Is there a plugin that suppresses the frequency below a threshold?

[whyterabbyt]

No, it's not what you said. You said one is volume control, the other is frequency control.

You're being disingenuous, Ive already made it clear to you that I was indicating the OP was looking for volume control of individual frequency ranges, differentiable from volume control of the entire signal, which are not the same thing.

Dynamics processors apply volume control over the signal; all frequencies. That's not what is wanted here.

The term dynamic has nothing to do with whether something is being applied onto everything or only specific parts.

That was 'dynamics' not 'dynamic'. Do you need the generic term 'dynamics' explained to you?

Dynamic EQ's work on very specific frequencies, not on the entire signal. Dynamic only means that something is responding dynamically to input.

I used a noun, not an adjective.

I'm not ignoring it, I said two posts ago that I don't believe that it is what the OP asked for.

And yet it is.

I don't see different bands in the sketch

Because its not bands, that was a simplification because you kept trying to pretend a single-band dynamics processor didnt apply amplitude changes across the entire spectrum of the signal. You were claiming that the threshold excluded frequencies from attenuation.

You can see the frequencies. On the frequency axis.
Do you not know how a graph works?

nor any mention of different settings for different frequency ranges.

That's the threshold. There's a different threshold for each frequency.

Instead, we see only one treshold line for everything.

Yeah, its one line. A line that indicates a threshold level for each frequency, which varies per frequency (because its a slope). Do you not understand lines either?

And the fact that you're throwing around definitions that don't mean what you think they mean, such as the frequency/volume equation or

Whit? You're just making stuff up now. Its not 'the frequency/volume equation' there's no equation in it at all, its a graph of frequency versus volume aka amplitude.

If you'd have used these correct terms from the beginning, we could've saved a lot of time.

Oh, the irony.

While I still don't know whether the OP needs multibands, I said earlier, that today's gate/expander plugins come with tons of features and could do the job as well, you objected that 99,99999% plugins couldn't do it.

And yet they cant do the job.
And what I said was that 99,99999% of gate plugins wouldnt do what the OP needed. Because they cant.

And then you claimed that the threshold excluded 'certain frequencies'

"Your assumption that Gate/Expander plugins would work on all frequencies is wrong. The threshold is there for a reason, it excludes certain frequencies which remain untouched."

Clue : the OP didnt ask for 3 bands. That isnt the use case. That's you finally realising that no, the threshold doesnt 'exclude certain frequencies' and looking for some way of still being right despite not having a clue of what these things actually mean.

Fabfilter Pro-G (the first plugin I threw in earlier) has a filter band that you can adjust, so that it works only on specific areas of the frequency spectrum. You could use it only from 0-500Hz. Then load another instance, use it only from 1k-1,5k Hz and so on.

And if they want to suppress 50 different frequency ranges, you'd suggest 50 instances, would you?
And if they want to suppress 200 different frequency ranges, you'd suggest 200 instances, would you?
What about 20,000 different frequencies. One at a time, with a different threshold for each.

Know what could do that in one instance? A spectral gate.

Which you'll find in about 0.0000001% of all the hits for gate.

"Multiband Gate" also has this feature and comes with 3 bands built in per instance: https://aixdsp.com/multiband-gate/

wow, three. And here was me thinking that if I made a massively simplified example as an illustration of something you were completely missing, then you'd be smart enough not to take it absolutely literally and run with it.

Lets try again. Imagine I had listed 250 separate bands. now tell me the plugin that does that, that isnt a spectral gate.

So yes, you can definitely find gate/expander plugins that allow control over frequency ranges, and knowing what these plugins are called certainly helps with searching for them.

still no.

[Scotty]

Mutiband dynamic Eq will do it.

[Foxedge]

You're being disingenuous, Ive already made it clear to you that I was indicating the OP was looking for volume control of individual frequency ranges, differentiable from volume control of the entire signal, which are not the same thing.

I can hardly be disingenuous when I quote the exact post you've written. Your statement didn't say "The OP wants to have volume control over individual frequencies", it said "the OP does want frequency control, not volume control".

I hope you're able to identify the difference between those two statements. The first one, that you didn't make, correctly communicates that you can either control the volume of all or some frequencies. The second statement - the one that you actually made - indicates that frequency control is something different than volume control, which is wrong. It's still volume control - just over specific frequency ranges.

That was 'dynamics' not 'dynamic'. Do you need the generic term 'dynamics' explained to you?

Doesn't change a thing, since dynamics processors such as multiband compressors also don't apply volume changes over the entire signal. You can easily spare frequency ranges. The statement is false, whether with the s attached or not.

Clue : the OP didnt ask for 3 bands. That isnt the use case.

The OP didn't ask for 250 or 20.000 different bands either. You're making up stuff now because you just want to remain right, and you're arguing against your own strawmans.

The OP said she wanted to get rid off the noise floor. The sketch shows 4-5 different peak frequencies that should be preserved. A job any plugin with 3 bands can easily do.

You also don't need to speak for others; the OP can decide on the plugins herself.
That's all I have to say, I'm considering this debate finished.

[whyterabbyt]

I can hardly be disingenuous when I quote the exact post you've written.

What, when you repeat and deliberately ignore any previous or subsequent posts which provided context, clarification explanation?
Yeah, absolutely you are.

Your statement didn't say "The OP wants to have volume control over individual frequencies", it said "the OP does want frequency control, not volume control".

Well obviously that statement didnt, because I had already said it. That statement repeated words from the differentiation I'd already made. I dont need to make the same differentiation over and over again for you do I?
I mean I know you claimed that this was my 'opening statement', but we both know it wasnt.

I hope you're able to identify the difference between those two statements.

Yes, one is the specific set of words I used that specific time, and the other isnt.

I guess you're maybe trying to pretend that I was talking about change of some other parameter of the frequencies? In a frequency versus amplitude process?
Want to propose what that would be?

The first one, that you didn't make

, correctly communicates that you can either control the volume of all or some frequencies.

Yes, I'd already made the differentiation that that was what I was talking about... like here... (which is emphasised for this time, im clarifying that so you dont go off on another one)

they need to gate all the frequencies of the signal that are below the threshold, rather than gate the whole signal
In other words, all frequencies would be gated equally by an ordinary gate, and they dont want that to happen, they want to preserve frequencies louder than a given level.

Oh dear, did I use 'ordinary' not 'simple?' or 'single-band' I guess I couldnt have meant either of those things, then, but some other thing.
Are you now going to argue that multiband gates are ordinary and simple and that they're under every hit for 'gate'?

The second statement - the one that you actually made - indicates that frequency control is something different than volume control, which is wrong. It's still volume control - just over specific frequency ranges.

[/quote]

You're trying to split a hair that isnt there.
You can pretend all you like that I wasnt making a differentiation between targetting the signal volume and something else, or that the thing I was differentiating from the signal volume wasnt individual frequencies (or bands of freuqncies in case you get pedantic yet again) even though the OP was looking for frequency-dependent gating, or that I was positing some other control than volume over the frequencies of something shown in a volume versus frequency graph, go ahead
But we both know exactly the differentiation I was making. And that I'd already made the differentiation.

And more to the point, we also both know that the kind of 'simple' gate that you said there were tons of hits for, that you were insistent would do the task, would not be capable of controlling the volumes of individual frequencies, only the signal volume, so you knew exactly why I was making that differentiation.

That was 'dynamics' not 'dynamic'. Do you need the generic term 'dynamics' explained to you?

Doesn't change a thing, since dynamics processors such as multiband compressors also don't apply volume changes over the entire signal. You can easily spare frequency ranges. The statement is false, whether with the s attached or not.

True. I forgot to specify single-band until the very next paragraph. And I suspect you know that that's what I meant. But I did specify, so really you're just being disingenuous again.

Either way I've still been saying all along that 'simple' processors, the ones that will turn up under all those 'tons' of hits if you search for 'gate' will not be sufficient, specifically because they'd almost certainly be single-band, gating the signal volume, not frequencies. Oops sorry 'individual frequencies'.
But you said they would be sufficient. And handwaved about use of the words 'frequency control' over and over again as a distraction.

And yet now you're saying that this is in the domain of multiband processors, which is what I'd been saying all along.
Interesting that.

Know what's multiband and lets you target individual frequencies like the OP asked for? A spectral gate. But not a 'simple gate' that you'll find 'tons' of.

Which Ive been saying from the beginning.

I mean you do know that a spectral processor is(*) multiband right down to the individual component frequencies, dont you? And that that means it can do as many bands as the OP could ever want, without having to find enough crossovers or multiple instances.

(* oops. 'can be' Dont want you going all 'that specific sentence didnt explicitly say single-band' on me again)

Clue : the OP didnt ask for 3 bands. That isnt the use case.

The OP didn't ask for 250 or 20.000 different bands either. You're making up stuff now because you just want to remain right, and you're arguing against your own strawmans.

No, they didnt specify. But it wouldnt matter how many it was to a spectral process, though it would for 'simple' single-band 'plugins' that there are tons of.

But 3 bands definitely wasnt their use case, just try counting.
Oh, and multiband 'plugins' typically use crossovers to separate the bands, so to gate 3 separate bands of interest, you'd probably need 6-band processing, not 3-band.

However they did provide a spectrum of frequency versus amplitude. So they could be after any number of bands, right up to one per frequency. Which a spectral process could scale to if that was what was required.

And you know as well as I do that the precedent for that, one band pre frequency in noise reduction, which they want to do, is very well established.

(Ive even got a diagram, though you'd probably compare it to a blurred photograph of Ariana Grande's butt-tattoo and claim they were 'the spitting image' of each other.)

The OP said she wanted to get rid off the noise floor. The sketch shows 4-5 different peak frequencies that should be preserved. A job any plugin with 3 bands can easily do.

Any plugin? I think not. No single multiband reverb plugin could do this easily.

You dont mind if I keep pointing out that you repeatedly say 'plugin' instead of being explicit about the type of plugin, do you? If I repeat what you actually say, and ignore the surrounding contextual information that makes it clear what you actually mean?

Bet you would, right?

You also don't need to speak for others; the OP can decide on the plugins herself.
That's all I have to say, I'm considering this debate finished.

Aw, that's a shame. Still its probably best that you focussed on the wording of a differentiation between signal and frequency content which allows you to get away with maintaining that you were right all along. That'll easily cover up the fact that you wind up explaining that the job could be done with multiband processing ie controlling frequency content not the signal as a whole, and ignoring the fact that you'd started off insisting that any simple gate would do and were still doing so.
Amazing that you finally got to 180 degree round to where I started, innit.

Then again, Im not surprised. Remind me again how the threshold 'excludes certain frequencies', and how a bipolar graph of amplitude over time is the 'very spitting image' of a unipolar graph of frequency versus amplitude or how 'tons' of 'simple' 'plugins' will do the job you say the OP meant (except mumble mumble something something multiband)

[swilow11]

What a surprise to see whyterabbit involved in a lengthy and pointless argument again.

[whyterabbyt]

what a surprise to see swilow11 being an arsehole again.

[swilow11]

what a surprise to see swilow11 being an arsehole again.

Coming from possibly the least pleasant person on KVR, this doesn't mean a whole lot.

[whyterabbyt]

Says the guy who turned up to contribute precisely nothing to helping the OP.

[nix808]

as far as I can tell this does involve frequencies,
as wr was saying

so it simply eq then?

if a combinance of amplitude envelope following and frequency,
a multi-band compressor or expander,
as suggested

or parallel eq and dynamics plug

wr- wot u sayin' 'bout swilow?

[whyterabbyt]

so it simply eq then?

I dont think so, as its not attenuation of a fixed set of frequencies; attenuation needs to be applied to frequencies with amplitude below a specific threhold (where that threshold may be per-frequency). so its gating, potentially right down to per-frequency.
the latter being a thing that spectral processing (ie FFTs) lets you do. so, if you do want per-frequency, then a spectral gate or similar, as was being suggested by myself and others at the beginning of the thread.
OP said this was for noise reduction, the application of FFT-based processing to that is pretty well entrenched.

wr- wot u sayin' 'bout swilow?

that they're not interested in helping the OP, basically just looking to have a go at someone. thinks joining a thread to say 'you suck' makes them morally superior.

[nix808]

actually noise reduction must be such a difficult art

-have a tiny bit of patience with swilow,
I think he/she is not after a disturbance
I think they don't want an argument really
to the point of dissonance from urself,
if you don't want that

--'ite you 2 u should both step back and breathe, I think you are both valid entities,
and i will hug u all up. So easy, ur both top at putting points


maybe like SoundForge nr is a spectral process

-there is SoundForge, Wavelab, Izotope RX, Adobe Audition for environments for this
Google 'noise reduction', and so many have really thrown themselves at it
coz we had tape for so long, and digital noise is still manifest

[whyterabbyt]

I think he/she is not after a disturbance

but they came to this thread specifically to create one. they werent in the thread, they didnt have any intention of contributing to the OP's question, and the thread had already been inactive for two days.

I think they don't want an argument really

I agree; they just want to do the moral posturing as a driveby.

maybe like SoundForge nr is a spectral process

-there is SoundForge, Wavelab, Izotope RX, Adobe Audition for environments for this
Google 'noise reduction', and so many have really thrown themselves at it
coz we had tape for so long, and digital noise is still manifest

Yup. Typically those NR things define their 'threshold' via a noise footprint...

https://github.com/timsainb/noisereduce ... reduce.png

[nix808]

my thought is
-how the {expletive} can they properly phase the print?
it is hit and miss, inverting the noise
...once more, fair play

and beside the phase, these noises are the most chaotic(random) sound
as far as repeating

edit- it of course spectral as wr said, but that is no simple answer imo
fair go, we have nr hmm -like a Thomas Dolby in our midst

[whyterabbyt]

my thought is
-how the {expletive} can they properly phase the print?
it is hit and miss, inverting the noise
...once more, fair play

clever dsp maths stuff

sorry, I did actually post the wrong link above beside the image, that image was from an implementation of a spectral gate noise reduction algorithm, which describes it as follows (my italics)

Noisereduce is a noise reduction algorithm in python that reduces noise in time-domain signals like speech, bioacoustics, and physiological signals. It relies on a method called "spectral gating" which is a form of Noise Gate. It works by computing a spectrogram of a signal (and optionally a noise signal) and estimating a noise threshold (or gate) for each frequency band of that signal/noise. That threshold is used to compute a mask, which gates noise below the frequency-varying threshold.

Here's the proper link

https://github.com/timsainb/noisereduce/

[nix808]

perhaps the estimated noise is subtracted, but how that doesn't create more noise is beyond me,
my noise print was noise heh

ta will read quickly, I have reduced noise before. Even dramatically.
Sound Forge 4.5 owner

My fft is very slowly any sort of device other than pretty