Muting the TV in the clip - Windows 10 - 2.1.1

[DeerBear]

Hello!

I am a complete noob at audio editing.
I understand what a track is, what a sound wave is but not much more than that.

So, I have a clip where the recording device was very close to a turned on TV.
Now, there are people talking, but the TV is taking over totally and I can't hear what they are saying.

I have already extracted the sounds as a sound clip (I am a noob at editing, but I know a fair bit
about computers).

So, I need help in understanding how to do this.
My first thought was to remove the background noise, but then it occurred to me that - actually - the TV might be the foreground noise,
so that is what I would probably have to do.

As you can see, I am not even sure what is the right approach and I need some schooling

Anybody's got the patience to walk me through the process please? I am not even quite sure what are the
right questions to ask or what information is really important, which is why I am being generic and not providing a lot of information.

Thank you for your time!

A

[Gale Andrews]

No, can't be done with conventional audio software.

If it's for a court case, you should use your favourite search engine to look for "audio forensics". Only a forensics practitioner recognised by the legal system would be regarded as able to provide legally admissible audio evidence.

Gale

[DeerBear]

Hi Gale!

First of all, thank you for your answer.

I hate it when things can't be done without the help of special software but, alas, I am software
developer so I know exactly what you mean there.
Heh. For users who may look to do something like this: NO means no. This is not like "no it's a long process",
this is more like "No, the bloody sounds overlap so to extract the exact ones you need you would have to have a
much more powerful machine with completely different algorithms and selling a software that does this may
even be illegal". So think about that before throwing a tantrum.

Anyway, with the help of a fellow user who's been using Audacity for quite some time, we went through the
motions: reduce noise, equalise, etc. So I was looking to see whether it was legal to buy forensics software for
personal use and I found a company which sells a "Speech clarification processor". It has a really complicated
download form (I suppose, to ensure that no rapist, blackmailer or otherwise bad guy gets his hands on it) but I
thought "hey, this could be the thing for me: if I cannot remove the TV, maybe I can clarify what is being said".
So I needed to find where in Audacity you would load the VST plug ins. Since this, to me, sounds like an effect,
I tried browsing down to see whether I'd find some "Load effect" or "Load plug-in" kind of thing.

I was met, instead, with a "Vocal Reduction and Isolation" effect - BANG!
I tried it on a clean copy and it's almost identical to all the work I'd done together with the more
advanced user. Bingo! This does not turn the TV down or mute it, but it helps tremendously to
clarify what is being said.

So... I have a much better understanding of Audacity now and also a much better sound

Now, the only thing that remains to do is: how do I put this new sound file in the clip?
Anybody knows?

Thanks!

A

[kozikowski]

how do I put this new sound file in the clip?

Video clip?
Export the work or prepare a WAV file of the work. Video works at 48000/16-bit/Stereo, not 44100/16-bit/Stereo, but most editing programs know how to deal with this.

Open the show in your video editor program and import the new sound file.

Koz

[DeerBear]

Hi!

Thanks!

I found a program called "AviDemux" and it was really easy:

- Install the program
- Open the video file
- Go to the "Audio" menu
- Click on "Select track"
- From the window that appears, select the first track
- Click on the arrow down
- You will see two options: one is the current track and then another one called "Add track"
- Select the 2nd option, which will show an open dialog
- Select the track you want to replace the current one with
- Press ok
- Keep pressing Ok
- Save

Done

A

[Gale Andrews]

I hate it when things can't be done without the help of special software but, alas, I am software developer so I know exactly what you mean there.

Koz here usually answers like this:

The Four Horsemen of Audio Recording (reliable, time-tested ways to kill your show)
-- 1. Echoes and room reverberation (Don't record the show in your mum's kitchen.)
-- 2. Overload and Clipping (Sound that's recorded too loud is permanently trashed.)
-- 3. Compression Damage (Never do production in MP3.)
-- 4. Background Sound (Don't leave the TV on in the next room.)

For users who may look to do something like this: NO means no. This is not like "no it's a long process", this is more like "No, the [...] sounds overlap so to extract the exact ones you need you would have to have a much more powerful machine with completely different algorithms and selling a software that does this may even be illegal".

From your description, which sounded as if the TV and the inaudible voices were both spread across the stereo spectrum, with no audio sample to listen to, and nothing about how the intended audio came to be like that, the answer did seem to be no, using Audacity.

Almost always the answer really is "no", if it's a covert recording, which many of these requests are.

Anyway, with the help of a fellow user who's been using Audacity for quite some time, we went through the motions: reduce noise, equalise, etc. So I was looking to see whether it was legal to buy forensics software for personal use and I found a company which sells a "Speech clarification processor".

This - http://www.salientsciences.com/dacaudio ... Manual.pdf?

It's interesting, it has a deconvolver that Audacity does not have.

It has a really complicated download form (I suppose, to ensure that no rapist, blackmailer or otherwise bad guy gets his hands on it) but I thought "hey, this could be the thing for me: if I cannot remove the TV, maybe I can clarify what is being said".
So I needed to find where in Audacity you would load the VST plug ins. Since this, to me, sounds like an effect, I tried browsing down to see whether I'd find some "Load effect" or "Load plug-in" kind of thing.

So did you try some demo of that enhancer? Was it better than what you had already done?

I was met, instead, with a "Vocal Reduction and Isolation" effect - BANG!
I tried it on a clean copy and it's almost identical to all the work I'd done together with the more
advanced user. Bingo! This does not turn the TV down or mute it, but it helps tremendously to
clarify what is being said.

So tell us more - did you use isolation?

the only thing that remains to do is: how do I put this new sound file in the clip?

Koz beat me to it. I too would recommend Avidemux which can do this without re-encoding the video, which would degrade it.

For anyone else reading this, depending on the format of the video, you may need the exported audio to be in the same format as the original audio in order to replace it without re-encoding.

Even if you re-encode the video to another format, the audio would have to be compatible with that video format.


Gale

[DeerBear]

Hi!

Yes, the TV and the voices overlap, there are not multiple tracks. If there had been, I'd have just muted the right one
There are some moments where there is no talk and the TV is still "talking", but my fear is that, if I use that, I will end up
muting the voices as well.
The software demo didn't come through yet, so I have no idea how it works.
I don't think it'll come before Monday - if all. I don't even know if it's legal to have something like that.

Voice Isolation works quite well actually, I was able to distinguish clearly a lot of the conversation - not all of it though.
Also, Voice Isolation did a lot of the work that I had done with the other user, but I would not advise other people to do the
same unless what they want to obtain is just that: isolating voices.

If you have a different kind of audio (say, nature or mixed) and it has lots of noises, the only way is taking a
sample and use Noise Reduction.

I do not think a lot more can be done. It is a real bummer that the TV and the voices overlap on the same track.
However, I had been trying to do this for a while now, so I am relatively happy with the results

As for Audacity: man, there's a lot of work that has gone into it and it shows (to me, at least).
I think in some areas there are way too many clicks, but it's generally a great program.
The most clear example where there are too many clicks is noise reduction: you first have to get a
noise profile, then you have to select the whole track (or the bit you're interested into) and then
click again to actually do it. It's a lot of clicks. I am not sure how you can reduce them without
creating a lot more problems for yourself in normal usage, but I would suggest looking into it.

A

[Gale Andrews]

The software demo didn't come through yet, so I have no idea how it works.
I don't think it'll come before Monday - if all. I don't even know if it's legal to have something like that.

If it's "QuickEnhance" I am sure it's legal, but the price tag would be the deterrent.

You can get the trial version on demand from http://www.oceansystems.com/forms/dl_re ... /index.php. The trial version silences the audio every few seconds. I don't find it works a lot better than "manual" methods using separate tools.

Another tool you could try (free, open source) is ISSE : An Interactive Source Separation Editor which doesn't rely on what you want to remove or isolate being centre-panned.

No tool is going to perfectly unmix and unmask the different sound components. You will likely not achieve any worthwhile improvement if what you want to reveal is inaudible.

As for Audacity: man, there's a lot of work that has gone into it and it shows (to me, at least).

I think in some areas there are way too many clicks, but it's generally a great program.

The most clear example where there are too many clicks is noise reduction: you first have to get a noise profile, then you have to select the whole track (or the bit you're interested into) and then
click again to actually do it. It's a lot of clicks. I am not sure how you can reduce them without
creating a lot more problems for yourself in normal usage, but I would suggest looking into it.

I don't know if Noise Reduction is suitable for translating to a real-time preview effect but I think Noise Reduction would be easier to use if capturing the noise profile did not close the effect.

However after capturing the noise profile you can use Effect > Repeat Last Effect to apply Noise Reduction at its current settings without reopening the effect.


Gale