double) blind testing.

All that considered the audio quality was extremely good. The
two main differences were a difference in base extension, and
a slight lack of precision in the sound. I'll explain what
I mean by these statements below, but I just want to make one
comment first.

The quality from the Squeezebox/DAC under this scheme was
pretty damn good. It was significantly better than my current
system at home, and my audiophile friend pronounced that it was
good enough to listen to all day. Generally well done to the
Slim Team.

So, what does "lack of bass extension" mean? When listening
to the same tracks, with plenty of bass content, you could
actually feel the low notes coming through the floor. With
the SQ/MP3 combination this really low frequency bass just
wasn't there. We both fully believe that this is entirely
due to the MP3 encoding rather than any SB issue. The bass
that was there (certainly more than my home system can
replay) was pretty good (subject to my next comment).

Now onto "lack of precision". This is a bit more subjective
but many aspects of the sound just weren't quite as clearly
defined as when coming from the CD transport through the
same DAC. The two key points here are that the instrument
location in the soundstage wasn't quite so accurate and
that the timing didn't seem quite so precise. Sharp drum
beats from the CD became slightly extended and slightly
merged into each other rather than being discrete beats.

In context all of these are a pretty fine points, and the
difference was small enough that given more time we should
have tried blind testing to see if it was really there. I
also suspect that most (if not all) of this was caused by
the MP3 coding rather than any other effects, but we can't
yet conclusively rule out clock jitter here. Once we get
Flac/Wav decoding working we should be able to take the lossy
encoding out of the equation and truly compare like with like.

In summary, it's starting to look like the SB has the
potential to replace a serious CD transport in a high-end