Beta Test: Multitone Loopback Analyzer software

[pkane]

Is the idea of a w&f for vinyl and tape a bit abandoned?
;-)

I actually started the implementation and it's in the current version... except that it isn't quite finished or fully functional I'll get back to it when I have a bit more time.

 

[pkane]

Found nothing broken

(is there any chance we can define our own file for putting the recordings? )

Export the recorded signal to a file, you mean?

 

[morillon]

I actually started the implementation and it's in the current version... except that it isn't quite finished or fully functional I'll get back to it when I have a bit more time.

suuuuper ;-)

 

[Sokel]

Export the recorded signal to a file, you mean?

No,the actual real time measurements which is now under MT/history

(of course I mean folder,not file ,my bad )

 

[pkane]

suuuuper ;-)

Here's a preview

 

[morillon]

thank you!

 

[NathanR]

Hi Paul (pkane),

Been a bit since I last checked in *grins sheepishly* (I bounce heavily between projects as I have energy; recently has been smart-home stuff).

Anyways. It seems you have implemented a number of my ideas (downplay my hubris thinking this, as likely ideas were contributed by others/yourself prior to my own). Sorry I wasn't there to support & encourage you. But they seem to have been useful/important enough to make it into your software regardless. (I am talking of the test plan, presets, voltage calibration).
Woot!

To wit, has anyone compiled a list of tests that approximate Amir/AP suite as a test plan?
I went back 10 or so pages and didn't see any.
I'd rather not reinvent the wheel so-to-say.

I have some random equipment I'm interested in testing.
An audio over fiber solution that's CHEAP
And Amir is slammed/backed up so far that It'll be next year before he *might* get around to testing it. Note: *No fault of his own*

Anyways, great work as always.
Cheers!

 

[pkane]

Hi Paul (pkane),

Been a bit since I last checked in *grins sheepishly* (I bounce heavily between projects as I have energy; recently has been smart-home stuff).

Anyways. It seems you have implemented a number of my ideas (downplay my hubris thinking this, as likely ideas were contributed by others/yourself prior to my own). Sorry I wasn't there to support & encourage you. But they seem to have been useful/important enough to make it into your software regardless. (I am talking of the test plan, presets, voltage calibration).
Woot!

To wit, has anyone compiled a list of tests that approximate Amir/AP suite as a test plan?
I went back 10 or so pages and didn't see any.
I'd rather not reinvent the wheel so-to-say.

I have some random equipment I'm interested in testing.
An audio over fiber solution that's CHEAP
And Amir is slammed/backed up so far that It'll be next year before he *might* get around to testing it. Note: *No fault of his own*

Anyways, great work as always.
Cheers!

Hi Nathan,

Don't minimize your contribution to the testing plans -- while I was thinking about implementing them for a while, your request pushed me to actually get it done

I didn't try to reproduce any of the specific AP tests that Amir runs, although I did put together a DAC test plan that collected various pieces of information that I normally find interesting about these devices. The test plan is attached to the initial post here:

RME ADI-2/4 Pro SE - Auto-measured with Multitone

I don't know if this post is about the extremely nice, new RME ADI 2/4 Pro SE interface or about the new automated measurement capability of the Multitone app :) Most of these measurements were done automatically using a Multitone test plan (I'll share in a separate post). This new feature...

www.audiosciencereview.com

 

[NathanR]

Hi Paul,

The test plan is fantastic! Huge time saver.
This is an incredible add!
I'm still tinkering with it and figuring out how best to use it for my own purposes (amp/electronic testing).

The two tests I am adding are:
1) an initial level/gain step that is super fast
2) log-sweep 'chirp' full frequency/sample rate response test.
These two tests on the front-end help to ensure there are no 'I'm pulling a dumb' moments

A few thoughts/notes.
Note: I am not tasking/questioning you, just thinking aloud

---

Improvements:

1) Global distortion stop parameter. (0.1%, 1%, 10%, custom)

2) Lvl sweep

Around interesting areas (mainly cuttoff) increase resolution.

• 1) 10dB increments [-100 to -20]
• 2) 2dB increments [-20 to -5]
• 3) 0.5dB increments [-5 to +1]
• Or just log instead of linear?...

3) Highest SINAD/THD+N (or maybe SNR?) vs level sets max level for all other tests. [variable gain, based upon measured value in previous test_step]
I guess this is a preference (2Vrms [8dBu] or 4Vrms [14dBu] nominal for measuring SINAD vs finding the optimal DUT gain)
-hmm...needs more thought. Interesting idea though.

Bug(s):
1) seems to be a bit cantankerous when trying to quit a test plan running. I am still working this out to exactly what is going on. Basically the level/freq sweep will not run again.
(don't bother looking at this until I have the specifics & repeatable steps)

2) Lvl Sweep > Settings pane > Chart units > Vertical units (dBFS, V, W, dBr, etc) = changes horizontal units [vertical units stay in dB]

 

[weme]

... we can also create such measurement series:
H2, H3, ... vs housing temperature of the ADCiso @1kHz
H2, H3, ... vs amplitude of the ADCiso @1kHz
...
To avoid having to manually transfer the measured values to a spreadsheet programme, it would be great if you had the time and motivation to create a text file that contains all the values of the measurement series in lines after a header line. This text file could basically be created for THD measurements and similar, so that no additional parameter would be required in the Test Plan.

I mean such measurement series:

View attachment 331011 View attachment 331012

For example, to analyse the effects of temperature and input amplitude of an ES9028PRO (Cosmos ADCiso)- related to H2, H3, ...

One text file (*.csv structure) per Test Plan contains all the data from the single *.html files after a few header lines.

You mean like this?

Beta Test: Multitone Loopback Analyzer software

Could you give these measurements (The RMS values): Like this?

www.audiosciencereview.com

@Sokel
The two exemplary test plan files I have provided can be modified as required using an editor and thus changed to the desired evaluation parameters. It is only time-consuming to manually transfer the desired values to a spreadsheet programme.
You can also view the test plan "MTA_Test-Plan_Measurement-Series_PlayGain(1)_Example.mtp" as a "film" on the monitor or browse through the graphic files after the end of the measurement. A dB reduction of 0.5 dB or even less would be even more interesting to see:
You get what you see: an addition of the signals (by magnitude and phase) from the DUT and the Measuring Device.

Here is a rough evaluation with REW, which needs to be analysed more closely in the range from -17 dBFS to -40 dBFS and from -53 dBFS and after H2 / H3 - when I have more time again.
BTW: The apparent measurement error in the range of -55 dBFS is none (was remeasured) - we have no time axis here.

 

[weme]

v1.0.97 - some remarks:

Same selection colour for driver selection/test selection:
The selection colour (blue) for the driver selection is darker than for the test selection, the name of the selected driver is therefore more difficult to read.

The background colour will be the same, but the different font colour (white, black) makes the difference.

Test Plan > Edit > Save Plots (none) | PNG | ...
Changes the file extension, but does not show the current status.
Those who work this way (do not change a running system) will have no problems, others need a quick test to see what they had changed.

File > Export Measurement > Spectrum as CSV
The separators used (no semicolon or other special character?) to separate the columns prevent the data from being read from the file:
e.g:
Frequency (Hz), Left Amplitude (dB), Left Phase (deg)
0,000,-84,2883987859148,180
1,346,-86,0663567853495,-7,21904558171598E-15
2,692,-91,4939843551453,180
4,037,-100,879247608949,-2,38009540227601E-14
...

 

[Rantapossu]

Hi @pkane

MultiTone plays with all the available output channels when selecting any ASIO 2-channel output combination, just like DeltaWave did before you fixed it on the DW version 2.0.12.

For example selecting the channels 3+4, 9+10 or any similar channel pair for the output makes all 16 output channels of my interface to output the signal simultaneously instead just the two selected channels.

Would it be hard to do the same modification for MultiTone to deactivate all other output channels than the selected ones?

 

[pkane]

Hi @pkane

MultiTone plays with all the available output channels when selecting any ASIO 2-channel output combination, just like DeltaWave did before you fixed it on the DW version 2.0.12.

For example selecting the channels 3+4, 9+10 or any similar channel pair for the output makes all 16 output channels of my interface to output the signal simultaneously instead just the two selected channels.

Would it be hard to do the same modification for MultiTone to deactivate all other output channels than the selected ones?

That’s already done. I actually implemented this in Multitone, first, then moved it over to DeltaWave. I’ll release this version soon.

 

[pkane]

Trying to finish up 1.0.99. @Sokel you may find this one interesting...

 

[pkane]

Multitone v1.0.99 is up for testing/preview: https://app.box.com/s/ue7ll9xmvwogst817x2l1xg09opvgy47

A few changes:

• Added support for Native DSD output on ASIO -- needs testing with different devices and drivers
• Fixed: CSV spectrum file export to use US decimal point notation for locales where comma is used, instead
• Fixed: two channel ASIO output with multi-channel devices playing on all available channels
• A few minor bug fixes and clean up

The None/DOP/Native DSD Mode setting is in the Device section. DSD Base Rate can be 44.1k or 48k, and a DSD Multiplier can be up to 1024x. I've only tested up to 512x48k, so no promises that it'll work at 1024x, and of course, you need a DAC that's capable of that rate.

I don't recommend listening to test tones when using DSD or DOP, at least not at any loud levels -- switching between PCM and DSD can often result in loud clicks, depending on how the USB interface is implemented. You've been warned!

 

[Blumlein 88]

Multitone v1.0.99 is up for testing/preview: https://app.box.com/s/ue7ll9xmvwogst817x2l1xg09opvgy47

A few changes:

• Added support for Native DSD output on ASIO -- needs testing with different devices and drivers
• Fixed: CSV spectrum file export to use US decimal point notation for locales where comma is used, instead
• Fixed: two channel ASIO output with multi-channel devices playing on all available channels
• A few minor bug fixes and clean up

The None/DOP/Native DSD Mode setting is in the Device section. DSD Base Rate can be 44.1k or 48k, and a DSD Multiplier can be up to 1024x. I've only tested up to 512x48k, so no promises that it'll work at 1024x, and of course, you need a DAC that's capable of that rate.

View attachment 333844

I don't recommend listening to test tones when using DSD or DOP, at least not at any loud levels -- switching between PCM and DSD can often result in loud clicks, depending on how the USB interface is implemented. You've been warned!

Sorry I've not been testing your recent releases. But loud pops, etc with DSD, just one more reason not to do DSD. Of course you know my feelings about DSD. While I've not done testing so much, I have used your latest releases and you have it so well developed kudos.

 

[Sokel]

Yesssssssssss!
Thank you!

It will be tested (a lot) and report back.

 

[Sokel]

Oh,DSD looks so much nicer



PCM




DoP



DSD native


DSD512 now doesn't start the counter either DOP or Native.just stays there saying DSD GEN and does nothing.

 

[Sokel]

I lowered FFT size and it plays DSD512 just fine,probably my super machine hasn't enough juice for more



Not so nice and clean as DSD256 but Tone's specs is up to DSD256 officially,so...

 

[pkane]

I lowered FFT size and it plays DSD512 just fine,probably my super machine hasn't enough juice for more

View attachment 333946

Not so nice and clean as DSD256 but Tone's specs is up to DSD256 officially,so...

Yes, DSD generation can take much longer on a computer without floating point acceleration. On my (non-accelerated) I7 laptop, 512x takes a few minutes, while 64x takes only a few seconds. On an I9 desktop, 512x takes only 10-15 seconds to generate.