Err, besides that it has way more than just three bands, I think the pic supports what @pratyahara is supposing. I applied a low and then a high cut and Gullfoss immediately started to compensate for this.

From manual: TAME treats the dominant components while RECOVER treats the dominated components.

 

When I use the provided mug they sent. Gullfoss sounds 6.3% better.

 

Yes, at it extreme setting it does. But try to erase the bell below 200Hz with a bell filter. You really have to go extreme to do this. Nowhere near those 3db that gullfoss tames.

 

I wish you were right. With Bell -6dB @2kHz:

 

Maybe it's because I do not recover. But I definetely hear less bass now.
(Master channel)

 

I used the pink noise file downloaded from https://www.audiocheck.net.
What pink noise file did you use? I would like you to post it.

 

This thread is not about tuning something to please one's ears.
It is anybody's private right and choice, of course, but it doesn't contribute to this subject in any respect, I believe.

 

a new version of gullfoss has been out for a while but sadly has never been released, a lot of refinements as well as oversampling

 

Reaper's own Pink Noise Generator. Have you measured your file with an analyzer?

EDIT: I just dl'ed that file. Same reaction of Gullfoss'. Like with the generator, I mean.

 

Last edited: Oct 26, 2020

I am sorry, I switched the screenshots. The one I posted was in fact with the white noise input.
Now the pink noise input screenshot looks pretty much like yours:

So, it is still showing three band EQ-ing. Very small deviations confirm that they use pink noise curve as reference (this screenshot shows it just the other way around than previous white noise delta curve in which it is drastically obvious.)
The deviations from the pink noise curve in low/@~60Hz/ and high /@~10 kHz/ bands are just up to -1.5 dB, and in the mid section up to +1 dB. Since the pink noise is monotonous, I don't know why the curve is changing, plus that changing is non-completely-periodic? The 300 Hz gain adaptation is prone to mistakes, or a random generator is used?
In theory, one can not hear gain differences smaller than 1.5 dB, so this hard working minute pink noise EQ changing does not make much sense. Maybe the point is to obtain the overall gain difference between bass/high and mid bands of 2.5 dB. So, in the end, the overall effect is raising the mid band gain by 2.5 dB (!?).

 

Sweeping a test tone



https://imgur.com/a/O5gJNDY
different curves with 1k test tone

 

Last edited: Oct 27, 2020

plugin doctor is not registering that its doing anything at all

 

you''l need to throw a little extra dB at it, you can do that in settings

 

What makes you think it's a three band EQ? The whole line is moving, so it's rather more than 31 bands than less.

 

Raise Recover and Tame and you'll see funny things happening. Then switch to Harmonic Analysis and you can see what it's really doing.

 

I did 5 iterations of the starting pink noise file, and Gullfoss showed zero IQ, because in every pass it treated the file as if it didn't render anything in previous passes (although it did), using always the same /average/ amount of 'recover' and 'tame'.
So, this is how it looks after 5 iterations:

 

What makes you think it's a three band EQ? The whole line is moving, so it's rather more than 31 bands than less./QUOTE]

Four, maybe:


And also, here's Sonible SmartEQ2 graph (+/-9 dB max.)

 

Last edited by a moderator: Oct 27, 2020

Nah, I added 16 bands of a 31 band graphic EQ (interestingly, below 200Hz nothing much is happening).
Really mate, this thing is far from being a three or four band EQ.

 

Nothing affecting the sound quality happened since v.1.4.1:

https://www.soundtheory.com/static/ChangeLog.pdf

Soundtheory Gullfoss Change Log
Known issues (2020-07-14)
• None
v1.7.0 (2020-07-14)
• Added “session bypass” affecting all instances accessible by shift-clicking
the bypass button.
• [WIN] Improved processing performance.
v1.6.3 (2020-06-05)
• Fixed an issue that could lead to the mouse pointer vanishing.
• [MAC] Fixed an issue with audio unit instantiation on older macOS
versions.
v1.6.2 (2020-05-09)
• Fixed a problem with mouse dragging with UI scaling enabled.
v1.6.1 (2020-04-28)
• Fixed an issue that could prevent proper parameter saving and loading
with VST 2.
• The license dialog now appears on every launch if the plugin is not activated.
v1.6.0 (2020-04-06)
• Added audio quality control for both realtime playback and offline rendering.
• Fixed a bug that could reduce the audio quality in rare cases.
v1.5.1 (2020-03-24)
• Fixed an issue that would cause illegal instruction errors on older CPUs.
v1.5.0 (2020-03-13)
• Added support for 384kHz and Pyramix DXD.
• Added user setting for interface scaling.
• Added user setting for display refresh rate.
• Improved compatibility and higher performance on non-Intel CPUs.
• [MAC] Improved compliance with Apple’s notarization requirements.
• [MAC] Improved graphics rendering performance.
• [WIN] Fixed a bug that could lead to a frozen EQ Graph display.
1
v1.4.3 (2020-02-13)
• Fixed a processing bug that disabled any sound processing.
v1.4.2 (2020-02-12)
• Updated external dependencies for better stability.
• Improved readability of the license agreement during installation