Having a problem with ACX RMS spec

Narrating and Producing Audiobooks.

ImageIf you require help using Audacity, please post on the forum board relevant to your operating system:
Windows
Mac OS X
GNU/Linux and Unix-like

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by HLC » Thu Aug 24, 2017 7:17 pm

OK, it's no mystery what Mosquito Killer does. The blue waves on the timeline don't change but the whine goes away.

The first step in Mastering Suite (Low Rolloff—Rumble Filter) shouldn't be visible, either.

Effect > RMS Normalize to -20dB.

Effect > Limiter: Soft, 0, 0, -3.5, 10, No.

Noise Reduction of the Beast (6, 6, 6) using the first 1.5 seconds of the clip as Profile or reference.


Followed in your tracks as above, got this:

Audacity sample 2 ACX check Aug 24 2017.JPG
Audacity sample 2 ACX check Aug 24 2017.JPG (72.39 KiB) Viewed 558 times


Not identical twins with your results, but at least fraternal, and the patched recording sounds good to my (admittedly untrained) ears.

THANK YOU!

Is my takeaway here that, assuming the creeks don't rise and no other mishaps occur, and nothing else goes awry, my new booth placement, microphone setting, and additional padding will produce future recordings to which I can apply similar corrections and get similar results?

And to verify, the Audacity Audio Book Guide says

Once LF-Rolloff is used, you have to use it on all chapters. It changes the voice just enough so the sound is not transparent. It is required the chapters match.


so consistency is required with that effect.
HLC
 
Posts: 10
Joined: Mon Aug 21, 2017 2:24 pm
Operating System: Windows 10

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by kozikowski » Thu Aug 24, 2017 9:25 pm

It was originally called "LF Rolloff" while it was in development. The release version is Low Rolloff. Exactly the same thing.

We would rather it did all its tricks without affecting the voice at all...but it does. If you have a lower pitch voice, you might notice Low Rolloff "tightens it up" a bit and is just different enough you can't ignore it.

That means you can't use it ad-lib.

It also does background electronic tricks with some microphones...like yours. For you in particular it's not optional.

Koz
kozikowski
Forum Staff
 
Posts: 38284
Joined: Thu Aug 02, 2007 5:57 pm
Location: Los Angeles
Operating System: OS X 10.9 Mavericks

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by kozikowski » Thu Aug 24, 2017 11:59 pm

Right then. The quote is "Good Lord willin' and the creek don't rise."

Try the correction suite on a longer passage. You should know that the Noise Measurement section of ACX Check can be faked out. If you have a presentation perfect in every way except you never have a half-second or more of clean "hold your breath" room tone, ACX Check will select some part of the show as "silent" and try to measure it.

For Some Reason, the noise measurement on that one chapter is through the roof, but it doesn't sound any different...??

The other thing you can do is "make" perfect silence. Generate > Silence. If the duration is over a half-second, ACX Check see that as room tone and try to measure it. In that case, ACX Check will give an impossibly good noise reading.

I'm sure you're up with this already, but ACX has specific recommendations for number of seconds of room tone and where, and they're all published. We had one performer who got bounced for the wrong number of seconds of lead-in silence. You do have to pay attention.

Low Rolloff is a permanent part of the correction suite. Nothing in the process is optional until you get to the noise page. Each step and tool depends on the tools before.

The goal in writing that was to be able to point a performer to that list and have it always the same. I'm sure you think this dance has gone on too long, but trust me, before we created the custom tools and processes, this audiobook thing used to go for months..

Did you figure out what you were doing wrong? Part of this exercise is to collect user experiences to help others. I think I told you one reader was using everything absolutely, perfectly correctly and kept getting insane results. I don't remember who thought to ask, but they weren't using Audacity 2.1.3.

Koz
kozikowski
Forum Staff
 
Posts: 38284
Joined: Thu Aug 02, 2007 5:57 pm
Location: Los Angeles
Operating System: OS X 10.9 Mavericks

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by HLC » Sat Aug 26, 2017 12:59 am

Did you figure out what you were doing wrong? Part of this exercise is to collect user experiences to help others.


I had to think about this because I know I followed each of your steps exactly when I made corrections to the first recording. I think the difference between my results and yours on that recording was due to the fact that when I prepared the clip for uploading, I highlighted part of the recording and exported. When I went back to follow the steps to make the corrections, I used the entire clip. So there was more to measure, and the measurements were different.

On the second clip, the length was identical and my corrections matched yours fairly closely.

I did another recording today, followed your recommendations, and had a passable recording in less than five minutes.

I'm going to start over recording the book I was having such a problem with. Once I have finished, submitted, and heard back, I will return and report the results.

Again, I thank you very much for your patience and your generosity in sharing your knowledge. I was really frustrated with my attempts to get the recordings right. Over the course of less than a week, you got me on track and have given me the confidence to proceed.
HLC
 
Posts: 10
Joined: Mon Aug 21, 2017 2:24 pm
Operating System: Windows 10

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by kozikowski » Sat Aug 26, 2017 1:29 am

I'm going to start over recording the book I was having such a problem with.

What happens if you just throw one of your existing chapters in the mill and see what happens? We know the noise is going to be high because you weren't using the appropriate microphone pattern. Which one were you using, by the way?
So there was more to measure, and the measurements were different.

I have a super hard time with that because of the way the tools work.

RMS Normalize and Limiter put the sound in very specific places and it doesn't matter if you're wearing a blue dress or had cornflakes for breakfast. I expect all your chapters to end up with peak and RMS values very close to mine and the only variation is background noise. When we built the suite we determine that noise was just too open-ended and had too many variations. It didn't even lend itself to one of those decision trees. "If you have This, Then do That."

It was a temptation to not publish Noise at all and just tell people to post on the forum if they had really high noise.

I think I put the TV in there, right? If your background noise is a TV -> start over.

It is possible to crash the tools. They have a length limit, but I think you have to be talking for five or six hours to hit it. Were you trying to process your whole book at once?

Do try processing an older, existing longer chapter.

Koz
kozikowski
Forum Staff
 
Posts: 38284
Joined: Thu Aug 02, 2007 5:57 pm
Location: Los Angeles
Operating System: OS X 10.9 Mavericks

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by kozikowski » Sat Aug 26, 2017 1:45 am

OK, here. Wear your developer hat for a minute.

I started with Sample 2, Aug 23 Raw. Effect > Equalize > Low Rolloff.

Effect > RMS Normalize to -20dB.

Analyze > Contrast > Measure selection (write it down) > Close.

Screen Shot 2017-08-25 at 18.36.11.png
Screen Shot 2017-08-25 at 18.36.11.png (19.21 KiB) Viewed 534 times


Analyze > ACX Check > Write down the RMS value.

Screen Shot 2017-08-25 at 18.35.50.png
Screen Shot 2017-08-25 at 18.35.50.png (23 KiB) Viewed 534 times


Same number?

Koz
kozikowski
Forum Staff
 
Posts: 38284
Joined: Thu Aug 02, 2007 5:57 pm
Location: Los Angeles
Operating System: OS X 10.9 Mavericks

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by HLC » Sat Oct 28, 2017 3:17 pm

I promised to report back when I finished the audio book and had it accepted. The specific project I was working on when I asked my question is ongoing. However, in the meantime, I recorded a much shorter project, did the adjustments exactly as you suggested, submitted the MP3 to ACX...and was accepted with no changes! Yay! The audio sounds superb! The acting part ... well that could use some work, but I tend to perfectionism and wanted to get at least one project finished so I could feel as though I am making progress.

At any rate, the audio was produced in a small home sound booth using a Yeti Blu microphone set as instructed in an earlier posting in this thread. Also as posted earlier in this thread, here are the adjustments I made, in the order I made them.

1. Mosquito Killer
2. Equalization -- Low Rolloff
3. RMS Normalize
4. Limiter
5. Noise Reduction
6. ACX check

I tend to esses and mouth clicks and so used the De-esser and De-clicker as necessary.

Koz, I cannot thank you enough for the advice you provided, and for your patience. The story is now listed on Audible and I have a coupon code if you would like to get the complete story for free. Just email me off-forum and I'll send the code to you.

Thank you again.
HLC
 
Posts: 10
Joined: Mon Aug 21, 2017 2:24 pm
Operating System: Windows 10

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by kozikowski » Sat Oct 28, 2017 4:14 pm

And just to be obsessive about this, better microphones can handle USB problems and you won’t need mosquito killer, and better environment may help with noise reduction which will make de-esser unnecessary.

Glad they accepted your work.

Koz
kozikowski
Forum Staff
 
Posts: 38284
Joined: Thu Aug 02, 2007 5:57 pm
Location: Los Angeles
Operating System: OS X 10.9 Mavericks

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by kozikowski » Sat Oct 28, 2017 4:16 pm

I’ll do that when I get back to the house.

Koz
kozikowski
Forum Staff
 
Posts: 38284
Joined: Thu Aug 02, 2007 5:57 pm
Location: Los Angeles
Operating System: OS X 10.9 Mavericks

Re: Having a problem with ACX RMS spec

Permanent link to this post Posted by HLC » Sat Oct 28, 2017 5:02 pm

Obsessive is good when it comes to making the audio the best it can be, and your suggestions have been on-target and very useful. I'm stuck with the existing mic and the sound booth for now...but there's almost always the opportunity for hope for the future, so we'll see how that goes. :)
HLC
 
Posts: 10
Joined: Mon Aug 21, 2017 2:24 pm
Operating System: Windows 10

Previous

Return to Audiobook Production



Who is online

Users browsing this forum: No registered users and 2 guests