Actually, a few months ago I was reading a lot about these tweaks.. But I did not have enough time to experiment.
I remember some one saying that the transient were somewhat lost with the freq. mode. Anyway I thought it wasn’t 100% true.
Then I had a quite big mixing session. I’ve used nebula for the EQ, tapes & pres (normal freq mode). No other plugin was used for the EQ. (Comps were from Waves+Brainworkx Vertigo, Reverb from Altiverb).
I’ve bounced I don’t know many times to EQ my tracks (Logic’s bounce in place+freezing,..)
The final mix was of an exceptional quality!!! (clean, defined, wide, just nice to hear, no digital zzszszszzsss if you know what I mean), thanks to Nebula!
But unfortunately, there was something missing. THE PUNCH! Especially from the drums.
Seeing this tutorial motived me to "clean kernel timed" all my pre+EQ. The sound is just better for me.
I took the two Example Flac files and imported them into a 88khz 24bit Cubase Project.
then I Phase flipped one signal.
The result you should be able to hear is the difference between the two files (one file phase flipped deletes all the other frequencies that are exactly the same, leaving only those frequencies that differ and dont have a counterpart that was flipped to cancel them out).
I wasnt able to hear anything. There is practical no difference in timed vs freqd in those two flac bongo loops or I did something wrong.
i don't quite understand what you're really saying here. you say you switch to timed mode. then you leave kern and come back and the freqd lengths have changed? but then you 're-match' the timed mode lengths? why would you be re-adjusting the timed mode lengths if they haven't changed? you want them at 50ms right?
i don't know if i'm just misunderstanding you or what, but i'll just go ahead and say that if you want to be using timed mode, and you have it set at the length you want, and you are set so that you're actually using timed mode, then whatever happens to the freqd length doesn't matter- because you aren't using it.
Yes ... I wanted to match the TIMED ms to what the library defaulted to in the FREQD setting.
You last sentence was exactly it!
THAT's what was confusing. The FREQD loaded in 50ms, I properly switch to TIMED [clean,even] setting those to the same 50ms ... all working fine .... only to come back and see that the FREQD values were now different.
I now clearly understand that 'those' values are NOT in use ... but it honestly threw me a curve that deemed to seek clarification.
I know you have dug deeper into the workings of NEBULA, and very much appreciate your response to my moment of confusion
I've had a chance to read information regarding TIMED and FREQD processing that concurs with the audio results. Thanks for clarifying the particulars off how the KERN page settings are wired in.
honestly, it is kind of weird though. you're saying that even though you're only touching the 'timed' length, for some reason the 'freqd' length changes after you do, and then switch pages and come back. ?
you definitely don't want to then go and change timed again to match whatever freqd jumped to. if freqd is first set at 50ms when you load the program, that's what you want to set and keep timed at. then you switch so that you're actually using timed mode, for the 2 kernel sets (clean and then either even or odd). your cpu use goes up, and that's that. as long as the timed value isn't changing i would say you're ok. but it is weird that the freqd value changes.
There is a funny little bug, sometimes when you change a kernel length, to for example 50ms, save it, come back and its turned into something like 49.7. You change it back, save again and eventually it sticks to the right number. Sometimes changing TIMED effects FREQD for no good reason, and vice versa.
This is annoying, but i doubt it effects the sound much having a few ms difference in kernel length, all a bit odd, hope it gets fixed!