PDA

View Full Version : RE: SB Sound quality (Danny Rego)



Phillip Kerman
2004-09-30, 08:44
Without getting on your case Rob, I do have to disagree with a few
points....

> If you have any concern about sound quality of the Squeezebox v.s. CD
> player, and are playing any kind of compressed (MP3) output, then you
> should investigate your encoder way way before you start
> worrying about
> external DACs and audophile digital cables. There are real
> non-at-all-controversial reasons to believe that compression
> affects the
> sound.

This is definitely a fact. I think it's actually safe to say
that--depending on the nature of the sound on the original CD--an MP3 will
always sound worse.



> I am told that 128 kbits/sec constant bit-rate MP3 encoding
> is audibly
> different from CD quality. I have been using lame with -v -b
> 112, which
> is a variable bit rate averaging about 160 kbits/sec, which
> supposedly
> is near-CD quality. I can't hear the difference, but what the heck.

Yes, it has been stated that 128 kbps is near CD quality--who ever came up
with that doesn't have very good ears however. I can definitely hear the
difference between raw WAV and 320kbps MP3 in every piece of music I've ever
tested. Sometimes it's very very subtle though.


> A louder source will always sound better. This
> can't really be
> done properly without using a meter and a test CD, but if the
> badness of
> the SB seems not to depend on the SB volume setting, even
> when the SB is
> clearly louder, then you probably are hearing a real difference.
>

Not sure what you mean... the A/B should both be louder or the one which
happens to be louder will sound better? I think---in addition to setting up
a fair comparison for volume--one should also pick out a good piece of
music. Generally, recordings of real live things: voice and accoustic
instruments are easier to recognize as accurate or distorted because our
brains know what they should sound like. Also a piece with a large dynamic
range is good to test. MP3s really get crappy when you listen to changes
from quiet to loud.


> So far as the effects of different digital cables or even
> external DACs,
> the effects are clearly only heard by a minority of people, and many
> people with relevant expertise in electrical engineering and human
> hearing performance tend to be skeptical that these effects exist at
> all.

I don't know. I just used an okay Dennon receiver's DAC and my Rotel's DAC
and in both cases the sound quality difference was striking (compared to the
internal DAC in the SB). Night and day almost.

Regarding the cables. The digital coax cable I had at first (~$25) is
notably worse at reproducing trebble... compared to the one with silver in
it (~$50). The difference was very subtle though. And, I didn't bother
looking at cables costing way more. The thought that these are low end
cables is almost funny.

Anyway, I'd say a separate DAC vs. the one in SB is not a subtle difference.

Regarding this link:
http://2eyespy.tripod.com/myaudioandhometheaterhomepage/id3.html

Looks like some fun reading. I can say with utmost certainty that some of
the statements on that page are absolutely false. (Which, by the way, does
not necessarily mean the rest are false.) For example, suggesting speakers
don't experience a break-in period is misguided.



Thanks,
Phillip

Bob Fish
2004-09-30, 10:06
>
> > Regarding the cables. The digital coax cable I had at first (~$25)
> > is notably worse at reproducing trebble... compared to the one with
> > silver in
> > it (~$50). The difference was very subtle though. And, I didn't
> > bother looking at cables costing way more. The thought that these
> > are low end
> > cables is almost funny.
>
> Yet more subjectivity!
>

Isn't it Digital? That would mean that if the digits are
getting across the cable, there would be no difference in
sound quality for any given cable. If you could pass the
digits across a phone line, you'd get the same results.

The only differences that I can think of are bad shields
letting outside noise leak into the amp via a ground or
other analog path.

It is difficult to justify expensive cables, but people love
to do it (so do marketers).


--
Robert Fish -> king (AT) bobfish (DOT) org
www.bobfish.org PGP available

Bob Fish
2004-09-30, 11:44
> > -----Original Message-----
> > From: discuss-bounces (AT) lists (DOT) slimdevices.com
> > [mailto:discuss-bounces (AT) lists (DOT) slimdevices.com] On Behalf Of Bob Fish
> > Sent: 30 September 2004 17:07 To: discuss (AT) lists (DOT) slimdevices.com
> > Subject: RE:[slim] SB Sound quality (Danny Rego)
> >
> > >
> > > > Regarding the cables. The digital coax cable I had at
> > first (~$25)
> > > > is notably worse at reproducing trebble... compared to
> > the one with
> > > > silver in
> > > > it (~$50). The difference was very subtle though. And, I
> > > > didn't bother looking at cables costing way more. The thought
> > that these
> > > > are low end
> > > > cables is almost funny.
> > >
> > > Yet more subjectivity!
> > >
> >
> > Isn't it Digital? That would mean that if the digits are
> > getting across the cable, there would be no difference in
> > sound quality for any given cable. If you could pass the
> > digits across a phone line, you'd get the same results.
>
> Not quite. Bad tranmission can cause degradation of the digital
> signal, particularly the timing information causing jitter.
>
> Devices are available to condition the digital signal prior to the
> DAC, for example the Perpetual Technologies P-1A
> (http://www.av123.com/products_product.php?section=processors&product=
> 1.1). One of the things devices like this do is to re-clock the
> bitstream to reduce jitter.
>
> So, even digital cables can make a difference.
>
> R.

Perhaps I misunderstand the spec, but the data rate on the
digital cable is only marginally related to the actual
sample data rate

Signal Sample
2.8224 Mbit/s = 44.1 kHz
3.072 Mbit/s = 48 kHz
2.048 Mbit/s = 32 kHz

and the interface is really kinda like a serial port.

You wouldn't feed this directly into a DAC, you would buffer
it and clock it into a DAC with an accurate clock. Jitter
would not be an issue.