KVR Audio

Aleksey Vaneev · Post by **Aleksey Vaneev** » Fri Jul 14, 2006 3:47 pm

martian, please send me to info@voxengo.com single spike processed by RenEQ (in 32 bit float WAV). I just want to draw this group delay plot on my own. I can't really believe a steeper filter has less group delay given it is a minimum-phase filter.

martian · Post by **martian** » Fri Jul 14, 2006 6:44 pm

ok aleksey.

i just did a clearer image for you too. gliss 24db slope with waves ren matched to -25db approx (around -30db and off the plot waves has the 'hump'). faint line is gliss, bold is waves ren.

heres the renaissance manual too which talks about the filter types: http://www.kswaves.com/download/pdf/REQ.pdf

and will send spike.

regards,

Christian Budde · Post by **Christian Budde** » Fri Jul 14, 2006 8:29 pm

martian wrote:

One thing about the images: You can also use the 'export' function to save the graph as PNG or SVG. Both formats should be read by normal webbrowsers.

martian · Post by **martian** » Fri Jul 14, 2006 8:55 pm

Christian Budde wrote:One thing about the images: You can also use the 'export' function to save the graph as PNG or SVG. Both formats should be read by normal webbrowsers.

yes i found this, but for multiple images decided its simpler to put things up with as few files as pos, so i just combine scn prints in paint.

is there a way to stop the analyser loading with the last opened vst? other than editing the config file each time i mean. simply because waveshell crashes it, and i always forget to change to another vst before closing.

finally, thanks you for the analyser is greatly appreciated!

Aleksey Vaneev · Post by **Aleksey Vaneev** » Fri Jul 14, 2006 9:09 pm

martian, that's mysterious difference to say the least.. Please send that spike - I'll measure myself (I'm still worried about that discontinuity on the RenEQ graph).

Christian Budde · Post by **Christian Budde** » Fri Jul 14, 2006 9:24 pm

martian wrote:is there a way to stop the analyser loading with the last opened vst? other than editing the config file each time i mean. simply because waveshell crashes it, and i always forget to change to another vst before closing.

Sorry this feature is only available in the pro version. I mean not that there is a pro version available. It's work in progress and mainly for internal purpose only.

Anyway, it would be nice if you could scroll down the upper graph a little bit further down, so that we can see the hump in the magnitude. Also you can Zoom to the interesting area.

Kind regards,

Christian

martian · Post by **martian** » Fri Jul 14, 2006 9:33 pm

up to 18db/oct there is no freq response hump, after which the hump appears and increases in size up to max slope (as shown in first image i posted, page4). above is the matched 24db to gliss again.

i wonder if there really is some fault when reading waves ren, and im hearing things? because in the last analyser i had (as mentioned in the electri-q thread) there were some wierd things.

anyway, hopefully aleksey can shed some light with the file i sent.

night.

[edit]

Aleksey Vaneev · Post by **Aleksey Vaneev** » Fri Jul 14, 2006 10:44 pm

martian, thanks for sending me that spike. It seems that RenEQ uses notch filter (instead of a classic high-pass filter) - that's why its group delay is a bit lower overall. And that's why you have that 'glitch' on those plots above (analyzer does not handle steep phase changes it seems).

Probably it's not bad overall considering group delay is lower, but as I've replied earlier, this may not sound as preferrable in all cases. It seems to me that linear-phase in that respect (the way you prefer it) is going to sound even better.

Christian Budde · Post by **Christian Budde** » Fri Jul 14, 2006 10:51 pm

Ok, I forgot to mention, that it would be nice if you could increase the resolution, but anyway: The gliss lowpass seams to be butterworth, while the other is something like chebyshev type 2 (ripple in the stopband)

martian · Post by **martian** » Sat Jul 15, 2006 2:33 pm

christian, picies above updated.

Christian Budde · Post by **Christian Budde** » Sat Jul 15, 2006 3:05 pm

martian wrote:christian, picies above updated.

Thanks

kylen · Post by **kylen** » Sat Jul 15, 2006 3:19 pm

Concerning Electri-Q "Transient" filter vs GlissEQ "Dynamic" filters it seems they operate differently and would therefore be expected to have a different sound as well:

GlissEQ defines a dynamic signal as "a signal with varying loudness, independent of the absolute loudness level."

Electri-Q states that "The amount of gain/attenuation depends on the current volume of the transients of the input signal. To achieve this behavior a simple auto-compressor is inserted in the detection chain."

I dunno - I guess if each EQ performed some type of rms or envelope detection then the methods might actually be pretty similar. Electi-Q might be alluding to peak detector but I don't know personally yet, it sounds like it might be though. I'm listening to each one. Any Comments?

defjamm · Post by **defjamm** » Sun Jul 16, 2006 7:09 pm

4damind wrote:For my ears GEQ and Gliss listens identical. This graph shows a (near identical) HighShelf with both EQs. So its not wrong to say that both playing the same game Can you show us a graph of Electri-Q ?

this is an asymmetrical bell-boost, no shelf. gliss can do it better when you enable high quality(oversampling). electri-q can give many different results, here are some in digital mode:

the first is an rbj-filter without oversampling(the asymmetrical one like in gliss in normal mode or kjaerhus), the one above is an oversampled rbj-filter(like gliss in high quality) and the third orfanidis(like in sonalksis).

Azucryneo · Post by **Azucryneo** » Sun Aug 20, 2006 3:45 am

i would like to start a interesting analytical posts.

About:

- Electri-Q
- GlissEQ

and the new ones:

- AirEQ
- SSL 4000

GlissEQ vs. Electri-Q