Thanks guys.
-
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
- Thread starter pkane
- Start date
OP
- Thread Starter
- #4,282
Hi Folks,
After a long hiatus, I've been able to dedicate a few quality days to testing and enhancing the previous version of Multitone (posted, but not released). The new version, 1.2.0, is being shared here for testing. Please try it out and report any issues or suggestions:
https://app.box.com/s/wof2od3idnspw36up1rn2es14bfmpzx6
There are a few changes in this version that are worth mentioning:
1. DC filter and clock drift correction is not going to have any major effect on most measurements, except for frequency/phase response and impulse response plots
2. Cross-correlated averaging requires that the same DUT output feed both inputs (L + R) on the ADC. By cross-correlating both inputs, the effect of the ADC noise on the measurement can be reduced significantly
3. Collapse button (up arrow) works in two stages. First click will collapse the input/output settings at the top, leaving more room for the plot:
A second click on the same button will result in collapsing of the plot area:
A third click will undo both settings, restoring the window, I/O settings and plot area to their original state. Multitone will remember the state of all the detached plots next time you start it.
An example of how the detached plots can be arranged, the position, size, and choice of detached plots is totally up to you:
Once detached, each plot will show a gear icon near top left (right for Spectrogram) to enable changes to settings/properties directly for that plot.
Right click on the detach icon provides the option to detach or reattach all the plots to the main window.
After a long hiatus, I've been able to dedicate a few quality days to testing and enhancing the previous version of Multitone (posted, but not released). The new version, 1.2.0, is being shared here for testing. Please try it out and report any issues or suggestions:
https://app.box.com/s/wof2od3idnspw36up1rn2es14bfmpzx6
There are a few changes in this version that are worth mentioning:
- Floating/detached measurement plots with magnetic borders
- Option to collapse some of the I/O settings at the top of the main window to make that part of the display smaller, as well as, the ability to collapse the plot area of the main window to make it as small as possible when using detached plots
- Cross-correlation averaging option to lower the effect of ADC noise on measurements
- DC filter measurement option will lower the effect of the DC offset in measured data
- Clock drift correction setting to eliminate measured clock drift due to independent clocks in the DAC and ADC
- Some reported bug fixes
1. DC filter and clock drift correction is not going to have any major effect on most measurements, except for frequency/phase response and impulse response plots
2. Cross-correlated averaging requires that the same DUT output feed both inputs (L + R) on the ADC. By cross-correlating both inputs, the effect of the ADC noise on the measurement can be reduced significantly
3. Collapse button (up arrow) works in two stages. First click will collapse the input/output settings at the top, leaving more room for the plot:
A second click on the same button will result in collapsing of the plot area:
A third click will undo both settings, restoring the window, I/O settings and plot area to their original state. Multitone will remember the state of all the detached plots next time you start it.
An example of how the detached plots can be arranged, the position, size, and choice of detached plots is totally up to you:
Once detached, each plot will show a gear icon near top left (right for Spectrogram) to enable changes to settings/properties directly for that plot.
Right click on the detach icon provides the option to detach or reattach all the plots to the main window.
Rantapossu
Addicted to Fun and Learning
- Joined
- Jul 21, 2022
- Messages
- 625
- Likes
- 448
Thanks!
Unfortunately I ran into a problem when installing (It's the same error with two separate computers):
There's an install log attached below...
OP
- Thread Starter
- #4,284
Can you please re-download and try again? I've rebuilt the installer one more time:
https://app.box.com/s/wof2od3idnspw36up1rn2es14bfmpzx6
Rantapossu
Addicted to Fun and Learning
- Joined
- Jul 21, 2022
- Messages
- 625
- Likes
- 448
OP
- Thread Starter
- #4,286
The automatic package builder is mystery that can't be explained by modern science and works in unpredictable ways
And that those inputs are not merged into Mono but kept separated...
Typical Microsoft
(Sorry Amir)
OP
- Thread Starter
- #4,291
Multitone displays cross-correlated result vs simple average. That's what you see as two graphs overlayed on top of each other.
That's a good one !
I added my comment because people using E1DA ADC have to know this only works in Stereo mode.
By the way, I updated the thread
"Using cross-corelation to lower influence of ADC for DAC measurements"
"Using cross-corelation to lower influence of ADC for DAC measurements"
Rantapossu
Addicted to Fun and Learning
- Joined
- Jul 21, 2022
- Messages
- 625
- Likes
- 448
Is this intentional?
During the measurement with the cross-correlation (The last second just before the measurement stops, still running it) I get these (Overly positive?) values:
And second later when the final calculation is over and the end result appears, I get these (Worse) values:
The noise gets 9.1 dB and TD+N 5.2 dB worse.
It's still better than without cross-correlation however:
Should there be this kind of drop when calculating the end result of cross-correlated measurement or is it just too positive during the measurement?
During the measurement with the cross-correlation (The last second just before the measurement stops, still running it) I get these (Overly positive?) values:
And second later when the final calculation is over and the end result appears, I get these (Worse) values:
The noise gets 9.1 dB and TD+N 5.2 dB worse.
It's still better than without cross-correlation however:
Should there be this kind of drop when calculating the end result of cross-correlated measurement or is it just too positive during the measurement?
What are you measuring, by the way ?
FrenchFan
Active Member
- Joined
- Mar 12, 2024
- Messages
- 162
- Likes
- 72
Great
Another candidate for cross correlation.
I would have liked to use your software, do you know if
it runs under Wine/Linux, I don't have Windows.
Does your software contain THD vs Frequency; THD vs Level
and IMD vs level???,
I develop my own software under octave it could be
interesting to exchange in case of blockage.
Another candidate for cross correlation.
I would have liked to use your software, do you know if
it runs under Wine/Linux, I don't have Windows.
Does your software contain THD vs Frequency; THD vs Level
and IMD vs level???,
I develop my own software under octave it could be
interesting to exchange in case of blockage.
Rantapossu
Addicted to Fun and Learning
- Joined
- Jul 21, 2022
- Messages
- 625
- Likes
- 448
1 kHz sine with Topping E50 and Cosmos ADC grade 0 (THD+)
OP
- Thread Starter
- #4,298
The final result is the valid one, the preview mode during recording may be taking some shortcuts to get the result quicker -- I'll have to double-check.
OP
- Thread Starter
- #4,299
I was able to launch it under Wine/Linux, but the interface to audio devices didn't work for me. I did use virtual machine to run it, so maybe it was related to that.
Yes
Sokel
Master Contributor
- Joined
- Sep 8, 2021
- Messages
- 8,865
- Likes
- 9,428
That was already happening in the previous versions,during the measurement what was shown was usually 2-3dB better noise.
At the end the results was normal as always.
Ok,new one now.
After collapsing and restoring the plot measurement shows up without the notch filter highlighting.Measurement is ok though,is just cosmetic.
