It means I think it could sound better but I don’t know why yet. I know you have sub-audible interference (thunder) that could be affecting the other tools.
Yes, continually stepping on original, live recordings with corrections is a very New User error. If the edit becomes damaged or the computer falls over without a safety backup, you get to go straight back to the beginning and read it all again.
Pretty much what you’re doing now, although I would record a test clip rather than a whole page.
As a very fuzzy rule, the microphone doesn’t matter as much as the room. I have a terrific room and I can almost get away with presenting into my laptop microphone.
I can’t listen until I get back home.
Noise Reduction doesn’t work so well on traffic. It works on constant noises, and traffic is always changing.
I may be able to give you a Noise bump without Noise Reduction.
But the more serious problem is you sound like you’re announcing into a bath towel.
– MultiNotch
You have an effect called Yeti Curse. That’s the frying mosquitoes whine in the background. You can’t get rid of it with normal Noise Reduction. That just makes you sound honky and talking into a wine glass.
– LFRolloff
That’s basically turning the bass control down in a very controlled way. This helps you with ACX Noise and gets rid of DC offset. You can’t actually hear most of this working, but it helps a lot with ACX.
– SetRMS
You know about this one. That forces your loudness to meet ACX.
– Limiter
gets rid of the peak problems caused by SetRMS.
– VE4VoiceEq
This brightens you up and gets rid of that talking into a towel effect.
– Noise Reduction
Those tools boost the noise a little, so this gets you under ACX noise compliance without sounding like it’s working. If you just hate that quiet frying sound, you can bump that up a little without causing problems.
After I got done getting rid of the mud (bath towel) sound, you can actually hear the room in there now, but it’s not very bad. and I think it’s good to go.
Put something on the timeline > Effect > Equalization. Look for VE4VoiceEq preset. Do you have that? I can’t tell if it’s something I added or not. I don’t think so.
Three of those are pre-baked effects. The other three are custom programs and will have to be installed or loaded as needed.
Every chapter.
And yes, I understand you said this was a quiet reading because of low traffic. These readings may not pass any more if the traffic picks up. That mud/towel effect covers up a lot of problems
Listen to the clip as you sent it to us, and then the one I corrected.
I’ll post back with details and settings. I had to write them down. You got lucky. Sometimes corrections don’t work.
Attached is MultiNotch, SetRMS and LF-Rolloff. You need to unzip LF-Rolloff. You’ll need to flop back and forth between those tools as needed. There is only one Nyquist Tool, so you’ll need to keep opening the different programs as you need them. Once you get Equalization working, that’s just a selector window to go back and forth.
Open the show.
Effect > Nyquist Prompt: Load > MultiNotch.ny > OK.
Effect > Equalization: Select Curve: LF rolloff for speech, Length of Filter, around 5000 > OK.
Adding Audacity Equalization Curves
– Select something on the timeline.
– Effect > Equalization > Save/Manage Curves > Import
– Select LF_rolloff_for_speech.xml > OK. (it won’t open the ZIP. You have to decompress it)
– LF rolloff for speech now appears in the equalization preset curve list.
Effect > Nyquist Prompt: Load > SetRMS.ny > OK.
Effect > Limiter: Soft, 0, 0, -3.5, 10, No > OK.
Effect > Equalization: Select Curve > VE4VoiceEq > OK.
Effect > Noise Reduction: 6, 6, 6 > OK. If you want a little more suppression use 12, 6, 6. Don’t go over that.
Missed one. SetRMS has settings but the default ones should be good to go. You should always be in mono (one blue wave), so the stereo settings don’t mean anything.
It’s looking more and more like VE4VoiceEq is an equalization curve I made for someone else a while ago. It’s a little honky in this case so I’ll probably see if I can stop that and still “crisp up” the voice a little—get rid of that flannel blanket sound.
Did the posted, corrected sample sound reasonable? The new voice quality equalization I left should sound more natural. Did I lose you in the process? Higher end microphones do better. I recorded a sound test in my super quiet third bedroom with a good microphone and I got one test to pass just by making it louder. We kept telling people you could do that so I wanted to see if I could.
Lower end microphones can start a death spiral. The volume is too low so you correct that. That brings the noise up, so you fix that. That makes the voice sound funny…
I did wonder what that mosquito noise was
We know what it is. That’s the USB data flashing back and forth getting into the little part that catches your voice. Higher end microphones and other devices don’t do that because they have additional protective shielding and filtering to get rid of it.
Expensive filtering. The difference between the Yeti microphone and the Yeti Pro is double the price. The whine doesn’t happen to everybody.
You are also missing performance and quality issues that many other people have with much better equipment, so you really got lucky.
MultiNotch written by Steve brutally smashes certain musical tones that we determine were the worst offenders. They’re just gone. We’re hoping nobody looks too hard at the show and you don’t try to send music through the system.
The tones are 1000Hz, 2000Hz, 3000Hz and 4000Hz. Those roughly correspond to the signals that USB uses to do its job. You might know what 1000Hz is. That’s one of the tones Broadcasters use, for example, behind colorbars as a test signal.
The more normal test tone is 400Hz. Never being one to follow the pack, NBC New York used 400 and NBC Washington used 440Hz.
You totally know what 3000Hz is. That’s a baby screaming on a jet. That’s one of the reasons this whine sound can be low level and still so irritating.
Normal piano tones don’t ever land exactly one one of those tones, but if you have an instrument with no fixed pitches such as a violin or a slide trombone, you could hit musical tones that don’t make it into your show.
I just got a chance to listen to the “fixed” version with no mosquitoes, and it sounds GREAT to me. Thanks for all your hard work.
I’m mostly interested (of course) in passing the ACX check robot, without getting over processed, so it’s been good to get a little feedback before committing to reading a whole book.
Each time we correct a performer’s work, we get better at it. This pass saw the debut of MultiNotch. Until now, that was the digital equivalent of writing a filter program on a napkin at a breakfast shop. I remember seeing it posted for someone else and I went down the listings until I found it.
ACX Check looks like a simple and perfectly obvious tool, but if I told you how you had to do it using the older manual effects, you’d fall on the floor laughing.
And etc.
I think I used 12, 6, 6 in the best version of Noise Reduction. Noise Removal of the Beast (6, 6, 6) isn’t quite enough for you.