True 78 cut on Lacquer master

[recordmaker]

I have a working and active process for making gramophone compatible 78 records and although I can derive my masters from a wax master blank I would like to have a 78 master/s cut with the full depth groove and true mono so I can have a nickle stamper made to use as mold.
I am working with a company that can process the master but have not yet located a rout to getting one cut. The groove would be 90 to 100 TPI and 150micron wide with no stereo component so making cutting through to the the substrate less of a danger I understand.

Duncan

Sheffield

[Dub Studio]

I see no reason why this can't be done. If the mastering engineer doesn't have a 78rpm turntable, you can always send in a sample-rate-converted master and have it cut at 45rpm. Assuming the audio is mono, it will come out mono, but even if it is stereo, it's easy to centre it without audio loss.

Variable pitch is the default these days but it should be an easy enough thing to cut fixed pitch.

150 micron wide would be fine, even on a DMM disc (since they are the shallowest at 100um, allowing up to 200um width cuts).

If you don't find anyone, contact me via my website, I would be up for a proof of concept on dub plate, with subsequent lacquers.

[boogievan]

Fractional speed cutting ideally would include a customized recording network with adjusted time constants for proper decoding on playback at speed. 1/2-speed-capable cutting amps with built-in RIAA, por ejemplo. have this extra feature.

[Dub Studio]

Fractional speed cutting ideally would include a customized recording network with adjusted time constants for proper decoding on playback at speed. 1/2-speed-capable cutting amps with built-in RIAA, por ejemplo. have this extra feature.

Is that just a fancy way of saying "EQ it a bit"?

In any case, "Gramophone-compatible" does not always mean RIAA, so unless this is a priority for the OP, I wouldn't worry too much.

[boogievan]

It's not just saying 'use a bit of EQ', Dub Studio. Arguably most 78 records were cut with a 300-ish-Hz high-bass turn-over frequency, and an 11 - to - 15 -dB low bass cut around 50 Hz. While there was some variety to the bass turn-over frequency used for wax mastering - the variety was relatively close together, compared to the situation you suggested when one cuts a 78 at 45 rpm. If doing that, the time-constants need to be _almost_ doubled...since 45 RPM is only about 57% of 78 RPM... That's why I pointed to 1/2-speed cutting and modified time constants in the record EQ to compensate for the big change in cutting speed versus playback speed that is twice that of the cut.

So, for an effective 300-Hz turn-over EQ on playback at 78 rpm, the turn-over EQ that one applies during cutting to a suitably time-expanded program should be at 173 Hz, if the 78's cutting is done at 45 rpm. The time constant for playback in that case would be [159.1549/F (in kHz) =] ~ 530 µs. Whereas, during cutting at 57% of the target playback velocity, the time constant would have to have been ~ 920 µs.

There also needs to be a boost at 3180 µs on playback, since the low bass frequencies are hidden on disk. If the 78 was cut at 45 RPM, then the slowed-down program (already well-eq'd for good sound when heard 'at speed') would have to be further processed with a low bass cut centered at ~ 5526 µs (or, ~ 28.8 Hz)....

For treble pre-emphasis, which was done for electric 78s by some labels (e.g., Decca), resulting in a quieter noise floor (since, on playback, the boosted treble is attenuated back to 'flat' - just as the later, RIAA curve does), the treble pre-emphasis time-constant used for playback at 78 would also have to be multiplied by 0.57 for cutting at 45.


- Van Boogius

[Dub Bull]

...the treble pre-emphasis time-constant used for playback at 78 would also have to be multiplied by 0.57 for cutting at 45.

True, but it should read, '_divided_ by 0.57' when the playback target is 78 and the cut is made to slowed-down music at 45.

f. José

[boogievan]

...the treble pre-emphasis time-constant used for playback at 78 would also have to be multiplied by 0.57 for cutting at 45.

...agree that the record-eq time-constant needs to be a bigger number (ergo, divide previous tau by 0.57 for a lower center frequency) if the cut is done at a fraction of the target pickup speed and the playback eq is (still) to be for standard cuts. However, by then I was sort of rhetorically fuguing, if I'm honest, and, technically, the phrase you gainsay could nevertheless mean that the _pickup_ eq time-constant needs to be made to have a smaller number (ergo, 'multiply by 0.57',(...as I wrote), for a higher center frequency) if cut, with normal-speed eq, though at a slower speed (which would be Dub Wise's way, in the first place. 0l;.)... This place. So, we're both right, and I'm glad you agree. smh (trolz);

What about the pickup stylus for the proposed 6-mil groove? 2.5-mil tip conical?
Wouldn't the traditional shellac-era coarse groove be cut with a cutting stylus possessing a ~ 110-degree included angle (rather than 88.4 or so?)...? (...not questioning the proposed 3-mils' depth, per se, but the angle of the V...etc...)

[boogievan]

Whilst 300 Hz is a reasonable approximation for many a shellac's bass-turn-over- F, 200 - 500 cps includes most of the variety of the turn-over frequencies that were used on electric-era 78s, and many were right around 375 cps (i.e., 424 µs (= 375 cps), rather than 530 µs (= 300 cps)), according to this ARSC .pdf:
https://arsc-audio.org/journals/v20/v20n1p14-23.pdf

Btw, for those newer to circuit mods, the mention of time-constants is valuable because it simplifies tweaking a circuit, such as the RIAA record equalization in a cutting amp, or the equalization in your phono pre?... This is because Tau (the Greek letter used to symbolize the phrase, 'time-constant') is the product of the resistance times the capacitance in a network, and most audio filters use those components, innit? First calculate where it is and then you can add capacitance by placing a second capacitor in parallel with the original, or subtract capacitance from the time-constant being made by adding a second capacitor in series with it. Just the opposite effect on resistance is seen by placing a second resistor in parallel or series with the original resistor. Also, it's interesting to note that 3180 µ seconds - the time-constant usually given for "50 Hz" - is actually the time-constant for 50,05 Hz (or, 50.0487..Hz). The explanation for the slight discrepancy (and how the number, 159, got into the formula for Tau) is given by Dr3d Sl3d in the post, here: https://www.discolathe.com/viewtopic.php?t=145

It's a hairy topic, but not quite a free-for-all. More importantly, a successful and professional-quality venture of the kind proposed by the O.P. should entail the deliberate use of some way to hide bass and mitigate surface noise via record-equalization, and choosing one that's available to most end-users for restoring the recording to the flat status of the nicely-EQ'd program may well involve just sticking to bad-old, but ubiquitous, RIAA time-constants (3180/318/75)... But to use RIAA for lacquer-cutting at 57% of 78 rpm (i.e., 45 rpm) should include modification of the RIAA encoder network to something like: 5584 µs / 558 µs / 132 µs, as in: 29 Hz / 290 Hz /1,210 Hz, simple.

[Dub Studio]

Thanks for the explanation, I am not questioning the value of what you are saying or the validity of modding time constants per se. I am sure plenty of people will find that information useful, and clearly this would be standard procedure when setting up a lathe for this task permanently.

But the OP isn’t setting up a cutting house here, they just want to get some lacquers cut, and what you are suggesting is not something you can just walk into a cutting house and ask for, is it?

I am trying to offer up a practical solution to this problem, one that would mean the OP can take their pick of studios, not just mine. I am happy to be corrected on this, but the response curves of the various 78s are in the public domain and EQs are ubiquitous, so you should be able to prep the file at home no matter what type of 78 you need. Right?

[Dub Studio]

What about the pickup stylus for the proposed 6-mil groove? 2.5-mil tip conical?
Wouldn't the traditional shellac-era coarse groove be cut with a cutting stylus possessing a ~ 110-degree included angle (rather than 88.4 or so?)...? (...not questioning the proposed 3-mils' depth, per se, but the angle of the V...etc...)

Yeah now this could an issue. I think from this post the OP wants steel needle compatible: https://lathetrolls.com/viewtopic.php?f=15&t=7122&p=63084#p63084

[boogievan]

If I were to cut the lacquers, I'd use an arbitrary function generator to create the reference clock signal for the turntable servo (and another for feed compensation), while cutting fixed-pitch, which maxes at 100 LPI on my machine, so that it can cut at full speed (since the servo's crystal's master clock resonance normally only can be divided down to the reference clock frequencies for '33', 45, and half-speeds). I'd probably bypass the 75-µs time constant on RIAA encode, since many 78 disks had 'flat' treble response, and often did use RIAA's bass turn-over frequency made by an RC network, the resistance x capacitance of which produces 318 µs (i.e., 500.5 Hz).

If you "bake" the 78-rpm disk-EQ into the cutting master(files), which would also have to be program-EQ'd ahead of time for best sound _after_ disk-EQ is decoded. during pickup), then you need a way to bypass disk-EQ inside your cutting amp, if that's where it gets applied (and it does, here, and I can only bypass the 75-µs time constant via pushbutton)...

Some Technics decks have the 78 option. Yet few phono pres have Decca or BIS or ffrr EQ, and, as you, and the OP already know, there are many others having minor to major variations - and the disk-EQ used was sometimes signaled with a sigil or letters in the dead-wax of the master. So, using RIAA for this would probably be better - allowing more convenience and better gear? - and having the playback be with a crystal-locked Technics that can do 78 rpm would help to demonstrate the superiority of high land-speed, compared to 33.3, for example, even though it makes a huge limitation on program duration for a given side.

A truncated elliptical stylus is sometimes recommended modern decks for safely tracing such a deep groove (as 6 mils, across the top).

But if the goal is to make a back-dated disk in vinyl to be played only on rare, old machines, at least let's pick a disk-EQ that many of them will possess, I guess. I mean, if it's just for a prop in a movie, of course, the sound on film can be (...wait for it...) dubbed! 0;