• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many 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

OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
I can't get a decent result with 8 (or 12, 16 and many other) averages. The last average always spoils it:

View attachment 232825

7 averages works all the time:

View attachment 232827

Is the test signal too short?

Here's a test with 96 averages....
1663964326877.png
 

Rantapossu

Addicted to Fun and Learning
Joined
Jul 21, 2022
Messages
507
Likes
355
Here's the waveform (I zoomed the waveform in and found nothing. I also listened the result wav and it was clean):

1663968526291.png


And the result:

1663968575966.png


Changing the ASIO buffers doesn't help.


Here's the 8 averages with the 1 kHz sine, no problems at all:

1663968742124.png



Your 96 averages example is too much to verify the problem, here I plugged the RCA off at the 94 averages of the 96 while still running the measurement and the last two averages were calculated completely without the signal and with the floating RCAs and the result it's still decent:

1663968207360.png
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
Here's the waveform (I zoomed the waveform in and found nothing. I also listened the result wav and it was clean):

View attachment 232895

And the result:

View attachment 232896

Changing the ASIO buffers doesn't help.


Here's the 8 averages with the 1 kHz sine, no problems at all:

View attachment 232897


Your 96 averages example is too much to verify the problem, here I plugged the RCA off at the 94 averages of the 96 while still running the measurement and the last two averages were calculated completely without the signal and with the floating RCAs and the result it's still decent:

View attachment 232891
Don’t know how to explain this, as I don’t seem to have a problem here. Looking back at @Sokel ’s tests, he used 8 averages also with no obvious errors. If you can, please export and share the history record with the error and I can see what I can find.
 

Rantapossu

Addicted to Fun and Learning
Joined
Jul 21, 2022
Messages
507
Likes
355
Don’t know how to explain this, as I don’t seem to have a problem here. Looking back at @Sokel ’s tests, he used 8 averages also with no obvious errors. If you can, please export and share the history record with the error and I can see what I can find.

Thanks, I will do that.

It takes some time, I'll have to setup some kind of cloud account for that.
 

Rantapossu

Addicted to Fun and Learning
Joined
Jul 21, 2022
Messages
507
Likes
355
Try wetransfer.com. You can upload up to 2GB without setting up an account. Just copy the link after uploading and post it here.

Ok, let's hope this works:


There's one working 7 average and one not-working 8 average measurement.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
Ok, let's hope this works:


There's one working 7 average and one not-working 8 average measurement.
Got it, thank you. The 8 average shows the problem, so I can see what you see. Now I need to figure out what’s causing it ;)
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
Ok, let's hope this works:


There's one working 7 average and one not-working 8 average measurement.

Interestingly, I found what's causing the problem when I try to load your file from history, but I don't think this explains why it's not working for you since it only has to do with counting averages when a history item is double-clicked. I'll look at it some more, but now that I have it loading correctly from an mth file, I'm even more puzzled as to why it might not be working for you when running live measurements...
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
Ok, let's hope this works:


There's one working 7 average and one not-working 8 average measurement.

Can you please test this version: https://app.box.com/s/qbrh3czrvudclkqs9n8lm806k3hzmjx5
This has some unrelated changes that are not fully baked yet (always working on the next thing!) but I think it might resolve the issue with the last average in your measurements. Fingers crossed!
 

Rantapossu

Addicted to Fun and Learning
Joined
Jul 21, 2022
Messages
507
Likes
355
Can you please test this version: https://app.box.com/s/qbrh3czrvudclkqs9n8lm806k3hzmjx5
This has some unrelated changes that are not fully baked yet (always working on the next thing!) but I think it might resolve the issue with the last average in your measurements. Fingers crossed!

It freaking works!

1663985066080.png


I tested with the following averages:

1
2
4
8
12
14
16

And all was ok!

There were problems with lower averages (Even with 1) too with the previous version, now the behaviour is completely different.

Do you know what was the exact problem or was this solution the just a shot in the dark?

I just tested the previous version with the virtual "ASIO Bridge and Hi-Fi Cable" and it averaged incorrectly with it too, so it isn't my sound card that was causing the problem. Just to let you know, if you want to replicate the previous result...
 

Rantapossu

Addicted to Fun and Learning
Joined
Jul 21, 2022
Messages
507
Likes
355
This has some unrelated changes that are not fully baked yet

And I found one of the new features very useful. I was going to suggest it to you, but now I don't have to ;)
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
And I found one of the new features very useful. I was going to suggest it to you, but now I don't have to ;)
Now that you found it, may as well test it :) although it's not finished yet (some of the sweep labels are not formatted correctly).
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
It freaking works!

View attachment 232924

I tested with the following averages:

1
2
4
8
12
14
16

And all was ok!

There were problems with lower averages (Even with 1) too with the previous version, now the behaviour is completely different.

Do you know what was the exact problem or was this solution the just a shot in the dark?

I just tested the previous version with the virtual "ASIO Bridge and Hi-Fi Cable" and it averaged incorrectly with it too, so it isn't my sound card that was causing the problem. Just to let you know, if you want to replicate the previous result...

Glad it's working, and I'm glad you reported this! There was a condition at the end of the capture where the number of remaining samples could be computed incorrectly resulting in only a partial FFT with some data missing :oops:. That caused the overall average to degrade right at the end.
 

Rantapossu

Addicted to Fun and Learning
Joined
Jul 21, 2022
Messages
507
Likes
355
Glad it's working, and I'm glad you reported this! There was a condition at the end of the capture where the number of remaining samples could be computed incorrectly resulting in only a partial FFT with some data missing :oops:. That caused the overall average to degrade right at the end.

I might have triggered this condition before too as far as I remember, but the latest test tone made it repeatable. I thought previously that it was the problem with the ASIO buffers, but now I know.

Thanks for fixing it!
 

Sokel

Master Contributor
Joined
Sep 8, 2021
Messages
5,840
Likes
5,775
I see you've been doing some work!
I did occasionally had some problems too with 8 averages but it was random.
Let's see...

images.jpg
 
Top Bottom