LibriVox Forum

I'm reading a novel where some of the characters get pretty worked up and holler, usually exclamations. I have been trying to sound holler-y while not being too loud while recording, with limited success, which means I end up way over the line in the audiograph in several places each chapter. I have been de-amplifying these individually, which has worked fine, but I'm wondering if there's a better way in recording, editing, or processing.

I've tried different things--more distance from the mic, turning my head away, holding my hand up slightly in front of the mic, but it all sounds sucky.

What suggestions do you have?

Thanks, everybody.

Try changing the tone of your voice, rather than the loudness.

Aside from Jo's suggestion, I think something like Limiter essentially de-amplifies peaks. I use it on all my recordings to even it out. (I have it set to Soft Limit, and usually use -4 dB, but one can experiment to find what they prefer.)

I think I read somewhere that Audacity doesn't actually lose the sound data that has been clipped (the part that goes above maximum); that it affects the file once it's been exported, but not really before. So then, if you limit (or de-amplify) before doing any exports, it shouldn't sound bad like it would if you left the clipping in or fixed it off an MP3 import.

Don't quote me on that, though - I read it somewhere, and I'm not sure if I read it or remember it correctly.

Try adding tension to your voice, rather than volume. Shouting requires stress, which comes out in the nature of the sound in addition to loudness. Go for the stress and tension, rather than volume. Turning slightly off-mic at the same time can add to the effect.

Thanks, everyone. I'll do some more experimenting.

vviera wrote: ↑January 30th, 2024, 11:01 am Thanks, everyone. I'll do some more experimenting.

I always apply Loudness Normalization set at -17 to the whole track and that always seems to get me at that 88-89db range without any clipping. I think that was something Phil made a video about.

flavo5000 wrote: ↑February 7th, 2024, 4:49 pm

vviera wrote: ↑January 30th, 2024, 11:01 am Thanks, everyone. I'll do some more experimenting.

I always apply Loudness Normalization set at -17 to the whole track and that always seems to get me at that 88-89db range without any clipping. I think that was something Phil made a video about.

I don't have any trouble being at around 89 without any normalization. It is just these few spots that are loud.

I just manually reduce those spots 2 or 3db. As long as they haven't clipped.

That's typically what I do also.

sjmarky wrote: ↑February 8th, 2024, 10:51 am I just manually reduce those spots 2 or 3db. As long as they haven't clipped.

What if they have clipped?

I thought about this problem of mine as I PL'd your latest. You do the "tense voice" really well to give that impression of volume. I'm still working on that.