abbey road d enfer

Soundcard for measurements
« on: February 27, 2019, 02:25:04 PM »
I've considered a number of alternatives, finally I consider AudioTester to be the best software for me, followed closely by REW.
The only nag is that AudioTester is not downloadable anymore, but I have it installed on 3 computers, but it may mean that it may not be maintained.
The only remaining issue is finding an adequate soundcard.
The one that looked the most promising, the ICON Cube Pro (192kHz, no knobs) turned out to be a pig in terms of frequency response with about 10dB variations in the passband!
The Behringer UMC202HD does not show anything above 25kHz at 96k SR - it is supposed to operate at 192 but it just doesn't.
So, does anyone can suggest a soundcard that would operate significantly up to 70-80kHz? Calibrated level would be a plus.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.


Gold

Re: Soundcard for measurements
« Reply #1 on: February 27, 2019, 03:57:12 PM »
I'd go with a Lynx E22 or E44. In my early days I used a Lynx II (no longer made). It had very good converters for the time, especially for a card.

john12ax7

Re: Soundcard for measurements
« Reply #2 on: February 27, 2019, 04:19:34 PM »
Lynx tends to make well engineered products. I use a Lynx Aurora  for measurements,  it's a pricier solution,  but had the lowest distortion  of a bunch I tested years ago.

I would also expect the E22 and E44 to be better than most.

At the expensive end would be the hilo at 2k+.

abbey road d enfer

Re: Soundcard for measurements
« Reply #3 on: February 27, 2019, 05:11:21 PM »
Lynx tends to make well engineered products. I use a Lynx Aurora  for measurements,  it's a pricier solution,  but had the lowest distortion  of a bunch I tested years ago.

I would also expect the E22 and E44 to be better than most.

At the expensive end would be the hilo at 2k+.
I forgot to mention I want to use it with a laptop that has no PCI slot, so I need USB.
The Hilo may be a possible answer, but at that price I'd rather buy a Lindos.
Actually I notice Lynx spec the frequency response only for Single and Double Speed, not QS.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

Gold

Re: Soundcard for measurements
« Reply #4 on: February 27, 2019, 06:31:28 PM »
I forgot to mention I want to use it with a laptop that has no PCI slot, so I need USB.

That's a "USB interface". The best quality stereo USB interface I know of is the Sound Devices USBpre II. They come up used on the bay if you don't want buy new. You want the II version. The original one is not USB 2.0 class compliant.


cyrano

Re: Soundcard for measurements
« Reply #5 on: February 27, 2019, 06:45:59 PM »
...but at that price I'd rather buy a Lindos.

Thanks for mentioning Lindos. I've never heard about this brand.

Checking out their site. Really interesting. Especially the measurement result database. I'll be reading for some hours now :D
Why is it people love to believe and hate to know?

abbey road d enfer

Re: Soundcard for measurements
« Reply #6 on: February 27, 2019, 11:38:43 PM »
That's a "USB interface". The best quality stereo USB interface I know of is the Sound Devices USBpre II. They come up used on the bay if you don't want buy new. You want the II version. The original one is not USB 2.0 class compliant.
That's an interesting tip, thanks. I'll check the Ebay.
I see the frequency response at 192kSR is specified as [email protected] It's a tad less than expected but I think I'll have to make myself content with it, since apparently there is not much choice in this area. That means I will keep my trusty old analog set when I want to study ultrasonic artefacts.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

trobbins

Re: Soundcard for measurements
« Reply #7 on: February 28, 2019, 12:19:23 AM »
If you can cope with 2nd hand/used, then I strongly recommend an EMU 0404 USB or variant, if you can track something down.  The latest beta firmware is fine with Win10, and ASIO drivers allow REW to use it up to 192kHz.  It can be battery powered if needed, and is USB2 to allow a simple USB isolator for more ground loop noise isolation.  It comes with 'knobs', a practical output headphone/line out for driving, and a 1 Meg unbalanced input (which I use for convenient scope probe connection to give 10:1 and 100:1 compatibility for high impedance valve amp work) or balanced.  It has been pulled apart by a few, so is modifiable if needed, and has been performance benchmarked.

abbey road d enfer

Re: Soundcard for measurements
« Reply #8 on: February 28, 2019, 01:00:51 AM »
If you can cope with 2nd hand/used, then I strongly recommend an EMU 0404 USB or variant, if you can track something down.  The latest beta firmware is fine with Win10, and ASIO drivers allow REW to use it up to 192kHz.  It can be battery powered if needed, and is USB2 to allow a simple USB isolator for more ground loop noise isolation.  It comes with 'knobs', a practical output headphone/line out for driving, and a 1 Meg unbalanced input (which I use for convenient scope probe connection to give 10:1 and 100:1 compatibility for high impedance valve amp work) or balanced.  It has been pulled apart by a few, so is modifiable if needed, and has been performance benchmarked.
Thanks for the tip. The 404 USB is quite rare actually. The review in Sound On Sound indicates that the frequency response is not consistent with the increase of SR. At 96kHzSR, it's only 37kHz, so I doubt it excesses 50kHz at 192k.
The problem with second-hand gear is you can't return it if it does not live up to expectations.
I'm more and more inclined to accept that I won't be able to do ultrasonic measurements with a soundcard-based rig.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

ruffrecords

Re: Soundcard for measurements
« Reply #9 on: February 28, 2019, 03:55:21 AM »
The Behringer UMC202HD does not show anything above 25kHz at 96k SR - it is supposed to operate at 192 but it just doesn't.
So, does anyone can suggest a soundcard that would operate significantly up to 70-80kHz? Calibrated level would be a plus.

I am sure you know this but your laptop needs a class 2 interface to support bit rates above 48K at 24 bits

Cheers

Ian
www.customtubeconsoles.com
https://mark3vtm.blogspot.co.uk/
www.eztubemixer.blogspot.co.uk


'The only people not making mistakes are the people doing nothing'


ruffrecords

Re: Soundcard for measurements
« Reply #10 on: February 28, 2019, 03:59:13 AM »
Thanks for mentioning Lindos. I've never heard about this brand.

Checking out their site. Really interesting. Especially the measurement result database. I'll be reading for some hours now :D

I have Lindos. It's sweep range is up to 45KHz and you can make spot measurements down to 10Hz (very handy for transformer distortion measurements). Its only downside is it only measures distortion at 1KHz.

Cheers

Ian
www.customtubeconsoles.com
https://mark3vtm.blogspot.co.uk/
www.eztubemixer.blogspot.co.uk


'The only people not making mistakes are the people doing nothing'

abbey road d enfer

Re: Soundcard for measurements
« Reply #11 on: February 28, 2019, 04:35:16 AM »
I am sure you know this but your laptop needs a class 2 interface to support bit rates above 48K at 24 bits

Cheers

Ian
Actually I was not aware of this requirement. Thank you for mentioning.

Wikipedia is not much help since they ignore "class 2". I suspect it's the same as ADC2.

My laptop is a brand new Acer Aspire 5 with W10. The datasheet does not mention any class regarding the USB ports; it just indicates:
Nombre de ports USB 2.0 Two
Nombre de ports USB 3.0 One
Nombre de ports USB 3.1 Gen 1 One

I have not noticed any difference between connection to a 2.0 or a 3.0 port.

BTW, the Behringer does not work at 192k with my desktop (also recent, W10) either.
To be more accurate, it works at 96 and 192, but the response brickwalls at 25kHz.
« Last Edit: February 28, 2019, 04:51:41 AM by abbey road d enfer »
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

abbey road d enfer

Re: Soundcard for measurements New
« Reply #12 on: February 28, 2019, 10:28:48 AM »
Still investigating.
I loaded a more recent driver for the UMC202HD, and meaurement goes up to about 90kHz!
BUT, there are two notches  at 25k and 75kHz!
EDIT: In fact they are not real notches, what happens is that aliasing happens at 24k (Fs/2) and again at 72k (3/2 Fs). Apparently it does not happen with REW... So again, even with the new driver, the card is still operating at 48k. Seems like the software is not capable of properly telling the soundcard what SR it should work at.
« Last Edit: March 01, 2019, 04:19:45 AM by abbey road d enfer »
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

cyrano

Re: Soundcard for measurements
« Reply #13 on: February 28, 2019, 10:29:32 AM »
USB class compliancy is software, as in a driver, not hardware on the computer's side.

USB audio class I worked perfectly, with all OS'es. True plug and play. But it's limited to 8 channels (in and out combined) at 48 kHz, 24 bit, because of the 12 Mbps bandwidth on USB 1.1. There are some rare 96 kHz ones, but with fewer channels.

USB audio class II wasn't included in Windows 10 at first, by Microsoft. The spec is a bit of a mess, with lots of options that might be included, but often are not. That's why it never took off.

USB audio class III is a complete mess, so there are hardly any interfaces out there that use it. And the few that are class III capable don't even announce it, probably to avoid overloaded help-desks.

There are a couple of third party drivers out there too, for USB audio class equipment. Not free and not everyone seems pleased with them.
Why is it people love to believe and hate to know?

Re: Soundcard for measurements
« Reply #14 on: February 28, 2019, 10:59:25 AM »
Hello

Just a point, not needed for all lab measurement I guess,  but a DC coupled converter can be useful (or mod/remove all capacitors in signal path).

Best
Zam

abbey road d enfer

Re: Soundcard for measurements
« Reply #15 on: February 28, 2019, 11:04:07 AM »
Hello

Just a point, not needed for all lab measurement I guess,  but a DC coupled converter can be useful (or mod/remove all capacitors in signal path).

Best
Zam
Most audio converters use Sigma/Delta conversion which, by essence don't pass DC. Removing capacitors in the signal path won't change a thing.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

abbey road d enfer

Re: Soundcard for measurements
« Reply #16 on: February 28, 2019, 11:08:00 AM »
More tests.
RMAA seem not capable of telling the soundcard to go above 24kHz.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

abbey road d enfer

Re: Soundcard for measurements
« Reply #17 on: February 28, 2019, 11:09:45 AM »
But REW seems to do better, although the -3dB is a tad disappointing at 60kHz.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

ruffrecords

Re: Soundcard for measurements
« Reply #18 on: February 28, 2019, 12:29:49 PM »
I looked at the SOund Device web site where it says:

"As a class-compliant audio device, the USBPre 2 is limited to a maximum data rate of 24-bit, 48 kHz in Windows, Linux, and Mac OS versions 10.4 to 10.5.7. Mac OS 10.5.8 and higher support the audio class 2.0 device and can address the USBPre 2 up to 24-bit, 192 kHz.

To enable higher sampling rates in Windows OS, download the available Windows ASIO driver."

https://www.sounddevices.com/support/downloads/usbpre-2-asio-driver-download

Hope this helps

Cheers

Ian
www.customtubeconsoles.com
https://mark3vtm.blogspot.co.uk/
www.eztubemixer.blogspot.co.uk


'The only people not making mistakes are the people doing nothing'

abbey road d enfer

Re: Soundcard for measurements
« Reply #19 on: February 28, 2019, 12:39:39 PM »
I looked at the SOund Device web site where it says:

"As a class-compliant audio device, the USBPre 2 is limited to a maximum data rate of 24-bit, 48 kHz in Windows, Linux, and Mac OS versions 10.4 to 10.5.7. Mac OS 10.5.8 and higher support the audio class 2.0 device and can address the USBPre 2 up to 24-bit, 192 kHz.

To enable higher sampling rates in Windows OS, download the available Windows ASIO driver."

https://www.sounddevices.com/support/downloads/usbpre-2-asio-driver-download

Hope this helps

Cheers

Ian
Yes, I read the same; that's why I made the search for a better driver. Curiously, it makes no mention of being class 2, just that it allows higher SR.
Even with standards, these things are complicated enough; I didn't write complex, because it is not, however, people talking about the same things but never using the right words makes things difficult.
« Last Edit: February 28, 2019, 12:44:20 PM by abbey road d enfer »
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.


 

Related Topics

  Subject / Started by Replies Last post
13 Replies
3221 Views
Last post September 07, 2004, 01:53:34 PM
by dmlandrum
8 Replies
2946 Views
Last post September 27, 2004, 04:36:56 AM
by Rochey
6 Replies
2034 Views
Last post March 07, 2005, 04:31:46 AM
by Mbira
19 Replies
4649 Views
Last post April 11, 2005, 03:39:45 AM
by dukasound