Thank God for limiters

[Dogoth]

No religious implications intended.

 

I'm currently working with a National act who's monitor guy hasn't a clue. He's a very nice guy but I gotta tell you there's a lot of erroneous info floating around out there. Here's an example:

 

Guy needs more vocal. He says it sounds like it's getting squashed or distorted. I look and say "you're hitting the limiters on the amp, that's all the watts you've got". Reply was "Turn the input of the amp all the way up, having it turned down is like having a governor on your car!"

 

Rather than explain to him, The limiter is a current limiter on the output, I just smile and say "OK turn the aux send down, I'll dime the amp and you can bring it back up slowly" To his ears that sounded so much better . As long as the limiter's on and doing it's job, I don't really care if someone wants to muck up the gain staging (I'll just put it back when they leave )

 

Sometimes you just gotta wonder!

 

I'm glad it their own monitor guy (I wouldn't want that position (unless it paid REALLY WELL)).

[gruvjack]

There are so many talented guys out there working for peanuts and a national act books this schmuck?

 

V.

[TimmyP]

It could be that with the amp down he was also over-driving the console, and with the amp up the compression/distortion in the console was eliminated, making for a better sound.

 

This is why I never configure a system "everything clips at once" - I leave at least 10dB of headroom in everything ahead of the amps, so that operators can bang on the limiters without clipping anything else in the system.

[agedhorse]

I'm sure JR has his gain structure set so that he has plenty of drive headroom. There would be distortion, not compression with the drive electronics overdriving, and withthe gear JR uses, it's all +26dBu stuff so there's plenty of drive capability even with the amps 10dB down.

 

But for the bottom-feeders, it's a good point.

[Dogoth]

 

It could be that with the amp down he was also over-driving the console, and with the amp up the compression/distortion in the console was eliminated, making for a better sound. This is why I never configure a system "everything clips at once" - I leave at least 10dB of headroom in everything ahead of the amps, so that operators can bang on the limiters without clipping anything else in the system.

 

 

No even close. The board's an M7CL. It has good always visible metering for every input & output. I've run this exact system for hundreds of shows so I know the gain staging pretty intimately (the limit light on the amp starts at somewhere between -12 & -18 on the aux output). The only increase in sonic quality was in this guys head (psychoacoustics :>) and in reality there was an small loss in S/N ratio (although at their stage volume I'm sure it wasn't noticable).

 

The post was just me grumbling about the disinformation and just plain lack of logical reasoning floating aroundout there. This is why I'm such a proponent of matching the amp to the speakers (even underpowering by a bit) and using the limiters built into the output stage. Had I relied on a DSP limiter preset earlier in the signal chain, this action would have resulted in more peak power being delivered to the speakers up to and including where the amps sent square waves out. Of course had The system been set up that way I never would have allowed them to change the gain structure.

 

It's also a post about how smiling and agreeing with the idiot steering the bus (as long as they aren't driving dangerously) helps everyone have a much more enjoyable trip (why point out that they're in the wrong gear most of the time - It's their bus (show)).

[SpaceNorman]

 

This is why I never configure a system "everything clips at once" - I leave at least 10dB of headroom in everything ahead of the amps, so that operators can bang on the limiters without clipping anything else in the system.

 

 

What's theory behind an "everything clips at once" configuration? Is there an existing thread that described the details?

[agedhorse]

 

What's theory behind an "everything clips at once" configuration? Is there an existing thread that described the details?

 

 

Complete and total lack of understanding of how limiting works in amplifiers and speaker processing.

 

It comes from the early days where system noise was a bigger problem and amps didn't have limiters so clipping was clipping.

[SpaceNorman]

Complete and total lack of understanding of how limiting works in amplifiers and speaker processing. It comes from the early days where system noise was a bigger problem and amps didn't have limiters so clipping was clipping.

 

Hmm... so reading between the lines ( ) - I take it you don't recommend this approach eh?

[agedhorse]

Hmm... so reading between the lines ( ) - I take it you don't recommend this approach eh?

 

Definately not. It's just wrong on all levels.

[RoadRanger]

Definately not. It's just wrong on all levels.

Pun intended ?

[TimmyP]

 

There would be distortion, not compression with the drive electronics overdriving...... But for the bottom-feeders, it's a good point.

 

 

A distorted signal has a greatly reduced dynamic range as compared to the same signal with no clipping. As such, in addition to being distorted, it is compressed. This is what gives distorted guitars their sustain.

[TimmyP]

 

Definately not. It's just wrong on all levels.

 

 

I agree, yet it is the method that is taught in every gain staging article I've read.

[NUSound]

 

No even close. The board's an M7CL. It has good always visible metering for every input & output.

 

 

Disagree. I love my M7, but the metering is horribly imprecise... only 7 LED's per channel. Not good enough to setup PFL gain. Which we can probably agree isn't an issue since all you have to do to bring up a full meter is hit cue.

 

That having been said, the metering is certainly good enough to recognize any channels that are clipping or something completely out-of-whack.

[agedhorse]

 

A distorted signal has a greatly reduced dynamic range as compared to the same signal with no clipping. As such, in addition to being distorted, it is compressed. This is what gives distorted guitars their sustain.

 

 

I was referring specifically to (delta G) linear compression, not to dynamic range reduction due to clippin, which is generally not desireable in pro audio equipment and operation.

[agedhorse]

 

Disagree. I love my M7, but the metering is horribly imprecise... only 7 LED's per channel. Not good enough to setup PFL gain. Which we can probably agree isn't an issue since all you have to do to bring up a full meter is hit cue. That having been said, the metering is certainly good enough to recognize any channels that are clipping or something completely out-of-whack.

 

 

7 LEDs are IMO plenty precise for setting up PFL gain and everything else for that matter. All the important information is there, what is necessary is where you are operating relative to your nominal and maximums. Who cares what's happening in between, something around -40 or so, -20, -6, nominal 0, +6, +12, +18 and actual clipping (could be anything between +22 and +26 on a pro console). Now depending on the level structure of the comsole, this might be scaled to the operating levels specific to the hardware but the concept is identical.

[agedhorse]

 

I agree, yet it is the method that is taught in every gain staging article I've read.

 

 

Not every article I have seen but I agree that certainly enough that it's disturbing. It's good practice up to the power amplifier or speaker processing (with limiting) stages but there must be a difference at that point to allow the limiters to work. You can't have delta G without headroom. You can't subtract something from nothing.

 

As a designer of the equipment being discussed, these guys have no idea what they are talking about. Trust me, it's an interesting conversation where they are telling me their interpretation of how MY limiters and gain structure of MY products work... without their having even the slightest idea what's really going on inside my boxes. It's incredulous, they really show their ignorance every time they open their mouths. General rule of thumb... if somebody doesn't know what they are talking about, keep they should keep their mouth shut. That way, at least the doubt will remain, but when they open their mouths, all doubt is removed.

[MrJoshua]

I generally try to set things so that the clip limiters on the amps start to operate when I have around 10dB of headroom left on the mixer ... but I'm running a pretty simple setup, just a MixWiz into QSC RMX and PL2 amps, so while I don't know that it's the RIGHT way to do things, it made sense to me to have some headroom left on the mixer so I don't have to worry about clipping something there before the amps are loaded. It isn't the absolute maximized lowest-noise gain structure, I suppose, but with a little care in the setup to keep power cables away from signal cables, etc, the noise in my system is normally very low and certainly at acceptable levels for the bar gigs I do.

[agedhorse]

I generally try to set things so that the clip limiters on the amps start to operate when I have around 10dB of headroom left on the mixer ... but I'm running a pretty simple setup, just a MixWiz into QSC RMX and PL2 amps, so while I don't know that it's the RIGHT way to do things, it made sense to me to have some headroom left on the mixer so I don't have to worry about clipping something there before the amps are loaded. It isn't the absolute maximized lowest-noise gain structure, I suppose, but with a little care in the setup to keep power cables away from signal cables, etc, the noise in my system is normally very low and certainly at acceptable levels for the bar gigs I do.

 

IMO, it's just about the perfect practical tradeoff.

[MrJoshua]

IMO, it's just about the perfect practical tradeoff.

 

It just seemed to make sense to me. I'm able to go right up to limiting on my amps without driving any of the rest of my system into clipping, leaving plenty of headroom while still keeping an acceptable noise floor. And if I need to be louder than that, I'm not using a big enough system anyway.

[agedhorse]

You can also drive your system even 6dB into limiting on peaks and still be ok throughout, provided you are somewhat conservative on sizing your amps to your speakers.

[MrJoshua]

 

You can also drive your system even 6dB into limiting on peaks and still be ok throughout, provided you are somewhat conservative on sizing your amps to your speakers.

 

 

I power my speakers at RMS rating. I figure if I need to be louder than that, I need more speaker anyway. (OK, my subs are actually at about 109% of RMS, 1300W into a 1200W RMS speaker, but that's pretty close.)

[agedhorse]

 

I power my speakers at RMS rating. I figure if I need to be louder than that, I need more speaker anyway. (OK, my subs are actually at about 109% of RMS, 1300W into a 1200W RMS speaker, but that's pretty close.)

 

 

So IMO, you are ok hitting the limiters which will raise your average SPL level without the risk of mechanical damage. Thermal failure is not common in pro audio where clipping does not occur and powering at or below the RMS level with limiters set and only 6dB of limiting.

[Tomm Williams]

 

Not every article I have seen but I agree that certainly enough that it's disturbing. It's good practice up to the power amplifier or speaker processing (with limiting) stages but there must be a difference at that point to allow the limiters to work. You can't have delta G without headroom. You can't subtract something from nothing. As a designer of the equipment being discussed, these guys have no idea what they are talking about. Trust me, it's an interesting conversation where they are telling me their interpretation of how MY limiters and gain structure of MY products work... without their having even the slightest idea what's really going on inside my boxes. It's incredulous, they really show their ignorance every time they open their mouths. General rule of thumb... if somebody doesn't know what they are talking about, keep they should keep their mouth shut. That way, at least the doubt will remain, but when they open their mouths, all doubt is removed.

 

 

For those of us who don't know much about your background. (and I'm certainly one) would you mind sharing some of the devices you've engineered? This is quite impressive.

[agedhorse]

One of the most recent products was the Genz Benz ShuttleMax 12.0 bass amp, which was voted best musical instrument amp of 2009 by Music & Sound Retailer magazine.

 

I have been (and still am) the Product Development engineer for Genz Benz for about 15 years, have done project work for many other manufacturers, most of which I can't discuss due to non-disclosure agreements. Plenty of power amplifier and speaker products, including 3 different line array designs using advanced DSP techniques. I have my EE degree with my field of study being primarily analog controls and power systems. I am doing a lot of work with switchmode power supplies and class D amps now that lightweight is being taken seriously. FINALLY!!!

 

Here's a picture of a monitor console I hand-built ~25 years ago:

[Dogoth]

 

Disagree. I love my M7, but the metering is horribly imprecise... only 7 LED's per channel. Not good enough to setup PFL gain. Which we can probably agree isn't an issue since all you have to do to bring up a full meter is hit cue. That having been said, the metering is certainly good enough to recognize any channels that are clipping or something completely out-of-whack.

 

 

Sure it's not like the metering on the input side of Soundforge that has resolution as fine as your monitor will allow and goes down to -90 but it's plenty fine for being "in the ball park" which is what live sound is all about. That last db or two of noise isn't going to matter but for an acoustic jazz trio, 20db will. It's accurate enough. and as you pointed out, the cue metering IS plenty precise, if you need it.