Limiters with FETs

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Limiters with FETs

Post by JR. »

Yup a limiter for a guitar is an "effect" so by definition does not need or even want to be transparent. Many a bass guitar limiter used the basic JFET shunt with simple rectifier to feed the gate.

I don't even try to design effects any more (last studio effect was in the late '70s early '80s. I do have a friend with a guitar pedal company, but I'm still waiting for him to listen to my GFCI stinger cap ground (for noise floor). Last I checked he was slammed with NAMM show orders, maybe he has caught up by now. I need to ping him about that.

My couple decades of bench work on dynamics processing was predominantly focussed on transparent compression/expansion for use in noise reductions. This transparent companding does not automatically mean the compression is transparent. I ended up with pre/de-emphasis in the audio path, pre/de-emphasis in the side chain, a hard (overall diode clipper across the feedback path so the compressor never lost feedback, etc.. lots of moving parts, that only had to sound good together, not by themselves.

Sorry, I was confusing this discussion with the location recording thread.

never mind... :oops:

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Limiters with FETs

Post by mediatechnology »

Yeah the location recording topic is spread out over a lot of threads...
Part of the veer is mine...
carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Limiters with FETs

Post by carlmart »

Yup, the location recording matter is my interest.

Unfortunately I can only carry a guitar, can't play it...


Carlos
ricardo
Posts: 262
Joined: Sun Apr 10, 2011 7:24 am

Re: Limiters with FETs

Post by ricardo »

mediatechnology wrote:The location recordist might want something like this with a higher threshold and a lower ratio.
Below threshold operation is linear.
Pretend the tone burst is an explosion.
The input is on top, output on bottom.
Image
Zero attack, zero release.
No overshoot, no recovery.
No diodes, no FETs, no Vactrols.
I've actually taken part in DBLTs on a power amp 'soft clipper' with similar characteristics. The comparison was with an even simpler clipper which was basically one resistor & 2 diodes.

I think I've still got the measurements somewhere in my junk. I'll post if I find it. Did a big clearout in preparation for Cyclone Nathan which hit us twice in 2 weeks and all sorts of good stuff is re-emerging from the previous Millenium :mrgreen:

It was a follow on from http://www.aes.org/e-lib/browse.cfm?elib=3793 Audibility of Amplifier Clipping which used a hard clipper. Loadsa interesting stuff .. like the discovery of music (?!?) where clipping 50% of the time was undetectable. :o

FWIW, the simple clipper sounded best and was used by KEF in a later series of tests at an AES Convention. These weren't as carefully controlled but came up with similar results.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Limiters with FETs

Post by JR. »

ricardo wrote:
mediatechnology wrote:The location recordist might want something like this with a higher threshold and a lower ratio.
Below threshold operation is linear.
Pretend the tone burst is an explosion.
The input is on top, output on bottom.
-clip-
Zero attack, zero release.
No overshoot, no recovery.
No diodes, no FETs, no Vactrols.
I've actually taken part in DBLTs on a power amp 'soft clipper' with similar characteristics. The comparison was with an even simpler clipper which was basically one resistor & 2 diodes.
Yup, a diode break approach (linear gain that drops above a threshold), can be arbitrarily gentle, with the trade-off being to completely avoid clipping requires lower above threshold gain. If you know or can make assumptions about the peak level of the input stream you can dial in the above threshold gain to completely avoid clipping.
I think I've still got the measurements somewhere in my junk. I'll post if I find it. Did a big clearout in preparation for Cyclone Nathan which hit us twice in 2 weeks and all sorts of good stuff is re-emerging from the previous Millenium :mrgreen:
cyclones suck... or is that blow?
It was a follow on from http://www.aes.org/e-lib/browse.cfm?elib=3793 Audibility of Amplifier Clipping which used a hard clipper. Loadsa interesting stuff .. like the discovery of music (?!?) where clipping 50% of the time was undetectable. :o
This kind of depends on the program and how much but for complex wide band material clipping of narrow HF transients riding on top of lower frequency material is hard to hear. When you overdrive the path hard enough that the LF envelope is saturated, that can be much more audible.
FWIW, the simple clipper sounded best and was used by KEF in a later series of tests at an AES Convention. These weren't as carefully controlled but came up with similar results.
Most of my listening tests were between simple clipping and a fast limiter. The elephant in that room was that the clipped signal contained more energy so was and sounded louder. Louder was perceived as better in this case, even though it wasn't cleaner by objective measures. A soft limiter will reduce energy some so that could be a perceptual factor.

JR
Cancel the "cancel culture", do not support mob hatred.
ricardo
Posts: 262
Joined: Sun Apr 10, 2011 7:24 am

Re: Limiters with FETs

Post by ricardo »

JR. wrote:This kind of depends on the program and how much but for complex wide band material clipping of narrow HF transients riding on top of lower frequency material is hard to hear. When you overdrive the path hard enough that the LF envelope is saturated, that can be much more audible.
The audibility and yuckiness of clipping does indeed depend on the programme.

For what its worth, the objectionable stuff is almost always Intermod. eg http://www.aes.org/e-lib/browse.cfm?elib=2476 Intermodulation Distortion Listening Tests. Stuff like drums can be heavily clipped w/o the listener complaining but if each whack cuts out the vocals even momentarily, the listener complains.

That's one reason why overload on 'small' (less than 50W amps) with conventional PSUs is nasty cos the overload causes the 50/60Hz sawtooth to (inter)modulate with the signal.

Piano is probably the most sensitive material .. even more sensitive than voices, cos voices can be naturally rough. ;)
Most of my listening tests were between simple clipping and a fast limiter. The elephant in that room was that the clipped signal contained more energy so was and sounded louder. Louder was perceived as better in this case, even though it wasn't cleaner by objective measures. A soft limiter will reduce energy some so that could be a perceptual factor.
If you can arrange 'overload' without Intermod, you can use this to make your stuff sound 'better'.

In da old days, i had this perhaps naive belief that if you made your stuff sound 'better', it would help sell stuff and keep you in business .. hence the obsession with DBLTs.

I used to recommend people always bring their own CDs to the shop to audition gear .. and this was one of the most important facets of our DBLTs.

This Millenium, it is almost impossible to hear stuff before buying so it might no longer be worth designing stuff to sound 'better' in the shop. :( Instead, telling everyone very loudly, your stuff is hand carved from Unobtainium by virgins is the way to go. JC will say, you used to have to do this with Reviewers but now its the only option you have with everyone. :?
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Limiters with FETs

Post by JR. »

ricardo wrote:
JR. wrote:This kind of depends on the program and how much but for complex wide band material clipping of narrow HF transients riding on top of lower frequency material is hard to hear. When you overdrive the path hard enough that the LF envelope is saturated, that can be much more audible.
The audibility and yuckiness of clipping does indeed depend on the programme.
and the amount of overdrive into that clipping.
For what its worth, the objectionable stuff is almost always Intermod. eg http://www.aes.org/e-lib/browse.cfm?elib=2476 Intermodulation Distortion Listening Tests. Stuff like drums can be heavily clipped w/o the listener complaining but if each whack cuts out the vocals even momentarily, the listener complains.
I find IMD and THD two sides of the same linearity coin, while they can measure dramatically different in an active circuit with non-flat transfer function. I figured out that THD was grossly understated in RIAA eq phono preamps because of all the HF roll-off (not to veer two threads together)
That's one reason why overload on 'small' (less than 50W amps) with conventional PSUs is nasty cos the overload causes the 50/60Hz sawtooth to (inter)modulate with the signal.
not sure why PS modulation on the signal envelope wouldn't occur for higher power amps too.... With PWM the PS voltage directly modulates the signal result so is involved in the NF. I haven't seen too many regulated supplies in power amps, expect perhaps for some big dog power factor corrected amps
Piano is probably the most sensitive material .. even more sensitive than voices, cos voices can be naturally rough. ;)
My favorite piano anecdote was from a tweako mgazine editor who was having dinner with some HIFI industry big shot when the wife complained about the unnatural piano sound in the background music system. It was a real piano in the other room and she had SH__ for ears.
Most of my listening tests were between simple clipping and a fast limiter. The elephant in that room was that the clipped signal contained more energy so was and sounded louder. Louder was perceived as better in this case, even though it wasn't cleaner by objective measures. A soft limiter will reduce energy some so that could be a perceptual factor.
If you can arrange 'overload' without Intermod, you can use this to make your stuff sound 'better'.

In da old days, i had this perhaps naive belief that if you made your stuff sound 'better', it would help sell stuff and keep you in business .. hence the obsession with DBLTs.
I got off that happy train on the '80s... after getting different results from different reviewers... I didn't have the stomach to arm wrestle everybody. I knew the products were good but that alone does not generate sales.
I used to recommend people always bring their own CDs to the shop to audition gear .. and this was one of the most important facets of our DBLTs.

This Millenium, it is almost impossible to hear stuff before buying so it might no longer be worth designing stuff to sound 'better' in the shop. :( Instead, telling everyone very loudly, your stuff is hand carved from Unobtainium by virgins is the way to go. JC will say, you used to have to do this with Reviewers but now its the only option you have with everyone. :?
Hopefully the electronics have matured to the point that there are not huge audible differences between different SKUs. Loudspeakers will still vary so deserve a listen before dropping serious money, but even then you need to separate the speaker from the listening space, or review it in situ at home.

JR
Cancel the "cancel culture", do not support mob hatred.
Post Reply