Hi all,
Can someone check my reasoning here?
When cutting at half speed, we effectively double all the time constants in the IRIAA circuit, so that in the constant amplitude region we add 6dB at any given frequency (Of course the frequency is now half what it was, so peak velocity remains same).
Thus if we consider say a 5kHz (full speed) signal, at half speed (2.5kHz) we get the same peak velocity but at half the frequency, which it seems to me raises the peak displacement by 6dB?
Does cutting at half speed imply turning down the input to the chain by 6dB? I have never done it, it so I am curious.
It would make sense as the velocity geometric limit will be 6dB lower due to the requirement that the peak stylus velocity cannot exceed the tangential velocity at the cutting radius.
Regards, Dan.
Re: Half speed cutting levels and limiter settings?
Hi,
-6db probably gets you in the ballpark, buy I don't think its that cut and dried. I think that depends on the program material and the open loop head response. If you look at the open loop response of a textbook moving coil head in terms of displacement vs. frequency, its a constant amplitude device (stiffness controlled) below resonance and drops off at -12/db/oct on the other side of resonance (mass controlled). If you want the resulting cut to be the same level when played back at normal speed, you want to cut such that the excursion vs the frequency/2 is the same as at normal speed over the full frequency range. For program material in the stiffness controlled region, dropping the cutting speed (and frequency) in half takes same power to cut as full speed since the current to achieve the required excursion is the same (F = Kx and F=BLi). However, any program material that lies in the mass controlled region requires less power at half speed to achieve the same excursion. If the frequency is lower by half but we keep the same excursion, the acceleration is reduced by a factor of 4. In this region F=ma. The average power to cut should be less, but it would take some calculation, measurement, or simulation to determine how much and would depend on the head response and the source spectral density. Does that make sense?
Mark
-6db probably gets you in the ballpark, buy I don't think its that cut and dried. I think that depends on the program material and the open loop head response. If you look at the open loop response of a textbook moving coil head in terms of displacement vs. frequency, its a constant amplitude device (stiffness controlled) below resonance and drops off at -12/db/oct on the other side of resonance (mass controlled). If you want the resulting cut to be the same level when played back at normal speed, you want to cut such that the excursion vs the frequency/2 is the same as at normal speed over the full frequency range. For program material in the stiffness controlled region, dropping the cutting speed (and frequency) in half takes same power to cut as full speed since the current to achieve the required excursion is the same (F = Kx and F=BLi). However, any program material that lies in the mass controlled region requires less power at half speed to achieve the same excursion. If the frequency is lower by half but we keep the same excursion, the acceleration is reduced by a factor of 4. In this region F=ma. The average power to cut should be less, but it would take some calculation, measurement, or simulation to determine how much and would depend on the head response and the source spectral density. Does that make sense?
Mark
Re: Half speed cutting levels and limiter settings?
Hi again,
My bad. I was thinking in terms of head power levels, not record levels into a head that has flat velocity response either via EQ or feedback. You can disregard my last post.
Mark
My bad. I was thinking in terms of head power levels, not record levels into a head that has flat velocity response either via EQ or feedback. You can disregard my last post.
Mark
Re: Half speed cutting levels and limiter settings?
Yea, thing to remember about a feedback head is that the feedback is (absent any LC resonances and interference from the drive coil, never a given) a pure velocity signal, moving wire in a magnetic field and all that.
My suspicion is that the major difference with half speed back in the day was actually the tape replay head bump ending up one octave higher then it normally is, probably combined with a reduction in the issues around the structure outside the loop not being completely stiff.
The actual dynamics of these things is surprisingly hairy when you try to model it.
Regards, Dan.
My suspicion is that the major difference with half speed back in the day was actually the tape replay head bump ending up one octave higher then it normally is, probably combined with a reduction in the issues around the structure outside the loop not being completely stiff.
The actual dynamics of these things is surprisingly hairy when you try to model it.
Regards, Dan.
-
dubcutter89
- Posts: 359
- Joined: Thu Oct 19, 2006 6:30 am
- Location: between the grooves..
Re: Half speed cutting levels and limiter settings?
Hmm, from my understanding the 6dB extra gain make sense!
A signal that at normal speed is let's say a pure sine with 0dB on the input will be inverse RIAA encoded to about +20dB velocity in the groove (and ends up as an amplitude A in the groove).
Cutting at half speed (with doubled time constants) will make the "same" 20kHz that are now 10kHz a sine wave with a velocity boosted by +20dB (and ends up as amplitude 2xA in the groove to get same velocity at half frequency as peak velocity V = 2 x pi x f x A).
Playing back that tone again at regular speed will boost the velocity by a factor 2 which is +6dB.
(Ok, basically repeated your explanation but sometimes it makes to read/think something twice...)
But that's all in theory for a perfect flat transfer system that translates the input signal into an (RIAA encoded) velocity on the record.
I don't have much experience with that stuff as well outside of one or two attempts but if you now throw in tape in that system everything will be different again...(will half speed tape playback also induce half the voltage in the preamps as the magnetic flux change rate is 6dB lower??)
The challange is to make that perfect flat transfer system!
And then after having it flat you can start turning knobs on big old mastering consoles to get the sound right...
Lukas
A signal that at normal speed is let's say a pure sine with 0dB on the input will be inverse RIAA encoded to about +20dB velocity in the groove (and ends up as an amplitude A in the groove).
Cutting at half speed (with doubled time constants) will make the "same" 20kHz that are now 10kHz a sine wave with a velocity boosted by +20dB (and ends up as amplitude 2xA in the groove to get same velocity at half frequency as peak velocity V = 2 x pi x f x A).
Playing back that tone again at regular speed will boost the velocity by a factor 2 which is +6dB.
(Ok, basically repeated your explanation but sometimes it makes to read/think something twice...)
But that's all in theory for a perfect flat transfer system that translates the input signal into an (RIAA encoded) velocity on the record.
I don't have much experience with that stuff as well outside of one or two attempts but if you now throw in tape in that system everything will be different again...(will half speed tape playback also induce half the voltage in the preamps as the magnetic flux change rate is 6dB lower??)
The challange is to make that perfect flat transfer system!
And then after having it flat you can start turning knobs on big old mastering consoles to get the sound right...
Lukas
Wanted: ANYTHING ORTOFON related to cutting...thx
Re: Half speed cutting levels and limiter settings?
Ain't that the truth!
And yea, tape at half speed would be about 6dB down in terms of the head output (Modulo the head bump moving), but all your metering will probably not, so I suspect 6db of gain was applied at the tape machine and the cutting level reduced by 6dB at the lathe input (Possibly even automatically in some cases).
Working this sort of thing thru in your head sometimes provides interesting insights into how the whole process works (In my case I was puzzling out how a limiter should react in a half speed environment for both velocity (Limit is on the cutting side) and acceleration (Limit is playback), as well as where in the chain at -6dB should be?
Assume a limiter with fixed differentiation time constants (Because that is probably the way to play).
The velocity limit is clearly 6dB lower because 'half speed' as is the acceleration limit (because the acceleration will double when you play back at full speed), but where then the 6dB for the changed time constants in the IRIAA?
Switching in a -6dB pad in the IRIAA network would make the Input level to that stage independent of half speed in terms of groove amplitude, and would this actually make the limiter independent of half speed mode as well?
Does doing that make the limiter threshold solely a function of PLAYBACK speed and cutting diameter then? Feels wrong somehow?
This stuff gives the grey matter a workout.
And yea, tape at half speed would be about 6dB down in terms of the head output (Modulo the head bump moving), but all your metering will probably not, so I suspect 6db of gain was applied at the tape machine and the cutting level reduced by 6dB at the lathe input (Possibly even automatically in some cases).
Working this sort of thing thru in your head sometimes provides interesting insights into how the whole process works (In my case I was puzzling out how a limiter should react in a half speed environment for both velocity (Limit is on the cutting side) and acceleration (Limit is playback), as well as where in the chain at -6dB should be?
Assume a limiter with fixed differentiation time constants (Because that is probably the way to play).
The velocity limit is clearly 6dB lower because 'half speed' as is the acceleration limit (because the acceleration will double when you play back at full speed), but where then the 6dB for the changed time constants in the IRIAA?
Switching in a -6dB pad in the IRIAA network would make the Input level to that stage independent of half speed in terms of groove amplitude, and would this actually make the limiter independent of half speed mode as well?
Does doing that make the limiter threshold solely a function of PLAYBACK speed and cutting diameter then? Feels wrong somehow?
This stuff gives the grey matter a workout.