Apply EQ Only Above a Specified Volume Level

I wonder if this has come up before? Is it possible to apply an affect, specifically EQ, only to a passage above a certain volume level?

I have a Classical Music recording, actually off youtube, which dates from 1948. The frequency spectrum is a little bright but in the loudest passages it seems to get really thin and strident. I tried applying an EQ to the entire file but the loudest passages still sound a little bright/thin. There’s something happening where once the wave form gets above a certain db level the music suddenly gets even brighter. Therefore, if I EQ everything even more and make those passages listenable, everything below that volume level will sound a little too dull.

I wish there was some way to put a “gate” around the waveform and apply an treble reduction just to everything louder.

Can anybody suggest a way for Audacity to do this? I’m using vers. 2.1.2.

Otherwise, I’ll have to go through the entire concerto, 37 minutes long, and select each loud passage manually and apply a stronger EQ there.

pete in oakland, calif.

…once the wave form gets above a certain db level

That would be nice, but it doesn’t work that way. Tools that affect character of the performance work by average and guesswork, not where the blue waves go. That’s why those tools have settings for attack and release times, level settings, effect correction, and sometimes more.

which dates from 1948.

That sounds an awful lot like Dolby, but Dolby didn’t start cranking until 1965.

It could totally be a bad tape transfer.

There is a Noise Gate. You could duplicate the performance and run Noise Gate on the duplication such that only the loudest portions make it through. I’m note sure if Noise Gate has control range that big, but it’s worth a shot.

Then equalize the duplication and mix the two together. Audacity will mix all your timelines into one show on export unless you stop it.

Maybe others will have other ideas.


You may discover by the time you get the automatic technique to sound right, you could have gone in and done it manually.


You can compress certain frequency-bands, so the peak-loudness in those bands is less as a consequence.
The tool for that is a multi-band compressor plugin, e.g. GMulti.

When the wave gets really loud, close to clipping, harmonics are generated, increasing the high-frequency content.

Hello Koz,
Thank you so much for your suggestions. Well, here’s my source: the Brahms Violin Concerto recorded in 1948 at the Concertgebouw in Amsterdam by Decca records (London Records, here in the US) I don’t know where this youtube poster got it. I had never heard of the violinist but apparently he died very young, in a car accident. He showed great promise. Here, with this conductor, a quite special performance emerged. Doesn’t always happen with young phenoms.

Anyway, other versions on youtube are from 78’s or who knows where. Overall this one seemed like the best I could find. But I have no idea how the climaxes got so thin and aggressive, at least on my home audio system. Perhaps some noise reduction or sound enhancement was used and it had this effect.

Thanks for your suggestion but the learning curve on mixing timelines, etc. sounds daunting!

Fortunately, I did find this plugin, It’s a sibilant surpressor. This free software plugin for Audacity seems like it pretty much leaves the quieter music alone and then tames the high frequencies on the climaxes. In this recording I had to turn the two dials up pretty high to get something I could live with.


I think you’re right there, ( the YouTube transcoding doesn’t help either).

SpitFish can only go down to 4000Hz, that’s not low enough for violin. GMulti plugin can go down to 1000Hz …
GMulti on v=s0GR49R2pcU.gif

SpitFish can only go down to 4000Hz, that’s not low enough for violin. GMulti plugin can go down to 1000Hz …

Hello Trebor,
Thank you for this suggestion. Is this another free plugin for Audacity? I get your point about 4K hz being higher than the pitch of a violin but I think what’s bothering my ear is the excessive harmonics above 4K hz which are giving such an edge to the sound of the orchestra in the very loudest sections.

If I used GMulti to compress high frequencies, would it have a noticeable effect, a negative one, in the passages that were less than loud? What seems useful here with this SpitFish plugin is that it leaves every alone that is below a certain dynamic level. Or at least I HOPE that is what it’s doing!

Anyway, what are some of the uses of GMulti? What other kinds of problems is it good for?


Yes GMulti is free*, you install it into Audacity the same as SpitFish.
[ * strictly speaking it’s donation-ware ]

You set the threshold at which the compressor kicks-in: it won’t touch anything below that threshold.

You set the threshold at which the compressor kicks-in: it won’t touch anything below that threshold.

Thanks again, Trebor. I have d/l GMulti [Windows 7 32-bit]. I’m still not sure where the .dll file goes, under Program Files (x86)/Audacity/Plugins OR under Program Files/VstPlugins. I put it in both and it works.

Could I ask you about a couple of parameters? How would I set the controls if I want to use GMulti for my specific need here to compress JUST the high frequencies, i.e. Band 3? In other words, how do I disable any action on Bands 1 and 2?

And then in Band 3, with Freq2 set at 2khz, I assume I have the Low Cut set to Off. But what is the Level control for? Would that raise or lower the volume level of the entire Band 3?

And what is Width used for? The manual says to “accenuate” the stereo image which I don’t follow. What setting is neutral with Width? 100% or 0%?

thanks so much for this powerful tool,

pete in oakland, ca.

As shown in the animation above: Thresh=0, Width=100%, Level=0.0dB for Bands 1 & 2 neutralizes them.

Yes “level” raises or lowers the volume of that band after any compression has been applied.

100% is normal (unmodified) stereo. 0% makes it mono.

Graham Yeadon deserves the thank$.

OK, Trebor, once I’ve found settings that I like for a particular music track/file, is there a way to apply GMulti, or for that matter, ANY Audacity Effect to a series of music files? This Brahms Violin Concerto is three movements, split into three music files. Otherwise I have to manually apply GMulti to each music file separately.

In another wave editor I use, Goldwave, it’s called “Batch Processing.”


You should save those settings in GMulti as a “user preset”.

Audacity’s version of batch-processing is called Chains