Meeting ACX Requirements w/ Audacity

That DIY stuff is right up my alley. But the Blue shock mount is inexpensive and most likely a good investment as well.

I’m ok with being impatiently patient! I’m just thankful for all the help I’m getting! Slow and steady wins the race , at least that’s what they say right?

I’m working on a parallel task to yours. We never codified a testing process for someone who produced a perfect reading. We got lost in the weeds of “What Happens If…”

“What happens if the reading is perfect” wasn’t in the list. It’s not useful to dig through pages of damage solutions if you have no damage.

It’s true that creating a perfect recording isn’t likely, but still. A paragraph of simple tests and solutions is on the drawing board. You’re the experiment.

Koz

OK. I ran that last clip through an experimental regimen and wrote everything down as I did it. It did require pushing things around a bit to make all the numbers work (you did still have a little popping in there). We’ll see what everybody else thinks.

http://kozco.com/tech/audacity/clips/BlackRacketCompliant.wav

That room tone segment is insanely handy. I don’t have to go digging between words to get the noise.

A note. This isn’t the best I can do. It’s the best I can do with normal tools and in a reasonable time. Even at that, two of the tools are experimental and have to be installed by hand.

The clip should be very nearly the same as the original. You will notice that your little pops are gone (Steve’s custom filter) and the Room Tone segment is going to be more “velvety black” than it was before. Almost all the other changes are exotic behind the scenes adjustments.

Koz

I like what you did, the echo that I hear on my end is almost non existent! :3

I believe I could start with that (since the first 15 minutes are due this Friday) and send it in to see if it’ fits? What do you think?

Your original clip is indeed very good for itself.
I suggest that you record in mono from now on. This gives you a lot of file space and calculation head room.
Interestingly enough, the two channels do not hold the exact same data. You can try the vocal removal tool (just press enter). You’ll hear (after amplification…) ordinary white noise and some speech traces. The noise is presumably dithering (and therefore intentional) noise added at the export stage. However, that’s nothing that should concern you.
The Rms value of Koz’ clip is slightly too low (-25 dB), -18 to -23 dB are expected.
There are actually only a few peaks that prevent us from making it louder, e.g. in “our perception”.
After the usage of my semi-automatic levelling tool, the values ar: -3 dBfs peak, -21 dB Rms average and -75 dB Rms noise floor.
You can download and install Steve’s Peak-limiter to achieve nearly the same values.

My hearing is not very “golden” at the moment due to a bad cold, thus I can not give any advice there (equalizer and so on).
Please post the 15 min clip before sending it off to Acx–just to be on the safe side.

Robert! I’ll do that promptly :slight_smile:

It is a graphic novel so I hope you all don’t mind!

I’ll be posting a minute clip soon.

I got a 23.6dB loudness RMS measurement when I did it. You have to be very careful what you select for the measurement. Given a woman’s voice has higher apparent loudness than a man’s, I expect that to work.

I couldn’t get the compressor to do what I wanted, so I used Effect > Leveler followed by Effect > Amplify to put the peak reading back to -3.

Koz

This is what Audiobook Noise Floor did when I ran it (attached). What’s supposed to happen?
Screen Shot 2014-08-24 at 12.11.55 PM.png

It is a graphic novel so I hope you all don’t mind!

Speaking Engleese for a minute, that’s graphic novel as in comic book or a novel with graphic content?

Koz

You can download and install Steve’s Peak-limiter…

From?

Koz

Yes, don’t post anything to them until we settle on a protocol and effects values. Raw clips will never pass popping and noise. You’ll have to be able to process work by yourself once we get one through the system. As above, some of the tools I used are brute force that happened to work in your particular instance, but may not be generally applicable.

Analyze > Audiobook Noise Floor and Effect > Equalization > LF Rolloff for Speech are custom designed tools and have to be hand-installed in Audacity.

I will post the recipe shortly.

Koz

Haha, my apologies! Although one day I can hope to narrate a comic (maybe?) this is a novel with graphic content :wink:

Homemade recipes are my favorite!

The rule is five to one. So 15 minutes of finished, delivered show is good for an hour and a quarter to an hour and a half of work. You’re not accustomed to the tools and shortcuts yet, so it’s a bit longer. Also, we haven’t applied the ACX Conformance tools yet.

“Alice slowly lay back in the warm, sweet-smelling hay, her heaving…”

I know it’s common to think you can record a show, slap it on the fanny and go. I don’t know that production has ever worked like that. You always want to correct your mistakes and polish the work.

Oh, wait. I can think of one violation of the 5:1 rule. I used to do this.

“We are live from the fourteenth street bridge where the Air Florida flight Palm 90 has plunged into the Potomac River. Rescue divers and emergency lighting…”

Koz

I had just deleted it once I realized that I hadn’t connected a good 3-5 seconds of clips. UGHHHH, I’ll post the new one here, that doesn’t have that blank space I forgot to connect haha

I’m uploading it as an mp3 because apparently that’s ACX standards?

I remember reading somewhere, maybe in this forum https://forum.audacityteam.org/t/manipulating-decibels/33629/111
that ACX requirements are a ball park and their really just looking for it to sound consistent, and not garbled. I’m not sure if that was ever confirmed, but nonetheless

THANK YOU AGAIN FOR YOU HELP! :stuck_out_tongue:

https://drive.google.com/file/d/0B3HW7yC_7qhmaDBqQ3BkWXZmbzA/edit?usp=sharing

There’s the file, with no weird gap, raw stuff, no recipes!

I’m uploading it as an mp3 because apparently that’s ACX standards?

True. That’s ACX Final Deliverable Standards. However you can’t Edit, Filter and Effect in MP3. That has to be done on the WAV or Project. You did keep a WAV or Project of the final before you made the MP3 right? … Right?

MP3 is a dead-end road. It’s what you make when you’re sure you will never have to do anything to the file ever again — except listen to it.

We have a number of things we need to do to it.

People who make podcasts from downloaded MP3s always complain that the Podcast MP3 is ratty and doesn’t sound very good. That’s correct. They have two MP3’s worth of compression sound damage in the show and it’s permanent.

Koz

So we need a WAV or Project posted if I never got around to saying that.
Koz

Where is “LF rolloff for speech?” Assume I want to hand someone a thumb drive with the filter on it.

Koz

Here: Fledgling voice artist seeking counsel - #118 by kozikowski

Thank you.
I’m using your “filter list” for Ian as a base for corrections in this case, and maybe many other cases as well.

https://forum.audacityteam.org/t/completely-locking-up/261/1

The first mod is in step 2, Normalize to 0 and remove DC thereby solving a possible popping problem in step 3, LF rolloff.

An additional task is building tests into the list. Audiobook Noise Test, Analyze > Compare, etc. Finding and collecting all the bits and pieces has been quite a chore.

Koz

Sorry I’m late,

I did save a WAV just in case! Hope you still need it!

https://drive.google.com/file/d/0B3HW7yC_7qhmQkJNa05JenJDSDg/edit?usp=sharing