Gain limiter question

Caruso wrote on 4/11/2004, 5:42 AM
First, a courtesy request in responding to my question – please don’t tell me to go out and buy anything – I’m working on a budget and am currently restricted to the equipment I own – I know it’s not state of the art, but has served me (and my family) well over the years – and saved (earned) us a lot of money that was put to use for music training/tuition/audition fees, etc.

When logistics allow, I lug one of my TEAC reel to reel machines to make live recordings of my son or daughter performing. In studio settings, I record directly to my hard drive (such bliss!).

When those solutions aren’t practical (my son recently performed in LA, I’m on the east coast), I carry what was a state of the art prosumer video cam that features an integral 8mm digital recording unit. It is relatively compact and does a great job – except!! – it has audio level circuitry that cannot be disabled. There are recording level dials that control levels when external mics are used (I always use external mics – two mono, one for each stereo channel), and these seem to lower the level, but I am confused as to how to overcome or avoid the effect of that damned limiting circuitry.

I recently recorded my son (a cellist) in a two-hour recital with piano. My strategy (I only guessed that it was right) was to lower the recording level to the point where (I hoped) the loudest passages would not invoke that limiting circuitry. The result is almost perfect – in two hours of recording, there are only two spots where extended silence followed by sudden loud piano attacks caused the limiting circuitry to kick in audibly.

I have corrected these two instances by altering the volume envelope – the result will be transparent except to the most discriminating audio engineer – and, perhaps, to the musicians themselves (perhaps, not necessarily – the result is pretty darned good, IMHO).

My question(s): Is my strategy correct, or is there some other technique by which to moderate the incoming signal so as to keep the cam from trying to protect me from saturation? If I raise the level, I’m know the result would be full of evidence of the machine groping for more signal in soft passages while compressing the signal in loud areas. Fortunately, none of that is present on my recording (except for the two instances of limiting cited).

Lastly, is there some software or physical wizardry that could be brought to bear upon this unit that would permanently disable that damned limiting feature? I’ve scoured the manual and have satisfied myself that there is no built-in switch or option to turn it off (such a shame – an illogical feature set for an otherwise amazing machine – who needs manual input control if it is compromised by auto-limiting circuitry?).

The unit is a Sony CCD-V220 and dates from the late ‘80’s.

When I can afford it, I’ll “upgrade” to some sort of portable digital piece (DAT, digital HD recorder, or a laptop). In the meantime, I’m limited to this piece.

Feel free to also comment on my corrective approach. What I did was to alter the volume envelope to eliminate the effects of the limiter by locating the point where it kicked in, increasing the volume at that point, then gradually reducing the volume to mimic what my ear hears as the natural decay that would occur from an acoustic piano. Is there any better way (a plug-in, perhaps) to accomplish this?

Thanks in advance for any replies.

Caruso

Comments

farss wrote on 4/11/2004, 7:35 AM
Given what you've got to work with I think you've done the best you can. I'd advise against disabling the limiter even if you could. You mightn't like what it's doing but you'll like even less what might happen without it.
Even though a piece of gear of that vinatge is probably not recording the audio digitally hitting saturation can give you audio that's nigh impossible to fix. It sounds like you've been able to correct the recording fairly well, something that's been clipped is much harder to correct.
The other issue you'll face is if you set the gain too low you'll be running into the noise floor so to some extent you're between a rock and a hard place.
As for what you've done to correct the limiting, well I'm far, far from being an audio geek but from the little I know I don't believe there's any magic solution. For anyhting to 'delimit' it'd need to be able to identify where the limiter kicked in and I don't see how anything could do that. There are compression / expansion systems but they worked as matched pairs, trying to decompress compressed audio without knowing the compression settings would take a fair amount of experience, something that's just been limited I'd imagine to be impossible.
As to using the volume envelope to restore the attack, only thing to watch for is having enough headroom to accomodate the extra level.
Cold wrote on 4/11/2004, 9:43 PM
You could experiment with an expander if available, set to where you hear the limmitter kick in and play with the settings to taste. Realistically, this may not be a better solution than what your doing unless the limmiter kicks in more than several times and you need to automate the fix.
Steve S.
gettig wrote on 4/12/2004, 6:27 AM
I don't really have anything to suggest, but I definitely offer kudos for maximizing the equipment you have at hand. No technology can replace thinking and experimentation like that.
risenwithhim wrote on 4/12/2004, 8:55 AM
Ditto what gettig said. Good job with what you're doing.
I'd suggest padding the mic even more than before (to prevent the AGC from kicking in), and then normalizing the audio when you bring it into Vegas to redeem the maximum level. Since you're capturing digital, a high noise floor (think analog tape hiss) with a lot of gain isn't really going to be a problem. Digital audio can be so forgiving with boosting low-level audio, as compared to analog.

Here's a super tip for optimizing your average levels in Vegas. If you normalize, and don't get much gain due to a spike at one or two places in the recording, drag the head or tail to get that spike trimmed out of the event and recalculate the normalization gain -- this time based on the normal "spike-free" dB levels of the event. Now drag the head or tail back out to make the event the original length. You'll see that the spike is now boosted off the scale, so just drop in a volume rubberband to pad that spot as needed.

Keep up the good work.
risenwithhim wrote on 4/12/2004, 9:18 AM
It also occured to me that the hard attacks after long periods of silence show that the AGC was being employed with the level that the mic was feeding it. The hard attack was probably preceded by a less-noticable gradual release. All this to say that turning the mic level down at the mic is the same thing the camera is doing at the AGC stage of processing. Doing it at the mic end will just result in more smooth, natural audio level throughout. In other words, as for the noise floor argument, it's really a moot point since ultimately the level going to tape will only be allowed to ceiling at a certain level (-12dBfs?) whether you control the level of the mic, or let the camera control the level with its unfriendly AGC. So keep experimenting with backing that mic level off until you find a level that the AGC will be less frightened by. Then, as I said, Vegas-normalize it in editing.

This is really the beauty of editing in Vegas. If your audio is going onto a digital medium, you can get away with capturing low-level audio (levels we're conditioned to stray from because of our analog sensibilities) because the levels can be redeemed with the real-time normalization without extraordinary noise floor problems.
Caruso wrote on 4/13/2004, 2:41 AM
I truly appreciate the compliments, and, of course, I value the advice. I hadn't thought to try attenuating the signal at the mic, and will re-read those suggestions - I'm a bit hazy on exactly how to go about what you're telling me to do.

Padding the mic - is that to say I should wrap something around it to reduce it's efficiency in picking up the signal? Would the AGC on my cam simply open up further?

Guess I'll have to run some experiments to really find out.

There are 12 gradations on the left/right input gain control where you plug the mic into the cam - #7 is colored green as Sony's suggested general use recording level. That definitely seems too high. I recorded this performance at #5 and still got one or two instances of AGC.

Perhaps the cam's gain control is even more limited than I thought - opening up completely in silence, no matter where you set the control knob - but, I'll have to experiment some more to know for sure.

I like the idea of limiting input at the mic - just need some detailed suggestions as to how you might go about that.

. . . and, of course, I fully agree that any digital format allows for more flexibility and less concern for the noise floor.

BTW, Risinwithhim, the problem spots involved areas where there were pauses in the music followed by loud entrances.

Thanks again to all for your valued input.

Caruso
risenwithhim wrote on 4/13/2004, 9:39 AM
"Pad", as in "attenuate". I was thinking you had said there was a way to do this on your mic -- to step down the signal level.

And now, to go against your pleas for us not to tell you to go out and buy something... When you can afford one on eBay or something, get a Nomad Jukebox 3 (20GB, $300 new - I got mine for $175 on eBay). It's built to be an mp3 jukebox, but it has a very good line input for recording directly to harddisk at 48kHz/16bit in .wav format. Then, you can hook up to PC via firewire or USB and dump files over fast! It's the size of a discman and it has a lithium ion battery that has a reliable 6-hour life-span per charge, plus another slot for an additional batt for double that. Or, you can run off AC. I use this for weddings and stuff where I get a board feed, and it is totally awesome. The firmware has a little meter on the LCD. It's not great, but it will show you where you're peaking, and that's what's important. Best of all, no AGC! Then if you really want to be the man, get one of the little Beachtek micro-mixers like this one: Linkie
And a nice pair of shotgun mics on booms. You could use that Beachtek device with the camcorder though (that's what they're made for), but it's kind of pricey. Still, you're mixing 2 discreet channels and supplying phantom power with a mere 9v battery supplying power. That, as opposed to buying one of those cute little Behringer mixers that fit in the palm of your hand, yet use a big-old AC plug to power them.
larry-peter wrote on 4/13/2004, 11:14 AM
I'm not familiar with your exact camera model, but many Sony prosumers of that era only had AGC enabled on the mic inputs. If you have line inputs you may find that an external mic preamp can be used to feed the line ins and you won't encounter the gain limiter circuit. Good Luck.
Cold wrote on 4/13/2004, 10:05 PM
Be aware of cutting your input level too much. While your noise floor may be low, the resolution of your recording will drop significantly. As a test, record at a lower level, say 3 or 4 on your camera and also record a chunk at 6 or 7 with the same volume of source material. now drop these both into vegas, normalize them both, and zoom in hard on the wave form. You should notice a large degree of detail missing from the lower level recording. listen to them both and see if you can live with this. Trying the line inputs, as atom12 suggested, is a good idea if you have anything to use as a pre, such as a small board. This may be your best solution.
Steve S.
Caruso wrote on 4/14/2004, 3:16 AM
Atom:
Interesting suggestion - I do have line inputs and will have to check out your suggestions.

Thanks.

Caruso