I’m getting fairly efficient at using programs like Audacity for voice recordings and Premiere Pro for video editing, but there’s one problem that I still have not figured out: my voice recordings come out as being very quiet. Every time I’ve done a video for YouTube the first thing I’d do is go into the channel with the voiceover in it and crank that sucker up, followed by going in and making sure any clipping is taken care of. I believe this is a problem with my recording via Audacity because it’s ONLY my voice files that have this issue (I can bring in clips from all kinds of different sources… clips from videos, recordings via Open Broadcast Software, etc). But I don’t know what is causing the issue. I record at 90% mic volume because using 100% makes the problem of clipping absolutely terrible and from there, like I said, i deal with audio levels in PP.
I have a Samson Meteor Desktop Mic but I don’t have any other fancy programs/hardware for recording. I just plug and play it.
This is the biggest issue I want to figure out with audio. However, I have a lot of questions about audio in general… I just can’t think of them all for a cohesive list. I suppose the best I can do for this is ask more questions as this thread goes along.
However, here is a list of other things I can immediately think of to add to what I’ve already asked about:
Why does clipping happen at all (from a technical standpoint)? I’ve noticed that volume isn’t necessarily the biggest factor in whether or not something will clip… for example, some letters are prone to clipping while others are not.
Why can my voice sound so quiet in my videos even after being cranked up while I can bring in a clip from something else and it will be HELLA LOUD but won’t clip? Like… wtf??? For example, view this thread where I posted my Evolve Stage 2 review. Notice the part where I put in a clip from Game of Thrones. That clip (and my background gameplay footage) was reduced A LOT because it was louder than my voiceover even after being amplified.
What are some quick and easy things to do in Premiere Pro to reduce clipping? I know of the “adjust peak volume by” trick in the audio gain options, but that seems to affect the whole track. I would like an option that automatically searches for anything in the track that clips and reduces just that.
there exists something called the RMS volume/amplitude. It describes how "dense" your signal is, meaning how small the difference between highest and lowest peaks are.
Your signal is perceived as quiet because of a low "density", also called compression. So the solution to your issue would be to utilize the magic workings of a compressor. Imagine it to be a little gnome that turns down the volume knob on loud (transient) parts at the beginning of a word and cranks it up at the more quiet end of the word.
Important technical figures here are: Threshold - when should the compressor start to work? Attack - usually in ms; After exceeding the threshold wait X ms before doing anything Ratio - between input and output volume. ranges from 1:1 (no compression) to infinity (limiter) (hold - do that for Y ms) Release - Z ms after attacktime (or hold time) stop compressing the signal
Now to your 3 questions: 1. Digital Signals have only a finite resolution (otherwise they would take up infinite space on your harddrive). Every peak above the 0 cannot be represented anymore and therefore result in distortion 2. RMS 3. Limiter (infinite compression) on the Masterbus is usual to catch all evil clipping transients. More moderate compression on each track to increase the RMS volume ("loudness")
Hope it does help
Maybe a good addition to this: Compression is the reason you fall out of your chair when the ad breaks jump into your face on TV. Also why music these days is perceived as louder than 40 years ago. It is good practice to not overdo it with the compression. Too much compression makes the track "dull"/numbing to listen to and pumping effects (like techno bassdrums) can appear.
I love this explanation. You've described exactly the technical parts about it needed to get the "why" and done it without turning into a software wizard. I haven't seen anyone explain so much so clearly in such a short space.
So essentially you're confirming my guess that one of the reasons the audio is so quiet is that my voice is so monotone and quiet sounding to begin with, that the recording software is perceiving it as being very... flat. This would explain why I can record gameplay with OBS in which I mic talk to other people and everything sounds fine... but when my voice is the only sound being picked up these problems occur.
Soooo... how exactly would i go about using this magical compressor you speak of to get my voice over tracks to be as loud as they need to be without clipping?
I've been producing music for many years now, I can verify this is all true...as this was almost exactly what I was going to say.
I'll add this this by saying to help compression, it helps to remove unnecessary data so that you can compress more to make it even louder (pop music...) . So, in this event, it helps to have something along the lines of an EQ. For my voice over track, my chain usually looks like EQ, compressor, exciter. (we can ignore the exciter for now, that's just something that ruffles the frequencies to make the audio signal a little more colourful) Typically I make a cut around 50Hz, subtrack about 2db from 400hz, a bump of about 2 db around 1700hz, and a high-shelf starting around 6000hz with a linear curve at about -1db.
"Where can I get these "compressors" and EQs?" you ask. I don't think Audacity supports VST plugins, but I do believe Premiere does. so here: http://www.reaper.fm/reaplugs/ I cannot live without the ReaPlugin buddle, it's free and just so....damn....perfect. Particularly the EQ, Unfortunately you need to download the bundle to get ReaEQ, but, oh well, it's free. You can use the built in compressor in the ReaBundle, but personally I've always had my own compressor of preference which is 112db's "Big Blue Compressor" - that costs $100, so let's ignore that. There is the MeldaPlugin bundle which comes with a great compressor - I use to use the MCompressor a lot before I got Big Blue. It's another free bundle: https://www.meldaproduction.com/MFreeEffectsBundle (there is a paid version of the Mbundle but I have never needed it. It unlocks a few extra features as well as removes the nag screen about purchasing the full bundle)
EDIT OH! and important note on compression. There are 2 types of compression: Peak Compression and RMS compression. Peak Compression triggers the compression when the transient of the audio crosses the threshold. RMS (root mean-squared) is somewhat similar but rather than being triggered by the actual transient, it's triggered by the rms volume. What does this mean to you? Peak compression is more for drums and instruments with very attacky sounds. RMS compression is good for quieter instruments...like you voice. RMS makes things much louder and is the favored type of compression by all things pop music now (it even works wonders on drums making Peak compression almost entirely obsolete!)
@Tek_Elf I wouldn't say that your voice is quiet and monotone. It is more a property of the recording itself. Btw. If I remember correctly Premiere Pro has its own effects bundle. Look for something like that on your audio tracks. I'd bet that you already have a compressor there.
Hmm. I'll have to look into these plugins and such the next time I start working on a video... which should be in about 18 hours. Till then, I won't have any more questions until I get some hands on with what you've mentioned so far. In the meantime, I'll post any more specific questions I can remember... like I said, there was a good few.
I'm ready to start dabbling with this mumbo jumbo but... I'd love a quick-start guide that explains how to use one of these things and what to look for. Do you know of a video, perhaps, that explains how to deal with those packages? I don't mean to be lazy but, well, it is much simpler to jump into a guide that specifically covers what I need out of it than trying to learn professional audio as a whole. Perhaps I am missing something, but it seems to me that Audacity should be able to record at a wider range to begin with, after which I can edit and compress the audio as I need.
That's not audacity, it's any recording interface. Microphones in particular are the hardest means to record anything. Mostly due to dynamic range, compared to signal output. Microphones typically output tiny millivolt signals with enormous dynamic range. Most of them need a preamplifier at minimum and if the human voice is involved a compressor as well. USB microphones rely on all of that to be done in software. Thankfully audacity can do them for free.
I just woke up, but, I got LOADS of free time for the next couple weeks. Give me a few minutes, I'll put together a short tutorial on install VSTs and how to use them. Ask any questions you want answered now.
Tutorial on how to use free software: virtual audio cable, vst host and reaplugs (FIR filtering to take out background noise, compressor and eq, gate to auto-mute when not speaking) etc.
@the193rd explained compression really well, will be borrowing from that in future
Audio quality on your youtube clip is good - you did well balancing the levels between your voice and the other audio clips. Even if the overall volume is a bit quiet, viewers don't get hit by a loud section
Samson Meteor sounds good with your voice, but is it just sitting on desk? If you can put it on a stand or other means to raise it up closer to mouth level you can experiment with microphone technique. Most microphones have ' proximity effect' when speaking really close - this is how radio DJs and voice-over artists get a rich bass sound if you are working a microphone up close, a pop filter is essential. Plosives (P and B sounds) and sibilants (S and T sounds) can cause clipping because of the air pushed out of your mouth microphone technique tips
I work with folks doing live radio, so post-production compression isn't an option. If you want to try a quick and simple compressor, get Voicemeeter and set the 'audibility' control to between 2 and 3. This app also has very good peak meters which might be a better tool for setting your levels without clipping
There are lots of opinions (sometimes conflicting) about audio production. Expect to spend some time trying different things to find what works for you
Often the length of an Analog mic can have an effect on the volume of your recordings, I use Apple headphones and the mic is picked up quite well in my opinion due to the (relatively) short cable length. Here is an example of my recordings:
Yes, it's just sitting on my desk right now. I don't have a boon for it or the money to get one. It sits about a foot away from my mouth, down a little. If I had the money to do a closer setup with a pop filter, I would. I have been meaning to do that for a while but... no money.
Sometime today I should get around to starting a new video and I'll start going through what people have said here.
if money is short, just stick it on a box or something to raise it up, and put it on something padded (towel, bit of bubblewrap etc) to cut down and vibrations from desk getting to the mic
I made a pop filter with panty hose, but actually spent some cash on a small embroidery hoop to stretch it on. You can try without a pop filter by talking across the mic, not directly into it