Script for intermodulation distortion testing of vinyl playback, setting VTA ,Zenith etc by test record

[Balle Clorin]

If you have a intermodulation test record with vertical modulation it can be used to set VTA for minimum distortion, or more..
The problem when doing this in REW is that wow and flutter cause the calculation to miss the peaks and calculation becomes incorrect..
This script has configurable frequencies to fit several records and are tolerant to some w&F...
Does calculate main IMD components, but not higher orders above 2nd and 3rd

Try it on your CBS STR-112 for example

PS : script is made using AI, after 26 revisions I decided to end with this version

Record is LB202 from East-Germany

A digital test track comparing script and REW

Spoiler: Another record

CBS STR-112 same cartridge &TT as above

 

[Palmer]

This Forum is astounding, at times - for the contributions it brings to the discerning audio guys. I have to get my REW going - I spend too much time fixing things, and not enough time following up with measurements. Always short on time...

 

[JP]

And my list of things to play with over the winter continues to grow...

 

[Thomas_A]

Have not checked but the IM test of the Hungaroton is? Ottherwise only the LB, STR-112 and DIN records?

 

[Balle Clorin]

If script misses identifying peaks , play with the "search bandwidth" it used to be 10, but 25 helped to find the true FFTpeak better, depens on your TT W&F..
Updated file in post 1

 

[Balle Clorin]

Have not checked but the IM test of the Hungaroton is? Ottherwise only the LB, STR-112 and DIN records?

not tried Hungaroton yet, but should work with any record since the two tone frequencies are by user input 300/3000 or 60/7k 315/3150 and so on

 

[Thomas_A]

not tried Hungaroton yet, but should work with any record since the two tone frequencies are by user input 300/3000 or 60/7k 315/3150 and so on

Is the Hungaroton vertically cut? I have my warped copy and will try unwarp it with recordPi when it arrives in about a week.

 

[Balle Clorin]

Look at the scope or phase plot

 

[Thomas_A]

Look at the scope or phase plot

Will do if the record can be unwarped.

 

[morillon]

If you have a intermodulation test record with vertical modulation it can be used to set VTA for minimum distortion, or more..
The problem when doing this in REW is that wow and flutter cause the calculation to miss the peaks and calculation becomes incorrect..
This script has configurable frequencies to fit several records and are tolerant to some w&F...
Does calculate main IMD components, but not higher orders above 2nd and 3rd

Try it on your CBS STR-112 for example

PS : script is made using AI, after 26 revisions I decided to end with this version

Record is LB202 from East-Germany
View attachment 481214

A digital test track comparing script and REW
View attachment 481215

Spoiler: Another record

View attachment 481227

View attachment 481228
CBS STR-112 same cartridge &TT as above
View attachment 481230

This needs to be checked....but Virtins is, from memory, very flexible and tolerant with IMD dual tone calculations....even without frequency precision.

 

[Balle Clorin]

I do not have 500usd to spend in Virtins.. when I already have Analog Magik

 

[Balle Clorin]

not sure how well Hungaroron till work it is IMD track LEFT only and RIGHT only, and some nasty pops. With only single channel I cannot say if it is vertically modulated or nut

Right channel main signal

LEFT main

plotting seems to struggle, but table data look reasonable

 

[morillon]

I do not have 500usd to spend in Virtins.. when I already have Analog Magik

Even the cheapest version might be enough for us... or else just the basic one in ASIO...
really no need for our uses of the 500 dollar one!
(And maybe just no need for virtins at all between rew, multitone, and your scripts for our uses... even if I've been very attached to Virtins for years)
(Analog Majik is just a simplified virtins " disguise"..can you try it for imd with another lp test biton?)
(the particularity was the creation of the wf module...and suddenly the particularity of having it only in version 500...for obvious commercial reasons in relation to analog majik)

 

[Balle Clorin]

Analog Magik gives results for Ultimate test record IMD stereo track too, not sure how correct it is..

script..

script picks a higher F2...4003 vs 3993
not same recording I use.. AM is online, script by file

Doubling FFT size makes better match with peaks..

 

[Thomas_A]

not sure how well Hungaroron till work it is IMD track LEFT only and RIGHT only, and some nasty pops. With only single channel I cannot say if it is vertically modulated or nut
View attachment 481366


Right channel main signal

View attachment 481367
View attachment 481368
LEFT main
View attachment 481369

plotting seems to struggle, but table data look reasonable

Ok. L and R separately means that both lateral and vertical IMD is measured. 300+3000 seems a better choice (stereo is present in music records). Will pursue this test in my setup later.

 

[calle_jr]

Great work!

script is made using AI, after 26 revisions I decided to end with this version

Sorry for my ignorance, but can you tell us a bit more what this means?
I assume the core is still a fourier transform and standard thd / imd calculation.
What is the AI contribution?
Is it adding or removing the sidebands coming from w&f? smoothing, averaging or what?

 

[Balle Clorin]

@calle_jr I do not know how to program in Python so I asked Microsoft Copilot to write a code according to my specifications. That is the AI contribution.

The program takes any 2 tone recording and analyse for IMD. Two method are used.
1. Same as CBS STR-112 test record , calculate IMD from Amplitude min and max on high pass filtered high frequency signal.this value is shown in the plot. This is inherently robust against wow and flutter.
2. FFT looking at second and third order sideband levels around the high frequency tone. Values given in table.

The purpose is to get IMD values to evaluate VTA and tracing distortion and IMD effects. I found that REW misses the peaks and gave wrong IMD due to wow and flutter and speed drift on the turntables. The script is more immune to this- it looks for the peak values in a frequency window that the user can define.

Nice to see you here Calle, I read and enjoyed your post on a Swedish forum when I had same Cayin amplifier as you ..

 

[calle_jr]

Nice to see you here Calle, I read and enjoyed your post on a Swedish forum when I had same Cayin amplifier as you ..

Wow, that must be some twenty years ago!


Do you have any source information or own statistics about acceptable IMD percentages?
I guess it’s both frequency and level dependant?

 

[morillon]

Wow, that must be some twenty years ago!


Do you have any source information or own statistics about acceptable IMD percentages?
I guess it’s both frequency and level dependant?

more than in absolute value...maybe especially to be considered as an aid to VTa settings, in optimization...same for Python for W&F..or ́this for frequency response for load optimization etc.
;-)
(Too bad. I don't know how to use Python scripts, otherwise, having a 112 or another with bi-tons*... I would have tried...)
(*In the bizarre and brutal... there's the famous Shure 103, which was quickly put in the drawers, "too much"...if some have practice of this one?)

 

[Balle Clorin]

Wow, that must be some twenty years ago!


Do you have any source information or own statistics about acceptable IMD percentages?
I guess it’s both frequency and level dependant?

If you look in old issues of Audio on worldradiohistory.com you find test reports on cbs STR-112,?usually with much lover values that ai get with present cartridges. The values vary wildly between records.

See also here
Post in thread 'Analog Magik'
https://www.whatsbestforum.com/threads/analog-magik.25735/post-1026436