Room Noise Floor -59.6DB Argh! [SOLVED]

I am now having a look at those reflection/isolation booths

Make sure you need it before you pop for that much money. They’re intrusive and not cheap.

Behringer UM2 seeing as you said it was pretty good

I said I liked it. It seems to do everything I need although that’s not my first choice.

I own a Shure X2U which does a very similar job, but the designers didn’t have me in mind. It’s too quiet and it’s noisier than I think is good. It’s lose/lose. I’ve had it for years and I’ve never used it on a job. I complained to Shure and they said, “That’s the way it is. Have a happy day.”

You should know that it’s nearly 100% certain you will not need the Yeti Curse filter in your correction suite. You won’t have a Yeti any more. That and you will have an XLR cable between the microphone and the UM2.

(You should make sure you have at least one XLR cable in your kit) The USB cable has a practical limit of about 6 feet or so (2M), But there is no practical limit for the XLR. 100 feet (30M). If you do that, I should avoid running the cable over high power electric motors or in front of broadcast radio transmitters.

Koz

Thanks for this extra advice. I spent half of yesterday rifling round the house finding old pillows, blankets and duvets to out on the walls. I even found a wooden board covered in material to cover part of the window and the rest is covered by boxes draped with duvet covers and pillow cases.

My recording studio now resembles a very small, very hot cave :wink: I can’t wait to test out the new set-up.

Do create a standard sound test and post it. I’ll tell you the list of corrections you probably don’t need any more (hopefully).

A note on sound-proofing material. There is a scam where someone passes off shipping foam blocks as soundproofing because they look similar. Those have very different goals. The goal of shipping foam is to take up space, cushion the work and not weigh anything (shipping weight, right?).

Generally the goal of acoustic foam is to break up sound so it doesn’t bounce between walls. It works better if it’s heavy. It’s “acoustic mud” in addition to that funny pyramid shape to reject echoes. Sound loses "zot’ (technical term) each time it has to bounce between surfaces. Pyramids guarantee at least two bounces.

I know this is fun to obsess over, but I don’t remember anyone complaining about “boxy sound” or echoes in your postings. Maybe isolation from neighborhood metrobusses, yes, I did hear that once, so it’s good to keep the goals in order. I considered installing a heavy curtain rod over my one window and drape a moving blanket over it before I went to the effort of cutting a wooden cover or seal.

Koz

You may find shifting goals with the duvet as well. Furniture moving blankets are designed to be gooshy, firm and heavy to prevent damage when your credenza falls over in a moving lorry. The goal of a duvet is to be decorative, light, fluffy and warm. It’s not unusual for a duvet to not affect sound at all. You can get a fuzzy idea on how it’s going to work by draping it over your head and listen. Did the metrobus sound vanish? Muffle?

The ladies at the cloth store are used to me by now. I go down the ranks and bolts of cloth listening to them. The last time I did that the goal was to find something to drape over a microphone to break up wind motion without affecting sound. That’s the JoAnn Fabrics version of “Dead Cat” fabric.
Screen Shot 2017-03-29 at 11.46.35.png
There was a YouTube posting of someone who found perfect fabric and cranked out several cheap and effective dead cats and posted how he did it. He did not post where he got the fabric from and it turned out after many messages back and forth the material was a manufacturing end and no longer available.

And you will have a happy day, won’t you?

Koz

Hya,

My sample has been beset by next-door neighbour DIY for the moment. I’m hoping to get one up today (how much more sawing can he possibly do?).

By the way, is there anything you recommend for editing sound on video. I’m putting together a charity project (with no budget) and have to remove a few extraneous sounds, increase volume of speech etc. But whenever I try to put the edited track back together with the video, it seems to slowly de-sync (this is using Windows Movie Maker).

I have Wondershare Filmora, but haven’t tried it on that yet as I only just got it to put titles onto the first video (which didn’t need sound editing). Maybe it#ll do a better job of it?

Hope you’ve been enjoying the break from my endless questions. Ah well, you knew it couldn’t last. And I am very grateful for all your advice, as ever :slight_smile:

I know what used to cause desync. In the US, analog video didn’t run at 30 FPS. It ran at NTSC rates 29.97. Constant problems with video and sound drifting apart. Some older TV shows are broadcast on digital TV at the old rate. There was the old joke that an hour and 60 minutes weren’t the same in the US.

Make up your mind. Which do you want?

But Europe didn’t have that problem, so that is a mystery.

You can “pre-bend” the sound to match. Measure the two events and put those two measurements in Effect > Change Speed. That should create a new sound track that will fit.

Neighbors over the fence have chickens and for some reason they never run out of chicken noises.

Koz

Ah-hah! Thank you. I seem to recall doing this using Virtual Dub as well. It was ages ago, but I will have a play with some of these tools and see what I can do.

Aaaaand, back to my audiobook! The DIY has stopped, although I can’t control the trains and aeroplanes…

I’m still using the Yeti as my audio interface has to be ordered and will take a month or two to come. But I want to get the first book recorded, and will use the Rode and interface to do further work hopefully.

So, for now, with a bit more padding in the room to reduce that boxy sound, here is my sample of raw audio:

The process we had before was:

EQ: LF Rolloff filter around 5000 (5013)

Run Nyquist Prompt – (SetRMS: Target RMS -20.00, Linked Stereo > OK)

Run Nyquist Prompt – (YetiCurse)

Limiter: Soft Limit, 0.00, 0.00, -3.50, 10.00, No > OK

Apply Noise Reduction
— Drag-select Room Tone, silence or the flat area between spoken phrases.
— Effect > Noise Reduction: Profile
— Select the whole clip or show by clicking just above MUTE.
— Effect > Noise Reduction: Settings 9, 6, 6 > OK

Optional:
De-clicker at defaults (will lower noise floor)
De-esser

With the new set-up, do I need to change anything? Many thanks for your endless patience Koz! :wink:

I need to settle between Life Tasks long enough to listen.

You are warned that ACX places very firm importance on matching, so if you start with the Yeti, chance is very good you need to finish that way, too. Don’t change microphones in mid stream, so to speak.

Koz

I applied your published correction suite and the clip sails straight through ACX-Check and I’m guessing it sounds exactly like you. There are no prominent background noises and even if there were, they’re well below the volume of being important.

I did not apply the de-esser/de-clicker thing. You join the religious community that swears by those two tools one way or the other. I didn’t notice any significant aural defects and you should note one good goal is to apply as few corrections as possible against ACX finding them and rejecting the work. Also, a long laundry list of corrections is going to get really tired about mid-way through the third book.

Also note RMS Normalize is a thing now. It’s a formal effect and doesn’t require constant copy/paste, load/save in Nyquist Prompt. It replaces SetRMS but only works in Audacity 2.1.3.

https://forum.audacityteam.org/t/rms-normalize/45334/1

Koz

Cool. thank you Koz.

You’ve given me so much help that I feel much more comfortable experimenting and changing settings to suit.

My sample clip with the new padding in the room, and the laptop placed outside, already passed Peak Level and Noise Floor for ACX and I was only a few DB below RMS level, so I also tried this:

Run Nyquist Prompt – (YetiCurse)

Apply Noise Reduction
— Drag-select Room Tone, silence or the flat area between spoken phrases.
— Effect > Noise Reduction: Profile
— Select the whole clip or show by clicking just above MUTE.
— Effect > Noise Reduction: Settings 6, 6, 6 > OK

Amplify: 2DB

Limiter: Soft Limit, 0.00, 0.00, -3.50, 10.00, No > OK

It sounds very clear and unprocessed, my voice sounds warm and not boxy, and there is no extraneous noise - and it passes ACX! :slight_smile: Is there any reason not to use this? It was only a short clip, and when I read the whole book, there may be a little more needed.

Thank so so much for all your help and advice. Through talking to you, I have been able to understand what good quality sounds like, forgive my cursed Yeti, and have learned so much about what all the tecchy stuff really means and how to use the settings.

I can not thank you enough. You are a superstar Koz :smiley:

The first time somebody writes you a check, see orange Donate button on the top of the forum.

Koz

Funnily enough, I was going to ask you if the donations were shared among the people that help on the forum as well, because I’m more than happy to donate if that is the case :smiley:

Maybe you can donate anyway. It’s still volunteer. We don’t get a check at the end of the month. We do things such as buying a Mac so Gale can maintain The Standard Mac Platform for troubleshooting.

It has to be a machine with no production duties. We can’t tell a client we have experimental software and their show is in flames.

Well, we could, but they probably wouldn’t take it very well.
Koz

OK, no problem. I will donate. And for you, and others who are volunteering, a big crate of virtual chocolate/beer/cake/whatever-you’re-into :laughing: and a whole load of thanks.

God, another Argh!

So, I recorded the audiobook in one sitting, ran the process I needed over the book, and it was all sounding great…except for a few bits that were a little distorted or had an aeroplane in the background.

So, I made a note of all the bits that had been affected and re-recorded them. I ran the same process over them, but they sound completely different to the original recording! Trying to put a part in, it is really obvious it is not the same track.

How has this happened?! I’m guessing the room tone must have been slightly different as I didn’t change any settings between recording and re-recording.

I’m guessing that you have to

  1. Record the whole thing and DO NOT MASTER
  2. Then listen through
  3. Re-record the mess-ups
  4. Paste the re-records in
  5. THEN master the entire thing together.

Or is there any way I can rescue the parts that have been re-recorded to make them sound the same as the original? They sound more boxy and muffled than the rest of the book!

I think I’m going to have to record the whole book all over again, which is such a shame as there are only a few minor errors where this happens, but I know they are too much to be acceptable. :cry:

  1. Re-record the mess-ups
  2. Paste the re-records in

Yes, but maybe not in the way you mean. The tools that work over the whole course of a chapter may get confused if you instead apply them to a tiny portion of the show. If you know you have a fluffer, pause very briefly, re-record the work right then and keep on going. The chapter-wide tools should correct both the fluff and the correction to the same sound. Then you go in later in post production and slice out the fluff.

We are working with some Felt-Tip Pen on Paper Napkin programming, so if you’re experiencing some actual bugs (undesirable, unintended actions…etc), you may be the first user ever to do so. And yes, a “bug” is a very precise, specific thing. It’s not just a screwup.

Did we determine you had a way to post large sound files? You can’t post much over about 20 seconds of mono voice on the forum, but some people use File Posting Services if they have to share large works.

We may be able to cobble together a correction so you don’t have to read the whole thing again. It would also be good to know what happens to the software if what happened to you happens to others. Nothing like recommending a process that fails. We try really hard not to do that.

We may have a recommendation for file services. As we go.

Koz

It’s good to wear headphones while you perform. There is a certain amount of self-correction that happens when you try to get expressive and theatrical and you can hear yourself going crazy in your headphones. Instant Quality Control Feedback. Unfortunately, you can’t listen to the computer to do that. Most computer services are delayed and the sound can have a very serious echo.

So without going back over many multiple forum chapters, does your microphone or microphone system have a place to plug headphones?

Koz

I have a semi-active DropBox account. They used to have a free account as long as you didn’t go over a certain amount of storage. They may not have that any more.

DropBox has a setup where they appear as another storage drive on your computer—and you could mount the same one on several different computers. It’s dead simple to push files between machines no matter where on Earth they are.

I know people who manage complex productions like that.

Koz

Just reading that again…

Are you trying to patch an error before or after processing? Voice quality should be relatively consistent if you do it before processing. But yes, if you like to wander or change spacing or worse, have to set up your studio and tear it down at every performance, the possibilities of matching go way down.

Koz

Hya,

Hope you had a nice Easter. Thanks for the advice.

I hadn’t been using headphones, and since the computer was outside the room, it was difficult to monitor exactly what was being recorded into the mic or where it might distort. (lesson learned)

If I made a mistake or heard an aeroplane, I would stop talking and rerecord the sentence again.

But listening back, there were little places I didn’t catch (probably because I was concentrating on reading) where an aeroplane has slipped in. There are also parts, that - after processing - sound a little harsh and distorted, like my mic couldn’t take my higher pitch when emphasising certain parts.

So, there were little bits that I had to re-record. I saved the corrections pre-processing, and they sound much better slotted into the raw uncorrected version than mastering them separately. Some of them are pretty much seamless. So most of it will be salvagable :smiley:

There are, however, a few parts where my voice sounds a bit different (it does, depending on what/how much I have eaten/drunk/smoked, what time of day it is etc.) Or where I have spoken in a different “tone” from the rest of the passage.

I have pasted all my corrections into the raw audio now and will listen to them all through thoroughly. Any bits that sound a bit out-of-place, I think I will have to save the original as a small clip on my tablet (seeing as I can’t have my computer in my room), and relisten to exactly the tone of voice and try to completely mimic it. We will see how it goes…

I am so glad I am learning all this on my first book, which is fairly short. I will wait for my proper kit and get some mic headphones before even attempting the second, which is far longer…