KVR Audio

stian · Post by **stian** » Wed Nov 12, 2014 8:24 pm

Tapehead wrote:Do you plan to implement rendering/quality options like in Voxengo and some other plugins?...In which you can select a working mode/lower quality and a rendering mode/higher quality.

Yes, that would be an option as well. I'll consider different alternatives and see what can be done.

Best,
Stian

stian · Post by **stian** » Wed Nov 12, 2014 8:29 pm

lesha wrote:Running 20 instances with 68% CPU hit measured inside Ableton Live (32bit on W8.1), using AMD FX 6300 with GUI closed.

I am having a lot of visual problems running 10 instances with GUI open.

Yes, the GUI probably consumes quite a lot, especially with the analyzers running. Using buffer sizes that are a power of two, such as 256, 512 or 1024 is also recommended.

Best,
Stian

Lesha · Post by **Lesha** » Wed Nov 12, 2014 8:54 pm

I am using a buffer of 512 samples.

ibanguitar · Post by **ibanguitar** » Wed Nov 12, 2014 9:16 pm

Very nice eq indeed. +1 on the re-sizable GUI and being able to select multiple bands. Also a side-chain input to view and compare another input on the analyzer would be nice. I believe pro-q and a few other eq's have this option.

stian · Post by **stian** » Wed Nov 12, 2014 9:18 pm

asksol wrote:Stian, will you add the ability to select and operate multiple bands at the same time? equick can do this and makes it very easy to sweep and modify presets.

I've received a couple of requests for this feature, so that's high up on the list...

Best,
Stian

heffus · Post by **heffus** » Wed Nov 12, 2014 9:21 pm

I'd be all over this if I didn't already have more than enough EQ's.

Mario-C. · Post by **Mario-C.** » Thu Nov 13, 2014 5:02 am

I played with the demo this afternoon ... On acoustic drums mono overhead (which is a real test for me because there are lots of transients and full range freq content) I found the sound to be really weird and un natural when cutting mids, it's a pity because the overall concept is really interesting but it sounded too processed for me, it's hard to describe, I mean to give some feedback so I hope this won't offend anyone, I'm just giving my opinons.

stian · Post by **stian** » Thu Nov 13, 2014 8:11 am

Mario-C. wrote:I played with the demo this afternoon ... On acoustic drums mono overhead (which is a real test for me because there are lots of transients and full range freq content) I found the sound to be really weird and un natural when cutting mids, it's a pity because the overall concept is really interesting but it sounded too processed for me, it's hard to describe, I mean to give some feedback so I hope this won't offend anyone, I'm just giving my opinons.

Don't worry, I'm not offended.

Which phase mode are you using? The following applies to clean EQs ("character EQs" that mimic harmonic distortion and noise in analog equipment is another chapter): Given the filter curve you see (amplitude response) and the phase distribution, there's actually only one mathematically correct solution. If you run Equalize in linear phase mode, the results will be very close to identical to other properly programmed linear phase EQs with the same filter curve. There will only be subtle differences due to kernel lengths (the number of coefficients in the FIR filter) and windowing techniques.

Equalize doesn't use the common biquad IIR filters in the minimum phase mode, but uses FIR filters here as well. That has a couple of advantages. The filter curve will be identical to analog gear since we don't have any problems with frequency warping when reaching the Nyquist frequency at half the sampling rate (some plug-ins use biquad IIR filters with oversampling to overcome this problem). Anyway, since the amplitude responses actually vary between minimum phase EQs, there will be differences. However, given the displayed filter curve and the minimum phase distribution, there's only one mathematically correct solution here as well.

The mixed phase mode is new and intends to overcome problems with both minimum and linear phase filtering. It provides a unique way of setting the amount of pre-ringing vs. post-ringing. This is something new, but if you prefer either of the other phase modes, that is perfectly fine.

Best,
Stian

stian · Post by **stian** » Thu Nov 13, 2014 8:34 am

ibanguitar wrote:Very nice eq indeed. +1 on the re-sizable GUI and being able to select multiple bands. Also a side-chain input to view and compare another input on the analyzer would be nice. I believe pro-q and a few other eq's have this option.

Thanks for the suggestions and feedback. I just checked and ProQ2 does in fact have the side chain analysis. I've duly noted all these suggestions and will see what's achievable.

Best,
Stian

stian · Post by **stian** » Fri Nov 14, 2014 12:05 pm

Hi!

I've done some optimizations and the results are very promising. I've uploaded a new version to our servers, but I need to do some more testing before I make them "official".

Here are the links:

http://acondigital.com/software/Equaliz ... _1_0_2.exe (Windows 32 bit)
http://acondigital.com/software/Equaliz ... _1_0_2.exe (Windows 64 bit)
http://acondigital.com/software/Equaliz ... _2.pkg.zip (OS X universal binaries)

If you experience any problems, please let me know and I'd also be grateful for your findings regarding CPU usage.

Best,
Stian

KingofBeers · Post by **KingofBeers** » Fri Nov 14, 2014 12:19 pm

HOLY SHIT! Much better - are you sure it's the same plugin lol!! - much much better. 32 instances minimum phase mode 7% on my i5 laptop, haven't even been on the studio rig yet, but It's pretty obvious it's a winner!! will do some more tests later, got a tooth filling at the dentist this afternoon lol!.

All the best.

KingofBeers.

sinkmusic · Post by **sinkmusic** » Fri Nov 14, 2014 12:30 pm

Confirmed : the cpu use went down from 3,5% per instance to less than 1%.
Well done !

Lesha · Post by **Lesha** » Fri Nov 14, 2014 12:42 pm

Much, much better. 20 instances with GUI closed consume about 24% CPU, previously it was 68%.

I still have problems with GUIs open, I can't read the CPU consumption because the indicator is frozen, but I hear crackles, so it's about 80% in Ableton from my experience.

VSTHost reports 0.5 - 1.2% per instance.

xtrax · Post by **xtrax** » Fri Nov 14, 2014 1:12 pm

stian wrote: Equalize doesn't use the common biquad IIR filters in the minimum phase mode, but uses FIR filters here as well. That has a couple of advantages. The filter curve will be identical to analog gear since we don't have any problems with frequency warping when reaching the Nyquist frequency at half the sampling rate (some plug-ins use biquad IIR filters with oversampling to overcome this problem). Anyway, since the amplitude responses actually vary between minimum phase EQs, there will be differences. However, given the displayed filter curve and the minimum phase distribution, there's only one mathematically correct solution here as well.

Best,
Stian

This EQ sounds so good, and different from others. Now I see here why.
Great job Acon.

Thanks

stian · Post by **stian** » Fri Nov 14, 2014 4:56 pm

Thanks to all of you who reported the CPU usage! I'm glad it's much better now and there apparently haven't been any issues so far. The usage depends on the settings now, so it will increase when using sharp filters in the low frequency region. However, even the worst case scenario is around 30% faster on my computer. The linear phase still has constant CPU usage, BTW.

Best,
Stian

Acon Digital Equalize