HELP with EQ! Audiobook due yesterday!

Hi, I’m hoping someone can help me figure my sound out. So I am doing my first audiobook (yes due yesterday) and have been having nothing but problems! I swear I didn’t leave it til last minute, I have been working hard at it all along but all the obstacles have made it take SO much longer!

First I got partway through recording, and my computer decided to delete everything! Turns out there are issues with my hard drive so at one point it logged me into a TEMP profile, I didnt realize I was saving everything into that, and when I restarted my computer (it did an update over night) it was all deleted.

Then I had a limited amount of time to record, got an hour in and it decided to update mid-recording when it wasnt saved of course, and that was lost too. went through a whole bunch with my uncle who is a computer guy and remoted in, and figured out how to save things properly.

Everything was lost so I started again, but then my neighbors (duplex) started renovating their place so it sounded like drills were going through the wall right next to me anywhere in the house, even closets. So I started recording in my car. So some of my recordings are in a smallish room with foam around me and a box lined with foam around my mic, but some are in my car with same foam accessories. Different sound. It was all I could do with the time I had and I was hoping the sound would be similar enough or I could do something with editing, but I’m not familiar enough with all the mastering to get it to sound right.

I have tried multiple attempts and trying to normalize and have smooth transitions between different recording sessions/spaces, but I’m panicking and need this done yesterday and don’t have time to re record any of it. I have learned a lot and watched lots of tutorials but I have really strong sibilance and once I apply all the effects it sounds off.

Please help!

This is a short clip before editing:

This is the same spot after applying noise reduction, and declicking

This is another transition raw:

And same section with noise reduction, added slight bass to second section to match more with 1st section sound, and declicked.

I am currently also running a macro I set up on it, De-esser (Paul_L’s), Compressor with Threshold at -20dB and noise floor at -40dB, then Loudness Normalization to -18, and Limiter to -3.2dB. I will post that once it is done. my laptop is also incredibly slow so it takes forever to apply effects. I actually already ran this macro on this audio, but was trying to run another de-esser at the end and was playing around with the settings to find the right sound and it crashed Audacity so I have to run it again.

I have a plan to replace my hard drive with an SSD one soon, but right now I just need to get this audiobook out asap so I am trying to make it as passable as possible.

This is the first section I posted but once I applied the macro I mentioned in the post:

And the second section, also with the macro applied:

Trying to match different recording spaces is impossible:
they have different resonances & reverberation times.

IMO that has too much noise for audiobook.
Yes the noise can be removed to pass ACX, but the speech will then sound computery.

BTW there’s a big gouge missing from the middle of the spectrum on “Ch 2 Sample Raw”
big-notch
Sound reflections in the location you’re recording are the prime suspect, (destructive interference).
Boosting the EQ at that frequency (~1100Hz) to fill in the gap is going to add a lot of noise, because the signal-to-noise ratio is particularly low there.

Hi! Thank you for replying promptly.

So with this due now 2 days ago, am I better off re recording the parts that were recorded in the small room, and recording them in my car instead? Should i start from scratch?

I’ve had so many issues here i planned to do it all in the same space but the construction noises were super loud so of course i couldnt continue there, and sometimes there were also loud noises when in my car that interfered with recording there.

I looked at this authors other book that they had made into an audiobook and it sounds awful and like AI, and they dont seem very concerned about how it sounds, i accidentally uploaded the wrong file for the 15 min checkpoint that i had accidentally trimmed a word while moving sections over, and before i could upload the correct one they had already accepted it.

I only have a few hours a day while my kids are in school so recording in one session is difficult. The book is only about 4 hours long but i have spent so much time on it with having to re record, so the later parts i was trying to rush through and didnt put as much time in.

I am going on vacation tomorrow and coming back Sunday so i wont be able to work on it at all after today, until monday. I feel awful being late on this, is it common for producers to be late? Im only making $55 pfh on this, I knew it wasnt much but was treating it like a training experience to learn how to do this and see if i want to do it long term.

With being late and it sounding how it does, I’m feeling like giving up and saying i cant do it but they’ve already been waiting since November 4th. Since their expectations arent super high im wondering if i can make it passable, but i dont know if thats possible or if it will be rejected by acx once i do upload it. I’ve got about 4 hours left today where i can work on it before i have to pick up my kids and then i have to get ready for vacation! :sweat:

AI can do a better job of removing noise than Audacity …
before-after Adobe Enhance.
It’s also doing de-ess, de-reverb & EQ.

Oh crazy! Too bad it sounds a bit more like ai, I was looking to see if there was some sort of ai i could use to fix it but couldt find any that wasnt going to cost a lot. the construction seems to have stopped so I was able to set up a better space. this is raw sound in there:

I found the acx check in audacity would say I was at -3.2db peaks after running the macro, and that RMS was around -22, but when I ran it through audiolab it was coming out at 0db peaks, and -14 RMS. I changed the levels in my macro and managed to get a fail of acx check but pass of audio lab.

Thank you so much for sharing that though, I will definitely check it out!

This is the same clip after applying the effects that had it passing audiolab

Edit*Its also crazy how muh exporting as mp3 distorts it! It sounds so so different in audacity. I’m trying to keep my sibilance under control too and its still coming out so strong.

I’m also wondering if I should be recording somewhere different than audacity? Is it audacity or the microphone? I’m using an aluratech rocket usb mic and recording in audacity. at this point im not in a position to pay for any paid DAW or a new microphone. But definitely surprised by how different my sound is when exporting from audacity! When I ran the raw recordings through acx check (from this new session) I was getting passing levels on everything with it raw.

The spectrum looks a lot better on that: no big gouges missing …
before-after
The free plugins I used on the “after” …
GMulti multi(3)band compressor can be used to de-ess.
[ Much better than Audacity’s native compressor].
Couture (free version): which is like a noise-gate,
attenuating the noise when you are not speaking, and can reduce reverb a little.
.

When you make an mp3 in Audacity it can have peaks which are +2dB louder than the track in the project, that can result in clipping-distortion in the mp3 which wasn’t in the Audacity project.
The solution is to turn down the track gain slider by -2dB just before you make an mp3 of it.

1 Like

OK. Ignoring most of that. I can’t help you if you changed studios. That’s really rough to deal with.

However. Starting with your first clip “before editing.”

I applied the Audiobook Mastering Macro which guarantees Peak and RMS (Loudness).

That was too noisy, so I applied gentle, but serious Noise Reduction.

NoiseReduction_2023-11-30

That left some “essing.” so I applied DeSibilator.

DeSibilator

and then tested with ACX-Check.

ACX-Check_2023-11-30

This is the track.

The only real mystery is where the “essing” is coming from. Microphones don’t normally do that on their own. I have a pro mic that sounds overly bright, but even it doesn’t do that.

I don’t remember where I got the deSibilator from, but the Audiobook Mastering Macro is part of the audiobook tool collection.

As we go…

Koz

One more. I have no idea where the MP3 distortion is coming from, but a simple File > Export at 128 Constant Quality MP3 should sound identical to the WAV audio master file or the show on the timeline.

Koz

A trip back through the archives…

Baaaaack…Baaaaack…

desibilator.ny (56.0 KB)
DeSibilatorNotes.txt (544 Bytes)

Koz

I’m doing all this in Audacity 3.4.2 on a Mac. The New stable release.

Do Not upgrade anything in the middle of a show. Windows presents problems because I believe the latest trick from Microsoft is forcing an update whether you want it or not.

And that brings us to not recording on the computer.

That’s a Zoom H1n stand-alone, high quality sound recorder. Any minute now I’m going to cut a fresh, new test voice track with that configuration

Koz

There are two different Noise Reductions now. I use the older one where you drag-select a clean sample of background noise. Effect > Noise Reduction > Profile.

Then select the whole show and apply the effect. If you have a good, quiet, echo-free studio, you may not need any reduction, but most people can make do with Noise Reduction of the Beast. 6, 6, 6. Do that at the end after mastering.

Koz

Thank you both so much!! I will be downloading all of those tools to try!

Thank you!! This is super helpful! Really appreciate that :smile: makes it much easier than failing the acx check on audacity and having to guess lol

Thank you for all that and the pictures, that definitely helped while still sounding more natural! I wonder if that desibilator works better than paul L’s de-esser? I was struggling to get the right sound on it but i also could have had the wrong settings. I appreciate you going and finding that for me because i was having trouble finding it! I will definitely check that against the other de-esser and see which one sounds better with my sound, but i appreciate you including pics of the levels so i can replicate that and go from there!

For noise reduction, yes that is the way i was doing it as well. With the new space maybe i can get away with that.

Do you mean that the upgrade can change the sound settings? So dont upgrade anything until fully done recording the whole book? And yea my computer updated on its own, literally mid-recording. Thought everything was recording fine, then peek at my computer and see its completely off… :woman_facepalming:t4: its only supposed to do updates at 2am so not sure what happened there.

But I have reached out to the author and explained that i have had some setbacks (didnt go into details) and asked if she would be ok with waiting a couple more weeks and she said that was fine! So now i have some breathing room to re-record in the new space and this time editing shouldnt take so long with these new tools and also having a consistent recording space.

Thank you both SO MUCH for your help!!! This has made a huge difference to me getting out a finished product that hopefully wont sound awful!

Essing is a little rough as well. This may be the older tool that got rid of the piercing SS sounds by turning them into shshsh sounds. The newer one may not do that.

Are you sure there are no apps or packages running on your machine in addition to Audacity? Needle-sharp SS sounds generally don’t happen right out of the microphone. They are a result of processing, or some software “helping you” with your performance.

A while ago, someone discovered heavy noise reduction could cause sharp SS sounds. Thereby trading one problem for another.

I’m going to go back up the thread to find the microphone model. Scarlett Solo microphone interface has an “Air” button which is supposed to “help you” with the voice tones. I think the less help the better.

Koz

AluratekScreen Shot 2023-11-30 at 5.15.53 PM

As near as I can tell, the “front” of the microphone is the side grill above the company name. Not the rounded top.

Select the cardioid (heart-shaped) pattern as that ignores sound arriving from the back. That can help with a live room or environment noises.

Koz

Thank you! I looked to see if there were any settings on the microphone trying to “help” and didnt see any, but im not sure if there are more places to look that i didnt see. I looked in my computers settings and looked in the microphone section for this microphone and didnt see anywhere where anything might be applied. I definitely annunciate my S’s a fair bit naturally, ive discovered lol. But it comes through so much stronger in the recording than it sounds when i speak. Maybe im too close to it? I worried being farther away would make the distance between the noise floor and my voice smaller and cause more issues with my voice after editing.
I do have it set to the cardioid pattern. :blush:

In case I missed it in all the fuss, I understand why you’re in demand. You have a terrific, natural presenting voice.

Koz

1 Like

desibilator is paul L’s de-esser, but with different default settings.

Desibilator / paul L’s de-esser is capable of precisely removing the whistle in your ess …

desibliator setting suggestion#
[Typical de-esser plugins are too blunt a tool to do that].

1 Like

Hi, I can’t figure out how to download the audiobook mastering macro. When I go to the linked page, all it shows is the audiobook mastering macro.txt which brings me to a web page with the code, but i dont know how to turn that into the actual macro plugin?