People always forget about TDR SlickEQ M nowadays

Tranqualizr 1 & G2 are also really good!

 

TDR SlickEQ M is fine, but I also like apqualizr 2 so much, punchy on lows, deharshy on mid highs, a good brightener on highs. Tranquilizrs, flux epure are both for smoothening, lighter...

 

The dev said the GUI is temporary because it would take them months to get the JUCE framework out or something. They wanted to do a quick release to get the plugin out, and will be launching massive updates on it in the coming months

 

wow! just wow!

 

I like how you present your foundings, but you never show any real world uses.
Just hope people would upload sound examples for use cases, like this gtr sound doesn't sound right after this plugin

 

This is what I was talking about, yeah, Dude? Earlier versions of Magpha, pre last time ones, do not have this. I don't remember which version it was, but search thru the GS thread.... Mastering thread maybe, 10 pages :D

 

If you're asking me.. it's a direct respond to this :

meaning, they (Kush Audio) didn't fix it before re-releasing it. kinda pointless for real-world uses IMO


“Give me six hours to chop down a tree and I will spend the first four sharpening the axe.”
― Abraham Lincoln

 

Problem since 2.1.0
Its brother eq (called SpreadFocus EQ) has this too.
Here is a link to the long post and measurements (post #221):
https://gearspace.com/board/mastering-forum/1210438-balance-analog-magpha-eq-8.html

 

Last edited: Nov 1, 2023

I mean, I don't really care about GUIs that much as long as they're intuitive, but I think Kush Audio is underestimating the need for them for marketing/advertising. People are gonna look at this and think "man this looks like shit" and not even bother trying it.

 

I find Kush stuff way too extreme for my purposes generally but I was fascinated by the description on this. Glad I read this thread...

 

Sorry to say but... Kush's DSP Code is pretty FUCKED UP! Analog EQs never ever create such a curve ...:

 

Something fishy going on here (and being a cat and all, I love me some fish!).. so the GUI is temporary until he has time to make a new one? Like totally professional dude! Not and lame AF FFS lol. And the dev states he is updating ALL Kush products using this new eq algo he recently licensed and is in this plugin, as in its better and fixes all past issues. But now you guys are sayin its still fucked and exactly the same as the old algo? What a FN mess

 

Last edited by a moderator: Nov 1, 2023

It is the balancemastering's code, as noticed, this was added in 2.1.0, and in his SpreadFocus EQ too, not fixed for several years. Now it is just re-released in new GUI, under Kush brand, new name, version 1.0 (based on MagphaEQ 2.1.1), and with no changes in dsp code at all. Probably will be fixed soon

 

Exactly the same as the old algo, read post 221 in GS link posted above, it is from 2019

 

But I thought it already sounds like a $9,000 eq, so it does not have to be fixed?

Or how shall this logic work?

 

LMAO

This is called FIR filters that have too small of a buffer length. There's a reason why virtually all digital EQs use IIR filters. Quite surprising to see since Kush has a history of excellent plugins. I guess that Gregory is trying to get back in the biz from his surgery with a lazy release unfortunately, copy pasting an already flawed work.

Edit : Greg talked about this issue on GS. This comes from Plugin Doctor that is unable to properly measure the magnitude response, because of the multiple filters working in tandem, with a first one doing the magnitude and another one correcting the phase. Plugin Doctor uses a smaller FFT size to what is required to properly measure the EQ.

Problem : This means that this EQ doesn't give the same response when fed steady-state signals vs transients. This is NOT what happens with analog EQ or normal digital EQ. Maybe this sounds good, but this is not what's being advertised at all.

 

Last edited: Nov 2, 2023

Heads up on that: Plugindoctor doesn't measure q.632 correctly, something about its measurement speed and/or window size and/or other things that go over my head. q.632's eq dsp happens in stages, with the freq being calculated first, then a phase model being applied to that freq, and something about the "looking ahead and back" aspect of the phase filter messes with plugindoctor's analysis.

It freaked me out at first, I thought "shit, we broke this eq!"

If you're curious about the actual curve behavior, Bertrom works well over here.

Gregory Scott - ubk

#20
https://gearspace.com/board/new-product-alert/1416518-q-632-kush-magpha-eq-reborn.html

 

Last edited: Nov 2, 2023

I tried the usual null method and with Pro-Q3 on both Natural and ZL mode and it nulled to -60db. Not being greedy here, but I expect atleast -150db since both are clean digital EQ. From your understanding of stuff, could it be the artifact as how those analyzers showing ? I know for sure that it's not comb-filtering, because the render would show it, but it didn't even at zoom. Not showing anything, just straight line suggesting it nulled completely, but why meter is reading -60db on the straight line rendered file is what bugging me. No idea what's happening.

Also I am today's years old when I learned that linked Gain+Q on Pro-Q3 only work if you drag the EQ point with mouse, if you enter the gain amount manually, it wont affect Q. Damn, feel stupid now.

 

There's also a button somewhere next to the Q knob that enables/disables the Q-link IIRC.

 

yeah. What I mean is, when Gain+Q enabled on start, an 8db boost on 1kHz done with clicking and then dragging the point around will give a Q value of 1.180, it's tightening but if you enter the the frequency & amount of gain manually with keyboard, the Q value stays at 1. I expect the Q to automatically get to 1.180 when Gain+Q link enabled and behave like a proportional EQ. I have them enabled by default, like it that way. Really hope Pro-Q4 will upgrade all of this silly workflow thingy.