Complete noob, looking to see if I'm close to ACX standards

Hi! I’m new to the forum, and new to audio recording. Before I even start I want to mention I’ve been lurking in these forums for a few days, and I’m impressed and thankful for the thoughtful, complete, and thorough responses from Koz and Steve. It sucks to go into something new knowing almost nothing and stumble your way around - it’s even worse when those that DO know talk over your head and make ya feel dumb. You guys are cool as hell and tremendously understanding. Next beer’s on me…

Ok, so I set up my “studio” and took a crack for the first time, tonight. I ran a short clip through the tests Koz has linked elsewhere in the forum, checking the Noise, RMS, and Peak values. At this point I have zero regard for the “theatrical” or even editing aspect of my clips - I want to know I’m in the ballpark technically before I put that time and effort in.

The short snippet, attached, seemed to be pretty close to hitting the marks. The peak measured -5.4db, the noise floor of a quiet section in the middle was -64db, and the RMS was around -25db. Could it possibly be that I stupidly lucked my way into being so close that I simply need to goose my levels up across the board by a couple db to hit the correct peak, and bring my RMS within standards as well?

Secondly, do your trained ears hear anything in the raw file attached that will be a problem as I keep going? Breaths and bad reading aside, it seemed ok to me, clear of pops and tinny-ness.

I am using Audacity 2.1.0
Windows Vista
Blue Yeti with pop filter
Gain on Yeti turned down to about 25% max, input on Audacity at 0.61

Any advice or corrections would be tremendously appreciated as I get moving…

Are you using the one-click ACX Check (Zip attached) or the manual process? Unpack that Zip and slide acx-check.ny into your Audacity Plugins folder. In use It appears under Analyze, not Effects.

You appear to be missing The Yeti Curse. The frying mosquitoes in the background that many people experience, so that’s a grand start right there. Good natural reading style without odd speech artifacts (that I can tell right this second).

If you prepare another test clip, you might do it this way.

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

That gives us all the needed parts of the testing and patching process in one short clip.

I got your clip to pass with minimum of fuss and I’ll publish the fuss when it’s daytime.

Koz
acx-check.ny.zip (3.19 KB)

See!!! The Blue Yeti can make nice quiet recordings!

Just so I can keep your system as a data point, what sort of computer do you have?
How long is the USB cable?
Is it the cable that came with the microphone?
Is it plugged directly into the computer or into a hub of any sort?

Otherwise your recording is going to need a touch of compression to meet the ACX RMS requirement. With practice you might be able to pick up the level of the trailing parts of your sentences without becoming monotone, and that will reduce the compression need. Acoustically it sounds good, no obvious echos or environmental problems.

Acoustically it sounds good, no obvious echos or environmental problems.

Exactly correct. You might pass on to us how you did that. Chances are good you’re not recording in your kitchen or bathroom. That one step kills people trying to read for AudioBooks.

a touch of compression to meet the ACX RMS requirement.

And that’s how I did it.

Step one is to get rid of rumble and low pitched sounds. This step does change your voice very slightly, so listen to it before and after you apply the tool. It’s a home-built filter and does not come with the stock Audacity.

Effect > Equalization: LF Rolloff for speech, 8191 Length > OK

See Attached: LF_rolloff_for_speech.xml.zip

Unzip it and apply it to the Audacity Equalization tool.

Adding Equalizer Curves
– Select something on the timeline.
– Effect > Equalization > Save/Manage Curves > Import
– Point at LF_rolloff_for_speech.xml > OK. (it won’t open the ZIP. You have to decompress it)
– LF Rolloff for speech now appears in the equalization curve list.



I artificially made your voice louder and denser (hopefully without changing the quality too much).

I work from top down so I force the loud bits to a standard value.

Effect > Normalize: [X]Remove DC, [X]Normalize to -3.2 > OK

That solved Peaks and Noise, but not loudness (in your case).

Effect > Compressor: Thresh -20, Floor -50, Ratio 2:1, Attack 0.2, Release 1.0, [_]No Gain Makeup > OK

That solves Noise and Loudness, but not peaks.

Effect > Normalize: [X]Remove DC, [X]Normalize to -3.2 > OK

Yes, that’s the same tool twice.

See attached, last sentence.

Koz
LF_rolloff_for_speech.xml.zip (326 Bytes)
Screen Shot 2015-07-03 at 11.20.04.png

Assuming your attached clip is raw and unprocessed, you are already reading at the recommended volume (occasional peaks into -6dB — the yellow zone), so the only way to improve it would be to dial back the vocal expression slightly. In real life, it’s good to vary your expression and volume to impart interest, but pushing volume variations too much will lower your RMS (loudness) and could demand higher processing.

You should probably not end sentences by fading out. We’re picking fine details here, your existing work does fit.

If you find yourself getting dangerously close to loudness (RMS) violations (-23dB to -18dB), try increasing the Compression Ratio to 3:1 from 2:1. Your voice should become slightly denser and louder.

Koz

Thank you sir!! Yes, I just did a manual check based on your 1-2-3 guide I saw elsewhere. I have since installed the plug in and will be using that.

The next you hear from me [tonight likely] I will follow the protocol suggested and actually think about my speech, and resubmit in this thread.

I can’t tell you how relieved I am that I’m not too far off. Getting this audiobook made is a big deal for me for a host of reasons. Thanks for being so quick to help.

It’s about time I got a lucky break on something! :sunglasses:

My setup is very low-tech :slight_smile: I patched together the best I could after hanging out here and watching the ACX videos. I actually have an older laptop. The details :

  • my DAW is a HP Pavilion g7 Notebook running Windows 7 [thought it was Vista…TIL]
  • the USB cable is the one that came original with the Yeti. My guess is it is about 10 feet in length [3m]
  • it is plugged directly into my PC, my PC is older so only has USB2.0 ports
  • I am recording directly to a Seagate external 500GB hard drive… I have seen arguments for and against recording externally. If this is a bad idea someone let me know…

My recording “studio” is my walk-in closet. I noticed a while ago when blabbing on the phone it sounded different in there. The hanger rods form a “U” shape, so I removed the clothes from the bottom of the “U”, hung a blanket there as a backsplash of sorts, and set a TV dinner tray in front of it, covered with a towel. The USB cord was long enough to keep my laptop PC outside the closet. The fan on this thing is way too audible.

Between the carpet, clothes, and blankets, it seems to work. Also helps that my walk-in is upstairs, at the opposite end of the house from the A/C unit and refrigerator, and I record at night when there’s no traffic or lawnmowers outside.

Photos of my super-professional setup attached…


The USB cord was long enough to keep my laptop PC outside the closet.

Where you can’t see it? You have to watch the Audacity recording meters while you read. The first time you get expressive and loud, you could overload the sound channel and that signals a do-over. Overloaded sound is one of the reliable ways to kill your show.

Number 2.

The Four Horsemen of Audio Recording (reliable, time-tested ways to kill your show)
– 1. Echoes and room reverberation (Don’t record the show in your mum’s kitchen.)
– 2. Overload and Clipping (Sound that’s recorded too loud is permanently trashed.)
– 3. Compression Damage (Never do production in MP3.)
– 4. Background Sound (Don’t leave the TV on in the next room.)

On the other hand, if, over the course of a chapter you start fading, that may kill your loudness match. There are ways to fix that manually, but it’s not pretty and you may fail noise even if you do get the voice to work.

People have been known to remote the monitor to the booth interior. It doesn’t have to be a 75" monitor with 3D and hot and cold running beer. It just has to be enough for you to see the meters flipping.

Koz

One of the ACX videos recommends recording to an external hard drive, I’m not entirely sure what their reasoning is. In any case as long as the hard drive is on a totally different USB bus from your microphone it should be fine. You will have to dig into the details of your laptop’s hardware to figure that out for sure. Unless your laptop is running very short on disk space I would recommend recording directly to the internal drive and then transfer the recordings to two different external drives (ie have a master and a backup) at the end of each session.

Not to worry, we’ve now see several artists using their walk-in closet, they seem to work very well. Typically the worst problem is the lack of ventilation.

You might check out local yard sales or EBay and pickup an older low-res LCD monitor cheap to plug into the external monitor port on your laptop. That will give you a quite way to be able to see the meters.

I’m not entirely sure what their reasoning is.

And they’re emphatic about it, too. Nothing wishy-washy here. Earlier on in video editing, it was highly recommended to have two internal drives. Put the OS and program on one and the show on the other. The machine had perfect, high-speed access to everything it needed.

Putting the show on an external USB drive just makes my teeth hurt. It’s not universally correct. I will put money that you can’t do a musical overdubbing session that way. The timings are just too sloppy. So it only works with straight vocal capture.

Koz

Ian in Hollywood has to come out of his “studio” about every twenty minutes and gasp for air.

We do tell people that they can do very well by just leaving the clothing in there. Don’t pile the suits and cardigans on the bed. Leave them hanging.

Also, you can get additional table isolation by parking the microphone on a book, then the towel and then the table. Put the book sideways; top of the book toward the right, for example. Any vibration coming up from the floor has to make it through the towel, then through the high inertia book and then finally the microphone. It doesn’t sound like you have any problems like that, so just keep that trick in your hat.

It looks like you have to hunch over to speak into the Yeti. Do you? How many hours can you do that? Will you have to be helped down the stairs after the second or third hour?

Do you have a file posting service you like, such as DropBox? How would you get an entire chapter to us if we asked for it?

Koz

First, I’m curious - what time zone are you in?

Anyway, point taken on the levels. I wasn’t super comfortable not seeing them, because as you mentioned over a longer reading I could trail off or not catch a track-ruining peak. Sooooo…I’m heading to the store now to get a long HDMI and an extension cord so I can put an LCD monitor in the space behind the mic. Dumb question from a technique standpoint - is there a trick to eye the meter levels at the same time you’re reading the text?

The chair sits kind of low in front of the mic, so I think I’m ok. Plus, I’m still on sabbatical, with no job prospects on the horizon, so time is a luxury I do have. I plan on going a chapter at a time [there’s 25, each about 2500 words] then editing… so my straight time in there won’t be too bad. I hope.

I’ll post another photo when I get the monitor in there. The monitor shouldn’t cause any noise in the mic, should it?

I can upload to Dropbox or Google Drive.

I think I’ll change my temp disk back to the internal drive.

You know those exceedingly rare times where sh*t just seems to go your way? I’m in one of those grooves…

Found a long HDMI cable I bought 4 years ago before realizing the palmtop I planned on connecting it to didn’t have HDMI. Found an extension cord I didn’t know I had.

Best of all, when I got the monitor attached, I realized I probably was close enough to the laptop outside the door to use my wireless mouse. Yep.

So now I can actually start stop and click around without getting out of my “studio”.

My only hope is that the monitor doesn’t create some kind of evil buzz in the noise floor now.
2015-07-04 13.28.15.jpg

Gentlemen - an update. And seriously, I cannot thank you enough for this and will be sending you gifts upon completion if it’s ethically acceptable :slight_smile:

So I’ve incorporated all your advice, recording locally [saving the projects and exports on remote though], I brought the monitor in the room, doesn’t appear to create a noise issue.

I recorded a second small clip, this time in line with Koz’s parameters. Attached is both the raw file [doesn’t meet ACX] and the file with Koz’s custom EQ and other touches [does meet ACX, of course].

The one change I made was upping the input on Audacity to 0.8 from the 0.61 it was at. Maybe it was my seating position or some other tweak, but I needed it a touch higher to get the peak values to dance closer to -6. Noise floor still seems ok.

So now I’m taking the full raw recording of Chapter 1, full of restarts and mistakes, and editing it down to one acceptable cut. I won’t do any mastering until what I’ve mentioned here makes sense to you guys.

Donations of any amount are always most welcome:
http://audacityteam.org/donate/.

Thank you.

Gale

Indeed I think you are good to go.

If you amplify the quiet section at the beginning of your latest posting there are some rather curious noises in the background. Maybe children playing off in the distance or something. In any case they are well down in the noise and are not going to be an issue.

Good luck!

There is a down side to having the monitor right behind the microphone like that. Some microphones will have an effect called “slap” or proximity or comb filtering where tone will change depending on how far away the flat surface is. If the Yeti is set for Cardioid, the effect should be minimized.

As long as you have a low background noise, you can get away with a lot of post production processing. High background noise just kills good quality recording and is a very common problem with home recording.

If you bury the meters high in the red or never see the bouncing lights higher than -30dB, chances of a good recording are zero and it doesn’t matter where the volume controls are.

Occasional peaks at -6dB (yellow zone) is chosen so on average, the recording volume is perfect. That gives you some leeway to get a little louder and softer for expression without creating sound damage or fighting background noise.

The sound meter and associated overload indicator take precedence. There is a common misconception that you set the volume controls once and put little pieces of tape with marking pen next to each one and that’s it for the next three years. Not really. The volume controls on your recording system are set to make the recording meters happy. As a practical matter, no option. Changing them in the middle of a performance is not good practice, but if you change your seating, vocal energy or anything about your environment, you may need to change the volume controls to make the meters happy.

It may take you a while to present with a constant voice except where expression is needed. After you get the swing of it, you only need to glance at the meters occasionally to make sure you didn’t wander off the mark.

There are puzzles. A recent posting inquired about the best way to announce a gun discharge. It’s safe the answer is not just get louder.

Koz

So, I have Chapter 1 edited to what I think is acceptable.

When I read " gating" in a discussion, is that referring to the quality of the edit, I.e. not including half breaths or the like?

Anyway, I’m not madly in love with it ( this first chapter) and I may come back and redo it later, but if you guys would be so kind I’d like to share a link here to the raw and mastered files to see if anything jumps out at you technically or artistically.

if there’s no glaring issues you find, I’ll be confident enough to get to work recording and editing the next 24 chapters and checking in with you cats again as I master and convert to MP3.

When I read " gating" in a discussion, is that referring to the quality of the edit, I.e. not including half breaths or the like?

What most people mean by Gating is an abrupt unnatural switch to silence between words.
I was about to say I wish I had an example of this…but I do. Rachel Maddow records her show in a very noisy studio. I once accused them of recording on the NBC loading dock and I said in one show I could hear somebody building an Ikea credenza with noisy hand tools in the background. One show I swear somebody was ordering pizza behind her voice. I think the show producer is in the same room with her and she occasionally makes cellphone calls. A very bad idea in my opinion.

Anyway, Rachel has a noise gate on her voice now. Listen carefully.

Hello shshsh …How are you shshshsh…

The extreme, tight version of this is happens when you get to the end of a word and the sound abruptly switches to The Blackness Of Space silence. That’s with it working normally. If you mess it up, it will actually slice off beginning and endings of words. There is no place for a noise gate in vocal narration. It usually sounds profoundly unnatural.

Number 3.

The other noises in your voice I have nearly no opinion about except to say if you have them when you’re speaking to someone over cups of hot tea, then you should leave them in. ACX does mention this briefly in their instructions, but I suspect strongly they do because performance artists kept doing it, not that ACX required it. Once you get past the QC Robot and human spot checks, it’s entirely between you and the rights-holder.

Also please note that taking out mouth clicks and other noises automatically could easily damage the performance quality or, if you decide to do it manually, take six months.

“I’ll have this for you right away. What is this? July?” [counting on fingers]

I’m not madly in love with it

Welcome to newbie heaven. It’s perfectly ordinary to get to the end of a book, listen to the beginning and start over. Nothing like performing for a month or so to tighten up your performance quality.

Also remember, the performance master is WAV. Lower quality MP3 is something you make from the WAV for delivery to the client (because they said so). You can’t easily re-edit an MP3.

Koz

http://www.pacificstandardtime.org/

Koz