• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Beta Test: Multitone Loopback Analyzer software

Grooved

Senior Member
Joined
Feb 26, 2021
Messages
461
Likes
289
Hey @pkane,

Since few days, Multitone stopped working, whatever the version I use (10.0.11 or 10.0.10).
View attachment 174201
This message occurs with either ASIO, WASAPI, DirectSound, whatever for input or output (E1DA Cosmos).

Just tested 1.0.12. Still the same. I've either this with ASIO (Matrix or Topping):

View attachment 174202

Or this with WASAPI:

View attachment 174203

Any thoughts?
Got the same thing yesterday, but I don't remember if it was coming from the "shared mode" or if it was another software opened and having exclusive mode on the same device.
I also got the ASIO devices only, no WASAPI devices appearing, which was fixed after rebooting the computer, so maybe it was from installing the last version without uninstalling first the previous one (which I did every time before this last one)
Sorry to not remember the problem but it was working after that and I didn't get it anymore
 
Last edited:
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
Got the same thing yesterday, but I don't remember if it was coming from the "shared mode" or if was another software opened and having exclusive mode on the same device.
Sorry to not remember the problem but it was working after that and I didn't get it anymore

That error is indicating that there are no audio channels available through this driver (channel range is from 0 to 0 :)), while Multitone is asking for a 2 channel input device. The reason, most likely, is that an output device was selected for input, or vice versa (@VintageFlanker -- please check)

index.php
 
Last edited:

VintageFlanker

Major Contributor
Forum Donor
Joined
Sep 20, 2018
Messages
3,045
Likes
9,730
Location
Paris
For ASIO, did you set the output sample rate in settings? It needs to be selected for the rate that's supported by the driver.

For WASAPI, try switching between shared and not shared modes to see if it makes a difference.


-Paul
It was WASAPI shared mode that caused trouble. Fixed ! Many thanks, @pkane !:)
 
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
Quick update to v1.0.13 based on the feedback from @IVX.

Please note that with this version, you'll need to set the system volume control to the desired level (usually 100%) for measurements. Multitone will no longer do this automatically to accommodate devices such as Cosmos ADC that use the volume setting for a different purpose.

Changes in 1.0.13

  • Change: Multitone will no longer automatically set the system volume control to 100%. The user will need to make that change manually, if desired.
 
Last edited:

jan.didden

Member
Joined
Apr 19, 2018
Messages
33
Likes
65
Paul, I may have missed it in all the previous pages, but I have a Q. How does the ADC-DAC combo impact the measured results? Are they the limiting factor or are you somehow measuring/compensating for their distortion?

Jan
 
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
Paul, I may have missed it in all the previous pages, but I have a Q. How does the ADC-DAC combo impact the measured results? Are they the limiting factor or are you somehow measuring/compensating for their distortion?

Jan
Hi Jan,

Ideally, the DAC/ADC is significantly lower distortion than the DUT you're trying to measure. That said, current released version doesn't have this feature yet, but I’ve been playing with a distortion pre-compensation feature. It works (sometimes) to reduce THD. This might help remove some of the harmonic distortions if calibrated with DAC / ADC loop, before inserting the DUT.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,150
Likes
4,713
Location
Berlin, Germany
Paul, yes, I was wondering about that pre-distortion or cancellation. There's a thread over at diyaudio by Pavel Hoffman who has been working on that and has shown interesting results. It's not for the faint at hearth though.
I wondered if you know about that project. I can search for it.

Edit: found it: https://www.diyaudio.com/community/...tion-applications.347854/page-65#post-6796435

Jan
The point is that in many cases any pre-distortion works only for a given level and frequency as the distortion mechanisms are often not a simple static transfer function nonlinearity.
 
Last edited:

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,150
Likes
4,713
Location
Berlin, Germany
This might help remove some of the harmonic distortions if calibrated with DAC / ADC loop, before inserting the DUT.
IME calibrating the DAC-->ADC (un-embedding the distortion) before inserting the DUT only works reliably to a high degree when
- the level is the same
- the frequency response is the same (mag and phase, and absolute polarity)
- the impedances (DAC load and ADC source) are the same, and the hookup (like mixing balanced and unbalanced)
- there are no warm-up drifts etc

It often may give an improvement but you cannot easily know to what extent the normalization still holds with the DUT in place.
 
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
Paul, yes, I was wondering about that pre-distortion or cancellation. There's a thread over at diyaudio by Pavel Hoffman who has been working on that and has shown interesting results. It's not for the faint at hearth though.
I wondered if you know about that project. I can search for it.

Edit: found it: https://www.diyaudio.com/community/...tion-applications.347854/page-65#post-6796435

Jan

I've done something similar, although I just read the first few posts on the DIY thread so far. As @KSTR pointed out, this kind of THD compensation works primarily with static/memory-less nonlinearities. Once frequency and level dependencies start to impact it, you'll need to make sure that the conditions used for testing at both, DAC and ADC are very similar to those that were used to create the compensation. Otherwise, you may just be adding more distortion into the mix instead of reducing it.
 
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
IME calibrating the DAC-->ADC (un-embedding the distortion) before inserting the DUT only works reliably to a high degree when
- the level is the same
- the frequency response is the same (mag and phase, and absolute polarity)
- the impedances (DAC load and ADC source) are the same, and the hookup (like mixing balanced and unbalanced)
- there are no warm-up drifts etc

It often may give an improvement but you cannot easily know to what extent the normalization still holds with the DUT in place.

Correct on all counts, Klaus! In my testing, this works best in a reasonably small frequency and level range around where the compensation was created. It is possible to find compensation that's similar or common to different frequency ranges. For example, I've created compensation at 100Hz, 1kHz and 10kHz, and then tried to find compensation that would reduce THD at all of those frequencies. It's not a simple or easy process, and there's no guarantee that such a compensation exists for every device. But it is somewhat fun to play with - almost like a game :)
 
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
Here is an example of THD compensation feature at work in Multitone. Not sure yet if it's worth releasing, but at least for some devices, it can make a difference. Blue trace is with THD and DC pre-compensation, white -- without. The DUT is just a Mac resampler, so not a physical device and so has no feedback, i.e., only static non-linearity.

THD goes down to -155dB with compensation from -132dB without @1kHz/-1dBFS signal:

1643502898934.png


THD goes down to -155dB with compensation from -132dB without @100Hz/-1dBFS signal:

1643503580213.png
 
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
Now, calibrating ADI-2 Pro FS as a DAC/ADC, i.e., a real physical device that's already of high quality.

Uncalibrated (THD of -115dB):
1643562379485.png


Now, calibrated (THD improved to -128dB from -115!)
1643562326244.png


And the same calibration used with a 100Hz test signal (THD of -126dB):
1643562497473.png


Slightly worse, but still a major improvement at 2kHz (-124dB THD):
1643562717124.png


SMPTE 60Hz/7kHz 4:1 also shows improvement in TD+N and IMD.

Uncalibrated:
1643562887196.png


Calibrated (same calibration as all the tests above). Some obvious differences visible in the distortions:
1643562824464.png
 
Last edited:
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
Now, calibrating ADI-2 Pro FS as a DAC/ADC, i.e., a real physical device that's already of high quality.

Uncalibrated (THD of -115dB):
View attachment 183200

Now, calibrated (THD improved to -128dB from -115!)
View attachment 183199

And the same calibration used with a 100Hz test signal (THD of -126dB):
View attachment 183201

Slightly worse, but still a major improvement at 2kHz (-124dB THD):
View attachment 183203

SMPTE 60Hz/7kHz 4:1 also shows improvement in TD+N and IMD.

Uncalibrated:
View attachment 183205

Calibrated (same calibration as all the tests above). Some obvious differences visible in the distortions:
View attachment 183204

And just to confirm that this is not a measurement artifact with Multitone, here's the same compensation measured by REW, THD at 1kHz shows even a slightly better result than Multitone (-128.9dB):

1643563236312.png
 

Grooved

Senior Member
Joined
Feb 26, 2021
Messages
461
Likes
289
And just to confirm that this is not a measurement artifact with Multitone, here's the same compensation measured by REW, THD at 1kHz shows even a slightly better result than Multitone (-128.9dB):

View attachment 183207
Hi @pkane
I just gave a quick look and I don't understand what calibration is it. How do you did that ?

For example, I get the same king of result with the Tone2 Pro/Cosmos ADC combo, but can I calibrate ?
 
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
Hi @pkane
I just gave a quick look and I don't understand what calibration is it. How do you did that ?

For example, I get the same king of result with the Tone2 Pro/Cosmos ADC combo, but can I calibrate ?
Hi @Grooved,

The calibration is designed to reduce THD in a DAC/ADC loopback. I'm still testing and working on adding this feature to Multitone. The plan is to make it so that by doing the calibration of your own system, you can reduce THD of the measurement devices (DAC + ADC). Until we test this feature with more devices, I'm not sure this will work for every combination. As you can see from the posts above, the reduction in THD for ADI-2 Pro was over 13dB at 1kHz, so this could help significantly when using ADI-2 for precise measurements.

Right now, I'm trying to simplify the calibration process (automate it) so that the initial corrections can then by fine-tuned manually.

As to how this works, think of it as a way to introduce the exact opposite of the distortion-causing non-linearity present in the DAC+ADC system. In effect, when done properly, it should help cancel out the distortions generated by the measurement devices. This is similar (but not exactly the same) as the calibration in the ESS chip of Cosmos ADC, except it's applied in software and so can work with any chip or device.
 

MC_RME

Addicted to Fun and Learning
Technical Expert
Manufacturer
Joined
May 15, 2019
Messages
636
Likes
2,671
This needs more testing as the principle might be flawed in itself. It works by adding harmonics with the right level with opposite phase. That way both ADC and DAC non-linearities are reduced. But its main purpose is to measure DUTs. A DUT in between DAC and ADC will not only add its own distortion, it can also phase shift the harmonics so the cancellation is no longer working, might even make the result worse.

Using such a method without the ability to verify the result is kind of lottery, IMHO. And if you can verify the result you don't need this method...

I still find it interesting. For example the ESS (aka Cosmos ADC) has a spray of harmonics and high frequency spurs that do not vanish at lower levels (with AKM the noise floor is clean at -20 dBFS, with ESS you need -70 dBFS. This also limits the use of simple notch filters for THD measurements with ESS). Hope is that Multitone would be able to remove the spurs from the FFT. And then using a DUT we could see if they come back...
 
OP
pkane

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
3,492
Likes
6,029
Location
North-East
This needs more testing as the principle might be flawed in itself. It works by adding harmonics with the right level with opposite phase. That way both ADC and DAC non-linearities are reduced. But its main purpose is to measure DUTs. A DUT in between DAC and ADC will not only add its own distortion, it can also phase shift the harmonics so the cancellation is no longer working, might even make the result worse.

Using such a method without the ability to verify the result is kind of lottery, IMHO. And if you can verify the result you don't need this method...

I still find it interesting. For example the ESS (aka Cosmos ADC) has a spray of harmonics and high frequency spurs that do not vanish at lower levels (with AKM the noise floor is clean at -20 dBFS, with ESS you need -70 dBFS. This also limits the use of simple notch filters for THD measurements with ESS). Hope is that Multitone would be able to remove the spurs from the FFT. And then using a DUT we could see if they come back...

To be clear, my method doesn't rely on adding harmonics. It introduces a new non-linearity that is the inverse of the one that is determined by the calibration method. Convolving such an inverse transfer function with non-linearities from the DAC + DUT + ADC shouldn't mess with phases and harmonics, except in the case of DAC and/or ADC having a non-static nonlinearity, in which case all bets are off. At least in theory ;)
 

MC_RME

Addicted to Fun and Learning
Technical Expert
Manufacturer
Joined
May 15, 2019
Messages
636
Likes
2,671
Thanks for clarifying.
 

G-K

Member
Joined
Jan 31, 2022
Messages
45
Likes
3
This needs more testing as the principle might be flawed in itself. It works by adding harmonics with the right level with opposite phase. That way both ADC and DAC non-linearities are reduced. But its main purpose is to measure DUTs. A DUT in between DAC and ADC will not only add its own distortion, it can also phase shift the harmonics so the cancellation is no longer working, might even make the result worse.

Using such a method without the ability to verify the result is kind of lottery, IMHO. And if you can verify the result you don't need this method...

I still find it interesting. For example the ESS (aka Cosmos ADC) has a spray of harmonics and high frequency spurs that do not vanish at lower levels (with AKM the noise floor is clean at -20 dBFS, with ESS you need -70 dBFS. This also limits the use of simple notch filters for THD measurements with ESS). Hope is that Multitone would be able to remove the spurs from the FFT. And then using a DUT we could see if they come back...
This needs more testing as the principle might be flawed in itself. It works by adding harmonics with the right level with opposite phase. That way both ADC and DAC non-linearities are reduced. But its main purpose is to measure DUTs. A DUT in between DAC and ADC will not only add its own distortion, it can also phase shift the harmonics so the cancellation is no longer working, might even make the result worse.

Using such a method without the ability to verify the result is kind of lottery, IMHO. And if you can verify the result you don't need this method...

I still find it interesting. For example the ESS (aka Cosmos ADC) has a spray of harmonics and high frequency spurs that do not vanish at lower levels (with AKM the noise floor is clean at -20 dBFS, with ESS you need -70 dBFS. This also limits the use of simple notch filters for THD measurements with ESS). Hope is that Multitone would be able to remove the spurs from the FFT. And then using a DUT we could see if they come back...
I claimed that distortions >h10 several times already at the E1DA dicord channel and l never got a clear answer whats the reason for that. For levels <20 dBFS the THD performance of AK5572 is better than COSMOS ADC. You can see my measuring results there
 
Top Bottom