Recording fireworks

[egproductions]

I tried recording fireworks with my h4n yesterday. I had the recorder set to manual gain and set the recording level so that that fireworks wouldn't over modulate. I found after that that volume difference between the fireworks and the general audio of people is so great that unless I have the speakers up very loud to emulate the actual loudness of the event, the audience audio is lost. What are some best practices for recording something like this or something similar such as a live concert or wedding dj? Is auto gain or setting a peak limited and exposing for the crowd something that should be considered? I've always been under the impression that manual audio gain is the preferred method. How would some of you approach this?

 

[Rippie]

So you're asking how to capture both the fireworks AND the audience?
You use multiple microphones pointed at the things you want to pick up.

 

[egproductions]

I'm trying to get an accurate representation of how the human ear hears the fireworks. The h4n has stereo mics with a 180 degree pickup pattern. I imagine I need something like an hdr for images. Autogain or a limiter would seem to to bring everything closer to range so that I get both detail in the quit and the loud tones. A microphone pointed at a crowd and levels adjusted to hear their audio would get over modulated by the audio of the fireworks because of the loudness. I basically want oos and ahs as well as good quality fireworks.

 

[Rippie]

You want an "accurate representation" yet you want to compress the hell out
of the signal to achieve a balance?

Many years ago back when I was doing a lot of production truck work for 5 or 6 years in a row I traveled to do a very large fireworks show
for a network.

I actually minimized the crowd noise because I had music tracks, talent, fireworks..etc
but throwing a few mics in the crowd would hardly be a big deal.

You choose the right microphone for the job AND
you MIX those microphones together to CREATE the illusion of
what the human ears can hear on their own.
Compressing a particular "source" is different than just slamming 1 compressor over
everything all at once.

Just as the human eye can adjust far better to extreme brightnesses compared
to a camera lens your ear is going to do the same.
Throwing ND on a window to create the illusion in camera of being able to see inside
and out is the same kinda thing as doing a simple mix.

By the sound of your reply you already have what you want to do worked out
so why bother asking the question.
If you have your mind set on only using internal mics on that recorder there really isnt much you can do
other than experiment with mic/recorder placement.

 

[Noiz2]

First off you should realize your chasing a myth. You can't accurately represent how the human ear hears in a recording for a number of reasons. First there is no "human ear", there are millions of different ears. They all hear differently, depending on age, damage, shape etc. And you don't "hear" with your ears. Your ears collect information that your brain processes and that is what you perceive as hearing. Your going to argue that point but the fact is different people with the same ear quality hear things differently. So trying to do one recording that sounds like "people hear it" is simply not possible. You could try to record the environment that people are listening too accurately. But that is also not really possible because it would take a huge system to have the dynamic range of life and even if you got there you still wouldn't "be" there because of that pesky brain. A listener isn't just hearing things they are experiencing things. They see the flash and there brain kicks in it's compressor because it knows that there is a BIG bang coming. Part of the reason loud noises that are a surprise sound SO MUCH louder than the same sound played back when you know it's coming.

Those stupid "head shaped" surround microphones are based on completely flawed logic. The only way they would work at all would be IF you had head phones that mimicked the mic placement. The closest you can get to sort of capturing a "I'm there" sound is to put microphones where your speakers will be. That will get a fairly accurate acoustic reproduction. For the perceptual part you need to make some choices. If someone is really into the fireworks the crowd is going to way back in the mix. If the fireworks are just background and your faux listener is into the crowd then the crowd is going to be up and not the fireworks. In the real world that is a constantly changing thing. Have an interesting fireworks, or one you really like and you will focus on it. If it's one you find kind of boring you may instead try to hear what those folks over there are talking about. If you get into that and they shoot off a big one you didn't notice that BIG pop is going to startle you and sound HUGE.

Your problem with recording is that unlike you the mic has no brain to process incoming signals and the guy talking about the quality of toilet paper in the port-a-potty is just as important to it as the fantastic firework that just went off.

I've recorded to get the explosions and I've recorded to get the people (figured it would be good for crown disaster kinds of stuff - the crown sounded panicky, though of course they weren't). A good fast limiter is your friend. 24 bit is also good. A mic with a high SPL ratting is a must.

Decide on the perception your going for and try to record that. Chasing the myth of "real" will just get you frustration.

I'm trying to get an accurate representation of how the human ear hears the fireworks. The h4n has stereo mics with a 180 degree pickup pattern. I imagine I need something like an hdr for images. Autogain or a limiter would seem to to bring everything closer to range so that I get both detail in the quit and the loud tones. A microphone pointed at a crowd and levels adjusted to hear their audio would get over modulated by the audio of the fireworks because of the loudness. I basically want oos and ahs as well as good quality fireworks.

 

[Cassius]

The H4n is not a microphone. It is a recorder that happens to have microphones built in. They are rarely if ever the right ones for the job. The best way to do this is with multiple microphones and multiple tracks, pointed at different things and set at different levels.

That said, the above post is also accurate- psychology is at least as important.