• Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.

My recorded sound is very quiet

Hey IndieTalk

So, I've finished up my first short documentary. I recorded my sound, using a wireless Sony microphone, Canon 550D t2i & Magic Lantern. I used 26 dB because that's where the sound would start going yellow in the sound bar (and I didn't want it to be squeaky). However, the sound is very quiet. I've completed the editing and I've made all songs and extra sounds go -10 dB and less. Then I've turned up the volume on my pc (my headphones can get really loud), that way it sounds just the way it should. Same as for when I've rendered and watched it on a television, it's really quiet, but when I turn up the television's volume it will be all good.

So my question is, how do I make the overall video become louder without making it squeaky and noisy? I've tried to increase the dB on every recorded clip themselves, but it became squeaky and noisy - it wasn't like when I just increased the volume on the tv/pc. I'd like to increase the volume of the whole video somehow (seeing that I've already adjusted every song and sound to be around the same volume).

I'd love to fix this, because I want to upload it to YouTube as well as submit it to festivals, and there I can't just increase the volume like if it was my tv or pc.

Thanks in advance! :)
 
Use better audio recording equipment would be my first suggestion. Something with a better noise to signal ratio.

Aim to record in the -6 to -12db range.

You could possibly go in and clean up the audio manually, but you're unlikely to have that turn out well.

You could always ADR it.

AA & APE, you two know way more than anyone else here. Your turn to chirp in.
 
Yellow is ok, you want to avoid red.

But yeah, turning up a track means turning up everything in the track. The sound you want and the noise you don't. Have you tried upping the gain in post, how noisy is it?
 
I'm talking about upping the dB in post, that's where I get the noise. I need to put it +10 dB in order to become normal. When I was recording it was only green on the sound bar, and just reaching the yellow sometimes. That's why I find it weird - shouldn't it supposedly be normal volume?
 
First, you mix by ear, not by meter. That's part of the problem with recording to a DSLR; the audio as it is being recorded cannot be monitored, you are hearing a "side feed", not the actual sound being recorded by the camera.

Meters in consumer devices, and even somewhat in prosumer devices, leave a great deal of headroom so that the audio will not distort. You can push almost into the red (and sometimes into the red) and not distort - most of the time. But you cannot be sure because you are not hearing the actual audio after it is recorded but the side feed of what is going into the camera.

You say that when you increase the gain in post the sound gets "squeaky and noisy." GIGO - Garbage In, Garbage Out. The sound you recorded is the sound you recorded; it was already noisy to a lesser or greater degree when you recorded it. Without hearing the raw audio I can't tell how noisy. BTW, you didn't mention if you used an entry level Sony wireless system or a professional system or how/where the mic was positioned. You also have made no mention of your audio monitoring equipment or your audio monitoring environment. Consumer audio reproduction systems tend to be very forgiving sonically and tend to make things sound "better" than they really are, but sometimes the reverse can be true depending upon which audio reproduction system.

You've made no mention of what audio post software you are using and what process you used to increase the gain. If you used a compressor you actually raised the noise level. If you simply increased the gain you are only hearing the sound as it is, but louder. Oh, and when you increased the overall volume of the mix did everything (music and "extra sounds") also become "squeaky and noisy" or just the dialog?

You can use noise reduction to, well, reduce the noise. You can use EQ to shape the dialog to be more pleasing. But you are not going to really know how the audio will sound on an uncalibrated system in an untreated room.
 
With audio on my first/only project so far, I too had to turn the DB way up in post to get the volume levels right. I also never once had the boom accidentally dip into a shot, and that makes me feel like I wasn't close enough. I ended up holding the boom a lot and I never even watched a video first, I was expecting a sound guy to show up but he didn't.

So yeah amateur booming but the noise isn't too bad most of the time so turning it up in most didn't ruin anything. Although there are one or two lines that are maxed and still weren't quite as loud as i'd like
 
I also never once had the boom accidentally dip into a shot, and that makes me feel like I wasn't close enough.

The boom-ops "trick".......

If the boom never pops into the frame, you're right, you're not as close as you could be. You start with the boom in the frame and slowly pull out until the DP says "Okay, you're out." You then pick a fixed sight line as an out-of-shot reference point.


Just for fun... Great boom-ops know the script inside/out so they can anticipate lines when switching from actor to actor. They also know their lenses. "Huh?" I hear you cry? Yes, just by knowing which lens is being used a great boom-op already knows 85% of the framing before getting a look at the monitor or conferring with the DP.
 
lol @ boom ops.

Horror story:
I just received a show to edit. I was told, "You can occasionally see the boom in shot, so keep an eye out for it."

The footage arrived. I picked a random piece of footage to watch. My jaw dropped when about 1/3 of the way through a scene I saw the boom appear from the top..... continue through almost all the frame and stop, leaving maybe an inch of the bottom edge of the frame without a boom. Luckily, it was the only time it happened and it was picked up from another camera.
 
Sweetie thats horrible lol.
I remember I saw the Ring 2 in theaters and there was a boom that falls into the shot. somehow they didn't care and sent it off to the theaters anyway.

I assume if it's just a little bit on the top you can crop that out and lose some resolution for the scene?

As far as my quiet sound goes, we used a rode shotgun plugged into a Panasonic AG-AC 160. There was no preamp, so maybe thats why I had to bump up the sound in post.
 
I've completed the editing and I've made all songs and extra sounds go -10 dB and less. Then I've turned up the volume on my pc (my headphones can get really loud), that way it sounds just the way it should. Same as for when I've rendered and watched it on a television, it's really quiet, but when I turn up the television's volume it will be all good.

So my question is, how do I make the overall video become louder without making it squeaky and noisy?

I'd love to fix this, because I want to upload it to YouTube as well as submit it to festivals, and there I can't just increase the volume like if it was my tv or pc.

You seem convinced that you need to make your mix louder, when this might not be the case. Youtube has no loudness specifications, so you can upload your video with any volume level you want. As for festivals, most of the smaller festivals also have no loudness specifications. The major festivals (those which screen in a commercial cinema), they require a mix to theatricals specs and the chances are that your mix is already louder than theatrical specs.

If you want to match a particular level on youtube, say TV broadcast standards then the only solution is to use a calibrated playback system and the correct tools for meeting the loudness specs. This almost always means the level of the dialogue needs to be increased and noise reduction processing applied to lower the inevitably high noise floor.

G
 
First, you mix by ear, not by meter. That's part of the problem with recording to a DSLR; the audio as it is being recorded cannot be monitored, you are hearing a "side feed", not the actual sound being recorded by the camera.

Meters in consumer devices, and even somewhat in prosumer devices, leave a great deal of headroom so that the audio will not distort. You can push almost into the red (and sometimes into the red) and not distort - most of the time. But you cannot be sure because you are not hearing the actual audio after it is recorded but the side feed of what is going into the camera.

You say that when you increase the gain in post the sound gets "squeaky and noisy." GIGO - Garbage In, Garbage Out. The sound you recorded is the sound you recorded; it was already noisy to a lesser or greater degree when you recorded it. Without hearing the raw audio I can't tell how noisy. BTW, you didn't mention if you used an entry level Sony wireless system or a professional system or how/where the mic was positioned. You also have made no mention of your audio monitoring equipment or your audio monitoring environment. Consumer audio reproduction systems tend to be very forgiving sonically and tend to make things sound "better" than they really are, but sometimes the reverse can be true depending upon which audio reproduction system.

You've made no mention of what audio post software you are using and what process you used to increase the gain. If you used a compressor you actually raised the noise level. If you simply increased the gain you are only hearing the sound as it is, but louder. Oh, and when you increased the overall volume of the mix did everything (music and "extra sounds") also become "squeaky and noisy" or just the dialog?

You can use noise reduction to, well, reduce the noise. You can use EQ to shape the dialog to be more pleasing. But you are not going to really know how the audio will sound on an uncalibrated system in an untreated room.

I'm using Sony Vegas as post-production editor. The music and the extra sounds don't become squeaky and noisy, because as I said, I made them -dB to balance them with the dialogue. I didn't turn them up.

Isn't it possible to increase volume without increasing the dB? Because I'm guessing that's what makes it noisy and squeaky?
 
Isn't it possible to increase volume without increasing the dB? Because I'm guessing that's what makes it noisy and squeaky?

Yes it is possible but it depends on a number of factors, including your knowledge and application of how human beings actually hear/perceive volume. Generally the simplest method of increasing perceived volume, and therefore the one most commonly used, is to increase the (peak) dB level, the next simplest is to increase the RMS dB level, through the use of a full range compressor but both of these methods increase the volume of everything on the tracks you're trying to make louder, including the noise.

Again though, what makes you so certain you need to make your sound louder? You can only know the answer to this question if you know what loudness level you are trying to achieve. So the question is: What loudness level are you trying to achieve and why?

G
 
Yes it is possible but it depends on a number of factors, including your knowledge and application of how human beings actually hear/perceive volume. Generally the simplest method of increasing perceived volume, and therefore the one most commonly used, is to increase the (peak) dB level, the next simplest is to increase the RMS dB level, through the use of a full range compressor but both of these methods increase the volume of everything on the tracks you're trying to make louder, including the noise.

Again though, what makes you so certain you need to make your sound louder? You can only know the answer to this question if you know what loudness level you are trying to achieve. So the question is: What loudness level are you trying to achieve and why?

G

It is for a documentary. It is very quiet, so people would need to turn up their volume completely in order to hear it, which saddens me. I'm trying to reach the standard volume of a YouTube video/documentary.

You can hear the volume here: http://www.youtube.com/watch?v=PA1S2nU8VfI&hd=1
 
This really isn't bad. I downloaded the clip for grins to see where it peaked. On average you're peaking at -12, except for a few spots. Her dialogue actually clips a couple of times between 3:00 and 3:30.

If you normalize the levels you'll be good to go. Leave the louder sections close to where they're at (unless you made her clip in post, then bring it down) and add 6db of gain to the rest and your golden. Even as it is, it's well in the range of youtube clips. Some are louder, others are more quiet.

Edit: I tried adding 6DB of gain too, and it didn't make it horribly noisy. It's fine.
 
For youtube, I recenly found there's no such thing as dynamic range. You have to assume everything is normalized to 0db and has no dynamic range, simply because a large percentage of clips are like that.

I'm scared to make this recommendation for fear of being pummeled... bring down any unusual peaks, then normalize it to 0db. Then compress it to go from -20 to 0 for youtube. Some are even more compressed. I was learning about this just recently, and that's the only way to make the sound fit in for youtube when people are playing it on their crappy laptop speakers or iphones. If that's the medium you're going to use, that's the levels where people tend to listen to it.

However, on other media, don't do the same thing! It'll be way too loud (except for maybe CDs).
 
I'm trying to reach the standard volume of a YouTube video/documentary.

There is no standard volume of a Youtube documentary! Professionals tend to aim for a loudness of somewhere around -16LUFS for Youtube video content but this is a professional convention rather than any sort of rule and you don't have the equipment to measure or create content to this loudness convention. I would say your documentary is already within the broad range of acceptable Youtube loudness.

For youtube, I recenly found there's no such thing as dynamic range. You have to assume everything is normalized to 0db and has no dynamic range, simply because a large percentage of clips are like that.

I'm scared to make this recommendation for fear of being pummeled... bring down any unusual peaks, then normalize it to 0db. Then compress it to go from -20 to 0 for youtube. Some are even more compressed. I was learning about this just recently, and that's the only way to make the sound fit in for youtube when people are playing it on their crappy laptop speakers or iphones. If that's the medium you're going to use, that's the levels where people tend to listen to it.

Prepared to be pummelled! :) The problem with Youtube is that it contains all kinds of content from features to CDs, both amateur and professional, of various commercial standards or to no standards whatsoever. Most CDs have, for various historical reasons, almost no dynamic range and it's a mistake to try to compete as far as volume is concerned with Youtube CD/music levels. It's also a mistake to normalise or aim for 0dB or as close to it as you can get, as is the case with commercial CDs. A full explanation of why would need to get quite technical regarding how digital audio works, I'll go into it if you wish but for everyone else who's not really interested, it's best to peak no higher than about -2dB and to aim for a level of loudness well below that of the average CD or Youtube music track.

G
 
There's no answer to this question because there is no set average or loud, and in addition, loudness cannot defined by dBFS. For example, which would sound louder, a bass guitar at -10dB or say a cymbal at -20dB? The answer is the cymbal!

If we are talking average (RMS) levels rather than peak levels then the dynamic range probably shouldn't exceed about 20dB RMS, but this is a quite wild guess because it depends on the exact material and various other factors.

G
 
I agree with the advice to normalize to below 0db. In some programs it might also be called amplify or energize.

I think the reason that you get the noise in your program but not in the tv or headphones is because they have eq filters.

You might be able to remove the frequency of the noise from your recording without affecting the rest of the recording. An old version of cool edit pro had a noise remove function that worked well.

You could also try removing the frequency manually using a eq filter. You can also boost the frequencies that you want using an eq, but be careful not to overdo it.

The sound in "The art of being old" sounds fine on laptop speakers.

Remove the noise frequency before raising the volume.
 
Back
Top