STP 3 / Levelator Combo

[Michael Gissing]

If you used send to STP, then the project with all the clip, FCP levels and handles would be imported into STP.

From there you can easily adjust track levels and dynamics processing which is all that Levelator does but with zero control over the gain normalising and compression. Having your clip separate and with handles is essential to doing dialog editing.

You will be able to automate your track gains and dynamics so that you can get a far superior result than just hoping Levelator will magically mix the show for you. Final autput for DVD or Bluray, you should aim for stabdard broadcast levels so the disk sounds as loud as normal professional programs. This means compression and peak limiting to -10dbfs

[Dave Bergan]

How I use Levelator is to run it through my main dialogue or VO track and then mix everything else according to that. Since I output for the web, Levelator’s default output to -1db works for me. After I’m done in Final Cut, where I do my mixing, I will export a 2 channel AIFF to STP to check the peaks, maybe manually adjust them if something pops a little to much, then do a peak normalizing to -1db again. In your case it would be -10db.

So mainly I find it helpful to use Levelator for the main track, the track I’m going to mix everything against, but not the whole mix.

[Kay Edwards]

Hi Michael,

Thanks for your response. I agree, that’s a thorough approach, but I’m running out of time 🙁

Cheers!

[Kay Edwards]

Thanks for sharing your thoughts, Dave! I noticed that the frequency waves of the output track from LEVELATOR has been trimmed from the bottom and are flat – am I doing something wrong?

Ok, I’ll set the Normalizing peak to -10dB (or -9).

[Dave Bergan]

Not sure what “trimmed from the bottom and are flat” means. I’ve never seen anything unusual with my output file from Levelator, and with no user controls I don’t think you can do anything wrong. They do use RMS levels, so if the waveform peaks looks flat, it shouldn’t be anything to be concerend about.

Of course you need to be at certain levels, but don’t be too worried about how the waveform looks. Really the bottom line is does it sound good?

[Kay Edwards]

sorry about the bad description – this is my first audio. Please see attached screenshots (original & After Levelator). Notice that the bottom peaks are trimmed at -0.90 Is this OK? Many thanks!!

[Dave Bergan]

I’m sure it’s fine, and if you zoom in on it my guess is the flat parts won’t be flat. It doesn’t look too pretty but as long as it’s not clipping, which that won’t be, and it sounds good, then go with it. I’ve seen worse and there was nothing wrong with the audio or how it sounded.

[Kay Edwards]

Thank you!!!

I noticed that my audio output settings of the timeline were set to “stereo”, since it’s only dialogue, I reversed the tracks to mono with “downmixing” off. It was -3dB by default. Shall I keep it @ -3dB?

You guys really help make work more enjoyable…again, thank you!

[Dave Bergan]

I would, it’s supposed to counter a gain increase when using two mono channels rather than one stereo, though I don’t know the technicalities behind it. I am betting you did see a jump in overall levels so watch your meters. You can also set the pans on the clips or tracks to 0 if you want them mono. But seeing that you’re at the end of a 2 hour piece, I’m not really sure you need to switch the dialog to mono. If you’ve been doing the whole project with stereo dialogue, and everything’s been fine, changing at the end might just add to the confusion. And remember, you’re making the whole timeline dual mono (mx and sfx etc.), not just the dialogue.

An answer to an earlier question too, I’d clean up between sentences after Levelator.

But I will say something else about Levelator, which may nullify my input, and certainly add to the confusion. It was created by two guys making a podcast so they could easily normalize, noise reduce and make it sound better. So it seems to have been created to be used on one recording, which I think it does a great job of. I use it for short pieces (5 minute web videos) with voice over that was recorded in one session, maybe two. I only use it on the VO track, I don’t use it on the live video with dialogue in the project. I do have a hard time imagining using it on a dialogue track for something that is 2 hours. In other work, even a short film, I don’t think I would use it. One reason being, if you have someone far off camera, you don’t want to bring them up to the same level as someone close to the camera, and since Levelator uses RMS peaks, it would. Which goes back to what Michael said in the beginning, you don’t have the control over it.

I’m sorry if I’ve muddied the waters on this, my mind went off on a tangent. You said you were running out of time, I’m sure by the time you’ve read this you are out of time.

[Kay Edwards]

Thank, Dave! I tried sending the timeline to STP and use its “normalizing” tool, but for some reason, the gap between sentences/words becomes a loud noise. LVTR does a better job. Also, sending the timeline (or even partially) to STP slows down the system. I noticed that the audio is unnecessarily recorded on 16 channels. I’ll try Exporting to AIFF from FCP, then return the AIFF files back to FCP – send to STP and try “fixing” from there without LVTR.

And yes, I have ran out of time…and all I do is start from scratch….very confusing to say the least.