Hello my name is Laura and I recorded the first chapter of an audio book that has failed the ACX check. The ACX QA Team suggested, “recommend increasing gain on microphone…” and this has helped me to pass the plugin ACX Check with a number of test files.
I would like to use the “Audacity Effect” tools (Amplify, Compressor, Equalization, Normalize, etc…) to see if I can bring my first chapter into ACX compliance. I have reviewed a number of related topics posted on this forum regarding ACX Check and used some of suggestions to no avail.
My setup is as follows:
-Microphone Shure SM7B Dynamic Vocal Mic w/ Mic Boom Stand, Pop Filter
-Zoom R16 - Recorder/ Interface/ Controller
-Windows 10
-Audacity 2.1.2
-with Plugin ACX Check
-with Plugin Wave Stats
I have attached a sample wave file and a pic of “ACX Check output(s)”
and I would greatly appreciate any suggestions on how I might be able to “Master” this recording to meet ACX guidelines if at all possible.
Thank you in advance.
Sincerely,
You should be working in Mono rather than struggling with two-channel stereo. Everything is easier or more efficient.
I got it to pass with relatively simple corrections. I made it louder and added very gentle noise reduction.
ACX is very particular about having your chapters match. If you added a ton of effects to the succeeding chapters, 1 and the others are not going to match.
Normalize
– Select the whole clip or show by clicking just above MUTE.
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.5 > OK
Noise Reduction
– Drag-select Room Tone, silence or the flat area between spoken phrases.
– Effect > Noise Reduction: Profile
– Select the whole clip or show by clicking just above MUTE.
– Effect > Noise Reduction: Settings 6, 6, 6 > OK
Don’t let “Normalize” scare you. That’s Audacity Speak for turn the volume up. It’s a sister to Amplify. They differ in how you tell them to change the volume.
I used the silence right after “And her commission” for the Noise Reduction Profile. I believe ACX makes you include a silent room tone segment and you should use that. It’s a grand mistake to “make silence” with the Audacity tools rather than using the actual room background noises. That will stop Noise Reduction from working.
By the way, making “fake” silence also gives you bogus ACX-Check noise readings. There’s no such thing as -206dB Noise. A normal WAV export from Audacity runs out of steam at -96dB and some of the processes prevent it from going even that low. I expect natural noises to come in at -60dB to -70dB which is why the ACX specifications are -60dB. I can hit that with a very good microphone, a dead-quiet room and no Noise Reduction.
Busted!
You can see from my ACX Check readings, -60.3dB, the actual, real-world noise just barely passes. If all Chapter 1 passes, we’re done. If it juuuuust fails noise, UNDO the noise tool and re-apply it with 8, 6, 6 instead of 6, 6, 6. That’s a slightly stiffer noise reduction and slightly higher chance of voice distortion
I just tried to upload the whole chapter twice. I believe I was kicked out of the post each time; possibly for too large a file. I do not show it completed on the site. If that is the case, do you have a suggestion and how I can get it to you. It is 106 MB.
Please note I changed your username to lwitek. It is a bad idea to expose your e-mail address in your username because nasty bots that harvest addresses for spammers will see your address then you will receive spam. Please log in with your new username.
The file size upload limit for Audacity Forum is 2 MB.
You can upload to an online file sharing service then give us the download link
Hello Gale,
Thank you for all your help.
I have uploaded Chapter 1 as requested as well as Chapter 2 which I just finished editing early this morning. These recordings were both recorded at the same old levels. These can be found here:
I was able to edited the wav file you edited and I was able to get your results.
I was unable to use the same instructions to edit Chapter 1 and pass ACX Check, as you thought I probably would.
Regarding your comment, “It’s a grand mistake to “make silence” with the Audacity tools rather than using the actual room background noises. That will stop Noise Reduction from working…” To the best of my knowledge I did not do this. I did copy and paste room silence to various places. Could this be the cause of what you observed?
In your opinion, should I go back to the old preamp setting (I raise mic level and output, small amounts) for the rest of the chapters to maintain continuity with how I recorded Chapters 1 & 2?
…
I got a correction to work on chapter 1. I need to pause and collect how I did it.
…
I have an odd complaint. Starting from the download (perfectly lovely, by the way), tell me who you are.
Your posting scrolled from the active help forum window, so I have to go looking for you. There’s no identifying information in the download page past what I can glean by actually playing the work and identifying the characters. Is “DB” the name of the show?
The elves juggle many multiple postings, complaints and suggestions and I for one can’t remember individual posts. I juggle two browsers, so there’s no return link to the forum.
Attach for Chapter 1.
You got expressive and theatrical at 9:50 or so and the blue waves poked up and threw off the overall average volume. “Rob!! You called him Rob!!” That’s OK. You didn’t break anything.
Audio Compressor
– Select the whole clip or show by clicking just above MUTE.
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.5 > OK
– Effect > Compressor: Thresh -20, Floor -40, Ratio 2:1, Attack 0.2, Release 1.0, > OK
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.5 > OK
The compressor removed the expressive peaks, pushed up the overall volume enough to pass RMS (Loudness). The settings are gentle and graceful, so I expect any damage to be minimal if you can even find any.
Attached: Before and After blue waves. After is slightly louder and has more “beef” to it. That’s not a bad explanation for RMS. Electrical RMS is a measurement of energy, horsepower or density of a wave. It tracks very well to overall show loudness which is why it’s used.
First, I am so sorry regarding the lack of “identifying info” in the “dropbox” link and the two files (Chapter 1 & 2). I am new to “dropbox” and did the bare minimum for lack of understanding the software and your needs. I will do my best to completely identify items in the future as requested. DB is the initials of the title of a book(s) (first of three I am told) I am narrating. Although I have years of announcing and commercial work experience, I am new to audio books, home recording studio, etc… All my other experience was at radio, TV, and production house studios. So I am trying to educate myself ASAP. I appreciate all your observations and comments.
Thank you so much for mastering Chapter 1. I will try to duplicate your process now.
I just tried to reproduce your fix and I still failed.
Audio Compressor
– Select the whole clip or show by clicking just above MUTE.
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.5 > OK
– Effect > Compressor: Thresh -20, Floor -40, Ratio 2:1, Attack 0.2, Release 1.0, > OK
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.5 > OK
Did you apply any noise reduction effects?
No, I didn’t. And RMS (loudness) isn’t the only bad number. In my correction, the noise is -70dB which is expected and rational. The -207dB noise reading from your correction is not. The only reason that number passes ACX Noise is a quirk in ACX-Check. The way it thinks, any number quieter than -60 passes whether or not it’s rational.
Are you working from DB Chapter 1 Edited.wav that I got from the hightail.com posting site?
What happens if you hide your original work safely in custom folders, download DB Chapter 1 Edited.wav and apply the correction suite to that? Let’s see what happens if we’re both working from the exact same sound file.
I downloaded the same file just as you suggested.
I ran these effects as follows:
Audio Compressor
– Select the whole clip or show by clicking just above MUTE.
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.5 > OK
– Effect > Compressor: Thresh -20, Floor -40, Ratio 2:1, Attack 0.2, Release 1.0, > OK
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.5 > OK
I see that Length of selection are the same… etc… Interesting that results are different. Your thought?
This combo is what lead to you low recording level. Think about changing the Shure to a higher output mic, or replacing the Zoom with a better interface or recorder. The result is noise (hiss) in your tracks. More work in post.
Most of these relatively simple interfaces fail if you combine them with a low output oldie like the SM7b.
See if you can borrow something like an Rode NT1 or a CAD E100S to test with your voice, if you feel like changing.
Open the DB Chapter 1 Edited.wav file you got from the download and Analyze > ACX-Check. I got the attached panel. I know of no reason they should be different.
You apply Normalize to -3.5 and remove DC and the blue waves get very slightly smaller.
Our compressors are the same. When I apply compression, the blue waves get very much smaller and the sticking up parts around 10 seconds almost vanish. (attached).
Then I applied Normalize again…and now mine fails, too. Son of a…!@#$%&
I can’t make it pass without multiple different tools that I know I didn’t need earlier.
