Hoping for input re: my hard-limiting of a Vox Only file

[espritbonne]

Note: I couldn't post them earlier, but the links for the audio files (before and after the hard-limiting) are now posted below.

Thanks so much for any input regarding my settings for hard-limiting a voice only mp3 file. I’ve been a voice actor for over 25 years and have produced most of my own audio during that time. Originally I voiced from the radio stations I was jocking at, but in 2001 I built my own studio. Because I was hosting a talk show in Washington DC at that time but living 90 minutes outside the city, I wanted to be able to do my show from home via ISDN whenever necessary. So I built my home studio to be similar to the set up in the On Air studio at the radio station: an ElectroVoice RE20 mic, through a Symetrix 528e processor, using a Mackie Board and a DAL Digital Deluxe soundcard.

Many of my VO clients like me to send them nearly flat audio, for them to tweak as they like. But I’ve started voicing audio books, and my clients now have varied requirements. I’m currently voicing a book that requires mastering, and since it’s voice only audio, I’m hoping it won’t be too difficult for me to figure out what sounds best. But I’ve never totally trusted my ear. I’ve always had a sibilant S challenge, and my voice is lower than many females, but if I work too hard to correct those, I create other, possibly worse issues. Hope you can spare a few minutes to check out the links below for my :40 excerpt from my original .wav audio (as recorded, and using their requirements: 44.1kHz, 16 bit, stereo) and compare it to the hard-limited version saved (as they require for finished audio) as an mp3, at 192kbps. For the hard-limiting, I kept the max amp at -6 dB and boosted by 6 dB, with the look ahead time at 7 ms and release time at 100 ms. Thanks again, I really appreciate your time. And please don’t worry about hurting my feelings, I’m tough.

Here's the link for the original .wav file:
www.debraleigh.com/AsRecorded_AudioBookExcerpt_44.1kHz16bitStereoPCMwavfile.wav

And here's the link for the mp3 file in the format the audiobook publisher wants, with the hard-limiting:

 

[espritbonne]

Here are the links for the original wav and the hardlimited mp3

Okay, I can post the link for the original .wav file now. Here it is, and below it I'll post the link for the hard-limited mp3 version. Thanks again for your thoughts. Hoping the hard-limited version sounds okay, (see my post above for specs) but if not, don't worry about being brutal.

www.debraleigh.com/AsRecorded_AudioBookExcerpt_44.1kHz16bitStereoPCMwavfile.wav

 

[DaleVO]

Debra,

The hard-limited version's quality sounds good to my ears. However, I am hearing the sibilance of which you spoke. Did you run a de-esser or EQ the mp3?
Dale

 

[espritbonne]

Thanks for checking out the files, Dale. I've wrestled with my processor settings for years regarding my nemisis. I mean, my sibilance. But nothing ever seems to solve the problem. Right now, my Symetrix 528e De-Ess frequency is set at about 3 o'clock, maybe 3:30. The threshold is set between noon and 1pm. When I try to De-Ess with software (any software I've ever used) I seem to only make things worse. By that, I mean the other problems I created while trying to lower the sibilance were even worse than the ssssibilance. (So you KNOW they were bad!) So I gave up trying. Do my Symetrix settings sound to you like they're exactly where they should NOT be? Thanks again.

 

[kip4]

Great voice btw. Im surprised it sounds quite so sibbillant given the mic your using (no offence). Maybe try using a differant mic. Idk maybe a ribbon or sm7b. Honestly i'm no expert but i'm sure theres plenty here that can help. Given that the text contains so many sibbillant sounds i still think it sounds good.

 

[espritbonne]

Oh wow, thanks for your kind words kip4. And thanks for the input. Over the years, I've tried different mics, but the RE20 actually works better for me than any others. So, it's just me being a bozo with my Symetrix settings. (When the engineers at radio stations I worked at controlled the Symetrix settings, my sibilance magically disappeared) I've been working on the settings this evening, and may have diminished it somewhat. Or may have worsened it. Thanks again for checking out the files and for your thoughts, I really appreciate it.

 

[grimtraveller]

And please don’t worry about hurting my feelings, I’m tough.

You might be, but I'm not ! Besides which, toughness is no defence against hurt feelings.

don't worry about being brutal.

But I do, I do....

I am curious about one thing though. Do you use a pop shield when speaking to the mic or do you ever speak off axis, that is, not directly head on at the mike but at an angle to it ? That might help a little. I found that it was quite sibbilant and the recordings were very 'bright'. In the old days, I would've called it 'trebly'.
But this is a new day !
Also, before you had your own studio, how were you recorded ? Was the sibilance noticeable then ? Is there any way of contacting any of the people that have recorded you and picking their brains ? It's just that you imply that you've had a sibilance challenge for a while and you've sent audio for others to tweak so they must have a way of ironing it all out to the satisfaction of all.
By the way, you have a very television voice.

 

[espritbonne]

Grimtraveller! What a GREAT idea! Duh, why didn't I ever think of this?! I've been supplying dry VO to the same 60 or so production houses for over 10 years, now - and clearly their sound guys have been working their magic successfully or they wouldn't have kept ordering audio from me! They probably have s-s-saved s-s-ettings just for me and my s-s-sibilance that they can pull up in a heartbeat and share with me. And I probably won't have to ask too many of them before I can determine what I'm doing wrong on my end. Thank you, THANK you!

And thanks for the other suggestions, too. Yep, I've got a nice windscreen on my RE20. And I've always angled my delivery to the mic because in addition to hissing, I'm also 'gifted' with the ability to pop not just P's, but pretty much any consonant. How in the world did I wind up a voice actor with all these 'gifts'? Jeesh.

From now on when I'm voicing at other studios here in LA, when there's a moment after a session, or during a break, maybe I could ask the engineer if he wouldn't mind briefly sharing with me what he's doing to correct my sibilance? And you've led me to another idea - during Source-Connect sessions occasionally the engineer on the other end will ask me to make a specific change to my De-Ess setting. Why haven't I kept notes about their requests?!? (Instead of just dialing back, after the session, to the "Duh Debra" settings that they were clearly trying to correct? Embarrassing, realizing how dumb this was) And finally, now I'm wondering if Symetrix has a service we can purchase, to hook up with them via Source Connect or ISDN and have them monitor us for a minute or so, then talk us through adjustments?

I know everyone's ear is different, so I'll probably get some conflicting info, but somehow, a pattern will begin to develop, I'm sure. I think I subconsciously or maybe purposefully really deadened my studio originally to counter the 'overly-bright' s-s-sound. But after getting suggestions to brighten it, I went to the other extreme? Several VO friends who work from their own studios recommended years ago that I trust my own ear, and that I never let people on the other end of a session make me change my settings. But my sweet friends were assuming my 'ear' was as good as it should be for my line of work, and boy is it not.

Thank you so, SO much Grim! And thanks also for your very kind words, and concerns about my feelings. But I treasure brutal honesty when it comes to something like this. You've been so helpful, though. As has everyone here. Thanks again!

 

[chuckduffy]

The solution of contacting an engineer you worked with in the past is probably the best. Barring that - when you are speaking into the microphone, how many dB of reduction is being indicated on the LEDs on the DE-ESS section of the Symetrix ? Repeat the same sibilant word (sibilance for example) over and over and watch the LEDs.

 

[DaleVO]

Debra,
I played your track in my DAW and found the "S"s hitting hard at 6200-6800Hz, if my graph didn't lie too badly. I am also curious about the DE-ESS reduction that Chuck asked about, and what are your EQ settings, especially the Parametric EQ High settings?

FREQUENCY
BANDWIDTH
CUT/BOOST

Dale

 

[espritbonne]

Wow, Chuck, how embarrassing. It's like finding out I've been driving a car for years without ever really understanding what the brakes, clutch, gearshift, etc actually do. Or how they affect each other. How have I even managed to ever drive from point A to point B all these years? I'm beyond Bozo Level with all of this - I'm at Imbecile Level.

When I tried the dB reduction test (using the same settings I had when I recorded the audio linked in my original post here) there was absolutely nothing - zero reduction. Jeez. Incidentally, I had changed my DE-ESS settings since recording that audio the other day, but to do this test you suggested, I put everything back just as it had been before. And here are those DE-ESS and Parametric EQ settings:

DE-ESS
1. Frequency knob @ 3 o’clock (6K ?)
2. Threshold @ about 12:30 (-13 or -12 dB?)

Low EQ
1. Freq @ 11 o’clock
2. Bandwidth @ 3
3. Cut/Boost @ noon

Mid EQ
1. Freq @ 2 o’clock
2. Bandwidth @ 3
3. Cut/Boost @ noon

High EQ:
1. Frequency almost all the way clockwise, @ about 4 o’clock, 21K?
2. Bandwidth @ 3 o’clock (roughly 3 octaves?)
3. Cut/boost @ noon

Seeing the High EQ freq setting boggles my mind. Using the car analogy I mentioned above, I should have my license AND my car taken away, right? It’s scary that I evidently arrived at these settings last week while twirling knobs letting my “ear” decide what sounded right.

During the dB reduction test: at original settings NO lights, ever. When I brought the DE-ESS frequency to noon (-15) and the threshold to about 11 o’clock (-18) I finally started seeing -2 dB, sometimes -4. I turned the frequency and threshold knobs both to about 11 o’clock, and still just saw -4. But with the frequency and threshold knobs both @ 9 o’clock, the -6 dB lit. I wish I would have thought to play with the High EQ frequency while I was at it.

Thank you so much, Chuck. And Dale. Man, I wish I had an 'ear' and could simply keep making gentle adjustments clockwise and counterclockwise until everything sounded 'right'. But clearly, the audio area in my brain has absolutely no clue how to process things. Pun maybe intended. You guys rock, thanks for your help.

 

[espritbonne]

Thanks, Dale!

Thanks so much, Dale. I've posted the Parametric settings along with the results of the dB reduction test Chuck suggested in a separate reply. I really can't thank you guys enough for your time and your thoughts. I didn't want to bother any engineers on a Friday afternoon, so I didn't contact anyone yet to ask them how they manage to tame my 6200-6800Hz worth of S Power, (thanks for that info, by the way!) but I spent time today pouring through the Symetrix website and 528e's user guide. But for me it's always like trying to read novels in French - I can understand just enough to be dangerous, you know? And to probably glean from it just exactly the opposite of what it was really trying to convey. I'm going to try using some other people's settings tomorrow and see if anything works for me: Where can I find some example settings from users for the 528E Voice Processor?

 

[DaleVO]

Debra,

I inferred that the DE-ESS tweaking may have knocked down the Ss a bit for you. I hope that we did help you find some magic tweaks.

IMHO, your voice quality, delivery, and acoustic environment sounds really good to me. The sibilance is not "horrible" in your recordings, although noticeable as you are already aware. We have a local (DFW) NBCTV, traffic reporter that shrills my ears, when she Ss. Yours is no where near that level.

You mentioned having your driver's license revoked... I believe you said that this is new territory for you. I assume that you have been doing what you were supposed to do as a VA, and "obviously" (from your good sound) focusing on your craft, rather than the mechanics.

I am at "Imbecile Level" too, still learning and tweaking. Anytime you want some ears, just holler.
Dale
PS: emailed the information that you requested.

 

[chuckduffy]

ok - here's lots of good news....

1. Sounds like you have a good dead room - the mic doesn't seem to be picking up any room sound or any high frequency reflections at all, even when compressed and hard limited. GREAT

2. With the EQ settings you describe (cut boost @ noon) you aren't actually boosting or cutting anything. GREAT natural speaking voice and mic combination doesn't really need any EQ - GREAT - just turn the eq section off.

3. You have no problem maintaining a very level speaking voice. GOOD NEWS - just turn off the compressor.

4. You do have some sibilance going, not horrible, but it's there. For now just turn the DE-ESS section off too - fix that after we get the levels right.

5. Now that the compressor and eq are turned off turn the MIC PRE-AMP Gain knob all the way to the left. SLOWLY turn the MIC PRE-AMP MIC Gain knob to the right while watching the meters on whatever software you record to. Stop when the meters peak at around -10.

6. Now re-engage the DE-ESS section. I agree with dale that the problem freq is somewhere around 6500. Set the FREQUENCY knob on 3 clock (should be about for 6500) and the THRESHOLD @ 9 clock. You should be getting really healthy reduction on sibilant stuff now. If it's too much reduction and saps too much high end simply move the threshold slowly towards noon until it sounds better.

 

[espritbonne]

Aw, man, thanks so much Dale! At first I thought you'd misread my analogy and you thought I'd actually had my driver's license taken away! Then I re-read your post, and realized what you were saying. Can't tell you how much that means to me, taking a moment to reconsider all this from the perspective that maybe I've just been focusing understandably more on the creative side of my work than the technical. Thank you! At the same time, I know I need a better understanding of what all these little knobs and meters really DO, and how they relate to each other. I'm so indebted to you and the others here who have patiently shared your ideas and knowledge. You've inspired me to stop being intimidated by the technical side of my work, and instead, just enjoy exploring it! Thank you for that, and also for your kind comments about my VO. And for the info you sent, much appreciated!

 

[espritbonne]

WHAT!?!? Are you kidding me? Get outa town. Wow, WOW!!! Thank you SO much, Chuck! Terrific news all around!! Ha! How funny is that about the cut/boost settings at noon? I actually read that on the Symetrix website this morning! I laughed out loud to think how I'd been all concerned about the frequency and bandwidth settings in my EQ, but meanwhile, I was cancelling it all out by keeping the cut/boost at noon!!

This is SO exciting!!! I can't wait to head into the studio now and get started. I've been reading, watching tutorials, looking up the terms they use that are like Martian to me --- and coming gradually closer to understanding it, but in a still-sorta-out-of-focus way. But now, something tangible and immediate. Thank you! The publisher is waiting for this audio-book, and I've fallen so far behind schedule, but I've been determined to get my audio corrected before I do this book. You've just saved me so much time, Chuck. And everyone else who has contributed to this thread. Thank you!

Okay, I'm off to the studio. Can't wait to see how it goes. I'm so glad I discovered this website, and you terrific people who hang out here. I'm forever grateful to you all.

 

[chuckduffy]

Don't thank me yet The funny thing about forums is someone else will always come along and say 'you told her to do what!!!?' The important thing is to get the recording done at a proper level. Outside of some sibilance I don't hear any of the other typical problems like picking up the room, uneven speaking volume, etc. I'm sure you could find someone to de-ess, compress/hard limit and eq tweak for a very reasonable price after the fact as it's a single mono track of spoken word. Might even take the pressure off you to focus on your work.

 

[DaleVO]

+1 Chuck
I pulled the manual on the 528 and scanned it briefly, last night, with intentions to dig a little deeper. I think you were spot-on with your tweaks, and getting her back in business. Good call!!

Debra:
Post some of the new recording, when you get it tweaked. Inquiring minds will want to know.

Isn't it fun? Don't forget to make notes, as you tweak.
Good luck,
Dale

 

[espritbonne]

Haha! Uh-oh. Be careful what you ask for, Dale! ( "post some of the new recording..." )

Some? How about links for 6 different reads, one with zero compression, the other 5 with various compression settings. But none of them with any EQ, I left that section off. (So this time I actually KNEW there was no EQ going on, Chuck! Haha!) Each file's title describes the settings. And I'll post the original .wav and hardlimited mp3 version (in the format this audio book publisher requires) for each of the 6 reads.

Chuck, I couldn't imagine the studio was quiet enough for zero compression - I hear noise in the room. But I gotta admit, it was WAY more quiet than I'd imagined it would be. Still, because of the huge boost audio book files require (I guess to be sure even the truck driver barreling down a noisy highway can hear every little whisper) I'm just not comfortable with how much extra work that might take to make it sound okay. Thanks for the idea of having it mastered elsewhere, good points you've made about that, but actually that's the whole reason I'm doing this. Since it's just vo, I really think/hope I can learn how to make it work. I'll be doing tons of audio books for these folks from now on, so I just really want to master this. Wow. Pun not intended.

And Dale, heck YES it was fun! I can't believe I just spent over 4 hours in the studio, and it felt like maybe 90 minutes! So cool to be knowing more about what I'm doing and hearing now!! Thank you guys! But I still have a long ways to go. Again, don't worry about hurting my feelings if not even one of these is acceptable. But there's one that I kinda think might be...



www.debraleigh.com/NoCompression_NoEQ_DeEssFreq3Thresh9-15_OrigWavRead.wav

www.debraleigh.com/CompThresh_11_ExpThresh_Noon_CompRatio_2oclock_OrigWav.wav

www.debraleigh.com/CompThresh_2-45_ExpThresh_Noon_CompRatio_2oclock_OrigWav.wav

www.debraleigh.com/CompThresh_2-45_ExpThresh_11-15_CompRatio_2oclock_OrigWav.wav

www.debraleigh.com/CompThresh_2-45_ExpThresh_10-30_CompRatio_2oclock_OrigWav.wav

www.debraleigh.com/CompThresh_2-45_ExpThresh_1oclock_CompRatio_2oclock_OrigWav.wav

 

[chuckduffy]

Hi,

I'm horrible at blind tests - but I can tall you that all of the samples have less sibilance than the original, and I don't the sibilance in any of them irritating in any way. The last sample seems to have the least. The second sample seems to have the best balance between dampening sibilance and losing high frequency content.

"Chuck, I couldn't imagine the studio was quiet enough for zero compression - I hear noise in the room." -

Somehow I think you got the impression that compression is a tool to lower room noise, sadly it's not - there is nothing (I know of) that can remove the room and ambient noise from a recording - which is why you hear people go on and on about 'treatment' and 'dead space'. Compression does *not* lower or prevent room noise in your recording. It functions by lowering the audio signal of the peaks, which allows you to raise the overall gain of the track without clipping, which in turn makes quiet things like ambient noise *louder*.

With all that said - there is nothing wrong with compression, I understand that it is the norm in the industry and if it sounds better with it on - use it

I also noticed that your voice has a suprising amount of low frequency content - especially with the sub engaged on my monitors. You may also want to engage a high pass filter in your recording software to filter out frequencies below 100 or so. I don't believe the Symetrix has a HPF. Take this for what it's worth though, as I don't believe people listen to audio books with sub-woofers