Phono Cartridge Response Measurement Script

[mfalcon]

Thanks @JP
I just rerecorded it and got it running. I normalized my files, is that ok?
Now clearly I have a problem but I don't know exactly what's going on. Does anybody see what I did wrong or if it's setup? The crosstalk is positive for the right channel.
This cartridge is 10-20 hours. Its my second Kleos.
Input File: Lyra_Kleos_CBS_STR100_L_2024_02_01.wav

Sample Rate: 96,000Hz

Channels: 2

Length: 72.988875s

In/Out (s): [ 0. 72.988875]

Min Freq: 0Hz

Max Freq: 0Hz



X-talk @1kHz: -22.33dB





Input File: Lyra_Kleos_CBS_STR100_R_2024_02_01.wav

Sample Rate: 96,000Hz

Channels: 2

Length: 64.97380208333334s

In/Out (s): [ 0. 64.97380208]

Min Freq: 0Hz

Max Freq: 20,200Hz



X-talk @1kHz: 23.47dB

 

[VinyLuke]

Now clearly I have a problem but I don't know exactly what's going on. Does anybody see what I did wrong or if it's setup? The crosstalk is positive for the right channel.

Reverse the channels in Right channel rip. That the sweep is on top (left channel).

 

[JP]

I normalized my files, is that ok?

Yep.

 

[mfalcon]

I feel foolish. Here is the fixed one. I need to optimize some more before posting it in the Library thread. @JP Thanks for your help.

 

[VinyLuke]

I feel foolish. Here is the fixed one. I need to optimize some more before posting it in the Library thread. @JP Thanks for your help.

Looks good! Also set file0norm = 0 to 1 in script settings.

 

[JP]

I feel foolish. Here is the fixed one. I need to optimize some more before posting it in the Library thread. @JP Thanks for your help.

Looks rather close to the one I measured, even the differences between channels. Too bad they don’t have that consistency in a good way.

 

[mfalcon]

That’s what I was thinking too. This is my second (discount for trade in). One of 2 MCs I have. I really like it but that channel imbalance is not great in those measurements. I improved the Azimuth earlier today but not expecting miracles. I’m going to going to remeasure in a bit

 

[mfalcon]

So I used the Clearaudio Break In Test Record to see if the sweep is any good. It is both channels at once so no Crosstalk.
My CBS test done right after for compare. I'll keep using the CBS but is the sweep on the Clearaudio useless? I'm thinking so.

 

[jjbunn]

Thanks for providing this Python utility - it's great! It just works!

Now I would like to understand whether what I am seeing is normal, or indicates I have an error in my setup. This is a result recorded from a CBS STR 100 test LP dating from 1961, so I suppose first version, played on a Denon DP-1200 table with a Shure V15 Mk III cartridge (2g tracking weight). The signal is passed to an ART DJPreII phono amp with 200pF load setting, via a cable that measures 140pF capacitance on each lead. From there it is recorded at 96kHz in 32 bits on a Tascam DR-05. The file is then trimmed to exclude the signal tones in Audacity and saved as two files, one for each left/right sweep.

The output from the script looks a bit strange in terms of X-talk and min/max frequencies?

Input File: LeftTest.wav
Sample Rate: 96,000Hz
Channels: 2
Length: 64.68661458333334s
In/Out (s): [ 0. 64.68661458]
Min Freq: 0Hz
Max Freq: 0Hz

X-talk @1kHz: 22.83dB


Input File: RightTest.wav
Sample Rate: 96,000Hz
Channels: 2
Length: 64.8793125s
In/Out (s): [ 0. 64.8793125]
Min Freq: 0Hz
Max Freq: 300Hz

X-talk @1kHz: -0.92dB

 

[mfalcon]

Thanks for providing this Python utility - it's great! It just works!

Now I would like to understand whether what I am seeing is normal, or indicates I have an error in my setup. This is a result recorded from a CBS STR 100 test LP dating from 1961, so I suppose first version, played on a Denon DP-1200 table with a Shure V15 Mk III cartridge (2g tracking weight). The signal is passed to an ART DJPreII phono amp with 200pF load setting, via a cable that measures 140pF capacitance on each lead. From there it is recorded at 96kHz in 32 bits on a Tascam DR-05. The file is then trimmed to exclude the signal tones in Audacity and saved as two files, one for each left/right sweep.

The output from the script looks a bit strange in terms of X-talk and min/max frequencies?

Input File: LeftTest.wav
Sample Rate: 96,000Hz
Channels: 2
Length: 64.68661458333334s
In/Out (s): [ 0. 64.68661458]
Min Freq: 0Hz
Max Freq: 0Hz

X-talk @1kHz: 22.83dB


Input File: RightTest.wav
Sample Rate: 96,000Hz
Channels: 2
Length: 64.8793125s
In/Out (s): [ 0. 64.8793125]
Min Freq: 0Hz
Max Freq: 300Hz

X-talk @1kHz: -0.92dB

View attachment 347725

I made tne same mistake. Swap the channels in your Right channel file. So the signal is in channel 1 of each file and crosswalk is in channel 2 of each file.

 

[jjbunn]

I made tne same mistake. Swap the channels in your Right channel file. So the signal is in channel 1 of each file and crosswalk is in channel 2 of each file.

Oh, wow ... thanks! Is there an easy way to do that in Audacity, or some other tool?

 

[jjbunn]

Oh, wow ... thanks! Is there an easy way to do that in Audacity, or some other tool?

I figured it out ... via Dropdown menu to the right of the Audacity track name.

Now I have something that looks more sensible, albeit pretty atrocious!

 

[MaxwellsEq]

I figured it out ... via Dropdown menu to the right of the Audacity track name.

Now I have something that looks more sensible, albeit pretty atrocious!

View attachment 347726

Does music sound like that with reduced top end?

 

[DSJR]

Thanks for providing this Python utility - it's great! It just works!

Now I would like to understand whether what I am seeing is normal, or indicates I have an error in my setup. This is a result recorded from a CBS STR 100 test LP dating from 1961, so I suppose first version, played on a Denon DP-1200 table with a Shure V15 Mk III cartridge (2g tracking weight). The signal is passed to an ART DJPreII phono amp with 200pF load setting, via a cable that measures 140pF capacitance on each lead. From there it is recorded at 96kHz in 32 bits on a Tascam DR-05. The file is then trimmed to exclude the signal tones in Audacity and saved as two files, one for each left/right sweep.

The output from the script looks a bit strange in terms of X-talk and min/max frequencies?

Input File: LeftTest.wav
Sample Rate: 96,000Hz
Channels: 2
Length: 64.68661458333334s
In/Out (s): [ 0. 64.68661458]
Min Freq: 0Hz
Max Freq: 0Hz

X-talk @1kHz: 22.83dB


Input File: RightTest.wav
Sample Rate: 96,000Hz
Channels: 2
Length: 64.8793125s
In/Out (s): [ 0. 64.8793125]
Min Freq: 0Hz
Max Freq: 300Hz

X-talk @1kHz: -0.92dB

View attachment 347725

2 grammes on a V15 III - I'm surprised the thing hasn't collapsed!!!

1g should be ample for a good VN35E original stylus and my two haven't gone off in tracking at 1g despite their fifty year unused life.

A V15III should measure pretty damned flat judging by copious reviews of the time.

I'll get my coat (again)

 

[VinyLuke]

Now I have something that looks more sensible, albeit pretty atrocious!

Show your settings in script. That drop-down looks suspicious.
And YES! 2grams of VTF?! What stylus are you using?

 

[jjbunn]

I've always run it at 2g, as that's what the chap I bought it from advised he used Looks like I should reduce it to 1g and see. Sounds good to my ears, and that HF decline on the graph isn't what I hear. Strange.

I need to measure my Soundsmith SMMC4, which has less than 50 hours use on it. That runs at 1.2g VTF.

The only settings other than filenames and descriptions I've changed in the script are:

riaamode = 1
riaainv = 1
str100 = 1
file0norm = 1

 

[VinyLuke]

I've always run it at 2g, as that's what the chap I bought it from advised he used Looks like I should reduce it to 1g and see. Sounds good to my ears, and that HF decline on the graph isn't what I hear. Strange.

I need to measure my Soundsmith SMMC4, which has less than 50 hours use on it. That runs at 1.2g VTF.

The only settings other than filenames and descriptions I've changed in the script are:

riaamode = 1
riaainv = 1
str100 = 1
file0norm = 1

Try with riaamode = 3.

 

[jjbunn]

Thanks - OK! Here is the result with riaamode=3. This seems to apply two filters in series to the data, correct? The result looks more believable in terms of what I hear.

 

[JP]

Thanks - OK! Here is the result with riaamode=3. This seems to apply two filters in series to the data, correct? The result looks more believable in terms of what I hear.

View attachment 347854

In the recording of the signal with RIAA, i.e. the same as if you were listening to a record?

 

[jjbunn]

In the recording of the signal with RIAA, i.e. the same as if you were listening to a record?

Yes, recording from the line outputs of the ART DJ PreII.