Correcting Cartridge Freq Response

[John B]

Does anybody have a good method for this?

I’ve been taking a pink noise band on a test record and building corrective EQs by looking at differences relative to a digital reference. It’s pretty straightforward but quality the of the corrective EQ when mapped to music content is variable.

I’m aware noise profiles, music energy dispersal, and the records themselves add variables which skew results here, even so, I’m curious if somebody has figured this out and if so, what they are using?

It’s mind blowing how “off” cartridges measure even when approximated. Any progress toward evening them however imperfect, could be useful IMHO.

Thanks!

 

[abdo123]

https://www.ortofon.com/test-record-p-707

 

[The Mule]

Does anybody have a good method for this?

I’ve been taking a pink noise band on a test record and building corrective EQs by looking at differences relative to a digital reference. It’s pretty straightforward but quality the of the corrective EQ when mapped to music content is variable.

I’m aware noise profiles, music energy dispersal, and the records themselves add variables which skew results here, even so, I’m curious if somebody has this figured this out and if so, what they are using?

It’s mind blowing how “off” cartridges measure even when approximated. Any progress toward evening them however imperfect, could be useful IMHO.

Thanks!

This is exactly how I do it. I can see changes with capacitance, different cartridges, etc. Very neat and useful.

I got a very nice used SL1200mk2 many years ago. FInally got it hooked into my system. Figured out very quickly, that it had crap (i.e., high capacitance) 3rd party interconnect cables on it. (6ft vs 4ft std, and high gauge/skimpy) Swapped those for some low capacitance Mogami's, and I can definitely see the high freq peak shift higher in freq. Very nice tool.

 

[USER]

This is a great thread on the topic: https://www.audiosciencereview.com/forum/index.php?threads/mm-vs-mi-vs-mc.18636/page-14#post-630125

It'll take you a while to process it but I was able to get the following results by doing so (with some help): https://www.audiosciencereview.com/forum/index.php?threads/mm-vs-mi-vs-mc.18636/page-28#post-787537

If you are able to get one of these https://www.ebay.com/itm/224396431537, send me a recording, and provide me all the variables (capacitance, etc.), I should be able to provide you with a graph. Each individual record has its own problems, but I was able to get enough of the CBS records to compare them and get a good idea of what the respective ideal FR should be.

 

[The Mule]

Pink noise works, I don't know why you think it doesn't. This is from my Sony PS-X600/Shure V15 Type V-MR/U Turn Pluto 2. I plotted vs *additional* capacitance added by the DB Systems capacitance kit. You can see the high freq peak shift lower as capacitance is added. This is unsmoothed data too.

 

[DVDdoug]

Pink noise works, I don't know why you think it doesn't.

Just like the sweep, it depends on the record. The nature of noise (randomness) probably doesn't help. Regular music records aren't perfect either.

Personally, I have a much more "casual" attitude than in the vinyl days... I don't actually listen to records anymore but when I occasionally digitize a record, I'll EQ it by-ear if it doesn't sound right. Older records varied a lot and IMO most had rolled-off highs.

I was kinda' dumb searching for "perfection" and hoping that the record I just bought would be one of the few gems, and then feeling guilty, or like my system was inadequate, or like I was cheating if I was adjusting the tone controls other than flat...

I assume things are better (and more consistent) now but I don't have any modern records. But I was still listening to records I wouldn't be looking for that "perfect" or "better" cartridge and chasing perfection...

 

[sergeauckland]

I have a problem deciding which of the four traces above is actually the best. They all indicate the typical response of a MM cartridge with different capacitance loading. In my view, they're all equally bad, possibly the green trace being the best audibly, given my by now limited HF hearing.
I think the best way of equalising a cartridge, if it's deemed necessary at all, is to use an external equaliser, possibly a combination of parametric and spot frequency (graphic) and adjust for 'best' fit, or indeed most preferred sound. Or just accept that all LP playback is flawed, possibly enjoyably so, but still flawed.

S

 

[The Mule]

I have a problem deciding which of the four traces above is actually the best. They all indicate the typical response of a MM cartridge with different capacitance loading. In my view, they're all equally bad, possibly the green trace being the best audibly, given my by now limited HF hearing.
I think the best way of equalising a cartridge, if it's deemed necessary at all, is to use an external equaliser, possibly a combination of parametric and spot frequency (graphic) and adjust for 'best' fit, or indeed most preferred sound. Or just accept that all LP playback is flawed, possibly enjoyably so, but still flawed.

S

The navy blue trace is the best. It pushes that high freq peak as high as it can go with this setup, which means it's less audible. Unless you can hear out to 18-19kHz. Btw, I consider this quite good. +/- 2.5 dB, 10 Hz to 20kHz.

 

[sergeauckland]

The navy blue trace is the best. It pushes that high freq peak as high as it can go with this setup, which means it's less audible. Unless you can hear out to 18-19kHz. Btw, I consider this quite good. +/- 2.5 dB, 10 Hz to 20kHz.

Except that the blue trace has more of a dip at around 10k, so will sound dull to me compared with the green trace, as I can't hear 15kHz, but certainly can at 8-10k. That's why I wrote that it's hard to decide which of the four is the best, it depends on one's hearing, and indeed one's preferences. Some like the HF peak lower, as it adds 'sparkle'.

S

 

[John B]

Thanks all,

I figured out a few things this weekend while transferring an old LP:

1. The quality of the EQ and matching feature is important. (ProQ3 was best and easiest. It also has fantastic "tilt" EQs to brighten or darken the overall LP sound with minimal fuss after you correct for cartridge variance)
2. Left and right channels benefit from their own corrective EQs
3. Left and right channels need to be balanced back to their original volume differential after correction
4. Account for amplitude addition of your test LP's rumble and subtract some of it. (measure it on your test LP using track lead-ins) - I throw out any correction below 60hz and tune the rumble augmented bands by ear.
6. More corrective bands can help if you want to tame some resonant nodes but the frequencies excite them can be different than reference so results may vary.
7. Using ~50% EQ strength sounds most natural, but this will depend on how far off your cart is.
8. Get a better performing cartridge. The EQ lifts signal AND noise which is noticeable.

Let me know if you've got any tips for improving this. This is definitely worth the effort IMHO, especially if you are doing digital transfers which you'll have to live with for a while.

 

[DVDdoug]

1. The quality of the EQ and matching feature is important. (ProQ3 was best and easiest. It also has fantastic "tilt" EQs to brighten or darken the overall LP sound with minimal fuss after you correct for cartridge variance)

In my experience the records are the biggest problem.

2. Left and right channels benefit from their own corrective EQs

Personally, I've never done that.

3. Left and right channels need to be balanced back to their original volume differential after correction

In Audacity you can run Loudness Normalization with the "independent" option to re-balance the channels (no matter what the cause). After that, you should run the regular Normalize (not independently) or Amplify effect to readjust the levels and avoid clipping when you export. GoldWave has an effect called Max/Match that matches the RMS levels and then normalizes/maximizes with one click.

4. Account for amplitude addition of your test LP's rumble and subtract some of it. (measure it on your test LP using track lead-ins)

I assume you aren't actually subtracting (inverting and mixing) because that won't work.

7. Using ~50% EQ strength sounds most natural, but this will depend on how far off your cart is.

That would depend on the particular equalizer but I'd say 6dB is "a lot" for digitizing records.

There is another effect called an "exciter" or "harmonic exciter" (inspired by the Aphix Aural Exciter) that adds high-frequency harmonics. There is an optional plug-in for Audacity called "Harmonic Enhancer". I'm not necessarily "recommending" this, but it's another way of "restoring" highs that you might want to try.

Let me know if you've got any tips for improving this.

Buy the CD (if it's available)!

Audacity has "regular" Noise Reduction (for constant low-level noise like hiss, hum, and maybe rumble) and there is an automatic Click Removal removal tool, a "manual" Repair effect and "worst case" you can manually re-draw the waveform. Wave Corrector is a Free automatic vinyl click removal tool. Wave Repair ($30 USD) works manually so it only "touches" the audio where you identify a defect. It offers several repair methods and it does an audibly perfect job on most (but not all) clicks and pops. The downside is that it's VERY time consuming.

 

[John B]

This odyssey has absolutely affirmed my faith in digital! ; )

 

[dlaloum]

Several things to keep in mind....

1) In many cases the ideal loading for a MM (and to a lesser degree a MC) cartridge, is NOT the "standard" - custom loading is needed to get anywhere near a flat FR, and very few phono stages allow for it....
2) digital EQ is a valid alternative if the base FR is close enough to not require massive boosts that compromise headroom - when using Digital EQ it is best to use EQ that use IIR filters that affect phase - these will naturally Correct the phase as required by the cartridge (Basically the aspects that affect cartridge FR, resonances, loading, etc... all have natural phase/frequency relationships, which need to be adjusted by filters/EQ that has the same relationship )

If you have enough Headroom in your ADC, you can use digital EQ for your RIAA (keep in mind the type of filters needed to correct phase!!) but you are likely to still need a very good signal amplifier to get the cartridge output up to the required levels for your ADC.

This is difficult and fiddly (or it was when I was doing it about 6 or 7 years ago)

FInally an observation - based on an experiment I did at the time:

I digitised the same 2 or 3 tracks, using about 10 different cartridges/needles - and doing basic (cartridge loading) adjustments to get all the FR's as close to flat as possible - without any digital EQ

Each cartridge had its own distinctive signature sound - they weren't massively different but they were distinct - variances in FR in the main audible range were circa +/-3db

I Then EQ'd them all so their FR was within +/- 0.5db - and the distinct differences between them basically disappeared

I honestly believe a properly conceived digital EQ based filtering system is the way forward to improved fidelity with vinyl - but it also requires appropriate tools... FR test records, FR measurement and correction based on the test records etc...
And the EQ applied is individual to the particulare cartridge and needle in use - if you change needles (for an MM) - or recantilever/retip for an MC - you need to re-test and re-EQ... as the changes will have affected the FR.

 

[dlaloum]

And lo and behold within a couple of hours of having written that, I read a review of the new Technics SU-R1000 Integrated amp, which includes a digital phono stage, along with a test record.... I would like more information on what they do exactly - as it seems they may adjust both the FR and use some form of cancellation processing to boost seperation....