Reducing Harshness of Booming Vocals

[Trebor]

It's good but I can hear a problem with plosive Ps and Bs ...

How? How are you listening? Headphones? I admit I didn't do the Worst Case Listening Experience with Computer Speakers, but I guess that's warranted.

I can hear the popping on headphones and even on tinny laptop speakers which do not reproduce low frequencies,
the P of "sPeaking" is really conspicuous, the B of "Before" is not so bad.

[kozikowski]

I gotta go back and listen again. I'm at work. The only thing that bothered me was the slightly "essy" sound of his voice -- which I assume is natural. He's a reptile in addition to being unable to pronounce "button."

Koz

[kozikowski]

This discussion has a more serious extension. Whatever we do to get the voice perfect is going to need to be applied to each and every performance forever. So fine tuning every phoneme may not be warranted faced with that. The Audacity "Chains" batch process is, I believe, not up to doing everything we've done so far at a button (sorry, Buh-On) push.

Koz

[monkeywisdom]

@Trebor I did notice the tiny pop with that initial P after listening to your comparison. I'll make one out of wire and panty hose today. Likely, I can just hang the pop filter from the top of the isolation box. Was hoping I wouldn't need one, but whatever works. I can live with the s's or perhaps experiment with pronouncing them. Ironically, the CEO of the company I'm doing this project with used to work in the recording industry and he's not nearly as picky about sound. He just wants good content.

[kozikowski]

He just wants good content.

That's his new goal. He's now a producer. His old goal was professionally good sound.

I shot one half of a radio show a couple of months back and they emailed me a short list of technical requirements. I already do most of them, so it was a piece of cake past the upset of doing it at all.

"Wait, they want us to do what, now...?"

Time was everything would have to be done in a studio somewhere to get the required quality.

I remember one of the clear requirements in the email was Do Not Post Produce or Filter. Send me the raw show. This is against the horror of trying to unscrew somebody's bad filters and effects before the broadcast.

By the way, one of the Movie tricks is never record anything below 100Hz. All the live recording sound mixers have the "100Hz" button pushed. You might try the Hi Pass Filter set to 100Hz and that may suppress some of the plosives in the performance. Again, I think it's just fine as it is. This is your product; this is what you sound like. If you mess with it too much, one, you're being obsessive/compulsive, and two, you lead the client into thinking the patched together job is the real you -- and remember you have to do all these tricks to every show from now on.

A respected sound guy friend of mine in Florida once told me I was going to hear a lot of things in my shoot headphones that were never going to appear in the show.

Koz

[monkeywisdom]

My goal is good sound and content. His is good content. Regardless, it doesn't hurt to know how to make recordings passable as professional, especially with the animation stuff or future recordings for pickier people. I made a ghetto pop filter out of a wire, pantyhose and duct tape. It works well for pops. As for the sibilant s's, it seems like there are some pronunciation tricks to make them less overwhelming. I Googled a couple of them. That 100 HZ thing is a good tip to keep in mind for the future.

[kozikowski]

Even though human hearing goes down to 20 or so, there's not much of value down at the last few octaves. I have three mixers (over time) that have 80Hz, 100Hz and one that switches between 40/80/160. 160 is very noticeable, but if the options is getting a voice track in a high wind or not, you use the filter.

If you lose the sibilance, the only thing we have left is pronouncing "button."

I want to hear you doing a real piece. Test clips are OK, but I want to hear you doing the full-on cupped hand over the ear announcer thing. You know that trick, right? It feeds some of your voice back to your ear so you have a better idea what you sound like without the mixing board foldback.

Koz

[monkeywisdom]

OK. I officially recorded the intro track, which is an announcement about why I partnered with that company. There was one obnoxious "s" that I truncated because it was quite obnoxious. Once I complete the third or fourth track, I'll post it to this thread in a day or two when it's recorded. Not sure how secretive the company wants to be before the launch, so this first track won't be the one I post. As for listening to my voice while I'm recording it, not sure if that's going to happen. Seems very annoying, especially with a very slight latency. I typically do a soundcheck first and assume it will sound approximately like that. By the way, those compressor settings someone suggested really came in handy. The first track is very even amplitude wise. Didn't need to even amplify it afterward. It just seemed to level things out and amplify it all as much as possible. Also, the panty hose is doing wonders for the p's and t's. It was definitely worth the funny work some old man gave me as I was selecting it in the grocery store.

[Trebor]

... those compressor settings someone suggested really came in handy. The first track is very even amplitude wise.

NB: Dynamic range compression can make sibilance worse.

Excess sibilance can be caused by compression,

https://en.wikipedia.org/wiki/De-essing

Set the threshold (aka "floor") for compression as high* as will produce the "even amplitude" level you require,
This will leave the low amplitude shh sounds at low amplitude, if you set the threshold low the shh sounds will be made louder.

[ * Confusingly highest threshold means the lowest -dB value, e.g. use a -20dB level if you can, rather than say -30dB ]

... the panty hose is doing wonders for the p's and t's. It was definitely worth the funny work some old man gave me as I was selecting it in the grocery store.

heavy-duty old-lady type "surgical stockings" aka "support hose" produce a good result, but are the most embarrassing to ask for.

[kozikowski]

As for listening to my voice while I'm recording it, not sure if that's going to happen. Seems very annoying, especially with a very slight latency.

Right. That's why you monitor either at the microphone...

http://www.kozco.com/tech/audacity/pix/ ... ctions.jpg

Or at the mixer or with your cupped hand, not at the computer.

but are the most embarrassing to ask for.

You're doing it wrong. When they present it to you, you need to ask, "Do you think this will go with my coloring?" and then ask about proper care and laundering instructions.

Regarding this picture...

http://www.kozco.com/tech/audacity/pix/ ... ndSilk.jpg

That's dress material and the lady at the cloth store asked me what I was going to use it for. I finally broke the charade of creating a fabulous wedding dress and admitted I was going to use it as a lighting scrim for photography.

Koz