USB Audio ADCs
 

In article , Huge
wrote:
On 2014-03-11, Jim Lesurf wrote:
In article , RJH
wrote:
On 10/03/2014 16:49, Jim Lesurf wrote:
In article , Eiron
wrote:
On 10/03/2014 15:15, alexd wrote:


No gold was harmed in the making of these tests. :-)


Just to make your blood boil:

http://www.theguardian.com/technolog...usic-converter

Seems fairly typical for 'subjective' reviews.

Is it really any surprise, given that the Guardian is arsewipe?

I don't think it would be fair to single out the Grauniad for 'subjective'
audio items. You can find the same sort of things in a wide range of
magazines and newspapers.

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

USB Audio ADCs
 

In article ,
Jim Lesurf wrote:
Is it really any surprise, given that the Guardian is arsewipe?

I don't think it would be fair to single out the Grauniad for
'subjective' audio items.

The Mail readers don't tend to know the meaning of 'fair'.

--
*It doesn't take a genius to spot a goat in a flock of sheep *

Dave Plowman London SW
To e-mail, change noise into sound.

USB Audio ADCs
 

In article , Dave Plowman (News)
wrote:
In article , Jim Lesurf
wrote:
Is it really any surprise, given that the Guardian is arsewipe?

I don't think it would be fair to single out the Grauniad for
'subjective' audio items.

The Mail readers don't tend to know the meaning of 'fair'.

It has swings, roundabouts, etc.

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

USB Audio ADCs
 

On 11/03/2014 22:36, Huge wrote:
On 2014-03-11, Jim Lesurf wrote:
In article , RJH
wrote:
On 10/03/2014 16:49, Jim Lesurf wrote:
In article , Eiron
wrote:
On 10/03/2014 15:15, alexd wrote:


No gold was harmed in the making of these tests. :-)


Just to make your blood boil:

http://www.theguardian.com/technolog...usic-converter

Seems fairly typical for 'subjective' reviews.

Is it really any surprise, given that the Guardian is arsewipe?


I don't think many newspapers give especially good technical reviews.

--
Cheers, Rob

USB Audio ADCs
 

On 11/03/2014 10:40, Jim Lesurf wrote:
In article , RJH
wrote:
On 10/03/2014 16:49, Jim Lesurf wrote:
In article , Eiron
wrote:
On 10/03/2014 15:15, alexd wrote:


No gold was harmed in the making of these tests. :-)


Just to make your blood boil:

http://www.theguardian.com/technolog...usic-converter

Seems fairly typical for 'subjective' reviews.

or https://tinyurl.com/q8ju7p9

Incidentally, can you *hear* any difference between all these DACs?

Well, in this thread I've been talking about ADCs rather than DACs.
However...


Ah, yes! I would only use them mainly as DACs, but take your point.

FWIW I don't personally write, or take very seriously, a lot of what
appears as 'subjective comments' in audio reviews. I can't comment as yet
on any relative audible differences between the ADCs I've been testing. So
far, I'd only prefer one to the others because it clearly has better
behaviour in terms of distortion, noise floor, etc. i.e. ye olde basic
injuneering.

Have I sometimes heard differences between various DACs, CD players, etc?
Yes, I think so. Generally slight to negligable when I do. Do I think that
should sway anyone else? No. Not unless I can give a concrete reason.

On that I can say things like:

I preferred the sound of my first CD player (Marantz using the first
Philips chipset) when I added an analogue low pass filter to its output.
(Toko 19kHz low pass filters meant for good FM tuners.) I think that may
have suppressed some garbage around 20kHz and up. May also have altered
the impulse behaviour.

I preferred a Meridian 263 DAC to other DACs and players for many years. No
idea why, but I guess it simply did a better reconstruct and filter job
than others of the period. There are various ways for DACs to seem fine in
some ways but have flaws that cause audible problems.[1] However I didn't
sort out why I preferred it, but it just sounded more pleasing on source
material I liked. [2]

I now prefer the Cambridge Audio DACs, particularly the one in the 851C
player/DAC. Sounds good, and unlike my old 263 and 563 will play higher
rate source material via USB. Also (praise!) has a digital balance control
as well as volume one. Why can't everyone do that!?

So in practice I'd probably still be happy with the decades-old Meridian
263 and 563 if it hadn't been for USB audio, and 'high rez' files, etc.

But I have no idea what I may imagine or what has a real physical basis,
and no idea if anyone else using different ears, room, kit, and sources,
would agree with me. So I avoid giving flowery poetic descriptions of 'why'
one item is 'better' than another. Leave that to people who feel they can
do so, but tend to skip over what they write. :-)

Jim

[1] I've looked at various examples over the years. e.g. the behaviour of
DACs when presented with sample series that require intersample peaks above
0dBFS. Which, alas, happens a lot with modern pop CDs. Yet almost no-one
ever bothers to test for this.

[2] Sometimes the reasons may pass by the standard tests people make. e.g.
I noticed in the ADC measurements that one ADC's noise floor rises quite a
lot when you apply an input signal. So simply measuring the noise level
with no signal doesn't tell you what happens when a signal is applied. I
don't *know* why this happens, only that I can measure it. I can *guess*
various reasons for it happening. But whatever they are, it makes me feel
it is a mark against the ADC. That said, I can of course also say that ye
olde analogue mag taps used to exhibit modulation noise, probably at a
higher level, and we all survived that OK. :-)


Well, you're at least making an effort to figure out which is
technically the most accurate. As to whether that means anything in
practice, I'd make the following points:

When I used to follow this NG I think maybe 15 years back, the consensus
was that any difference in digital-analogue conversion was inaudible.
Differences could only be explained in the analogue amplifier stages.
These were unlikely except in very cheap or poorly designed kit. Also,
differences because of volume anomalies in the A-B testing.

FWIW (and I would have to accept that's not a lot!) I've never been that
happy with the notion that DAC is a done deal. Simply too many unknown
variables, not least those associated with hearing and the perception of
sound. The availability of higher bit rate audio files might perhaps
suggest there's still work to be done on ADC and DAC. Although I do
roughly appreciate the advantages of high resolution ADC as a more
robust method of recording music.

For myself, I can't reliably differentiate between devices. I have a
couple of DACs, and computer-based ADCs. I might think sometimes that
one or other is variously bright or smooth, or somehow 3 dimensional. I
suspect any difference I might claim to detect could be explained more
readily through environmental variables (say temperature) than the
conversion process.

Having said this, I'd rather use hardware that does as good a measured
job as possible. It would just get on my nerves if I knew I was
listening to music that wasn't 'translated' to a technically high
standard. Even if I couldn't reliably detect the contribution. I would
add that's only on my hifi - I'm not prissy at all in the car or via a
portable, where I do the bulk of my listening.

--
Cheers, Rob

USB Audio ADCs
 

In article , RJH
wrote:
On 11/03/2014 10:40, Jim Lesurf wrote:
In article , RJH
wrote:
[big snip]



When I used to follow this NG I think maybe 15 years back, the consensus
was that any difference in digital-analogue conversion was inaudible.
Differences could only be explained in the analogue amplifier stages.
These were unlikely except in very cheap or poorly designed kit. Also,
differences because of volume anomalies in the A-B testing.

In general I'm inclined to agree, but I'd say "usually negligable" rather
than "inaudible" for the sake of caution. In part because I can measure
defects in some cases, and in part because I'm quite happy to accept that
there will be situations where someone else can hear a real difference
which I didn't hear or notice. So "inaudible" for one person may not ensure
it is "inaudible" for everyone else.

For me one of the areas of concern is the tendency for DACs (and ADCs) that
oversample or use digital filters, etc, to have clipping (overflow) or
underflow problems. The most gross type being a failure to handle
intersample peaks about 0dBFS when a *lot* of popular music will generate
these.

Hence it is easy for two DACs to measure 'fine' on standard tests but
behave totally differently when you play commercial popular discs.

FWIW I'm also very wary of both SACD and HDCD for different reasons. Both
have their own areas which can cause problems that standard measurements
don't show.

[snip]

Having said this, I'd rather use hardware that does as good a measured
job as possible. It would just get on my nerves if I knew I was
listening to music that wasn't 'translated' to a technically high
standard.

I am similar. I can't be certain that optimum measured performance will
guarantee superb sound. But I can suspect that poor measured performance
may degrade or colour the sound. Since my main interest is in 'acoustic'
material (jazz or 'classical') want the sound to be like the real
instruments in the real venue. Not coloured or 'sexed up'. But this is a
matter of what the individual user wants. Hence another reason for my
avoiding telling others what something 'sound like'. Technical info is
useful, but in the end people choose what they like for their reasons, even
if they aren't the same as mine.

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

USB Audio ADCs
 

On 2014-03-20, Jim Lesurf wrote:
In article , RJH
wrote:
On 11/03/2014 10:40, Jim Lesurf wrote:
In article , RJH
wrote:
[big snip]



When I used to follow this NG I think maybe 15 years back, the consensus
was that any difference in digital-analogue conversion was inaudible.
Differences could only be explained in the analogue amplifier stages.
These were unlikely except in very cheap or poorly designed kit. Also,
differences because of volume anomalies in the A-B testing.

It depeds on the converter. Some are good, and the result of the D-A/A-D
is inaudible, some are pretty terrible. I designed a testing scheme for
sound cards a number of years ago, and got a range of results. Often the
onboard (motherboard included) sound cards are pretty poor. But some
even cheap sound cards can be pretty good.
(http://www.theory.physics.ubc.ca/sou...soundcard.html)


In general I'm inclined to agree, but I'd say "usually negligable" rather
than "inaudible" for the sake of caution. In part because I can measure
defects in some cases, and in part because I'm quite happy to accept that
there will be situations where someone else can hear a real difference
which I didn't hear or notice. So "inaudible" for one person may not ensure
it is "inaudible" for everyone else.

Unfortunately "others can hear" can be a real swamp. It has been shown
that people's hearing is hugely sensitive to expectation, and other
psychological effects. People hear differences where there are none.
Of course another question is what tests correlate best with difference
which are audible.



For me one of the areas of concern is the tendency for DACs (and ADCs) that
oversample or use digital filters, etc, to have clipping (overflow) or
underflow problems. The most gross type being a failure to handle
intersample peaks about 0dBFS when a *lot* of popular music will generate
these.

Hence it is easy for two DACs to measure 'fine' on standard tests but
behave totally differently when you play commercial popular discs.

FWIW I'm also very wary of both SACD and HDCD for different reasons. Both
have their own areas which can cause problems that standard measurements
don't show.

[snip]

Having said this, I'd rather use hardware that does as good a measured
job as possible. It would just get on my nerves if I knew I was
listening to music that wasn't 'translated' to a technically high
standard.

I am similar. I can't be certain that optimum measured performance will
guarantee superb sound. But I can suspect that poor measured performance
may degrade or colour the sound. Since my main interest is in 'acoustic'
material (jazz or 'classical') want the sound to be like the real
instruments in the real venue. Not coloured or 'sexed up'. But this is a

Unfortunately that is impossible, since at the least the rooms are
completely different. The resonances in your room and in the jazz venue
differ hugely and even in the venue, if you move a foot, the sound is
completely different.


matter of what the individual user wants. Hence another reason for my
avoiding telling others what something 'sound like'. Technical info is
useful, but in the end people choose what they like for their reasons, even
if they aren't the same as mine.

Jim

USB Audio ADCs
 

In article , William Unruh
wrote:
On 2014-03-20, Jim Lesurf wrote:
In article , RJH
wrote:



In general I'm inclined to agree, but I'd say "usually negligable"
rather than "inaudible" for the sake of caution. In part because I can
measure defects in some cases, and in part because I'm quite happy to
accept that there will be situations where someone else can hear a
real difference which I didn't hear or notice. So "inaudible" for one
person may not ensure it is "inaudible" for everyone else.

Unfortunately "others can hear" can be a real swamp. It has been shown
that people's hearing is hugely sensitive to expectation, and other
psychological effects. People hear differences where there are none. Of
course another question is what tests correlate best with difference
which are audible.

FWIW I agree. But also accept there will be times when I (or someone else)
can't hear something that others can genuinely hear. And some effects can
be subtle or not be what someone is 'listening for'. So this isn't an easy
thing to decide at times.



I am similar. I can't be certain that optimum measured performance
will guarantee superb sound. But I can suspect that poor measured
performance may degrade or colour the sound. Since my main interest is
in 'acoustic' material (jazz or 'classical') want the sound to be like
the real instruments in the real venue. Not coloured or 'sexed up'.
But this is a

Unfortunately that is impossible, since at the least the rooms are
completely different.

Yet in practice you can set up a system well and your hearing can tend to
'tune out' the effect of the home room simply because you are so
habituated to it. Can take a lot of work and good speakers, but does
happen. Stereo is an illusion after all.

In my case after many weeks of experiment, shifting speaker, furnishings,
adding wall rugs, etc, using Quad ESLs I can get a pretty convincing
imitation of sitting in a venue like the RFH or RAH or ye olde Maida Vale
that I used to know reasonably well. One of the advantages of human
perception doing subconscious processing on what we 'hear'.

An advantage of the ESL63 and its descendants is the directional behavior.
That - with some wall treatment of the far and rear walls - does help a
lot.

It won't be exactly the same, but can be convincing enough to let you
forget the room you are in and hear a decent result. [1]

The resonances in your room and in the jazz venue
differ hugely and even in the venue, if you move a foot, the sound is
completely different.

Again agreed. I usually have to adjust my ear locations to about a cm in
space to get the required results with good source material. e.g. for Proms
via R3.

Jim

[1] This is why I'd be interested in the BBCs current surround stream
tests. The best results I've ever heard was listening to 'quad' in ye olde
days at Kingswood Warren. But how close you can get at home is another
matter, even having fiddled with the room acoustic, etc.

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

USB Audio ADCs
 

Hi,

I've now put up a webpage at

http://www.audiomisc.co.uk/HFN/ADC/USBrecording.html

showing the results I got from three different external USB ADC/DACs. I did
the tests with Linux and RO and got much the same behaviour. Only added
variable is some sensitivity to the choice of PSU for the 5Vdc line of the
devices.

If someone knows of a 192k/24 device I can try that complies with the open
USB audio class standards, please let me know. Alas, All the ones I tried
to find out about either gave no info or I was told 'not compliant'.

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

USB Audio ADCs
 

On 03/26/2014 02:52 PM, Jim Lesurf wrote:
Hi,

I've now put up a webpage at

http://www.audiomisc.co.uk/HFN/ADC/USBrecording.html

showing the results I got from three different external USB ADC/DACs. I did
the tests with Linux and RO and got much the same behaviour. Only added
variable is some sensitivity to the choice of PSU for the 5Vdc line of the
devices.

If someone knows of a 192k/24 device I can try that complies with the open
USB audio class standards, please let me know. Alas, All the ones I tried
to find out about either gave no info or I was told 'not compliant'.

Is this the sort of thing you're looking for?

http://www.whathifi.com/blog/m2tech-hiface-dac-hands-on-review

No driver needed for Mac and Linux, although you might need a fairly
recent kernel.

--
╔�╦�╦�����╦���╗
â•‘ â•‘ â•‘ â•‘ â•‘
╔�� ║ ║ ║ ║ ║ ╔��
╚���╩�╩�╩�╩�╩�� -- JimP.