Topic: Dealing with pyrotechnics and explosions while recording shows

[bluntforcetrauma]

I know of a show that does have a bang or too in the performance that happens to clip the recording. Its only the explosion that clips the sound.  Especially recording seemless with the explosion clipping makes for a normalizing nightmare in post.  Since the normalization will occur in reference to the loudest part--the explolsion .The music is at a good level.

How do you handle that?

please advise

[vanark]

I've never had explosions, but audience applause has done the same on some acoustic sets I've recorded.  I take another section of the recording that appears to be the loudest (I might check a few spots) and see what kind of amplification I need to get it up to -1 dB.  Then I amplify the entire recording, allowing the spikes to clip.  If it is a single instance, I may just amplify before and after the loud noise.

[bluntforcetrauma]

I've never had explosions, but audience applause has done the same on some acoustic sets I've recorded.  I take another section of the recording that appears to be the loudest (I might check a few spots) and see what kind of amplification I need to get it up to -1 dB.  Then I amplify the entire recording, allowing the spikes to clip.  If it is a single instance, I may just amplify before and after the loud noise.

would you normalize? can you normalize everything but the one boom?

its a loud rock show, no acoustic or audience is as loud as the music or the boom.  The recording is in safe parameters, just a big boom.

[vanark]

would you normalize? can you normalize everything but the one boom?

I amplify, not normalize.  By amplifying, I determine how many dB to increase the sound.  Same result, just a bit different method.  Normally, I can select the entire file and amplify to a peak of -0.1 or -1.0 or whatever - this gives you what a normalize function would.  But, with the spikes, I can't use the max peak level and have to specify the actual dB amplification I want applied.

its a loud rock show, no acoustic or audience is as loud as the music or the boom.  The recording is in safe parameters, just a big boom.

No different than setting levels for acoustic music and enthusiastic applause spiking and clipping.  It is all relative to the recording level of the music you are trying to capture.

[Brian Skalinder]

Rather than simply amplifying the whole waveform, and therefore clipping even further the spike(s), consider clip restoration (a feature offered in some audio editors) and then compression or a volume envelope to bring down the level of the spike before peak normalizing.  Lots of threads on compression and/or volume envelope in the Computer Recording forum - try a search, and if you still have questions, ask away.

[rokpunk]

can't you just select and compress the offending "BOOM"?

[jamroom]

Rather than simply amplifying the whole waveform, and therefore clipping even further the spike(s), consider clip restoration (a feature offered in some audio editors) and then compression or a volume envelope to bring down the level of the spike before peak normalizing.  Lots of threads on compression and/or volume envelope in the Computer Recording forum - try a search, and if you still have questions, ask away.

Thanks Brian, I had thought about this and it's useful info. +t

[mmadd29]

Why not just grab the waveform of the explosion and hit it with a -3 or so deamp to bring it down to the level of the music.

I've never done a show with explosions, but have had issue when someone drops a mic or guitar cord, and it makes a large pop.  I just deamp and then normalize and everything seems good.

[BC]

Rather than simply amplifying the whole waveform, and therefore clipping even further the spike(s), consider clip restoration (a feature offered in some audio editors) and then compression or a volume envelope to bring down the level of the spike before peak normalizing.  Lots of threads on compression and/or volume envelope in the Computer Recording forum - try a search, and if you still have questions, ask away.

agree, my 1st thought was to compress the explosion down to below the level where the music peaks then go and do your normalizing.

[DSatz]

I suspect that what you really want is limiting, not compression. My suggestion would be to record the shows with 24-bit settings that allow you to capture the peaks without distortion. Then when you get back home, use a limiter effect in your editing software, with very short attack and release times, so that the explosions are a few dB above the loudest musical levels, but still not quite clipping. That way the explosions will still be startling, but they won't accidentally damage your loudspeakers or anyone's hearing.

The main differences between a limiter and a compressor are that (a) a compressor may raise gain to boost the quiet parts of a program (in addition to reducing the gain during the louder parts), and (b) in general, compressors adjust gain slowly while limiters adjust it more rapidly--sometimes almost instantaneously, depending on how you set their controls.

Limiters are normally used to duck the levels of momentary peaks so as to prevent overload. They're like a backstop--normally they're set to have no audible effect until/unless something really requires the intervention. Compressors on the other hand are used to maintain a relatively constant long-term average level, so that listeners don't constantly feel that they have to raise and lower their volume controls. A compressor may well be varying its gain all the time in normal use. Compressors and ALCs (automatic level controls) are close relatives.

My local NPR FM station (WNYC) abuses its compressor badly, flattening out the dynamics of all the music they play. Judging by ear I would estimate that their compressor reduces the gain of the highest-level ignals by as much as 15-20 dB. Imagine a Beethoven string quartet or symphony in which the pianissimo passages and the fortissimo passages are about equally loud--that's what comes out. If they play Ravel's Bolero, the orchestra sounds as if it's getting farther and farther away the whole time, since the full orchestra can hardly get any louder than ostinato percussion part and the flute solo with which the piece begins.

This is NOT the effect you want on your "two buss" (the main, mixed-down stereo channels) for music recording! And it has almost nothing to do with avoiding overmodulation of their FM signal (FCC regulations)--I'm sure they have a fast-acting limiter for that purpose.

--best regards

[guysonic]

Recorded many explosive sounds like close-up/distant thunder and fireworks.  Usually best to set REC level for what's important just before/after the impact letting the short duration impulse clip without any limiter or compression. 
Later in post you can use clip restoration like Brian suggests or not as if recorded close up will sound the same regardless if being actually clipped or dropped in level in post. 

This is because there is mostly NO SOUND inside near position short duration detonation, just the sound before, and the ambient reverberation immediately after the impulse that gives all the unique characteristics of what the detonation sounds like. 

If at some considerable distance from the source like that of distant thunder, then multiple sounds of a single impulse can arrive in short space of time from thermal layers and surrounding hillsides to give an audible character to the detonation's initial impulse and post clip reduction or setting a lower recording level (like with the rolling thunder recording and fireworks at the dock linked below) can help make recording impulse sounds more detailed.

Many recordings of fireworks, thunder, and other explosive sounds posted on page: www.sonicstudios.com/mp3.htm

[Petrus]

How about using a good mixer and limiter when recording a show with pyrotechnics? Like having a SD302 in front with 1:20 limiter set at -3 dBFS (+17 dBVU). With this setup you can ride levels quite high linearly, but with explosions or applause you have an extra 60 dB headroom before recorder clipping.