With my field recorder the loud noises I recorded, are loud but they are very quiet. I had to turn the trim and gain down, cause otherwise it would peak, and I remember exactly that it recorded in between 12 and 3, going up and down of course. But when I play it back, it's quiet. How do you tell if it's loud enough if the numbers don't necessarily mean that it is? Thanks.
What am I doing wrong when recording loud noises?
- Thread starter harmonica44
- Start date
Huh?
Please be very specific in excruciatingly boring detail.
What was being recorded? What type of mic? What recorder? What mixer? Gain staging levels?
What is the perspective with the rest of the material in the mix? What is it layered with? What are the levels? How hot is the rest of the mix? What processing/EQ are you using?
Can you post an example on SoundCloud of someplace similar?
Please be very specific in excruciatingly boring detail.
What was being recorded? What type of mic? What recorder? What mixer? Gain staging levels?
What is the perspective with the rest of the material in the mix? What is it layered with? What are the levels? How hot is the rest of the mix? What processing/EQ are you using?
Can you post an example on SoundCloud of someplace similar?
Last edited:
Recording mostly screams and loud thud type noises. I used a hypercardioid. Really loud screams, but it seems they are too quiet. I tried doing it louder, but it would just peak, and had trouble finding a good middle ground.
You still didn't answer:
You haven't gotten to the mix yet. If everything is loud, nothing is loud. If there is quiet before a loud sound it will seem louder. For instance, if you're spending a lazy afternoon snoozing in a quiet backyard and someone lights an M-80 three feet away from you, it's LOUD; if that M-80 explodes during Fourth of July fireworks you probably wouldn't notice.
There are lots of "tricks" for making sounds seem louder, such as brick wall compression/limiting and using EQ; however, these tricks take a while to learn how to master, as does the entire mix process. And, as I have delineated before, sound FX - like your "thud" - are really a combination of sounds that are individually manipulated to give them more sonic impact.
Recording loud sounds is an art unto itself. Charles Maynes specializes in recording weapons. His recordings are awesome and highly complex - multiple mics, preamps, recorders (often including analog tape), and perspectives. These multiple tracks are then mixed together to create the proper "loudness" and perspective needed for the sound track and its affect upon the audience.
http://www.gamasutra.com/view/feature/1746/weapons_sound_effects_recording_.php
http://designingsound.org/2010/06/charles-maynes-special-gun-recording-thoughts-for-2010/
The described techniques are useful for more than just gunshots and explosions.
And, of course, as we have discussed endlessly, a proper listening/playback environment is very important as well.
You haven't gotten to the mix yet. If everything is loud, nothing is loud. If there is quiet before a loud sound it will seem louder. For instance, if you're spending a lazy afternoon snoozing in a quiet backyard and someone lights an M-80 three feet away from you, it's LOUD; if that M-80 explodes during Fourth of July fireworks you probably wouldn't notice.
There are lots of "tricks" for making sounds seem louder, such as brick wall compression/limiting and using EQ; however, these tricks take a while to learn how to master, as does the entire mix process. And, as I have delineated before, sound FX - like your "thud" - are really a combination of sounds that are individually manipulated to give them more sonic impact.
Recording loud sounds is an art unto itself. Charles Maynes specializes in recording weapons. His recordings are awesome and highly complex - multiple mics, preamps, recorders (often including analog tape), and perspectives. These multiple tracks are then mixed together to create the proper "loudness" and perspective needed for the sound track and its affect upon the audience.
http://www.gamasutra.com/view/feature/1746/weapons_sound_effects_recording_.php
http://designingsound.org/2010/06/charles-maynes-special-gun-recording-thoughts-for-2010/
The described techniques are useful for more than just gunshots and explosions.
And, of course, as we have discussed endlessly, a proper listening/playback environment is very important as well.
Oh sorry, when I got the message before, not all of it was there. I am using the Fostex FR2LE and I tried some different gaines, but they are all quiet. One the gain was turned high up, but the trim was turned down, otherwise the scream would peak. The other, the trim was turned up more, and the gain turned down. I tried both on the screams and loud noises to have different options of which would sound better, but they all came up quiet. Not sure which EQ, I am still learning that one. When you say the rest of the mix, do you mean the whole soundtrack and the other tracks?
Last edited:
Obviously.
You're way out of your comfort zone with audio; it's going to take LOTS of time to figure it out. Mixing is the hardest part. When it comes to sound-for-picture it's all about perspective and context. I've been mixing audio for a long time and I still learn something new every time I do it.
Oh okay. Yeah I'm trying to learn it all. With the rest of the mix the volume is at a similar level, but it just seems like it should be louder since it's a scream. Like it sounds like a scream but further away.
Last edited:
I'll try and answer your question in 3 parts, First Part (why it sounds quiet):
You are falling into the common trap of those new to audio, in misunderstanding what level (the numbers) actually means. You take a sound and you increase it's level and it sounds louder, you therefore logically conclude that level = loudness. Unfortunately, your logic is letting you down because you are not in possession of "all the facts". In reality level does NOT = Loudness. In order to avoid explaining "all the facts", I'll try using an analogy. Let's say we own a Fiat 500 car: If we push the accelerator pedal exactly half way down, we are going to get one rate of acceleration and if we push the accelerator pedal flat to the floor, we are going to get another, faster rate of acceleration. Without any knowledge of performance differences between cars it, would seem logical to conclude that pedal position = rate of acceleration. However, if we take a different car, say a Ferrari 428, the halfway position of the accelerator pedal will give us more acceleration than the pedal to the floor in our Fiat 500. We all know that in all cars, the further we push the accelerator pedal the faster we accelerate but we also know there is no correlation between the accelerator pedal position and the actual acceleration when comparing different cars. The same is true with levels and loudness, there's no direct correlation when comparing sounds. In this analogy accelerator pedal position = Level, acceleration = loudness and different cars are different sounds. In other words, take two different sounds (cars), recorded with levels peaking at say -3dB (pedal position) and they could be of completely different loudness (acceleration).
Part two (how do you tell if it's loud enough if the numbers don't indicate loudness?):
While there are solutions to this problem, there is no simple answer. Science completely understands and explains what sound waves are and how they work but when it comes to the human perception of sound, science as yet only has a partial understanding. Loudness is in fact not a property of sound, it's a perception and as such it varies from person to person and indeed changes even within the same person. "How loud is loud enough?", is even more of a problem because there are a number of variables, such as the listening equipment/environment of the consumer, in addition to the fact that can't we accurately measure loudness. The best solution devised for theatrical content is to minimise the variables of listening equipment/environment (through acoustic design and calibration) and rely on the subjective judgement of experts (experienced re-recording mixers). The best solution devised for broadcast (where consumer equipment/environment can't be controlled) is again calibrated environments and the subjective judgement of experts but for this judgement to be constrained by increasingly sophisticated scientific methods of approximating loudness measurements.
Part 3 (how do you make something louder?).
Loudness is a relative term, so the first and most obvious option is to lower the level of every thing else. This option may not be viable in a calibrated environment, so you may need to manipulate the sound to appear louder, using tools such as EQ, compression, excitation, divergence, etc. How you apply these tools without damaging or loosing the essence of the sound depends on the properties of the sound, what you want to end up with, and the distribution format of the final product. There are no specific answers because we're trying to create a perception (loudness), not a scientific property. That is why others are asking to hear the sound before they can advise the best method of creating this perception.
G
