RIAA, Questions only please

Atmasphere 2-28-2019

At one point you were insisting that the cutter’s constant velocity characteristic was compensated by that of the cartridge....

Hi Ralph,

Actually it wasn’t Roger who asserted that. I stated that based on what was said in the following writeup:

http://pspatialaudio.com/displacement.htm

What Roger said is that the velocity characteristic of the cutter is compensated for by a 60 db EQ applied in the cutting process. And that the 40 db playback EQ compensates for the velocity characteristic of the phono cartridge, as well as for a pre-emphasis of approximately 12 db.

Based on comments you subsequently provided in the now deleted thread, though, backed up by the extensive first-hand experience with cutting equipment that you have, both statements are incorrect. And of course you and I both disagree with Roger’s statements to the effect that pre-emphasis covers a range of approximately 12 db, rather than approximately 40 db, over the 20 Hz to 20 kHz frequency range.

Roger, FWIW, while many of us here generally try to adhere to ground rules that may be stated by the originator of a thread, as you can infer from most of the responses that have already been posted a ground rule which excludes the possibility of dissent or disagreement, assuming it is presented in a civil manner, is not normal practice here.

Best regards,
-- Al

Tammy, here is a link to the deleted thread that has been referred to above by ct0517 and others:

https://forum.audiogon.com/discussions/how-does-riaa-pre-emphasis-work-on-reducing-record-noise

Best regards,
-- Al

Ramtubes 3-1-2019

Here is something to consider. What if we didnt use RIAA or any EQ in making a record. If we recorded a constant voltage sweep from 20 to 20Khz and looked at it under a microscope what might we see? When we played the final product what would a voltmeter connected to the speaker (or elsewhere) indicate from 20 to 20 KHz.

Choose your cutter wisely and state if it is amplitude or Velocity. Even better do both.

In modern times, at least, cutter heads and the vast majority of cartridges are nominally considered to behave as velocity transducers. I believe that you (Roger) attested to that in the deleted thread.

**If** both transducers were to conform to a theoretically ideal model of a velocity transducer, the response of the cutter head (in terms of groove excursion) would fall at 6 db/octave, the response of the cartridge (in terms of output voltage) would rise at 6 db/octave, and the falling response of the cutter head would be cancelled out by the rising response of the cartridge. So in that hypothetical situation, with no equalization having been applied either in the making of the record or in playing it, the voltmeter you referred to would indicate constant voltage from 20 Hz to 20 kHz. And the excursions of the groove on the record, as viewed under a microscope, would decrease in amplitude as frequency increases.

However, Ralph stated toward the end of the deleted thread that the behavior of both transducers differs considerably from those idealized models, and I certainly have no reason to doubt that. I am not in a position to comment, though, on the manner or the degree to which those deviations from idealized behavior would change the results I’ve described for idealized models.

Best regards,
-- Al

Yes, Imhififan has provided us with an outstanding reference on the subject. (Just as he had done, btw, in an unrelated thread not too long ago in which I and Roger had also been participating, which helped bring all parties to agreement). Thanks!

For one thing, the article appears to support statements Roger has made indicating that between approximately 50 Hz and 500 Hz, and above approximately 2122 Hz, records are cut at constant amplitude. ("Approximately" because as I think we all agree the transitions between constant velocity and constant amplitude are gradual, and those frequencies represent "3 db points" as Roger and the article have stated).

What the article leaves me uncertain about, however, given that in modern times a cutting head is to at least a loose approximation a velocity-based transducer (meaning that it converts signal amplitude to groove velocity), is **how** the cutter is caused to cut at constant amplitude at those frequencies.

If I recall correctly Roger had stated toward the end of the deleted thread that that is accomplished by means of an EQ applied to the signal prior to its application to the cutter, the EQ extending over a 60 db range (in addition to the "pre-emphasis"), between 20 Hz and 20 kHz. That would seem to be appropriate if behavior of the cutter corresponds to an idealized model of a velocity-based transducer. However, Ralph responded that cutter characteristics deviate considerably from the theoretical ideal, and with his cutter the required EQ is only about 15 db, and also is not linear over the frequency range. And I had pointed out, among other things, that neither EQ appears to be consistent with application of the approximately 40 db curve that I think we all agree is applied in playback.

Finally, and related to that, although the reference provided by Imhififan refers to the need to reduce groove excursion at low frequencies when a velocity-based cutter is used (specifically between approximately 50 Hz and 500 Hz), how that is accomplished is not made clear as far as I could tell.

So while the article is certainly helpful, and appears to me to be likely to bring us closer to settling these matters, there are still some basic questions that need to be answered before agreement can be reached.

Best regards,
-- Al