First of all, thanks to administrators buazo and steve for working me through the tortuous path to success. My thirty-eight novel chapters are well on the way to becoming the mp3’s that will go to Audible.
There are good videos to help newbies … try The 43-minute video, How to Pass QA Every Time, by Brian Solarz for specifics on ACX.
I had software problems that this forum helped me correct, finally resulting in a complete reload of the Audible software. If you’re just starting, the current download of Audacity is different from some of the descriptions in the instruction manual. For instance, “chains” … used to replicate series of effects … have become “macros” and moved to a dropdown called “tools”
An issue to watch out for: ACX wants up to 1 sec of room noise at the beginning of a file (generally, chapter) and up to 5 sec at the end. Most of us will have to apply some compression to raw files to meet ACX’s RMS requirement. (…be sure to download the Nyquist effect ACX Check and use it). I use Compress Dynamics 1.2.6 (also Nyquist) … apparently, it has trouble with the long room noise passage at the end of each file and adds a rushing sound. Someone may know how to prevent this … my solution is to clip the noise and substitute a clip of room noise I have assembled as a separate project.
We publish an ACX Mastering suite of tools which does not have a formal compressor and does not cause pumping or other damage. If you find Chris’s Compressor useful (and I do) you should know that his look-ahead algorithms don’t take kindly to falling off the ends of the show. Apparently, either end. So it’s desirable to “pad” the ends of a valuable show with “Something” for Chris to chew on so it doesn’t damage the show and then you can cut off the extra later. I used to use it to prepare a download podcast for listening while hiking. I always cut off extra trash beginning and end, so I never noticed that end-pads were required.
However, our ACX Suite doesn’t do that. No compressor. This is the write up. It’s pages of obsessive instructions for three tools.
Yes, certainly, several readers have fallen afoul of the submitting requirements for Room Tone. You do have to pay attention.
There’s one thing about that video I don’t like. They use the “Speaking to the Credenza” technique. I don’t know how that became popular, but it makes me nuts. I find you so repulsive that I’m not going to make eye contact with you.
If they’re speaking to someone in the room, there needs to be cut-away shots of that person even if they were recorded two weeks later. Changing a shot from 3/4 to looking at his ear is called a jump cut and is to be avoided. Those shot changes should cut to illustrations reinforcing the point.
This actually came up in the comments about describing microphone spacing.
It’s not hang ten, it’s hang loose. But I digress… > >
You should digress more. It’s based on the Hawaiian Shaka.
The lead guy does have good points. He doesn’t like condenser microphones, either. “They produce sibilents and and prone to P-Popping.” Dynamics are good microphones for presenting.
Yes! [pumping fist].
He also reinforced my belief that normal human breathing noises are natural and desirable.
Thanks for the redirect to the … ummm … manual. I had read that some time ago, then of course forgot it. Should be much easier to use the Audacity manual approach. I’ll run that process against my chain/macro and report.
Record your voice and don’t do anything to it. We can’t take effects and filters out of a clip, so if you put them in, we end up fighting each other.
The beginning of the ACX video did stress recording in a quiet room, no echoes and no noises. I posted several times the original people in my house had a kid that played drums. They soundproofed the tiny third bedroom to keep him from getting out. Of course, it also keeps neighborhood noises from getting in. I got insanely lucky. I can record with many ordinary and some trashy microphones and still pass ACX with only a little work. I can’t find it right this second, but I did a sound test using the built-in microphones on my laptop. I would not do that for a book only because it was physically awkward, not because it didn’t work.
I’m still dying to know why my iPhone test failed. That would have been intensely desirable. Forget the microphone. Grab your iPhone and a quilt/duvet and record in the closet.
I’ve attached a 18-sec sample of a place where Compress.dynamics 1.2.6 put quite a bit of rushing sound, both at the end and in the last phrases. The picture attached shows raw vocal on top, then a track using the macro I had been using, then (3rd track) the Audacity approach you suggested. Short story: the Audacity approach is much better … sounds identical to my approach, minus the noise.
My macro was EQ (voice rolloff with +4 db of bass 100 hz - 400 hz), Compress Dyn (.95,.5,-49,2,.99,0), Normalize (-3.1). I used the same EQ settings in the Audacity approach. I used this particular clip because it demonstrated the rushing best
That was a company that did their first promotion recording in an actual studio and then decided what a “waste of money” that was, and to do later ones at home. You both have that robust recording-in-a-bathroom sound.
The up side is you have good theatrical pacing. You will probably not have to worry about the storytelling part.
Prominent on my To-Do list is making a sound cave out of hardware store plastic pipe and furniture moving pads. And then write down how I did it. It’s this without the high cost.
First off, your help has been crucial to this project. Thank you.
My own studio is not set up now. I sent you a piece of chapter 2 of a 38-chapter novel. I used it because it had particular problems with that rushing sound using Compress 1.2.6. The base audio was recorded two years ago at my son’s audio/video studio, which is about 24’ x 30’. It’s quite possible that I did attempt some initial modifications during a false start to turn the novel into an audiobook about a year ago (see the *note below, though).
You have given me a wonderful lead on getting the audio right. The suggested EQ/RMS normalize/limiter combination gives the same good sound and ACX qualification my EQ/Compress 1.2.6/Normalize did, but without the artifacts that chain produced. I’m not in a position to do extensive rework on what I started with, so I’ll plan to go with your suggested chain and spare you the effort of more analysis.
I queried my son on equipment, and he responded:
The mic pre is an api 512c
I think we kept the compressor chain on so the signal would have been going through a dbx 160A*
I don’t remember which mic you used. It was a Neumann U87 or an AKG 414c
Do you remember what it looked like?
It was most likely the Neumann U87 which I tend to like more for vocals.**
Hope that helps!
My recollection is that I recorded a couple of chapters, then talked with ACX, who gave me the impression that the ‘audition’ should be uncompressed, then took off the compressor. I THOUGHT I’d reread this section w/out compression, but maybe it is still compressed. It DID behave differently than other chapters in the mastering process.
** I believe I did use the U87.
The base audio was recorded two years ago at my son’s audio/video studio, which is about 24’ x 30’.
I think I have the problem. Audio/Video studios in general are rooms that look great…full stop. I bet there is little or no soundproofing and it has a polished wooden floor. That’s where your echoes came from. Even the legendary U87 will sound funny in an echoey room.
For about a year, The Rachel Maddow Show was recorded in New York and looked great. It was only after you saw the show for a while you started to pick out set elements that didn’t quite fit. Tiny visual problems.
But it sounded like it was recorded on the Rockefeller Center Loading Dock. There was one show that sounded like someone building an Ikea credenza on the Rockefeller Center Loading Dock. That’s when I wrote the letter to the producers. “Come on, people…”
Now you know why ACX hates processing. I can’t underline that enough. Once you burn in compression or dynamic processing, you’re dead. Those never come out. If they’re clearly audible, you may get to read the piece again.
If you decide to do something at home, post here before you write any checks.
I may have caused a tempest in a teapot by sending you the chapter I did. I have attached a clip of what I’m pretty sure is raw speech. The full chapter passes the ACX Analyze test except for ceiling (attached).
For now, I’m going to follow ACX’s instructions at https://www.youtube.com/watch?v=p156u60gWdM and use the chain I made based on your good advice. At about the middle of that ACX instruction, the technician assumes use of compression or a limiter to hit the RMS criterion.
We’re trying to critically diagnose your work blind and over multiple time zones. We can only go with what you supply.
It’s not that unusual for us to go totally down the wrong road by the poster and the elves using one word to mean two different things, inability to describe problems, badly expressed goals, and even the rare troll or hostile poster.
So that’s what we’re doing and why this forum thing sometimes doesn’t work as well as it should.
Assuming the above clip is the raw recording, I had no trouble meeting ACX with just the Mastering Suite. I didn’t even need Noise Reduction. I can’t listen critically right now. That will have to wait for daylight.
There’s a “dong” sound at 11 seconds. Did your alarm clock app sound?
We don’t usually have to say this, but you do have to leave your phone in the car or someplace where you can’t get alerts. I think you can’t just turn the sound down. There is a new service in California which warns of earthquakes. I think you can’t turn that off. It could be argued if there’s a quake coming, you probably should not continue to record your book.
The background noise is low and well behaved (shshshshsh), you’re still recording in a big empty room. Is this the audio/video studio?
This is the business where you may pass the ACX Robot which is looking at the same things ACX Check is, but may not make it past Human Quality Control which worries about presentation and theater…and echoes.