Basic question on the logic of Effects in Presets

Report · May 12, 2024

Sorry maybe this is a stupid question since this should be obvious but I'm not 100% sure. If I open a track in Wavefrom view, and select all or part of the track, and then select an Effects Preset do all the Effects e.g. 1, 2, 3, 4 & 5 inside that Preset get applied to the track one by one in that order once I click Apply?

Report · May 12, 2024

That is the general idea, yes, and the order can be significant.

I should point out that presets are only supposed to be a starting point, because no two voices are the same, so 'fixed' processing as such, may not always be correct or relevant. The most extreme example of this is the Podcast session template; use of this is extremely likely to completely screw your results (as many have found), and should be avoided at any cost. In the example you've shown, things don't appear to be too violent, but you should always be critical of the result - the preview button is your friend!

Report · May 13, 2024

Thank you @SteveG_AudioMasters_ that helps a lot!
The example I showed in the screenshot was just an example, I have my own effects presets that I created from some video tutorials I followed (as I'm no expert) and I was wondering how exactly things worked.

A follow-up related question if I may. I like to try and add some bass to voice recordings, or at least when just recording with the raw microphone through Audition it doesn't sound so good. I should note that I just use a USB mic connected directly to a PC. I've been using a third-party virtual amp application for PC which can add some nice tone to the voice while live recording but the application is extremely complicated to use and following an update is now even more complicated so I would like to remove it from the workflow if possible.

I found a clip on YouTube that suggested Effects > Filter and EQ > Paremetric Equalizer > Loudness Maxmizer and then drag the scale up below 100 Hz. I need to try and test it fully. Can you recommend any options or techniques and is it at all possible in Audition to add such bass effects while you are live recording or do these effects only ever work in post?

Thanks again for the help!

Report · May 13, 2024

As good professional practice, Audition only records flat. The simple idea behind this is that there are some effects which you simply can't undo - like reverb. If you decide later that there's too much, then you'd have to re-record. The basic idea is that you record flat, then make a copy of the file and play around with that - you've always got something to go back to. As an aside, I should mention that if you record in Multitrack with an effect present, that effect is only in the monitoring chain - it doesn't get recorded, so effectively the same thing applies.

Parametric EQ is fine for producing a bit of a bass lift to the voice, but generally you need to start the lift at a higher frequency than 100Hz. A gentle lift from about 250Hz will probably serve you better, although you'll have to try it and see, as I mentioned earlier. Incidentally, ou'll get far more control over what you are doing if you switch to a 30dB dynamic range rather than the 96dB one. Also, it's worth noting that you won't need the boost going right down very low for two reasons: firstly the human voice doesn't produce really low frequencies - even though you think it is, what you are hearing is based on harmonics with virtually no fundamental, and secondly you'll find that you are boosting anything else around as well as the voice. But there is a fix for this - use the HP filter as well, and whilst you're about it, don't go overboard on the amount of lift! So I'd say start from something more like this:

And experiment from there. You don't really need to alter the levels until after you've got the recording to sound how you want it, because boosting the bass will increase them anyway. You can leave sorting out the levels until after you've got the sound right.

Report · May 13, 2024

As good professional practice, Audition only records flat. The simple idea behind this is that there are some effects which you simply can't undo - like reverb. If you decide later that there's too much, then you'd have to re-record. The basic idea is that you record flat, then make a copy of the file and play around with that - you've always got something to go back to.
Yes good point on applying things only in post. And yes copying the original files to test effects makes sense and it's what I've been doing 😄

As an aside, I should mention that if you record in Multitrack with an effect present, that effect is only in the monitoring chain - it doesn't get recorded, so effectively the same thing applies.
If I add an effect in the Multitrack view how do I then apply it to the track? Or at least I don't recall seeing an Apply button.

Parametric EQ is fine for producing a bit of a bass lift to the voice, but generally you need to start the lift at a higher frequency than 100Hz. A gentle lift from about 250Hz will probably serve you better, although you'll have to try it and see, as I mentioned earlier. Incidentally, ou'll get far more control over what you are doing if you switch to a 30dB dynamic range rather than the 96dB one. Also, it's worth noting that you won't need the boost going right down very low for two reasons: firstly the human voice doesn't produce really low frequencies - even though you think it is, what you are hearing is based on harmonics with virtually no fundamental, and secondly you'll find that you are boosting anything else around as well as the voice. But there is a fix for this - use the HP filter as well, and whilst you're about it, don't go overboard on the amount of lift!
I tried the effect you described and it was interesting so I'm going to try and create my own preset from that 😉

And experiment from there. You don't really need to alter the levels until after you've got the recording to sound how you want it, because boosting the bass will increase them anyway. You can leave sorting out the levels until after you've got the sound right.

I have to admit I know nothing about using levels in Audition.

Report · May 13, 2024

The whole point about Multitrack view is that it's non-destructive - so you can change anything right up to the point where you export it or mix it down - it's during this process that effects get applied. So if you don't like your mixdown, you can go back to the session, change whatever you want, and then mix it down again.

The levels issue is a bit more complicated, and difficult to sum up in a few words, because it involves a basic grasp of the implications, at least, of something that a lot of people regard as tantamount to magic. Last time I explained this I got accused of over-complicating it, but it's not simple, whatever that person thinks. Anyway, here's an attempt at the explanation of the results, not the process:

Audition uses a 32-bit Floating Point mix (the magic bit) engine, and this really doesn't care what levels it operates at, because it's a 'virtual' system. Unfortunately this doesn't apply to your sound device, which can only operate up to a fixed ceiling that we refer to as 0dB, because it's an integer device with a fixed dynamic range, and above 0dB, it's going to sound distorted. But that really doesn't matter, because you can pull the mix level up and down all you like, and if you keep the peaks below 0dB, it will sound fine whilst playing back in Multitrack. But, even if you completely overload the output and mixed it down so that everything looks like a solid block of green, you can still fix this in Waveform view after the event - just lower the level. And this process is lossless. If you did the same thing with an Integer mix, it isn't lossless at all - you'd end up with noise build-up, and all sorts of other issues. Audition used to allow this, but it doesn't any more - fortunately.

So to get a hot mix back to a sensible level to save in a file, we use a process called Normalization (it's in Effects>Amplitude and Compression) and with this we can set where we want the highest peak to reach. Generally just under 0dB will do fine, although if you are going to save to a compressed format like MP3 then -1 or -2dB might give you a cleaner result. Normalization doesn't affect the sound at all - just the level of it. Effectively it's just a volume control for files.

I don't think I can make it simpler than that. There's loads more that can be (and has been) said about audio levels and speech, especially in relation to percieved loudness and compression, and a search around the forum might reveal a few threads about it...