MEASUREMENTS: Do Bit-perfect Digital S/PDIF Transports Sound The Same?

[Archimago]

The round up of Squeezeboxes used as SPDIF transports including the
Receiver, Touch, SB3, Transporter :-).

Some measurable characteristics of SPDIF - jitter, data rate/timing
differences between devices.

http://archimago.blogspot.ca/2013/06/measurements-do-bit-perfect-digital.html

Thanks again to "fordgtlover" (SB Receiver) for helping to make the
family measurements complete!



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Mnyb]

Good article :)

..and thats why I think the unmodded Touch is a sufficiently good
transport for any and all systems and listeners , It's "limitation" to
only 24/96 is of no particular concern either .

If you have ultra expensive high-end the design may be a bit off , but
it actually match pretty well with my gear (my Meridian system is gloss
black)



--------------------------------------------------------------------
Main hifi: Touch + CIA PS +MeridianG68J MeridianHD621 MeridianG98DH 2 x
MeridianDSP5200 MeridianDSP5200HC 2 xMeridianDSP3100 +Rel Stadium 3
sub.
Bedroom/Office: Boom
Kitchen: Touch + powered Fostex PM0.4
Misc use: Radio (with battery)
iPad1 with iPengHD & SqueezePad
(in storage SB3, reciever ,controller )
server HP proliant micro server N36L with ClearOS Linux

http://people.xiph.org/~xiphmont/demo/neil-young.html
------------------------------------------------------------------------
Mnyb's Profile: http://forums.slimdevices.com/member.php?userid=4143
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Archimago]

Post by Mnyb
Good article :)
..and thats why I think the unmodded Touch is a sufficiently good
transport for any and all systems and listeners , It's "limitation" to
only 24/96 is of no particular concern either .
If you have ultra expensive high-end the design may be a bit off , but
it actually match pretty well with my gear (my Meridian system is gloss
black)

Thanks for the note Mnyb.

Just posted a couple pix of the Squeezebox family before sending the
Receiver off :-).

http://archimago.blogspot.ca/2013/06/musings-squeezebox-family.html

Happy summer as we head into July, everyone (at least those in N.
hemisphere).



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[darrenyeats]

Archimago,
Erm ... if I understand your blog correctly this is quite a
mind-expanding set of evidence! You measured digital transports
exhibiting distinct frequency responses?! If this is true ... I think
your explanation about clock -speed- (as opposed to clock dirtiness) is
probably the prime suspect. But on the face of it a result contrary to
intuition for bit-perfect transports.

It would be interested to see results using a different DAC e.g.
Benchmark.
Darren



Check it, add to it! http://www.dr.loudness-war.info/

http://www.amazon.co.uk/gp/richpub/listmania/byauthor/A3H57URKQB8AQO/ref=cm_pdp_content_listmania/203-7606506-5721503.

SB Touch
------------------------------------------------------------------------
darrenyeats's Profile: http://forums.slimdevices.com/member.php?userid=10799
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Archimago]

Post by darrenyeats
Archimago,
Erm ... if I understand your blog correctly this is quite a
mind-expanding set of evidence! You measured digital transports
exhibiting distinct frequency responses?! If this is true ... I think
your explanation about clock -speed- (as opposed to clock dirtiness) is
probably the prime suspect. But on the face of it a result contrary to
intuition for bit-perfect transports.
It would be interested to see results using a different DAC e.g.
Benchmark.
Darren

Yeah, that would be interesting with the ASRC in the Benchmark.

Actually, maybe I can do it with the TEAC using upsampling to 24/192 and
seeing if the slight frequency response difference disappears as well...



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[soundcheck]

Comment:

What you measure is your measurement DAC (e.g. WM8805 ) receivers
capability to cope with the incoming signal.

The better the audio DACs digital interface, with all its error
corrections implemented, the better the source issues will be
suppressed.

It doesn't say much about the actual and absolute interface quality.

A WM8805 SPDIF receiver is specified with a constant 50ps output jitter.
No matter, if you use Toslink or SPDIF.

Toslink can exhibit several 100ns of jitter, magnitudes higher then
Coax. I don't think that's shown with your measurements. That requires
real world measurement equipment,

However. Manufacturers like Wolfson promise 50ps fed to your DAC. Your
measurements say that these 50ps are not always achieved with your DAC.

To me this (your) exercise is useless and misleading.

Cheers



::: ' Touch Toolbox and more' (http://soundcheck-audio.blogspot.com) :::
by soundcheck
------------------------------------------------------------------------
soundcheck's Profile: http://forums.slimdevices.com/member.php?userid=34383
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Archimago]

Post by soundcheck
What you measure is your measurement DAC (e.g. WM8805 ) receivers
capability to cope with the incoming signal.
The better the audio DACs digital interface, with all its error
corrections implemented, the better the source issues will be
suppressed.
It doesn't say much about the actual and absolute interface quality.
A WM8805 SPDIF receiver is specified with a constant 50ps output jitter.
No matter, if you use Toslink or SPDIF.
Toslink can exhibit several 100ns of jitter, magnitudes higher then
Coax. I don't think that's shown with your measurements. That requires
real world measurement equipment,
However. Manufacturers like Wolfson promise 50ps fed to your DAC. Your
measurements say that these 50ps are not always achieved with your DAC.
To me this (your) exercise is useless and misleading.
Cheers

Where are you saying is the WM8805 SPDIF receiver in this post? That was
the SPDIF receiver in the old AUNE X1, not the ASUS Essence One... I
assume therefore you are the one mistaken.

"Toslink can exhibit several 100ns of jitter" - do you mean "ps"? I'm
pretty sure I show 100+ps difference with the Transporter's TosLink vs.
coaxial if one were to calculate the peaks.



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[SBGK]

Post by Archimago
Where are you saying is the WM8805 SPDIF receiver in this post? That was
the SPDIF receiver in the old AUNE X1, not the ASUS Essence One... I
assume therefore you are the one mistaken.
"Toslink can exhibit several 100ns of jitter" - do you mean "ps"? I'm
pretty sure I show 100+ps difference with the Transporter's TosLink vs.
coaxial if one were to calculate the peaks.
BTW: What's with the "To me this (your) exercise is useless and
misleading" comment? Such an offensive tone is barely constructive in
this or really any domain don't you think? Suppose you were correct in
your comments above... Don't you think it's still useful to document
the (slight) differences between transport devices and encourage others
to explore the question since results could vary depending on the SPDIF
receivers?

based on your measurements, you seem happy for Mynb to draw the
conclusion that

"..and thats why I think the unmodded Touch is a sufficiently good
transport for any and all systems and listeners , It's "limitation" to
only 24/96 is of no particular concern either ."

which is not the real world experience of hundreds of touch users,
perhaps that is why they are useless and misleading.



Touch optimisations http://touchsgotrythm.blogspot.co.uk/
------------------------------------------------------------------------
SBGK's Profile: http://forums.slimdevices.com/member.php?userid=52003
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Archimago]

Post by SBGK
based on your measurements, you seem happy for Mynb to draw the
conclusion that
"..and thats why I think the unmodded Touch is a sufficiently good
transport for any and all systems and listeners , It's "limitation" to
only 24/96 is of no particular concern either ."
which is not the real world experience of hundreds of touch users,
perhaps that is why they are useless and misleading.

Well, the Touch as transport does perform pretty darn well in comparison
to the direct USB input considering the limitations of an SPDIF
interface! Sounds very good to me as well in stock form, so why should I
disagree?

I'm going to fundamentally disagree though that "real world experience
of hundreds of touch users" is:
A. A fact... Where's the survey? Who was asked - apparently they missed
me and maybe Mynb? Did they ask the average music lover or only
card-carrying "audiophiles"?
B. Substantiated... What's inadequate about the Touch as transport? How
was the conclusion arrived at by the "hundreds"?

That comment is truly what is "misleading" - in itself presented as fact
when it appears at best as opinion if not just some kind of vague
impression.

It's like saying that there are "hundreds of members in the Flat Earth
Society" (who knows... there probably are...). Am I supposed to be
impressed?



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Julf]

Post by SBGK
which is not the real world experience of hundreds of touch users

Would you happen to have any evidence supporting that statement?



"To try to judge the real from the false will always be hard. In this
fast-growing art of 'high fidelity' the quackery will bear a solid gilt
edge that will fool many people" - Paul W Klipsch, 1953
------------------------------------------------------------------------
Julf's Profile: http://forums.slimdevices.com/member.php?userid=42050
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[mlsstl]

This post might be inappropriate. Click to display it.

[soundcheck]

Post by Archimago
Where are you saying is the WM8805 SPDIF receiver in this post? That was
the SPDIF receiver in the old AUNE X1, not the ASUS Essence One... I
assume therefore you are the one mistaken.
"Toslink can exhibit several 100ns of jitter" - do you mean "ps"? I'm
pretty sure I show 100+ps difference with the Transporter's TosLink vs.
coaxial if one were to calculate the peaks.
BTW: What's with the "To me this (your) exercise is useless and
misleading" comment? Such an offensive tone is barely constructive in
this or really any domain don't you think? Suppose you were correct in
your comments above... Don't you think it's still useful to document
the (slight) differences between transport devices and encourage others
to explore the question since results could vary depending on the SPDIF
receivers?

Just take a Cirrus CS8416 or a TI DIR 9006 instead of WM8805. They all
have similar specs -- on the intrinsic jitter part. The CS8416 is the
worst here. The Wolfsons WM8805/4 seem to attenuate jitter best.
I'd guess that you'll find one of those three in 95% of all the devices
out there.


No. I mean Nano Seconds. Have a look at the datasheets or e.g. Phillips
TORX Toslink transmitters. Just ONE side exhibits theoretically up to
15ns ( n as in nano) jitter. Now take the airgaps, the 2nd side, the
cable asf asf.

It's the WM8805 ( or something else ), which gets that SPDIF jitter
down. We should not mixup SPDIF jitter specs and I2S jitter specs. On
I2S we need pico second jitter. On SPDIF you can have 100ns and the
receiver will still lock to it.


Besides those transceivers you'll find in quality DACs even more jitter
attenuation methods ( FIFO reclockers, ASRCS asf, the Sabre does it
inside). All these will mess up your measurement results by improving
the digital input signal.

Measurement equipment or setups are not allowed to change a signal!
Otherwise you'll measure the performance of your source + your
measurement equipment performance.





You might want to read this 'Wolfson'
(http://www.wolfsonmicro.com/documents/uploads/misc/en/Jitter_performance_of_spdif_digital_interface_transceivers.pdf)
document.



People who still think and claim digital is digital just have not
understood how things work.



::: ' Touch Toolbox and more' (http://soundcheck-audio.blogspot.com) :::
by soundcheck
------------------------------------------------------------------------
soundcheck's Profile: http://forums.slimdevices.com/member.php?userid=34383
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Julf]

Post by soundcheck
People who still think and claim digital is digital just have not
understood how things work.

Sorry to pick nits, but people who still think and claim digital is
digital are totally right in doing so - per definition. Your issue is
that sound is not digital, and in converting the digital data (that is
digital - by definition) that represents the sound into an analog
signal, the result is influenced by other things besides the digital
data.



"To try to judge the real from the false will always be hard. In this
fast-growing art of 'high fidelity' the quackery will bear a solid gilt
edge that will fool many people" - Paul W Klipsch, 1953
------------------------------------------------------------------------
Julf's Profile: http://forums.slimdevices.com/member.php?userid=42050
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[soundcheck]

Post by Julf
Sorry to pick nits, but people who still think and claim digital is
digital are totally right in doing so - per definition. Your issue is
that sound is not digital, and in converting the digital data (that is
digital - by definition) that represents the sound into an analog
signal, the result is influenced by other things besides the digital
data.

If the shoe fits...

Sorry for stepping on your toes.

Simplifying things won't help here. Sometimes things are just more
complex than just counting 0 and 1. People who're not accepting analog
problems originated from digital domain issues are nothing less then
clueless.

But that's a different story.


I'm just saying that measurements and conclusions related to the
digital domain of the OP are wrong and misleading. A different thing are
his analog domain (DAC outputs) measurements. Perhaps not perfect, but
usually good enough to be relevant. Though also with those rather useful
analog measurements the OP has a tendency to draw "final" and "general
applicable" conclusions that would explain everything.
That's at least from my perspective a bit of a simplistc view. He should
avoid to put opinions into his measurements. Let the numbers speak.
That's enough.



::: ' Touch Toolbox and more' (http://soundcheck-audio.blogspot.com) :::
by soundcheck
------------------------------------------------------------------------
soundcheck's Profile: http://forums.slimdevices.com/member.php?userid=34383
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Julf]

Post by soundcheck
Simplifying things won't help here.

But silly empty (and patently illogical) phrases like "kumquats are not
kumquats" don't help either.

Post by soundcheck
People who're not accepting analog problems originated from digital
domain issues are nothing less then clueless.

Jitter and noise are not digital domain issues in my book. They only
become issues at the point the data is converted to analog - that is in
the analog domain.

Post by soundcheck
The entire "serious" audio industry knows that common measurements are
not telling everything. Otherwise we wouldn't have seen any progress in
the last 20 years.

Actually, most of the progress in the last 20 years comes from being
able to measure - and correlate the measurements with audible results -
better.



"To try to judge the real from the false will always be hard. In this
fast-growing art of 'high fidelity' the quackery will bear a solid gilt
edge that will fool many people" - Paul W Klipsch, 1953
------------------------------------------------------------------------
Julf's Profile: http://forums.slimdevices.com/member.php?userid=42050
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Archimago]

Post by soundcheck
I'm just saying that measurements and conclusions related to the
digital domain of the OP are wrong and misleading. A different thing are
his analog domain (DAC outputs) measurements. Perhaps not perfect, but
usually good enough to be relevant. Though also with those rather useful
analog measurements the OP has a tendency to draw "final" and "general
applicable" conclusions that would explain everything.
That's at least from my perspective a bit of a simplistc view
undermining his efforts. He should avoid to attach opinions to his
measurements. Let the numbers speak. That's enough.
The entire "serious" audio industry knows that common measurements are
not telling everything. Otherwise we wouldn't have seen any progress in
the last 20 years.

Oi... I think the numbers do say enough in themselves and support the
opinions. I don't think I overstepped anything and made the case
reasonably clear based on the particular setup. To me all the transport
devices sound essentially the same subjectively and although measurable
differences exist, they are small and most likely beyond normal hearing
abilities to differentiate. Since I'm measuring the analogue out from a
reasonably good DAC (ASUS Essence One), the final output is all I really
care about and specifics about the internal SPDIF/I2S jitter tolerances,
etc. aren't as important unless I can detect the differences with my ADC
which is better than human hearing as far as I can reasonably tell!

As for "Otherwise we wouldn't have seen any progress in the last 20
years" - I'd say it's pretty *obvious* the improvements made in the last
20 years just looking at the "common measurements" like frequency
response, dynamic range, improvements in jitter compared to stuff in the
early 90's...

Here's a question for you - which "uncommon" measurements do you think
are important to consider then that could have significant impact on
audible sound quality? Which of these have the "serious" audio industry
made major improvements on?



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[soundcheck]

Post by Archimago
Oi... I think the numbers do say enough in themselves and support the
opinions. I don't think I overstepped anything and made the case
reasonably clear based on the particular setup. To me all the transport
devices sound essentially the same subjectively and although measurable
differences exist, they are small and most likely beyond normal hearing
abilities to differentiate. Since I'm measuring the analogue out from a
reasonably good DAC (ASUS Essence One), the final output is all I really
care about and specifics about the internal SPDIF/I2S jitter tolerances,
etc. aren't as important unless I can detect the differences with my ADC
which is better than human hearing as far as I can reasonably tell!
As for "Otherwise we wouldn't have seen any progress in the last 20
years" - I'd say it's pretty *obvious* the improvements made in the last
20 years just looking at the "common measurements" like frequency
response, dynamic range, improvements in jitter compared to stuff in the
early 90's...
Here's a question for you - which "uncommon" measurements do you think
are important to consider then that could have significant impact on
audible sound quality? Which of these have the "serious" audio industry
made major improvements on? I would love to learn about these factors...

Yep. I know.

That's why I brought up your "digital" measurement problem first.
Something to digest. I hope that's enough as a first lesson. ;) By
reading and digesting the Wolfson doc you'll very quickly gain some more
background information.

Enjoy.



::: ' Touch Toolbox and more' (http://soundcheck-audio.blogspot.com) :::
by soundcheck
------------------------------------------------------------------------
soundcheck's Profile: http://forums.slimdevices.com/member.php?userid=34383
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[soundcheck]

One more:

Mr. Jeff Rowlands 2ct:

http://jeffrowlandgroup.com/kb/questions.php?questionid=445



::: ' Touch Toolbox and more' (http://soundcheck-audio.blogspot.com) :::
by soundcheck
------------------------------------------------------------------------
soundcheck's Profile: http://forums.slimdevices.com/member.php?userid=34383
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Archimago]

'Assessment of the Importance of Audio Measurements in High-End Design'
(http://jeffrowlandgroup.com/kb/questions.php?questionid=445)

Why relate this discussion with rather vague comments from the maker of
primarily analogue preamps and amps? While the testing procedures they
speak of may be useful, it looks like an excerpt from a 1999 article in
"Ultimate Audio" - is that a well respected journal and has the testing
procedure talked about been substantiated? Is this test method being
used by others due to its utility?

I read this review of their recent Aeris DAC:
http://www.hifi-advice.com/Jeff-Rowland-Aeris-review.html
and wish that more on-line reviewers at least did some of the "common
measurements" to make sure the basics are done well - especially for a
DAC costing >$11K Euros and has USB input limited to 96kHz at this
price. It does look very pretty though!



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[soundcheck]

Nice try. Wrong subject. Nobody is discussing Aeris DAcs. We're or
better I'm discussing your problems and your partially limited view of
looking at things. You don't want me to dig out more examples!?!

I'd like to add some more stuff to this thread.

The OPs attempt of measureing "digital" software players such as JRMC
JPLAY FOOBAR etc. lead to no results or better neglectable differences.
You'll find some lengthy blogs about it on his blog page.

Why do I bring this up over here. I do see a kind of relation.

The OP doesn't understand that his high quality Teac DAC ( I do also
have one at hand) tries everything to manipulate the incoming data.
Exactly that what a measurements chain shouldn't do.

There's a lot of signal recovering, refreshing, reclocking and filtering
involved in a DAC device. That signal refreshing is pretty much similar
to above WM8805 example. The better it's done the less problems remain
on the output.

The Teac DAC probaly does an even better job then the the WM8805.


The OPs SW player measurements just show that his Teac+EMU
(digital-analog-digital-analog ??) maesurments chain is doing such a
good job that his EMU is not able to show significant differences.

It doesn't say anything about the SW player performance and a potenially
relevant USB layer 1 impact. Relevant measurements would have to be
conducted on the physical USB interface.
Otherwise no reliable results can be expected.

Cheers



::: ' Touch Toolbox and more' (http://soundcheck-audio.blogspot.com) :::
by soundcheck
------------------------------------------------------------------------
soundcheck's Profile: http://forums.slimdevices.com/member.php?userid=34383
View this thread: http://forums.slimdevices.com/showthread.php?t=99073

[Archimago]

Post by soundcheck
Yep. I know.
That's why I brought up your "digital" measurement problem first.
Something to digest. I hope that's enough as a first lesson. ;) By
reading and digesting the Wolfson doc you'll very quickly gain some more
background information.
Enjoy.

Methinks you misinterpreted my response.



Archimago's Musings: (archimago.blogspot.com) A 'more objective'
audiophile blog.
------------------------------------------------------------------------
Archimago's Profile: http://forums.slimdevices.com/member.php?userid=2207
View this thread: http://forums.slimdevices.com/showthread.php?t=99073