Been trying some new stuff in NAT, and things are getting closer. The behavior is much improved, the controls are now accurate. Now my problem is that the Gain Reduction meter says it's compressing 6db or so even when no actual compressing is happening.
Even with that small issue, I'm really pleased with the outcome! Now, onto figuring this issue out, then some more experiments. I got a lot of toys I want to share with you guys!
tunca wrote:Thanks Botus.But Why Nebula's compressors are unsuccessful?
Any others are good.Tapes,PREs,reverbs...But compressors are dud
Only some people say that; people, I suspect, who have by and large not purchased or used many of the compressor libraries that are available. There certainly are some caveats (the inability to match the very fastest attacks gets brought up a lot and it is true) but there are also a great many compressors that don't live and die on their reputation for microsecond attacks.
By now there are actually many Nebula compressor libraries that are quite a bit better than just being "usable". I think at worst you could say that some libraries may not stand up to being as faithful to their hardware counterparts as the other categories you mentioned..
The question remains, though, as to whether those limitations are inherent to Nebula or simply the library they represent. I can't say that recent compressor offerings have dispelled the notion completely but the general reaction to the Rayphlex, the Kultcomp, and several others seems to be that it is indeed possible to get the kind of results out of compression that people desire including reasonable attack (it seems like there are still some limitations in terms of the very fastest attacking compressors). And that's a situation that's only going to improve as developers get more practice with sampling and tweaking compressor programs; but it seems from the forums (I'm not a sampler of anything so the nuances are a bit lost on me) that sampling compressors is quite a bit more complicated than sampling saturation-y circuits. So it's the kind of thing that practice will surely benefit as time rolls on and cool tricks are discovered. This whole method of sampling compressors hasn't been around nearly as long as many of the other templates iirc; many of the older compressors had to be hand timed in terms of the envelope followers.
I would also add that the sound of the circuit has a huge impact on the sound of the compression action itself; to me this is why so many other compressor emulations fail to get "the sound" despite spending lots of apparent time and money measuring, listening and coding; and that's something that no other plugin maker that I can think of has ever even come close to Nebula. In that respect pairing Nebula with any other algo compressor out there makes that compressor better by giving you that extra ability to not only color but also enhance that clamping.
In that respect pairing Nebula with any other algo compressor out there makes that compressor better by giving you that extra ability to not only color but also enhance that clamping.
this why i put one of the vintage tube collection programs before my compressor on busses all the time, the difference of the compressoraction even with the cleanest programs is so significantly better to my ears, sounds just more musical.
but i suspect, that the internal sidechain of the algocompressor is an important part. i think it is very different, how accurate they analyze the incoming signal, to generate the envelope for the compression. i get the greatest benefits from the nebulaprograms together with my psp mastercomp. with others, the effect is less significant.
now back to the original topic!
thanks for this freebie!
i got some good sounds out of it.
it is a little unpredictable, but with tweaking by ear, i got some nice glue on a mix. longer attacks seemed to work better in my case. short attacks made the compression jump around too much, even with low threshold and ratio.
system 1: windows 8 32 bit - samplitude prox/x3, tracktion6/7, reaper system 2: mac osx yosemite - reaper(32+64bit), tracktion6/7(32+64bit)
both systems on: macbook pro (late 2009), core 2 duo 3,06 ghz, 4 gb ram, graphic: nvidia geforce 9600M GT 512 MB
mathias wrote:it is a little unpredictable, but with tweaking by ear, i got some nice glue on a mix. longer attacks seemed to work better in my case. short attacks made the compression jump around too much, even with low threshold and ratio.
Thanks for the kind words everyone!
I've been working on sampling this thing further and its getting even better! The unpredictable attack thing has been sorted out, its drive has been improved, and its getting more consistent in behavior and sound. I'm trying to reach a "final version" for my next upload and its getting close!
Again, thanks for everyone with any input and thanks to all who have downloaded. I'm getting really excited about the possibilities here! Once I get this thing mastered, I have some other toys I want to share here
I feel like I'm THIS CLOSE now! Throw out that old one! The compression characteristics are there, the jumping around was because I had a dying power supply that made some inopportune artifacts in the gain reduction.
The color is right where it needs to be. This is a fairly clean compressor, but it does add some weight around 50hz. It sounds thick though, I like it!
Now, I feel like the fast attack and release times are ALMOST there. Like centimeters away, its REALLY close! I honestly just don't know where to go for that (this version uses 2.0ms kernels so I'm assuming it can only handle 2.0ms or longer attack/release, even with a generous lookahead?).
If ANYONE has any advice on how to lower that, I'd be PUMPED! Also, if anyone can clarify things about lowering the Program Rate in the program itself (not in the MAST page of Nebula), I think I will have everything where it needs to be.
For those who know more than I do (and maybe those with less knowledge than me... I hope this helps), here's the settings of how I made this. I made 3 screenshots to show the EXPERT settings. Everything else stays the same.
Can anyone explain "Micro-Kernels" (I enabled them thinking it would be better for fast attack/release times, but I really have no idea)? That whole section baffles me. I put 10 Volterra Orders thinking it would make it 10 kernels... nope haha.
I only did 15 second sweeps to get a quick test done. I'm still VERY surprised how well it turned out regardless.