Moho Forum

O.k. this is a question for those that have audio skills in film work... Here it goes.

1) Can anyone tell what the best format for sound is (file types,Ect...) what are the best HZ's.

2) In your work flow if your pulling from different sources, how can you makes sure all the sounds have the same volume. Is there a good way to measure volume.

3) Is there a way to import sounds other than wav. files when doing dialogue in anime studio?

4) Is wav. a bad format? and what is exceptable for good quality (8, 16, 32 bit?)

5) Are there problems when mixing formats in your edits? (Wav. with Mp3.)

6) Can a bad recording be cleaned up, and how?

Hi Neomarz,

I do a lot of work with audio, but strictly from a hobbyist perspective...I spend far more money on my music than I ever make from it. So take my answers with a grain of salt:

1. Avoid compressed audio files until the very, very end of a project (so no mp3s), and even then, only use it if you absolutely have to for space or bandwidth reasons. In these days of high bandwidth and massive storage space, compressed audio files are hardly necessary (it's a tiny fraction of your video file size in your projects).

So most people tend to use WAV or AIFF files. If you're on a PC, just use WAV. AIFF files at one time had an advantage of storing marker locations in the file itself, but I doubt there's a real distinction with newer software.

As for Hz, I assume 48 Hz is used when people are doing surround mixes or doing lots of post-production trickery, but 44.1 Hz is fine IMHO. The important thing, though, is to use THE SAME setting through your entire workflow. Don't convert your audio files during the course of the project.

2. Amazingly there ISN'T a good way to measure our perception of volume, it always comes down to your ears: does something sound too loud? Then reduce its volume. But if you want to bring everything somewhat more in line you can apply normalization to all the files. You'll still need to tweak it afterward, though.

3. I'm using AIFF files in Anime Studio. I'm not sure which other formats it supports.

4. WAV is a fine format. CD audio is basically a WAV file at 44.1 Hz, 16 bits. Don't EVER go below 16 bits (unless you want to introduce deliberate distortion!), but unless you're doing pro mastering later on I can't think of any reason to use 32 bit files.

5. There's no problems mixing formats in an edit as long as your audio software can handle it, but if you're mixing audio of different quality (compressed file vs. uncompressed file) you'll hear a big difference. So really, really avoid MP3. Also, if you're applying effects (reverb, vocal distortion, compression), do it during the final mix...don't apply it to individual sources. Otherwise you may find that the sources just don't sound right together. Reverb is particularly important as a "last stage" effect.

6. That depends on what you mean by "bad recording." Getting a good, clean, reverb-free recording right from the beginning is paramount, but few people can afford the equipment to do that (let alone build a sound-damping booth in their living room!)

Record your audio in as quiet a place as you possibly can -- no air conditioner, no traffic noise -- and preferably where the surfaces are porous (tiles and concrete bad, wallpaper not as bad, foam walls good). Use a good microphone with a pop filter and a preamp...you can rent these things from most audio stores.

If your audio has a hum, you might be able to nullify it by finding the hum's frequency and lowering the volume of the frequency with a good EQ plugin. If the hum is variable or it intrudes on the frequencies of the audio you want to keep, try a denoising plugin, and spend HOURS tweaking it, coming back, listening to it again, tweaking some more...these plugins can easily add unwanted "robotic" artifacts.

If you can't remove background hums or noise, remember that you'll probably be adding music or sound effects later: these can hide lots of audio sins.

If your speech has lots of "pops" from somebody who keeps Popping their Pees, again, there are plugins that help remove those problems, but sometimes you need to manually edit the waveform to remove the pops.

Hopefully this helps! If you can afford it, book some studio time and record the audio there: having clean sources is the difference between a pro sound and a "recorded in the basement" sound. But if you can't do that, get creative with your plugins, buy a good pair of headphones, invite other people over to offer their opinions, and try to have some fun with the frustration: audio engineering isn't a science. There's no single answer.

Cheers,
Muffy.

In case you're working with Audacity, I've found that trying to remove static noises with the "remove noise" effect that uses a noise profile creates the robotic sound Muffy mentioned. The noise gate plugin causes hardly any distortion for me, as long as the voice or sound you're recording is a bit louder than the static.

Muffy--I've seen that some people recommend throwing a blanket over your head while you record for some sound dampening. I did some tests with this and it seemed to work well. So I thought maybe making some kind of blanket tent would have the same dampening effect but provide enough structure so that the blanket doesn't cause rustling noises. I planned to make the frame from some PVC pipe that I reconfigure all the time for my various schemes. Do you think that would work? The insights you've already shared are incredibly helpful, btw! :D This went straight to my bookmarks.

A blanket tent! That would probably work, but I think you'll want to make it as big as possible. If it's too small the blankets might absorb too much sound before the microphone gets a chance, and you'll wind up with a muffled sound. I think that if you have a large enough tent and you do some experiments with mic placement you could have a very effective "clean sound" room.

Just don't pass out from heat exhaustion or oxygen starvation!

I did see your test clip of one of your characters speaking, and it seemed that the audio was really clean...it didn't have a "living room" sound. Did you use the blanket-over-the-head technique for that?

Oh, I didn't realize that it's good to keep a little bit of resonance. Come to think of it, the tests I did were a little dead sounding. Maybe I'll hold off on the blanket tent until I've experimented more without it. Plus it might weird out my friends if I herd them into some creepy fort I built. But on the other hand, the prospect of building a fort is almost irresistible. If I go for it, I'll definitely try making it on the larger side.

The audio for my sync test was not using the blanket over the head technique. I recorded it on my old cheapy mic and the room I was recording in probably has a lot of reverb because it has hard wood floors and not much furniture. Also, I wasn't talking loud enough to overcome the static. I didn't realize that the volume made a difference. I did try to clean it up in Audacity, though. But now that I have a better mic and more information to work with, hopefully my audio will improve. I plan to do some tests with my voice actors this weekend.

lwaxana wrote:In case you're working with Audacity, I've found that trying to remove static noises with the "remove noise" effect that uses a noise profile creates the robotic sound Muffy mentioned. The noise gate plugin causes hardly any distortion for me, as long as the voice or sound you're recording is a bit louder than the static.

Muffy--I've seen that some people recommend throwing a blanket over your head while you record for some sound dampening. I did some tests with this and it seemed to work well. So I thought maybe making some kind of blanket tent would have the same dampening effect but provide enough structure so that the blanket doesn't cause rustling noises. I planned to make the frame from some PVC pipe that I reconfigure all the time for my various schemes. Do you think that would work? The insights you've already shared are incredibly helpful, btw! This went straight to my bookmarks.

Yes Aduacity causes robotic sound but can still be usefull if you start with the least amount of removal and hitting preview and keep adjusting it till you notice the robotic sound then move it back down. I've found in most cases that getting the 1st few seconds of recording before u start to speak as the sample for noise removal helps alot.

Also you can record the room u are in with out speaking and look at the wave to see what noises your mic is picking up. I had a noise i couldnt figure out what it was and when i did this i realized it was my ceiling fan; a noise i have toned out over the years.

Other than that, Muffysb has said it all.

sbtamu--Yeah, for some reason I was still getting a noticeable robotic sound on the lowest setting. Maybe my voice is just at a static-y frequency. :D But you're absolutely right that the lower settings work better than the higher settings.

That's funny about your ceiling fan! I had a ticking clock that I didn't notice until I started recording.

I've read that you can do acceptable voice recording inside of a parked car. (Timothy Albee, CGI Filmmaking: The Creation of Ghost Warrior, I think.)

muffysb wrote:Hi Neomarz,

...But if you can't do that, get creative with your plugins, buy a good pair of headphones, invite other people over to offer their opinions, and try to have some fun with the frustration: audio engineering isn't a science. There's no single answer.

Cheers,
Muffy.

Wow, Headphones are good? I just had someone tell me that you need good speakers instead of headphones. I thought that sounded kinda funny. My Pc speakers are not that great. I figured headphones would allow you to hear better. So does it sound like he doesnt know what he's talking about?

I recorded and edited the voices for Happy Land with al el-cheapo PC mic, a stocking stretched over the bottom of a CD spindle with the middle cut out for a pop-screen, and Audacity. I spent a lot of time editing out clicks and pops in Audacity and adding quick fade ins and outs to silence because there was some background noise to sudden cuts to silence sounded abrupt. Here's the result:

http://www.youtube.com/watch?v=_okVLZutkl8

I should have clarified: headphones are good when you're trying to eliminate noise in your project...otherwise you're liable to get the noise in your project confused with the noise in your workspace. Otherwise speakers should be used when you're approaching your final mix, preferably different speakers in different places. I burn a CD and take it to different places and play it on their stereo and computer systems, and I always find a spot where the mix can be improved by doing that.

And I totally agree with Mikdog that you can produce great results with homemade solutions and consumer products, especially if everything else about your project is top-notch. While recording a series of product demos recently I used a cheap microphone with a sock over it, stuck into a tissue box.

But if we'd had better equipment I would have had to do 1/10th the post-production work to make it sound good, and I'm convinced the final result would have been leagues better.

muffysb wrote:Otherwise speakers should be used when you're approaching your final mix, preferably different speakers in different places. I burn a CD and take it to different places and play it on their stereo and computer systems, and I always find a spot where the mix can be improved by doing that.

Ah! That's brilliant. I never would have thought of that.