Hello! I am (hopefully!) going to attach a sample of audio from my home system. I am hoping to record my new sci fi novel. Is the audio too bad? Just using a homemade sound box with those pyramid squares, and a Samson mic.
Thank you for your feedback, if I can follow the directions to upload it.
Your upload works fine. And your recording is fine too.
No annoying traffic noise, no hum. Maybe just a tiny bit of de-essing and background noise removal…
You have a good reading voice and the clip arrived in good order.
Your upload works fine. And your recording is fine too.
That’s the “good cop” version. The bad cop version is the clip always fails two of the three ACX AudioBook standards. As submitted it fails Noise and Peaks. If I fix the peaks, it fails Noise and Loudness. If I fix the Noise, etc. etc. etc.
There is a lot of background noise in the clip. Do you have an air conditioner or fan running with you? Are we listening to the computer fan noises? Do you have a computer tower or loud laptop?
Putting the microphone in a box is half the problem. If you have the microphone set on directional (and you should) then it’s also recording everything behind you. Like sound reflections from the bare wall or the air conditioner in the window.
I agree we can get there with patches and fixes, but if we do it that way, you’ll have to remember to do each correction step to each chapter of the whole book. Much better to fix problems at the recording stage.
As we go.
Koz
I was going to publish the list of tools and filters to fix it, but I can’t do it without serious surgery. You’ll need to quiet down your room. This can be difficult because most people “tune out” normal day-to-day noises. For instance, each time I record I get my wall clock in the show. It’s insanely noisy, but I can’t hear it at all, I tune it out, until it shows up in a recording.
ACX has an interesting trick. Record 30 seconds of Room Tone (don’t say anything, just record). Now play that with the volume turned up (attached).
We had one poster who, over the course of testing identified a ventilation fan and a refrigerator running in his room. He got rid of those and most of his problems went away. High noise level makes applying the other tools much more difficult and extreme Noise Reduction can make your voice sound funny.
Koz
Just using a homemade sound box with those pyramid squares, and a Samson mic.
That can work. I do it with furniture moving pads.
That was a multi-actor shoot, so you won’t need all that.
Please give us details of your computer, audacity version and mic model numbers? See pink band on top.
If you can, post a snapshot of the microphone in its foam enclosure. Sometimes we can make simple changes to improve it.
Koz
Hello and thanks to you all! I like the good cop best, of course! To Koz, I have a big old Dell, using the latest Audacity, I just downloaded it, mic is a Samson CO1U (I think it’s always omni-directional, as I recall, and I don’t see a switch anywhere), I just have it on a regular tripod, not the shock stand. I am attaching pics of my little set up, which of course I found on the internet. I used big slabs of styrofoam left over from some project, taped them to make a ‘hinge’ so they’ll stand alone, and drape a nice old piece of zebra patterned fleece over it to sit behind me, as the internet DIY piece suggested. No AC is running in the room, but two are running elsewhere in the (small) house. I am sure the Dell makes noise–you’ll see I stood some left over pyramids along side of it in pure superstition. But I don’t have an extra closet! And I just don’t see how I can make the house much quieter. But if there’s any way you could help me with the settings to maximize what I’ve got! Oh, notice the finches. I am covering their cage and will try to be alert to edit out any fleeting song, but I haven’t heard any in the samples I’ve been trying, whereas my son says he can!
I’m doing the audio first to find any last second errors in the editing, am not planning to use Amazon thank you very much, and actually don’t know what I’m going to do with the files, I presume turn them into MP3’s and sell the download on the book’s website using Paypal. All three or four sales!
I’ll attach the pics. You’re very kind to help!
I have to add the pics in batches, sorry.
Hello again, here are a couple more photos, of the Dell and the offending finches.
The C01u is not an omni-directional microphone. It’s non-adjustable cardioid (heart-shaped, or kidney-shaped in Britain). It receives best by talking into the logo, it has some reception from the sides and almost nothing from the back. Attached is a similar microphone that you work by talking into the end instead of the side like yours.
I know you can’t do that, but the joke is to point the back toward the birds. There is no microphone in the back, right? It’s dead. That doesn’t work for a number of reasons, but it’s fun to think about.
It’s not serious, but I was wondering where the slight echo was coming from. It’s a modern room with polished oak floors and bare white walls, right? In the trade those are “Hostile Recording Environments” and some recordists refuse to work in them.
Am I making you warm and fuzzy yet?
I’ll take your word that you don’t have to pass ACX AudioBook standards and just do my best.
As we go.
Koz
Enclosed is the patched version of your clip. I’m writing up the list of how I got there. It’s very close to audiobook standards, I just wasn’t as obsessive as they are.
Enclosed, please find LF_rolloff_for_speech.xml.zip. Unzip it and that should leave you with LF_rolloff_for_speech.xml. That’s a custom effect designed by Steve to work with Effect > Equalization. Remember where you put it.
**Adding Eq Curves
Now on to your show.
— Open the show, chapter, clip, whatever. Select the whole thing by clicking just above MUTE.
Effect > Equalization > Select Curve: LF rolloff for speech, Length slider all the way up (8191) > OK.
Drag-select a second or so of pure room noise, no voice. In that test clip, I got it from 1.8 to 2.2 seconds. If your room noise never changes, you may be able to get the profile from that test clip but use it on your show. If you change anything like move the microphone, you’ll have to get the profile from the actual show. Anyway.
Effect > Noise Reduction: Profile.
Select the whole show by clicking just above MUTE.
Effect > Noise Reduction: 12, 6, 6 > OK
Effect > Normalize: [X]Remove DC, [X]Normalize to -3.2 > OK
That’s it. If you do that to each chapter, export as WAV (safety) and then re-export as MP3, you should be good to go.
It doesn’t hit AudioBook exactly, but it’s ACX Adjacent.
Post back if I lose you.
Koz
LF_rolloff_for_speech.xml.zip (326 Bytes)
Hi, Koz, I don’t understand it at 5 in the morning, but when I can I will go through the directions and try to do it.
But in the meantime, since my mic isn’t omnidirectional, should I take it out of that box? It makes it hard to get the script in front of my eyes. Or does the box help?
Thank you so much! I may have to ask more about the steps, too, if I can’t actually handle it! But I will try.
Hello,
I think my question about ‘inside the sound box’ vs ‘outside the sound box’ got buried in my gratitude. I was wondering which set up was ‘better’ in terms of fewer problems to fix at the end, in spite of the fact that my mic isn’t omni-directional at all and so probably doesn’t need the sound box (I think the internet directions I used were for an omnidirectional Samson). There’s a dif in the sound of these two selections, but I can’t tell which one is ‘better’ in the terms mentioned, that is, more amenable to the corrections Koz gave me. Outside the box the mic is more sensitive and there was clipping but I didn’t get that in the section I copied and saved.
Thank you!
Janet
I know it’s probably inconvenient, but you still need the box. The microphone receives from the sides, too, just not quite as well and I expect your room echoes to go way up if you stop using it.
I can’t listen to the clips right now. A bit later.
Yes, I’m not surprised that the volume went up when you took the box away. That’s not clear, quality voice. That’s all the extra echoes.
If you have trouble following the corrections, just say so. Give me an idea where you got stuck.
Koz
Enclosed, please find LF_rolloff_for_speech.xml.zip. Unzip it and that should leave you with LF_rolloff_for_speech.xml. That’s a custom effect designed by Steve to work with Effect > Equalization. Remember where you put it.
Adding Eq Curves
Open the show or Generate > Noise > OK. It’s just to get 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)
From that point forward it appears as a standard curve.
Now on to your show.
— Open the show, chapter, clip, whatever. Select the whole thing by clicking just above MUTE.
Effect > Equalization > Select Curve: LF rolloff for speech, Length slider all the way up (8191) > OK.
Drag-select a second or so of pure room noise, no voice. In that test clip, I got it from 1.8 to 2.2 seconds. If your room noise never changes, you may be able to get the profile from that test clip but use it on your show. If you change anything like move the microphone, you’ll have to get the profile from the actual show. Anyway.
Effect > Noise Reduction: Profile.
Select the whole show by clicking just above MUTE.
Effect > Noise Reduction: 12, 6, 6 > OK
Effect > Normalize: [X]Remove DC, [X]Normalize to -3.2 > OK
That’s it. If you do that to each chapter, export as WAV (safety) and then re-export as MP3, you should be good to go.
Hi, Koz,
Those are the directions you gave me, I pasted them above. Question #1: Are there actually two sets of directions here, one that begins with the word ‘Enclosed’ and the second that begins with the phrase, ‘Now, on to your show.’ I did what the first section says, I downloaded it and unzipped it and put it in my doc folder for Audacity stuff. But the rest of that section lost me. I don’t know what you mean by ‘point at.’ Also, I don’t know what ‘it’ is as in ‘from this point forward, it appears as a normal curve.’ What does?
Second question: Then in the second section, I selected the First Chapter, just completed, and tried to do the rest of the steps but only got to Effect/Equalization, because I don’t have any words on the screen like Select Curve LF Rolloff for Speech. And there’s only a straight green line dead center of the dialogue box that opened with Effect/Equalization, no curves.
The rest of the steps, I think I see those elements and will be able to do it, but I’m stuck at the beginning part.
Thank you, if I made enough sense to help me,
Janet
Koz, may I please ask a separate question? If I want to be able to present a download containing multiple chapters of a novel which will play in succession, how exactly do I do that? I am presently editing the first chapter of the novel (the noise reduction you gave me made visible tidying and I was just going along taking out too long pauses) and was thinking of how I would soon begin the second, and it dawned on me I don’t know how it works. Shall I save Chapter Two as a separate file? And in the end, will they all go in one folder, and that folder shall–unfold!–for the listener? : D I know nothing! I should be asking this of a producer of a website that would do this, but no, I just set up a wordpress site and expect to do it myself, using some kind of app, and Paypal. Someone mentioned Spotify.
Do you happen to know how it works?
Thank you, Janet
OK. The top half of that instruction list is to make a special sound filter available in the second half. It’s a special custom voice filter that’s not included with normal Audacity. That’s why you couldn’t find LF_rolloff. Unless you installed it, it’s just not there.
I need to push details off until tomorrow. Sunday night and Monday mornings can be a little challenging.
I should be asking this of a producer of a website that would do this, but no, I just set up a wordpress site and expect to do it myself, using some kind of app, and Paypal. Someone mentioned Spotify.
And what are you going to do after lunch. 
Koz
Enclosed, please find LF_rolloff_for_speech.xml.zip. Unzip it and that should leave you with LF_rolloff_for_speech.xml. That’s a custom effect designed by Steve to work with Effect > Equalization. Remember where you put it.
Where did you get stuck with this? Did you have any trouble “unzipping” LF_rolloff_for_speech.xml? It may not say “.xml” on the end. This is Windows trying to “help you” by hiding part of the filename.
You can force Windows to show you the whole, complete filename. Sometimes that’s a good idea.
Hidden File Extensions - Windows
– Start > My Computer > Tools > Folder Options > View > [ ] Hide Extensions for Known File Types (deselect)
– Apply (to this folder) or Apply to All Folders
– OK
Let us know.
Koz
Hi, Koz!! Boy, you made me laugh. I have overcome tougher obstacles than Paypal and Spotify. For example, I–me, not a company–brought my entire household of goods across the border from Mexico to the US at Nuevo Laredo. I will not elucidate. : D
Here’s where I got stuck with the --the roll-out thing (can’t go get the name of it, am in a new window!). I did the steps but now don’t know what to do with it. I see it gives settings. I think I downloaded it and unzipped it ok (the unzipped file has the name you gave me to recognize if I’d unzipped it right), and I can open it and read those settings.
But for the other section, I have applied the steps, normalize and equalization–I’m going from memory here due to lazy unwillingness to close this window and go look!–to my whole first chapter and judging from the change in the silence parts (from squiggly to darn near straight) I think a lot of noise got eliminated. Am grateful as all!
But what about the roll out thingy?
Janet
“Rolloff” and many of its cousins like “notch,” “clipping,” shelving filter," “brick wall” and “tilt” are all visual words describing something aural.
Clipping is the easy one. That’s when the real life sound becomes so loud that the recording system can’t handle it. The sound keeps getting louder but the recording doesn’t. If you look closely at the blue waves on the timeline, the tip of each wave is damaged and it looks like someone came by with scissors and clipped them off. Clipping is important because you can’t fix that later and it can permanently damage your show.
In the specific case of rolloff, many sound systems make sounds and tones that are very low pitch (thunder, church organ, earthquake) either accidentally or as an actual part of the show. Unless you are shooting an orchestra, most of these sounds aren’t helpful. Not only that, they take up room in your show and can mis-direct tools and filters, and a lot of the time, you can’t even hear them.
“For Some Reason” Audacity thinks my show is a lot louder than I can hear.
There’s a Hollywood joke about filming earthquakes. You can’t. You can’t hear them, either, so the best you can do is the vibrating whine glass. “Oh, my goodness, is that an earthquake?” switch to a closeup of the wine glass.
To avoid that problem, Steve wrote the equalization plugin: LF_rolloff_for_speech. When plugged into Effect > Equalization, it removes all those low-pitched tones and for most vocal performers, doesn’t affect the voice.
Noise Reduction, for one example, has a lot less work to do with those tones gone and you’re likely to get a cleaner and clearer show.
When you unzip the tool, do you see LF_rolloff_for_speech, or LF_rolloff_for_speech.xml?
Oh, “rolloff?” When Effect > Equalization is used with Steve’s plugin (and the “length” slider all the way up) the pattern on the screen is like a roller coaster and you are about to roll off the top. Steve specifically designed this tool to do little or no damage to the sound that is left. There are sharp cut-off or cliff filters that aren’t graceful at all and they can create sound damage.
Koz
Hi, Koz,
You did so much work on your reply, but it isn’t what I need! I can’t get it to work. I downloaded and unzipped the file, I stored it in my documents in my Audacity folder, and the name appears thus: LF_rolloff_for_speech, XML document. That’s the name I believe you gave me to prove I had unzipped it okay. I can open this files and it gives a list of settings. There’s no indication to me that such a small document would need to be zipped at all so perhaps I don’t even have the right file–maybe this is the documentation, not the program.
Anyway, here’s the place in your directions that things don’t work, because the choice Curve: LF rolloff for speech does not appear anywhere on the screen. I can’t make the selection:
— Open the show, chapter, clip, whatever. Select the whole thing by clicking just above MUTE.
Effect > Equalization > Select Curve: LF rolloff for speech, Length slider all the way up (8191) > OK.
Can you help me with this part? Thank you!
Janet
— Open the show, chapter, clip, whatever. Select the whole thing by clicking just above MUTE.
Effect > Equalization > Select Curve: LF rolloff for speech, Length slider all the way up (8191) > OK.
