Effects Order for Audiobook Narration

What is the best order to apply the following Effects for my Audiobook recording?

  1. Noise reduction
  2. EQ
  3. Compression
  4. Normalize
  5. Limiter
  6. De-Esser
  7. De-Clicker
  8. Low-Pass Filter
  9. Change Tempo
  10. Change Pitch
1 Like

High-Pass Filter should come before effects that have a threshold setting, e.g. compression & De-Clicker.

Compression should precede de-essing, as compression with make-up-gain will increase sibilance.

Noise-reduction should precede compression, as compression will cause the noise-level to fluctuate.

Where did you get that list?

I’m curious about 9 and 10. Do you talk fast in a high, squeaky voice?

Fair warning, if ACX Quality Control can tell you applied effects, they may reject you without getting any further than that.

There is a global test. They hate Distractions. Anything that takes the listener’s attention away from the show is to be avoided.

There is a YouTube presenter who has terrific graphics, visuals, production values, and does good research on the subject. I can’t watch him because he sounds like he’s announcing in a bathroom. He’s getting better at it, but I still can’t watch because he sounds like he applied effects to get rid of the bathroom.

I have some notes: One of my favorite announcers is Sarah Vowell who has a voice about as far from Broadcast Presenting as you can get. She’s also on her fifth or sixth audiobook.

We authored tools and processes for Audiobook Mastering. One of the requirements of tool design was that you can’t tell what we did. The quality sounds exactly like you, only now it passes ACX.

If you start messing with speed and pitch, that will dig you a hole very quickly.

Koz

Post a voice sample on the forum. I think the system will allow that now.

https://www.kozco.com/tech/audacity/TestClip/Record_A_Clip.html

Koz

Thank you Trebor. This is very helpful.

The list is my own. I thought I needed these effects. Your comments are extremely helpful. I realize I should not mess with my original recording so much. From the sample I’ve provided, can you tell me what you think is needed. I’m recording with a broken (missing) front tooth therefore in many places there is a lot of sibilance. (I’m waiting for a transplant but that takes 4 months and I have a deadline for the book). I feel a need to increase the tempo because it feels like the recording needs a bit more energy. I have tested a 1% or 2% increase and that sounds much more energetic and alive to me. What do you think? THANKS SO MUCH. All of this is brand new to me. Your advice is deeply appreciated.

Here’s the same sample with a 2% Tempo increase. What do you think?

You’re right. Your native presentation drags a bit. This is much better, much snappier.

I’m listening to the rest. Some tones in your voice are causing problems in my sound system, so you have graduated to being a diagnostic tool.

Koz

One obsessive engineer note. You should avoid “exotic” characters in filenames. You are going to hit a client that has no idea what to do with a percent sign (%) in the name.

As a general rule, upper case, lower case, numbers, dash and underscore are safe. Being obsessive, I don’t use spaces, either. Hello_how_are_you.wav

The slashmarks are particularly troublesome. Those may mean theatrical paragraph divisions to you, but to the computer, they mean folder and directory structures.

Koz

One downside of a laundry list of effects and filters is the need to remember to apply them to every chapter until the sun cools off. ACX insists your chapters match.

Another side issue is patching, correcting, and editing. You have to remember the order of your filters.

Koz

Found it. I had some metallic office tools parked on top of a speaker and certain tones in your voice were causing the tools to rattle.

That’s only partially good news. You have Room Resonances in your voice. Some bass notes are artificially stressed.

I removed the boomy resonances, but that made your missing tooth worse. You may get your laundry lists of effects, yet, but just different ones.

First phrase is original.

You may like the boom, but as mentioned, that caused problems in my sound system, plus, it messes up audiobook mastering.

As we go.

Koz

As I rip this up, please know that in general, I would have no trouble listening to a story in that voice. You have good storytelling.

Two odd things you should know. I need to be able to buy your book on Amazon right now, and the book can’t be on this forbidden list (scroll down).

https://www.acx.com/help/200878270

They won’t publish a cookbook.

Koz

This is where we ooze out of my comfort zone. I have almost no experience with sibilance. I think I had good luck with the DeSibilator. I’ll look.

Koz

Well, Ummmmm.

I can do it, but it doesn’t make your tooth go away. It just turns a harsh SSS sound into a softer shshsh sound. It still doesn’t sound entirely natural.

They were exSSSited about joining uSSS.

They were exshshited about joining ushsh.

You may have orthodontically pushed yourself beyond our ability to help.

Koz

I’m really close to the Hollywood Solution. Get somebody else to read it.

Koz

Thanks so much Koz for spending so much time on this. You must really love tweaking audio for strangers or coaching newbies. Either way, THANKS!

I have read that Low-Pass Filter plus Compression can help Sibilance.

I plan on re-recording this once my tooth is repaired at the end of the year. But I have a special event in October where I want this available. At the beginning of the book I will apologize and explain. This event is with a friendly audience so they will be forgiving. I won’t go public until after I can record properly. I won’t use Audible, I’ll go with a different platform.

Thanks again Koz.

That’s essentially what DeSibilator does in one package. The tools, no matter which ones you use, leave an SSS-shaped hole. The timing gets distorted. The tools don’t address that.

I skated right over your loudness problem. You could use some more volume in the raw reading.

Your raw submission looks like this.

Since you are expected to edit and do mistake corrections with that, it’s good if it was louder. Like this.

The fuzzy rule is occasional tips and peaks about 0.5 on the timeline.

Low recording volume can affect microphone noise and compression settings > see sibilance correction.

If this was easy, anybody could do it.

Koz

There is stereo reverb(eration) on both the samples you’ve uploaded.
Unless you have a stereo microphone, that’s been added as effect.
Reverberation (stereo or mono) undermines intelligibility.

Even processing by Artificial Intelligence (AI) cannot fix the effect of a missing tooth, but it can make the voice clearer: e.g. removing the reverberation …

before-after “Adobe Enhance” AI

There may be a prize for a natural announcing style that Artificial Intelligence (AI) can’t fix. I’m checking.

Koz