Phono Cartridge Response Measurement Script

[VinyLuke]

What’s the issue with 17? The only change is in the plot layouts, not the data.

I thought that my weird plots were related to that Version.
But tested again and works fine - so no issue with 17.

 

[narud]

windows users- any idea what may be going on here? i dont really understand how any of this works and i was surprised i ever got the script to work in the first place, but my laptop had an update and restart occur and now version 17 of the script doesnt work. thats what i was using prior to the update. 16.5 works though. heres the message im getting
Traceback (most recent call last):

File "C:\Users\User\.spyder-py3\untitled0.py", line 458, in <module>
fo1, ao1, fox1, aox1, fo2h1, ao2h1, fo3h1, ao3h1 = createplotdata(input_sig, Fs)

File "C:\Users\User\.spyder-py3\untitled0.py", line 330, in createplotdata
aoutx = signal.sosfiltfilt(sos,aoutx)

File "C:\Users\User\anaconda3\lib\site-packages\scipy\signal\_signaltools.py", line 4353, in sosfiltfilt
edge, ext = _validate_pad(padtype, padlen, x, axis,

File "C:\Users\User\anaconda3\lib\site-packages\scipy\signal\_signaltools.py", line 4123, in _validate_pad
raise ValueError("The length of the input vector x must be greater "

ValueError: The length of the input vector x must be greater than padlen, which is 12.

 

[JP]

Something’s wrong with the input file. Is there any silence?

 

[narud]

Something’s wrong with the input file. Is there any silence?

no silence. i was using 17beta for everything i posted and had no issue with any of the files. being a pessimist, after the update and restart i thought id try running the script and sure enough it didnt work. i had 16.5 saved and that worked.

 

[JP]

Nothing at all was changed in that part of the code between 16.5 and 17.0. The issue is with the right channel of File_1.

 

[narud]

nvm, im dumb. user error! i forgot to set the riaa settings correctly. my saved 16.5 file had the settings correct. sorry for time waste!

 

[JP]

No problem.

 

[tpd]

This is a pretty great project. Thank you.

After recording the sweeps into Audacity, I'm struggling a bit with the next step. Is the script expecting two stereo files with the channel under test in the "top" track, or can it read mono files?

I assumed the latter (two mono files, one for each channel) but I'm getting inconsistent results or outright errors with that strategy. But it could also be user error. I am not very familiar with Audacity.

Documenting a best practice Audacity workflow would be really useful. I'm happy to help with that once I understand what's expected.

 

[JP]

Stereo or mono, one or two files. Make sure there's no silence and that you also trim off the 1kHz pilot tone. If stereo, it expects the left channel to be the swept tone, and the right channel will be the crosstalk.

 

[tpd]

Ah got it. Thank you. I will make my next recording with the crosstalk. Perfect.

Also discovered that one of my problems was exporting channels with the panning set incorrectly, so there was not actual sound on one channel. Fixed that. Ignoring the fact that I probably need to dial in setup a little, does this version seem in the ballpark for expected results?

 

[JP]

Sans crosstalk, looks like it’s working

 

[Balle Clorin]

For some reson Anacinda/Spyder shuts down after a few minutes of inactivity, How do I stop this annoying "feature". Ir did something once to stop this but cannot remember where to change this

 

[n3mmr]

A possibly very silly question:
Can/should you use this script (&lib) with a normal Riaa-compensated phono playback amp or would it be better to disable all RIAA-correction and use the flat signal with this script?

 

[JP]

A possibly very silly question:
Can/should you use this script (&lib) with a normal Riaa-compensated phono playback amp or would it be better to disable all RIAA-correction and use the flat signal with this script?

It has settings for either case.

 

[n3mmr]

It has settings for either case.

Ok, but are there technical reasons to choose one over the other, under some conditions?

 

[JP]

Ok, but are there technical reasons to choose one over the other, under some conditions?

Always prefer flat if that's an option, as often a given phono stage RIAA compliance is an unknown. I think most are reasonable, but that's not always so.

 

[Bob from Florida]

First, I would like to thank JP for all the work on this measurement script!

I have a question for the group. Has anyone had success running the script on an IOS device. If so, can you share which app was used and any "gotcha's" encountered? Thanks.

 

[JP]

I think the two dependencies which may be problematic are librosa and maybe matplotlib. In the next version I intend to drop librosa as it won't be needed with the new automatic file parsing feature.

 

[mfalcon]

@JP I'm trying to run the V17 of the script using two separate stereo L and R files. I get this error. I looked and looked and can't find the silent sections which is what I thought was the problem. Do you have any ideas?


Traceback (most recent call last):

File "/Volumes/Music/plots/Freq Response/Lyra_Kleos_CBS_STR100_2024_02_01.py", line 437, in <module>

fo0, ao0, fox0, aox0, fo2h0, ao2h0, fo3h0, ao3h0 = createplotdata(input_sig, Fs)

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/Volumes/Music/plots/Freq Response/Lyra_Kleos_CBS_STR100_2024_02_01.py", line 299, in createplotdata

aoutx = normstr100(foutx, aoutx)

^^^^^^^^^^^^^^^^^^^^^^^^

File "/Volumes/Music/plots/Freq Response/Lyra_Kleos_CBS_STR100_2024_02_01.py", line 247, in normstr100

for x in range(find_nearest(f, fmin), (find_nearest(f, fmax))):

^^^^^^^^^^^^^^^^^^^^^

File "/Volumes/Music/plots/Freq Response/Lyra_Kleos_CBS_STR100_2024_02_01.py", line 155, in find_nearest

idx = (np.abs(array - value)).argmin()

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

ValueError: attempt to get argmin of an empty sequence

 

[JP]

It’s not able to get FFT values for the frequencies it needs for some portion of the file. Make sure the RIAA settings are correct, and that your levels are high enough.