Removing echo and enhancing audio

Hey everyone

I’m recording some audio for my youtube channel. I’ve got a better mic recently (nothing fancy, but way better, than the headset I used before). I’ve noticed, that it’s catching a lot of echo from all around. I’m using a DIY solution, by putting it in a box lined with cloth to reduce the sound and covering with a blanket in front. I can’t seem to get hang of it, so I’ve figured out, that maybe I can improve it with some post processing? And maybe make it sound more “meatier”, if you know what I mean? I’ve attached a sample with no effects on it (I usually use noise reduction and some bump-up of the volume). Do anyone of you have any guides, how can I improve it in the future?

Exactly correct. If you post a “clean” recording we have a fighting chance to help you. If you post a clip after post production, all we can tell you is, “you did it wrong.” We can’t take effects and filters out.

I can’t listen to it right this second, but yes, this is the magic place that kills most people. Live recording is harder than everyone thinks.

What is the microphone? The more information, the better. Depending on the headset, you may have been better off with that. I have a broadcast headset that I used with my podcasting sound test. I also have a cheap USB headset whose microphone is terrible. I think it’s intended to be a gaming headset, not podcast quality.

Which computer and OS, Which numbered Audacity?

If it’s Windows, sometimes Windows will try to help you whether or not you want the help.

Koz

my youtube channel.

Which is where?

Koz

As it’s a true stereo-recording , using Audacity’s “isolate centre” will remove some of the room reverberation [echo] from the recording , ( the result will be mono ).

I can suppress the hiss noise (you probably got that far), but we can’t do anything with the “recording in the kitchen” sound.

Modern interiors of polished wood bare floors, bare walls and ceiling pretty much guarantees a terrible voice recording. These rooms are aggressively hostile recording environments.

There are ways to get around this with fancy microphones, but nothing normal people are going to be able to afford.

So you are probably doing the right things with your sound-proof tunnel, only you may need to make it a lot more soundproof. “Cloth” is not the right word. You should use words like Heavy Blankets, Quilts (Duvets), and Fluffy Towels. I am a proud user of furniture moving blankets.

That’s overkill. I built a soundbooth that I can transport to different locations and supports two people. It knocks down to the back of my pickup truck.

So fill in the blanks. What’s the microphone and computer, and where are your existing podcasts? Describe the room.

Koz

Also, recording your voice in stereo adds to the problem. So I’m equally curious what mic you are using.

I’ve tried the box model, but it doesn’t work for me. It always sounded boxy, unless it’ a real BIG box. Koz’s method is easier to start with, as it surpresses the parts of the room you don’t want (reflections from walls, ceiling, floor and furniture) and you can move it if you need to.

Setting up a recording environment in a new location takes me two to four hours and that’s not because of the gear, it’s simply because I need to walk around with the mic I"ll be using, listening through headphones to the room. That’s the way to find the sweet spot where the reflections cancel out as much as possible. Once you’ve found that spot, it’s easy enough to add some absorption to calm down the “air”.

If you’ll be recording in the same room all the time, take a week or so do some test recordings in different spots. Store these and listen to them the next week, together with some podcasts you like as comparison.

You’ll notice that mic placement is much more important that gear. Even a lowend simple mic can sound quite good and even an expensive Neumann or DPA will sound bad if you don’t take care of placement.

That fat, heavy, American voice style can easily be created afterwards, with some gentle compression and some bass lifting EQ. A plugin like LoudMax will do that in a jiffy.

Oh, and one other thing. Turnaround on this forum can be minutes, not weeks. So, don’t post a complex question and then go on holiday to Côte d’Azur.

Koz

Ok, so in order of asked questions:

This is the microphone that I use:
http://www.samsontech.com/samson/products/microphones/usb-microphones/meteormic/

I run Windows 7 x64 and use audacity 2.1.2 (portable)

My youtube channel you can find here:
https://www.youtube.com/channel/UCReanM5W_neKww6Fwh1Pohg

@Trebor wow, that is actually a lot better

I’m recording in a living room, that is quite crowded. Sofa, table, chairs, big wardrobe and few smaller furniture pieces.

Do you suggest, that I try recording in mono? Also, how should I place myself and the mic? is it better to face it to have as much distance to the wall, or maybe the oposite - the closer is better?

If you start recording in mono, Trebor’s echo trick will stop working. It works by comparing Left with Right assuming most of the differences might be echo.

That’s a good and bad thing. Normal directional microphones have a thing called Proximity Effect. you get bassier and more “American Announcer” as you get closer. On your microphone, that’s going to mess up your stereo imaging and kill echo cancellation.

But yes, close is good. The limit is when you start distorting the quality and popping your P sounds.

http://kozco.com/tech/audacity/clips/PPopping.mp3

This may not sound like much on a whimpy sound system, but on a large system the P impacts are enough to scare the cat.

I need to make it home before I listen to the samples.

Koz

The Samson Meteor is a mono mic. Where is your stereo effect coming from? :open_mouth:

I’m no professional, but I think that sound bounces back from the walls and comes back to it.

Welcome to the paradox, ladies and gentlemen.

Echo is the single voice from your mono microphone coming back from the walls and ceiling late. This is the effect that makes your clap come back to you after a short pause from the building across the street. This is the big auditorium sound.

Stereo Effect is two different voices. Stereo effect gives you violins on the left and trumpets on the right. In your case, it gives us slightly different versions of your single voice, one on the left and a slightly different one on the right.

Since this is a mono microphone, something in the computer is apparently creating two different voices.

Even weirder, Echo cancellation normally only happens with two microphones or one stereo microphone. You don’t have either one. So we may have discovered the long-sought tool to suppress echoes in any voice recording.

You’ll be famous!!

Koz



Possibly kjubus is recording from the Samson USB mic and [unintentionally] from the computer’s built-in microphone ?

Still, echo cancellation happens by processing the difference between left and right. Both the Meteor and the built-in microphone would have recorded as mono with no difference between left and right.

No, I don’t believe the built-in microphone ended up on the left and the Meteor on the right. You would only get that by breaking too many things at once. If there is a Meteor driver and the driver fails, and if the built-in has environment processing and if.

Sorry. No.

I do have a microphone that can appear as either a one-track mono recording or “two track mono.” It appears to be perfectly content generating either one or two blue waves. I also have a microphone that will vanish if you get the stereo/mono switching wrong.

@kjubus
Was there a software package associated with the Meteor? Did you download anything? Install special software?

Koz

No, there was no software included and I didn’t download anything. I’ll check later, if maybe built-in speaker was recording as well, but I don’t think so

Ok, I’ve checked - I probably didn’t change the input microphone on audacity that time. I’ve recorded a new sample for you, attached to this post.

Again, I didn’t apply any post-processing. Also, I’ve improvised some sound proofing similiar to the one kozikowski showed behind my back. Please let me know what do you think now and if you would enhance this recording anyhow (apart obvious background noise reduction).

Whatever you’ve done it’s reduced the reverberation a lot.

There is a faint constant whine, which can be removed by putting the code below into Audacity’s Nyquist prompt

(let* ((q 50)                 ; set the base Q for the filter
           (freq 1000)            ; set base frequency
           ;set the number of iterations
           (iter (truncate (/ *sound-srate* (* 2 freq)))))
      (dotimes (i iter s)         ; start the DO loop
      (setf s (notch2 s (* freq (1+ i)) (* q (1+ i))))))

i.e.
Nyquist code to remove 1kHz + harmonics whine.png
IMO the high frequencies benefit from a boost. Try importing the equalization curve into Audacity’s equalizer, then applying it to recordings made with that particular microphone in that location …
kjubus.xml (17.6 KB)

Thank you, but I’m not very advanced in audacity and can’t seem to get it working. I need to select the audio track, go to effects → nyquist prompt, paste the code and click OK, right? Because that doesn’t seem to change anything. The noise reduction almost does the trick, but I’d like to hear if I do anything wrong here.

You may have omitted to tick the box in Nyquist Prompt , ( which I coloured yellow in by previous post).

The whine isn’t too bad on your original , so the effect of the DeWhine code is subtle …

A great system to shut down that pesky echo build a portable box to surround yourself and use egg carton dividers
they work great and cheap most any restaurant will have the larger 30 count style. built a four room studio for seperating singer from band. enjoy
egg crates in studio.png