Old 08-01-2006, 07:24 AM   #1
lbass
Mortal
 
Join Date: Aug 2006
Posts: 4
Default Automatic Gain Control (AGC)

Is it possible to get an AGC facility in Reaper?
It would be VERY nice to have it apply to individual channels and/or on a global basis(master). If it was applied to channels it would be GREAT if the controls could be "ganged" so that the relative aamplitude of the ensemble channels is maintained.

This product is already outstanding, this would make it by far best in class.

Thanks.
LBass
lbass is offline   Reply With Quote
Old 08-01-2006, 07:45 AM   #2
maa
Mortal
 
Join Date: May 2006
Location: Hamburg
Posts: 1,262
Default

Where did you get that idea from ?
Auto Gain Controll is not exactly becomming of high quality digital audio. Maybe you're in Radio ? Radio compressors have such leveling features. A compressor which has a release time upwards of two seconds would probably work for you - right?

Cheers

M
maa is offline   Reply With Quote
Old 08-01-2006, 08:01 AM   #3
Spon
Mortal
 
Join Date: Jun 2006
Posts: 447
Default

This is already available in REAPER - after a fashion.

ACG is compression with a slow attack time and SLOW release (and fancier stuff, if you like...) - some compressors do it better than others. You can use a plugin, in channels and/or master.

Anyone willing to recommend favorite compressors for AGC? Anyone here ever use such a thing?

ReaComp's settings limits may be marginal for AGC, but it's free, already in Reaper, and is a nice compressor otherwise - try attack, release, knee up all the way, then tweak everything.

http://www.reaper.fm/~pipeline/REAPERVideos/basics/7 ReaComp Basics 3bm.wmv

In addition to radio/tv/film users, I've personally used ACG-type compression (though not recently) in composing/arranging for live instruments. Sometimes you'd rather the section just hear themselves right now than set up the nice recording you'd use for a finished piece. You can change instruments, move people, run over to the piano, etc. Handy tool for band teachers too. (Do band teachers use DAWs yet?)

But no, probably not a default channel insert.
Spon is offline   Reply With Quote
Old 08-01-2006, 08:28 AM   #4
Rednroll
Mortal
 
Join Date: Jan 2006
Posts: 1,021
Default

Open in Sound Forge....goto Process>Normalize, chose a Normalize RMS setting that works best for you and save it as a preset.
Rednroll is offline   Reply With Quote
Old 08-01-2006, 08:58 AM   #5
Spon
Mortal
 
Join Date: Jun 2006
Posts: 447
Default

Fine, if A. you own Sound Forge, and B. you don't want real-time (i.e. AGC in recording, which is where it is usually used).

With a 24bit soundcard, though, truth is you don't really need AGC, as Justin pointed out somewhere else (the 192K thread, I think). You can just compress afterwards, and the musical data is still there. Can't necessarily do that with 16bit.

There is broadcast sorts of things where you might want a nice leveling compressor on the mains, though, if you use REAPER as a live mixing board.
Spon is offline   Reply With Quote
Old 08-01-2006, 10:04 AM   #6
pipelineaudio
Mortal
 
pipelineaudio's Avatar
 
Join Date: Jan 2006
Location: Wickenburg, Arizona
Posts: 6,225
Default

hey, just in case noone's noticed, ReaComp's RMS window time has been upped to a maximum of 1 second. This may help quit a bit for having a sensible analysis for an AGC circuit
pipelineaudio is offline   Reply With Quote
Old 08-01-2006, 11:42 AM   #7
lbass
Mortal
 
Join Date: Aug 2006
Posts: 4
Default AGC request and need

Thanks ALL...

I anticipate using REAPER as the recorder in a remote (unattended and unwatched) site. As no one will be there to compensate when the sound levels get much higher or lower than expected, some type of AGC is required as there is NO correction for the clipping that occurs during overload. If we set the gain to allow for empassioned screams and shouts, the calmer passages (almost) fall into the noise or sound like they are in a big cavern.

I will try the compression schemes to see which one in REAPER add the least amount of color.

I still think this a good feature.
lbass is offline   Reply With Quote
Old 08-01-2006, 12:05 PM   #8
olzzon
Mortal
 
olzzon's Avatar
 
Join Date: Jul 2006
Location: Europe
Posts: 705
Default

Quote:
Originally Posted by lbass
Thanks ALL...

I anticipate using REAPER as the recorder in a remote (unattended and unwatched) site. As no one will be there to compensate when the sound levels get much higher or lower than expected, some type of AGC is required as there is NO correction for the clipping that occurs during overload. If we set the gain to allow for empassioned screams and shouts, the calmer passages (almost) fall into the noise or sound like they are in a big cavern.

I will try the compression schemes to see which one in REAPER add the least amount of color.

I still think this a good feature.
Buy an old casette recorder with this feature and route this to your computer.
TO DO THIS, IT MUST HAPPEN IN ANALOG WORLD, AND REAPER HASNīT STEPPED OUT OF THE COMPUTER YET. :-)
__________________
Light travels faster than sound. Thatīs why some people appear bright until you hear them speak

Kasper Anhøj Olsson
olzzon is offline   Reply With Quote
Old 08-01-2006, 12:13 PM   #9
Youn
Mortal
 
Youn's Avatar
 
Join Date: Jun 2006
Posts: 999
Default

Yeah if you want to avoid clipping the A/D converters then that has to happen analog. You could slap a brick-wall limiter on it, crank the gain up so the quite stuff gets in clear, then adjust the limiter's release.
Youn is offline   Reply With Quote
Old 08-01-2006, 01:03 PM   #10
sstillwell
Mortal
 
Join Date: Jul 2006
Location: Cowtown
Posts: 1,064
Default

Bingo.

All the compression in the world won't help you if you've already clipped the A/D. You need a hardware AGC (or at the least a brickwall limiter) ahead of the A/D to avoid that.

Scott

Quote:
Originally Posted by Youn
Yeah if you want to avoid clipping the A/D converters then that has to happen analog. You could slap a brick-wall limiter on it, crank the gain up so the quite stuff gets in clear, then adjust the limiter's release.
sstillwell is offline   Reply With Quote
Old 08-01-2006, 02:45 PM   #11
Rednroll
Mortal
 
Join Date: Jan 2006
Posts: 1,021
Default

Quote:
Originally Posted by lbass
Thanks ALL...

I anticipate using REAPER as the recorder in a remote (unattended and unwatched) site. As no one will be there to compensate when the sound levels get much higher or lower than expected, some type of AGC is required as there is NO correction for the clipping that occurs during overload. If we set the gain to allow for empassioned screams and shouts, the calmer passages (almost) fall into the noise or sound like they are in a big cavern.

I will try the compression schemes to see which one in REAPER add the least amount of color.

I still think this a good feature.
Exactly what everyone else said....Reaper is Post A/D converter, which means this would have to happen prior to the signal hitting the input on your sound card. You should look into purchasing a hardware Limiter/Compressor, for what you're trying to do.
Rednroll is offline   Reply With Quote
Old 08-01-2006, 03:24 PM   #12
Art Evans
Mortal
 
Join Date: Feb 2006
Posts: 6,582
Default

http://www.soundalot.com/freeware/record/record-b.html (ALC Record near the top of the page) might be of interest - depends on your soundcard I think.
Art Evans is offline   Reply With Quote
Old 08-02-2006, 12:27 AM   #13
Panatrope
Mortal
 
Join Date: Jun 2006
Location: Melbourne, Australia
Posts: 25
Default A broadcaster's view

For location recording, a common strategy is to record without compression and high headroom on multitrack, but also to produce in real time an air feed or a safety/check/audition CD copy that is air-ready, most likely using an external CD burner, fed either analog or digital.

In these circumstances it would be useful to be able to put some real-time compression in the master out, using a moderate slope (1.5:1 seems to work well) and gain make-up so that dynamic range of 30dB or more and headroom of -6dB is converted to a dynamic range of 20dB and a peak limit of -1dB. It assumes you are using good 24bit ADCs. The compresser limit should have suitable time constants and in fact have a bit of look-ahead to minimise overshoot. Better than having to manually gain-ride the output when there are many other things on your mind? Anyone know of a suitable plugin.

Oh and while you're at it, it needs to be followed by a suitable dither and reduction to 16 bit if you are burning a CD.

The alternative is to shoot it out of the SPDIF output on your sound card and pump it through something like the TC Electronic M350, then into the CD ....

You'll have to excuse me. I'm a retro broadcast engineer who still tends to trust dedicated hardware, rather than something that has Bill Gates' hands all over it ...
__________________
Pan: (Gr) all, everything
Tropos: (Gr) turn.
Loose translation: all rounder!
Panatrope is offline   Reply With Quote
Old 08-02-2006, 02:46 AM   #14
malcolmj
Mortal
 
malcolmj's Avatar
 
Join Date: Jan 2006
Location: Australia
Posts: 1,613
Default

Quote:
Originally Posted by Panatrope
In these circumstances it would be useful to be able to put some real-time compression in the master out, using a moderate slope (1.5:1 seems to work well) and gain make-up so that dynamic range of 30dB or more and headroom of -6dB is converted to a dynamic range of 20dB and a peak limit of -1dB. It assumes you are using good 24bit ADCs. The compresser limit should have suitable time constants and in fact have a bit of look-ahead to minimise overshoot. Better than having to manually gain-ride the output when there are many other things on your mind? Anyone know of a suitable plugin.
What about trying ReaComp?

Cheers,

Malcolm.
__________________
Documentation
Videos
malcolmj is offline   Reply With Quote
Old 08-02-2006, 06:02 AM   #15
lbass
Mortal
 
Join Date: Aug 2006
Posts: 4
Default New views/data

All of the comments and suggestions have been most helpful.

It has come to my attention that Samson now sells a USB-microphone that attaches directly to a PC/MAC. I know, I know, Samson does NOT make THE BEST mics, BUT, with this device "the digital domain has now breached the analog shores"!

As the mic's built-in amp gain is directly under software control, it should be possible to directly implement an elegant "analog" AGC. If REAPER could start to decrement the gain based on a settable attach rate when a channel reaches a set threshold and restores the gain after the channel "settles" based on a decay rate, I think an elegant solution could be had. If I'm not mistaken, this is just the approach that is required as the analog signal (taken at the source-the MIC) is never clipped and thus the digital samplers never see overload.

Is this possible/desireable?
__________________
LB
lbass is offline   Reply With Quote
Old 08-02-2006, 07:02 AM   #16
Art Evans
Mortal
 
Join Date: Feb 2006
Posts: 6,582
Default

That mic is a pretty dodgy bit of kit. A colleague bought one and was I think rather disappointed - the hiss was very noticable.

If you have good quality mic and preamp (if required) and soundcard, then you simply set it so that the loudest possible sound cannot clip, then use compression afterwards (or while monitoring) to squash the dynamic range. Effectively you will be turning up the quiet bits a lot, but if the kit is good you won't hear significant amounts of noise. If the kit is not good (eg Sampson mic) you will. Doesn't matter how you do it, you can't get around the fact the system has a certain signal to noise ratio and compression will expose it, whether hardware or software.
Art Evans is offline   Reply With Quote
Old 08-02-2006, 07:30 AM   #17
lbass
Mortal
 
Join Date: Aug 2006
Posts: 4
Default Thanks. Advice taken.....

Quote:
Originally Posted by Art Evans
That mic is a pretty dodgy bit of kit. A colleague bought one and was I think rather disappointed - the hiss was very noticable.

I guess I've always suspected that there was no better solution available. I'll use our MOTU and Sennheisers as you described and get on with it....

Thanks.
__________________
LB
lbass is offline   Reply With Quote
Old 08-03-2006, 11:11 AM   #18
chip mcdonald
Mortal
 
chip mcdonald's Avatar
 
Join Date: May 2006
Posts: 1,927
Default

Actually, though...


.... if someone were to come out with a high-speed Firewire spec (ahem, Justin) that would enable software control of a preamp, that would be a very cool thing. USB won't cut it no matter how it's configured, the way USB works isn't suited to pro audio.

BUT, a software controled preamp tied to "knowing" the *exact* clipping point of your A/D converter would be very cool. It would be like having the most transparent compressor on the planet, plus you could make it do things like overdrive the frontend in certain ways without clipping your converters...

Seems like something Rick Naqvi/Presonus would possibly be into, or maybe McQuilken/RNP. It'd be cool to have a RNP (I think he uses a DSP in the RNC) controlled by Reaper to work on a LA2 sort of ballistics but without the extra stage... You'd still have to worry about overs, but it would be super clean and the gain staging getting to your A/D converter would be great.
__________________
]]]]]]]]]] www.chipmcdonald.com [[[[[[[[[[[
note: for future references, anything and everything I say may be considered OPTIONAL. Thanks (the management)
chip mcdonald is offline   Reply With Quote
Old 08-03-2006, 11:32 AM   #19
olzzon
Mortal
 
olzzon's Avatar
 
Join Date: Jul 2006
Location: Europe
Posts: 705
Default

Quote:
Originally Posted by lbass
All of the comments and suggestions have been most helpful.

It has come to my attention that Samson now sells a USB-microphone that attaches directly to a PC/MAC. I know, I know, Samson does NOT make THE BEST mics, BUT, with this device "the digital domain has now breached the analog shores"!

As the mic's built-in amp gain is directly under software control, it should be possible to directly implement an elegant "analog" AGC. If REAPER could start to decrement the gain based on a settable attach rate when a channel reaches a set threshold and restores the gain after the channel "settles" based on a decay rate, I think an elegant solution could be had. If I'm not mistaken, this is just the approach that is required as the analog signal (taken at the source-the MIC) is never clipped and thus the digital samplers never see overload.

Is this possible/desireable?
Thereīs no way you can avoid clipping this way.
The computer can not turn down the volume before it know that itīs to high. So this solution is a no go.
Why donīt you just buy some cheap behringer compressor and use that before the converters.
__________________
Light travels faster than sound. Thatīs why some people appear bright until you hear them speak

Kasper Anhøj Olsson
olzzon is offline   Reply With Quote
Old 08-03-2006, 03:08 PM   #20
Spon
Mortal
 
Join Date: Jun 2006
Posts: 447
Default

If you're *really* recording (A/D as well as file) at 24 bits you can record from 110db down to the Brownian motion of the air bouncing off the mic diaphragm. Isn't this enough?

This is from "gunshot loud" to "dig it out of the noise via fancy DSP". Unless you have to record 180db explosions cleanly, you shouldn't need a compressor.

Maybe more mics, or closer micing, so there is more near-field and less reverb recorded? BTW it sounds like you have possibly the oddest use of REAPER ever. What do you plan to do with these recordings, if I may ask?
Spon is offline   Reply With Quote
Old 08-03-2006, 07:33 PM   #21
Youn
Mortal
 
Youn's Avatar
 
Join Date: Jun 2006
Posts: 999
Default

Quote:
Originally Posted by olzzon
Thereīs no way you can avoid clipping this way.
The computer can not turn down the volume before it know that itīs to high. So this solution is a no go.
Why donīt you just buy some cheap behringer compressor and use that before the converters.
LOL yeah some behringer will really help with the noise!
Youn is offline   Reply With Quote
Old 08-04-2006, 05:14 AM   #22
willowhaus
Mortal
 
Join Date: Jun 2006
Posts: 315
Default

Quote:
Originally Posted by olzzon
Why donīt you just buy some cheap behringer...
Sorry, you must now be taken outside & beaten severely.
willowhaus is offline   Reply With Quote
Old 08-04-2006, 05:34 AM   #23
Rednroll
Mortal
 
Join Date: Jan 2006
Posts: 1,021
Default

Hey!!! Some of my Behringer equipment isn't noisy.......at least not when I have the power switch in the OFF position. LOL!!!
Rednroll is offline   Reply With Quote
Old 08-04-2006, 06:54 AM   #24
Art Evans
Mortal
 
Join Date: Feb 2006
Posts: 6,582
Default

Well.... I bought a cheap Behringer mixer today for a particular purpose, and connected a reasonable high-output low noise condenser mic to it, and was frankly impressed. At any realistic level monitoring through DT770 cans, I couldn't hear any noise at all. I know they have a reputation for lack of excellence but I do wonder whether it's partly internet folklore. People do speak well of their 8 channel mic pre for adats, and I imagine the mic preamps in that are shared with other models.
Art Evans is offline   Reply With Quote
Old 08-04-2006, 07:03 AM   #25
Panatrope
Mortal
 
Join Date: Jun 2006
Location: Melbourne, Australia
Posts: 25
Default

The Samson C01U is a great concept executed too cheaply. 13mm back electret, inadequate sensitivity and too high a noise for interview work. OK for close mic'd vocals and loudish instruments. But there are others on the horizon.

For other digital mics see Neumann Solution D and a VERY sophisticated gain-ranging ADC. For compressors, the issue is always overshoot and/or distortion of the rising waveform. For an elegant olde-worlde analog solution, see the BBC limiter from the late '60s (DEL Shorter et al) which had a 23 section passive delay line (2 or 3 ms) prior to the control element. By the time the peak hit the control element, the sidechain had already calculated the gain reduction required and had time to implement it inaudibly. In the digital world, you just need to fill a lookahead buffer, assuming the chain has good quality ADC so that you can leave enough headroom.

For AGC, which tends to take a longer term view of the world compared to a peak limiter, the length of the buffer would need to be seconds rather than ms. At which point it is probably better to use an experienced hand on a fader (and the other hand turning the pages of the score) ...

Now what was the question?
__________________
Pan: (Gr) all, everything
Tropos: (Gr) turn.
Loose translation: all rounder!
Panatrope is offline   Reply With Quote
Old 08-04-2006, 08:02 AM   #26
kejkz
Mortal
 
kejkz's Avatar
 
Join Date: Jan 2006
Location: Belgrade, Serbia
Posts: 226
Default

Quote:
Originally Posted by Art Evans
Well.... I bought a cheap Behringer mixer today for a particular purpose, and connected a reasonable high-output low noise condenser mic to it, and was frankly impressed. At any realistic level monitoring through DT770 cans, I couldn't hear any noise at all. I know they have a reputation for lack of excellence but I do wonder whether it's partly internet folklore. People do speak well of their 8 channel mic pre for adats, and I imagine the mic preamps in that are shared with other models.
I have a behringer cheap small mixer ub 802, and their condenser mic B1. It's a great sound for that money, and for my purpose of ocasionall demo recordings a real steal. Behringer is far from bad, it's just not the top range, but their devices just do the job every time...
kejkz is offline   Reply With Quote
Old 08-04-2006, 10:46 AM   #27
willowhaus
Mortal
 
Join Date: Jun 2006
Posts: 315
Default

Actually, aside from the dodgy quality the biggest beef I have is their blatant ripoffs of various audio products. Many of the things they build are cheap knockoffs, and sometimes they don't even bother to change the layouts - it's completely obvious as to what product they "borrowed" from. I just don't think it's right to support that kind of behavior.
willowhaus is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 10:37 AM.


Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2010, Jelsoft Enterprises Ltd.