Hum and Other Problems

Hello, I am a newbie and clueless about anything having to do with audio. But I wrote a book, published it through Amazon, and thought I’d try making an audiobook of it.

And nothing seems to be going right!

My #1 problem is the hum that’s always in the background, consistently, no matter what I try.

I started out in one room with my son’s huge gaming PC which makes a lot of noise. Then I moved the gaming PC into a closet. Everything sounded a lot quieter to me then but the same hum was still there on test recordings.

Then I bought a Mac Mini - noise still there.

I then took all electronics out of the room, even took off my Fitbit wristwatch. (Other gear: NT1-A microphone hung on one of those arm things and UR12 audio interface.)

Hum still the same with Mac Mini so I moved everything to another room. Same hum. Tried hanging up blankets in front of and then in back of mic. Same hum.

My sound floor is never, never lower than -42 db. This is with input gain knob on UR12 turned up all the way because, if I start to turn it down, my “testing 123” becomes too low very quickly.

It seems to be the low-hertz, ground loop type of noise that everyone talks about.

The only equipment I still have left to put up is the foam type stuff for deadening sound, but I don’t want to buy and put that up because I won’t be able to return it if I have to give up this project, and I don’t have any reason to think foam will help the hum. It’s in the electrical lines, probably, right?

I got so frustrated yesterday, trying to deal with not only the hum, but also the changeover from PC to Mac and other such problems, that I disassembled everything and put all the gear back in their original boxes for return. But today I’m reconsidering. I don’t want to be cowardly and give up. And I haven’t been patiently and carefully eliminating noise sources one-by-one, so I may have missed something.

I also haven’t tried one of those noise eliminator boxes yet.

So I thought I’d ask:

Should I just ignore the hum for now and soldier on? My son suggests I just do the high-pass thing - he knows how to do this himself and can show me - and suggests I just deal with the hum that way and not worry so much about it.

Or should I continue to work on eliminating the hum before I go any farther?

Or am I not projecting into the microphone properly, and failing to turn down gain as much as I should?

Or something else?

It’s discouraging because I read of so many people who have noise floor problems and then just “hang up some curtains and throw a rug down” and get great quality audio.

My son and husband are actually engineers (mechanical) so they can help me with some of the electrical stuff but they don’t have a lot of time to troubleshoot, especially since they don’t believe the hum is a big problem.

Any advice would be greatly appreciated.

Hello Therese,

Have you tried using the “Noise Reduction” Effect Plugin? It should be under the “effects” tab. Highlight a sample of static or hum noise and then click on “get noise profile”. Now the system recognizes the type of noise that you want removed. Now go back to the noise reduction effect, and now you can activate it. If my description wasn’t accurate, there are plenty of videos on Youtube on how to use the Noise Removal effect. I hope this helped :slight_smile:

If you can hear it, it’s not a “high pass thing,” and may be difficult to remove.

No, don’t send anything back yet.

I like to start with a test clip so we’re all working from the same ideas.

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

If you get stuck, just tell us where.

All that detail and I don’t think you ever told us what your microphone was or how you have it connected. That can make an enormous difference. There’s a line about everybody wanting to “improve their recordings by buying just the right microphone” when in fact everything else in their environment is broken, but in your case, the microphone may actually be broken.

And yes, totally, gamers wanting to record gaming commentary have to overcome their boiler factory gaming computer noise to do it. It’s not easy.

Congrats on the book. Unless you have some significant vocal problem, I don’t see any reason not to perform the vocal version, although just to scare you, it’s not unusual for first-timers to get to the end of the book with such improved technique and sound through experience and practice that they want to start over again.

Do you put people to sleep like I do? My vocal tone will go through soundproof walls and put people to sleep on the other side. It qualifies as a non-lethal weapon by the State of California.

I know all the ads show a Blue Yeti microphone parked on the kitchen table and somebody cranking out audiobook after audiobook, but that hasn’t been our experience.

Koz

Have you tried using the “Noise Reduction” Effect Plugin?

Yes, totally, but we found a notch filter works best for hum and that’s only if it’s not very bad. If you have multiple tone hum (normal), Noise Reduction ends up deleting large parts of the performance and is very difficult to set properly. ACX AudioBook is totally against the idea of using Noise Reduction at all.

Be careful about YouTube videos. If you get an older one by accident, you may get stuck with the older Noise Removal instead of the newer Noise Reduction. Totally different tool and not as effective in spoken word as the newer tool.

That and we don’t verify or record videos because it’s easy to make mistakes and it always takes far longer to correct the mistakes than many people are willing to put in. So the mistakes stay in.

We do keep up with the on-line manual.

http://manual.audacityteam.org/

http://manual.audacityteam.org/man/noise_reduction.html

Koz

published it through Amazon

What’s the book?

Koz

@Koz: The OP seems to use an NT1 and a UR22 as input. Hardly known for hum problems.

Could this be a natural hum location? I mean one where the hum is audible, but locals don’t hear it, because they’re used to it? :astonished:

locals don’t hear it, because they’re used to it?

Or much more likely, something else being called hum by accident [on edge of seat to hear test clip].

Another possibility: The elves can’t hear it…

Koz

THANKS SO MUCH to everyone for the kind and encouraging replies. They alone have helped me so much. I will keep trying.

As cyrano says, I am using an NT1-A microphone (hanging on one of those arm thingies) and it feeds into a UR12 interface.

Also, I could hear the hum when I was running my son’s gaming computer, but when I switched to the Mac Mini, I couldn’t hear the hum in the room itself. I heard it through the earphones when listening to the recording play back. I hope that makes it a candidate for either the Noise Reduction Effect Plugin that texastoast mentions or else for a notch filter.

I will get you an unfiltered sample as soon as possible. (I have to take everything back out of the boxes and set it all up again…)

Koz, the name of the book is “Ghosts and Other Unpleasantries.” It’s a collection of ghost stories. Thanks for asking. As for my voice, it sounds weak, hesitant, and yes, soporific. But it improves with either coffee or beer. (At least, it seems to improve…)

But it improves with either coffee or beer.

The performer or the listener?

Many alarm bells sound when a poster neatly wraps up the whole thing and presents a mostly solved problem, dusts off their hands and goes home.

Last time that happened, the poster was the only one who could hear the problem. What’s wrong with that picture?

The time before that, the sound was not even close to what the poster thought it was and all their assumptions were wrong.

We learn to pay attention, but not be swept up in the drama of the chase.

The goal is to walk into your “studio” (however you do it), present some vocal work, maybe cut it to length, set the volume and we’re done. It’s not to create a massive damaged sound file and then spend the next forty-five minutes emergency rescuing it. You have enough equipment in boxes right now to build a respectable announcing/presenting studio. We just have to walk you through setting it up and using it.

Koz

Hardly anyone likes their own voice, especially when new to recording. My son, who is seven, likes to juggle a mic around. And he likes to sing. Whenever I play back a recording, he runs.

And when it improves with beer or coffee, I’m certain it is all psychological anyways.

Besides, weak and soporific (had to look that one up) might be just what’s needed for ghost stories. I used to read those before going to bed :laughing:

Paper books stay open to the page where you nodded off. Audiobooks, however…

Koz


Koz, Audacity gave me the option of saving a 32 bit float wav file. If that’s what you want, I’ll upload that.

Yes, cyrano, I use audiobooks to fall asleep to a lot. My stories would be just the ticket.

This is fine. A lot of the Audacity default settings are remarkably effective. the Audacity default 16-bit WAV export has the same quality as an Audio CD.

See sentence about 2/3 of the way down.
Screen Shot 2016-07-08 at 5.31.49 PM.png
Attached I posted the corrected clip. It didn’t take much. I applied a custom vocal filter Steve designed and adjusted the overall volume. Dust off hands and go home. They were right. Suppress many of the low pitched tones.

This is the custom filter. I saved it on my web site. It’s also available on Audacity, but I can never find it there.

http://www.kozco.com/tech/audacity/Documents/LF_rolloff_for_speech.xml.zip

Unzip it to LF_rolloff_for_speech.xml

Adding Audacity Equalization Curves
– Select something on the timeline.
– Effect > Equalization > Save/Manage Curves > Import
– Select 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 preset curve list.



Using LF Rolloff (rumble filter)
– Select the whole clip or show by clicking just above MUTE.
– Effect > Equalization: LF Rolloff for speech, 8191 Length > OK

Length is an activity or strength number. The filter should also work with values in the 5000 range and cause less sound damage.


Follow that with:

– Select the whole clip or show by clicking just above MUTE.
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.2 > OK

Flynwill developed a special analyze tool that mimics the ACX AudioBook “Guard-Robot”—the first thing you hit when you submit to ACX.

http://wiki.audacityteam.org/wiki/Nyquist_Analyze_Plug-ins#ACX_Check

Download and install it in your Audacity plugins folder.

– Select the whole clip or show by clicking just above MUTE.
– Analyze > ACX Check



Make sure you have View > Show Clipping selected. You should not record so loud that you have red stripes in the blue waves (as you did on this clip). The general, fuzzy, overall goal is to record so the bouncing sound meter is around -6 and the blue wave tops show about 50% (0.5).

If you’re going to be expressive and theatrical, aim lower. You should not cause the sound meter to go all the way up and turn red. You can get permanent crunchy distortion by doing that. You should not record too low, background noise will be a problem

Koz

So - Koz - what you’re saying is that, with the fixes you applied, my sample passes the ACX test? And if I follow the recipe you gave above I should be able to publish the whole book?

Sorry to be so dense, but I just want to make sure. Can’t believe there’s a light at the end of the tunnel!

That last post had a lot of technical detail. The last part of the post describes “ACX Test,” the special analysis tool. Highly recommended.

Whoever said “get rid of all the low pitch tone rumble” was right. Most of the trash is 100Hz and below. The custom rumble filter gets rid of everything from 100Hz and down.

You can try the corrections on a full chapter to see what happens. Fair warning ACX AudioBook has specific requirements for silences, file sizes, formats, etc. and further, our ACX Test will not work if there isn’t at least a half-second of room tone (natural silence) somewhere in the clip.

http://www.acx.com/help/acx-audio-submission-requirements/201456300


So before anybody globally blesses your book, download and install those two specialized tools and make sure all three tools work. See if you can replicate what I did on your copy of the clip, and do post back when/if you get stuck.

ACX AudioBook has the bad form to require submissions in MP3. We would really rather they didn’t do that, but it’s their show. Audacity will not make a new MP3 without adding yet even more custom software. You should download and install the Lame software package. Did we ever settle on which computer you were using? Some of these tools change with the computer type.

Koz

I just started using a Mac Mini with OS X El Capitan 10.11.

http://www.audacityteam.org/download/mac/

Scroll down to where it ways Optional Downloads > LAME MP3 Encoder.

Install it and see if you can make an MP3 sound file out of that short clip we’ve been using.

Koz

Yes, Koz, it worked great!

Thanks so much for all the help. I really was close to giving up. I was really stupid, not listening to my son who told me the file would be fine with tweaking, and not checking with you folks earlier.

I was just freaked out by the ACX guidelines that seemed to say I should have a -60 noise floor while recording, before even starting to edit.

Now I will work at improving my recording room (and my voice technique) without having to worry that I will end up with a “garbage in, garbage out” problem.

Once again, thank you all for sharing your knowledge with a newbie. You can bet I will check in soon with the next problem that comes up.

I should have a -60 noise floor while recording, before even starting to edit.

They do and you should. That’s studio conditions. They also say under no circumstances should you use Noise Reduction, and it’s been our experience that many people can’t make ACX Conformance without a little.

In my opinion they forbid use of Noise Reduction because most people use it wrong and destroy their voice. Recorded work makes it past the automated robot and crashes immediately when Human Quality Control listens to it.

They also have shaky advice on making your MP3 submission and I will post notes on that…as soon as I find them.

Koz

We recommend very strongly to do all your personal production in WAV (Microsoft)—the Audacity default sound format—and only stoop to making an MP3 just before you submit. I know some people think they need to do everything in MP3 because that’s what ACX requires.

No.

There is a trick to submission.

be a 192kbps or higher MP3


192 is not a requirement, it’s the minimum. You can go higher than that.

You may upload 256kbps or 320kbps files if you’d like…


…and it’s recommended as long as you don’t go past the 120min chapter count or the file size limit…

Each uploaded file must be no larger than 170MB.



Those are all direct quotes from the ACX published specifications.

Submissions higher than minimum will help minimize distribution and product generation damage.

Koz