The default setting is tasty. High frequency is great at TIMED mode. But when the release is low, bass heavy stuff seem to distort easily. Is it caused by the fact that the compressor sounds fatter??
yeah i forgot to mention that.
ok with the attack, the default position the program loads gives the attack timing like the hardware i used (it doesn't allow attack adjustment). it's labeled in nebula as 5ms although it could be a bit faster than that in reality. that means that the settings above and below that are timings that i added in/allowed for, which don't really represent the hardware's ability anymore. but we are just talking about timing, it's not like going away from that default setting will totally destroy all remnants of authenticity to the hardware. further, in my online research for the unit, i've found a few places where people were talking about the attack time on their units, which they actually measured and said was around 20ms or more. this could mean- a) the unit i used had a faster attack than an average piece, for whatever reason. b) the unit(s) those guys had were slower than average. c) a bit of both
it's hard to say because i couldn't really seem to find much direct mention of the attack time or where it 'should' be at, online anywhere. the fact still remains that the default setting on these programs reflects the hardware i used. i've allowed up to a little over 20ms to allow for the possibility that mine was for some reason faster than an average unit. i'm thinking about increasing that to a little over 30ms.
the fastest attack setting, i set to be 0ms, but of course nebula can't have instant attack unless you use look ahead to compensate for the prog rate time at least. also, since this is a feedback compressor, that also will prevent it from being instant, regardless of how much look ahead you add. as i mentioned in the txt for the demo, the look ahead might not even be doing anything here (due to the feedback detection) anyway. i'm about to test to be sure, and if not i'm removing it. anyway, the actual hardware is not instant so this is NOT a flaw.
now to the release time. the hardware can be adjusted, between 100ms to 1500ms. that's represented here (demo version is limited up to i think 400ms though). i've also allowed for faster release times, going down to being close to instant. i didn't allow it to go all the way down to 0ms because obviously there was bad distortion at that point. this is pretty much true of about any software comp out there. if it allows you to set it at 0ms you will likely get nasty sounds out, especially with bass. so i didn't allow all the way down to 0ms, but somewhere close. i still noticed distortion with some inputs at the fastest speed, but in some cases i liked it.
the programs all load at 100ms (including the demo), which is the fastest speed you can get with the hardware. if you go below that you are going into territory that doesn't really represent the hardware, but again that's not necessarily a bad thing. as with any compressor, equalizer, etc, whether plug-in or even hardware, you can't always expect better or even good results out of it, with any type of input and any possible settings on the thing. it's a commonly known thing that really fast release times (under 50ms) on probably any comp with bass will lead to distortion.
long story short- if you get bad sounding distortion from using a fast release...
dont do that!
and again, since those times below 100ms release are more extreme 'what if' speeds that i allowed, which were not possible with the unit, it's not a flaw of the program's ability to recreate the hardware if they distort. it's just that i allowed extreme settings which can and will produce distortion on some inputs, as happens with compression in general. but sometimes they might sound good and be useful, as i found in my testing, which is why i kept them in.
one other thing- if you increase the input slowly with a signal that has some low bass, you will find that it starts to distort the bass more and more at a point (with any release speed), bringing it up a fair amount in level. the demo is only 3k so it's even more pronounced with the 10k version, and the actual hardware did this also. you can see people discussing the low freqs distorting with these things if you look. it could be useful in some cases...
With regards to attack and release. I have not yet used NAT so don't quite understand how it is working.
So if the compressor starts with its default 100ms release it is 100% authentic to the hardware, but if we move it down to 50ms it is not. I get that part, but is there something different in the 'quality/shape' of the release at non standard timings? Is the 100ms using a measurement but the 50ms just some kind of coded approximation?
Not sure if that makes sense, put another way, is there a quality difference in setting it from 100ms to 95ms? Is the 95ms interpolated?
The super high quality mode sounds incredible! Without giving any secrets away, is there a new breakthrough that enables us to have 'dirty/character' compression without the dreaded Nebula artifacts? I was under the impression that your new techniques were only for ultra clean compression?