I can’t hear any artifacts on the one labelled “48 sample”, but the one labelled “96 sample” has LOTs of drop outs ,
( the computer can’t keep up with that high-rate and it’s missing little bits ).
I can’t hear any artifacts on the one labelled "48 sample
I can, particularly at 14 seconds, when I say “only”. There’s one between 8 and 9 seconds as well, just not as noticeable. I need to find the phrase I read for the first sample, see if that brings out as many.
I can I would’ve run into this before now?
I can’t hear any artifacts…
…I can
And there you have the insidious nature of the problem. I was on the fourth pass listening to my recording before I heard it.
Koz
And there you have the insidious nature of the problem
For real. The first time I noticed, I had literally gotten through about 14 minutes worth of audio, then suddenly was like…“what’s that squeak?”. Go back to the beginning and I can hear it every 10-15 seconds, and in bad areas, every 3-5. If it didn’t occur too often, I might just cross my fingers and try to shove it through ACXs qc on the hopes they won’t notice, but it’s one of those things that once you begin to notice it, you can’t _un_notice it.
I need to find the phrase I read for the first sample
One of the possible causes advanced earlier was a reaction to sibilants. Throw a bunch of those into a test. You have 20 seconds of WAV posting on the forum before it cuts you off.
Announce which sampling rate. It’s endlessly fascinating if there’s a significant difference between the two rates. The lower is Audio CD and the upper is Television. Both perfect formats.
“Sister Suzy’s Assistants reproduce a trance.”
Koz
The “and only” has got problems: your voice is cracking,there,
but that’s not the weird clarinet artifact problem you were trying to fix.
Going to 48kHz seems to have got rid of the clarinet, which was the brief.
There is a DeClicker plugin for Audacity at the bottom of this post that can smooth out voice-crackles.
[ get the DeEsser plugin while you’re there ].
So to recap, at 44100, the system appears to be digitizing your voice wrong giving you the clarinet beeps. At 48000 and higher, your system can’t keep up giving you gritty sound from tiny holes in the audio.
Just to be clear, neither is an Audacity problem although we’ll be happy to stick with you for testing.
It’s depressing that computer systems keep coming up with unexpected, exciting and innovative ways to screw up audio.
Koz
The 48kHz is clarinet-free. That particular 48kHz recording has crackles in places, but it’s regular wet-mouth-noises , not artifacts generated by hardware/software. Running a DeClicker over it makes the crackles disappear …
The “and only” has got problems: your voice is cracking,there,
I legitimately think it sounds just like the two part horn sound I had in the original files I uploaded, but it occurred today at 48khz also when I did sibillance and plain read testing at both sample rates. Although thank you for the link to the declicker. Editing out mouth noises is the bane of my existance and I would love something that could make it easier.
Just to be clear, neither is an Audacity problem although we’ll be happy to stick with you for testing.
Yeah, after I realized that it wasn’t an editing thing that made the issue appear, I knew it likely wasn’t Audacity. Thanks for sticking with me.
I did your Sister Suzy sibilant test at both 44 and 48khz. Then I did a plain read at both 44 and 48khz, reading the same bit I uploaded originally. I don’t notice anything major (surprisingly) in the sibillance testing, but its fairly clear in the plain read that I still clarinet at both 44 and 48.
Then I went to a different room and desk in my house. I recorded my last book at this location, but stopped using it and built the blanket fort specifically because there is a lot of computer hum in this room that I unfortunately cannot avoid. It still passed ACX quality control with absolutely no problems, but I wanted to sound better. In here, I don’t seem to have issues at either 44 or 48.
Apologies for plane noises. I have no idea what possessed me to get into audiobooks when I live near an airport.
So, here’s four files for your perusment. I had to use mp3 again as even under 20 seconds, all my files were too large for the forums.
[edited to add: I can’t upload the other two files that were produced in the other room at the moment. It won’t allow more than four files, but more than that, won’t allow me to post again. But they’re honk free! I’ll post them later.]
File uploading is something of a moving target. This is The Last Word:
https://forum.audacityteam.org/t/how-to-post-an-audio-sample/29851/1
It’s 2MB total. That works out to 10 seconds (12 officially) for 44100, 16-bit, Stereo. 20 seconds (24 officially) for mono.
If you use a higher quality export (48000 rather than 4100) you’re going to run out faster.
Sometimes, if forum surgical file deconstruction is irrelevant you can use MP3 at Really High Quality like 256. In that case, I think I measured the file submission runs out at about a minute.
Editing out mouth noises is the bane of my existance
The other bane.
there is a lot of computer hum in this room
That at least is predictable and steps can be taken. It’s not a “magic” problem. I would kill to be able to create the honk sounds at will.
“Every time I say the word ‘chilibean’ on Thuresday afternoon, it creates the problem.”
I’m going to drop a note to Flynwill. He has more engineering juju than I do.
Koz
Paul-L’s DeEsser is one of few which is capable of removing your whistling-f on “from” …
To reduce the incidence of cracking you could try …
increasing the distance between you & the mic , and/or adding some additional layer[s] of fabric to your pop-screen.
increasing the distance between you & the mic , and/or adding some additional layer[s] of fabric to your pop-screen.
In the past week, I’ve been doing something of this. I always move back when doing loud characters/emotional bits, but whats seemed to help more was adding a windsock. It does seem to be happening during big sibilant moments, although not typically “s” sounds, but th’s and even j’s. I still can’t make it reliably occur. I can say the same sentence 6 times, and it may occur two or three of the times. I’ve been recording with a headset on to help catch when it occurs, so I can try resaying the sentence. I’ve also gotten a lot better at learning to spot them in the spectrogram, which is helpful.
After this book is over, I’m taking a bit of time off books to build a ‘real’ booth (although I may miss my blanket fort), and to get a different mic and a proper set up. I like the warmth this mic lends, but if it turns out to be the mic making me squeaky, it ain’t worth it.
There does have to be a gap between the pop-shield and the mic for it to work, 2-4 inches, (that’s another variable you can try experimenting with).
get a different mic and a proper set up.
It bothers me not knowing what’s causing the problem. You have no guarantee you’re not bringing the problem forward into the new studio.
It’s most concerning that in a multiple pass test the problem comes and goes. I meant to say up the thread if I didn’t, it counts if you can make it worse or better. I’ll take any predictable change either way.
Any chances of visiting somebody else’s studio?
Koz
It gets worse while you’re editing?
I thought it did, but I’m pretty sure I was wrong. I think I just didn’t notice it until edits, when I was listening to things really closely.
I’m now 95% sure it’s me. If I wear headphones and listen to it live as I record, I hear it happen sometimes.
I do have a question though - on your spectrogram, I can see them so clearly (beyond the fact you helpfully highlighted them) - when I look at the spectrogram myself, I can hardly ever find the buggers.
I’ve listened for a fairly long time to all the samples in this thread. First, on laptop speakers. No go. So I changed to the audio computer. No go. Got out my best cans. Still no go. Pumped up the volume. Nothing. Imported into Audacity. No squeaks.
I do hear a difference in Trebor’s before/after, but I don’t hear any squeaks.
I just can’t hear the squeaks.
Of course, it could be my tinnitus acting up. But that’s only in one ear and I even pulled out a one ear “pair” of cans.
Are we dealing with a case of Sympathitis, the little niece of Mass Hysteria? 
You’re not all using the same playback system, by any chance?
@MarieM. Have you effectively listened back to one of your previous recordings to see if you could hear squeaks NOW? Could be you never noticed them before…
It’s something of a common theme that nobody hears these problems on first pass, and sometimes multiple passes. So unless a paying client hears them, you’re good to go.
Koz
Are we dealing with a case of Sympathitis, the little niece of Mass Hysteria?
If Trebor hadn’t been able to highlight the exact points it’s happening, I wouldn’t disagree! In fact, I worried after posting that it was just me. I’d had another person here listen and they couldn’t hear them. Hell, I did my first run edits on the first chapter I noticed them in without really noticing them. Going back to do my second pass, I noticed it, and that’s why I initially thought it was an audacity problem, and a product of editing.
Truth be told, I thought I was gonna get a slap on the hands for editing within a project, since I’ve been told before by koz that they were somewhat brittle
Have you effectively listened back to one of your previous recordings to see if you could hear squeaks NOW? Could be you never noticed them before…
My previous chapters in this book do not have it. I essentially recorded chapters 1-9 of this book with the exact same equipment and setup. Chapter 10, bam, start squeaking. I’m still squeaking now, but this book is due tomorrow, so I’m powering through. Some squeaks I can manipulate out through finding them in the spectrogram. If all else fails, I can occasionally amplify that spot a bit lower to make it less noticeable. I’m running on the hope that it won’t be as noticeable to a consumer, versus someone like me who sits here and listens to a passage repeatedly.
They’re actually somewhat more noticeable to me now - I don’t know if this is because I’m attuned to hearing them now (likely) or if the issue in the end, is me and not my equipment, and I have something up with dem vocal cords.
I’m still suspecting the gear used…
Don’t take this as advice to run out and buy something else.
A few years ago, I was called in by a mastering engineer. He had a faint whistle in his studio chain. I’m used to not hearing it on the first session. So I start checking the gear and can’t find anything wrong with it. Meanwhile he is trying to make me hear the whistle by applying gain and eq. And suddenly, I hear it. We start eliminating every device in the chain, by bypassing each and every one of them. Nothing…
After hours, finding no evidence at all of gear failure, I leave.
About a month later, he calls me in again. The poor fella looks like he’s dying. I ask him about his health, noticing he seems oddly happy. He assures me he is fine and he found the issue. It’s not the gear an sich, it’s the order of the gear. And could I rearrange the rack, please?
He’d spent every free moment rearranging all the devices and listening if and when the whistle would vanish. And he found the magic order, after a lot of sleepless nights. He was feeling mighty proud of himself and wasn’t up to unscrewing about a dozen boxes from the rack. He also made a big spaghetti of my neat cabling that was there before.
I was glad to help him with that. Made me feel a little less useless…
Don’t concentrate too hard on this problem. If it’s you, it’ll probably go away and you might never know what caused it. And even if it’s the gear, it might also vanish one day, the same way it came.
[Big spaghetti…]
And that’s what did it. Not the equipment order.
It’s possible to be too neat. If you offend the cable gods, you can get cables talking quietly to each other. Do Not clean up the cables.
Koz
has anyone found an actionable solution for this problem?
i’m helping my sister produce audio books and we’re having this same issue 