Page 1 of 6

Nebula Compressors

Posted: Fri Sep 07, 2012 1:27 pm
by SWAN
How do people feel about Nebula compressors now-better?

I notice developers are making more compressor programs - was there an update in technology?

I am trying to re-arrange my set-up...I may need to sell my loved Drawmer 1968 for money reasons - but also - I start to prefer less bus compression now - more individual channel compression. So maybe I buy 2 D*x 160x (to use with Overstayer FET) - and maybe I use a Nebula program on the master just for a little compression for glue. For mixing punchy electronic music...maybe this is a mistake I dont know...

Was wondering about user opinions on:

Alex B Fenix

Alex B MWD

Alex B German Mastering Dynamics

Also - I am intrigued by THE Drum Compressor...

I like a master compressor to smooth the signal and glue - not to grab and change the transients too much...

They are not much to buy I know - but I am moving flat (maybe need to sell gear and downsize) and dont have time to test at the moment...thanks!

8-)

Re: Nebula Compressors

Posted: Fri Sep 07, 2012 3:21 pm
by piperboy
The Drum Compressor is a HUGE (in size and achievement) library that I just can not say enough good about! If I could only keep a few, this would be one of them. Haven't tried the others.

Re: Nebula Compressors

Posted: Fri Sep 07, 2012 8:28 pm
by richie43
Sorry to be vague, but all of the ones you have mentioned are good. You may want to use them in conjunction with your favorite algo compressor plugin to get some of the behavior you are used to, and use the nebula after it for the intense character. But they are all worth buying, for sure. I hardly ever need extreme compression, so Neb compressors without the algo are usually enough for me.

Re: Nebula Compressors

Posted: Fri Sep 07, 2012 9:15 pm
by SWAN
richie43 wrote:Sorry to be vague, but all of the ones you have mentioned are good. You may want to use them in conjunction with your favorite algo compressor plugin to get some of the behavior you are used to, and use the nebula after it for the intense character. But they are all worth buying, for sure. I hardly ever need extreme compression, so Neb compressors without the algo are usually enough for me.
cool - so when you advise using an algo compressor as well - is that in the case of heavy compression? Gabriel from Gemini made a demo of the Kultcomp which managed -10db compression very similar to the hardware...I cant imagine needing much more...

What Im interested in - is understanding what the limitations are. What are the circumstances where Nebula compress like the hardware? Why should you need to use an algo compressor also? I know there is a limitation on fast attacks which doesnt concern me much tbh...

Re: Nebula Compressors

Posted: Fri Sep 07, 2012 10:32 pm
by norburybrook
I'm interested in the replies too as a long time nebula user.

I still think nebula EQs are hard to beat when it comes to the hardware sound.I havent got on board with compressors yet.


MC

Re: Nebula Compressors

Posted: Fri Sep 07, 2012 11:29 pm
by Formbank
It'd be great to have more opinions on the neb comps, current and past. I've got a few and I'm still not sure where changes have been made regarding current (and future) libraries.

Also, the neb comp/algo comp thing, is it a bit of a 'bodge' ?

I'm a big fan of Nebula, but more feedback across the spectrum would be very welcome sometimes.

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 12:15 am
by ngarjuna
Formbank wrote:Also, the neb comp/algo comp thing, is it a bit of a 'bodge' ?

I'm a big fan of Nebula, but more feedback across the spectrum would be very welcome sometimes.
I'm not entirely sure what a "bodge" is but I can tell you that more than one developer had in mind that people would pair their Nebula programs with algo compressors.

To the OP, of the ones you mention I find them all very usable (with the exception of German Mastering one because I don't own it).

I get a lot more mileage out of the 4KD library (which really does sound better than any of the 4000 series emus I own, which are several) and, most of all, the 76D. The old revA blue stripe is nice but Alex's set is pretty great. The behavior is not identical to an 1176 but these programs consistently get me much closer than any others I've ever tried (Waves CLA, UAD, etc).

If it's buss compression you're after I'd consider the Fatso (you can buy just that part of the TSX bundle) that Alex did, I have been preferring it to the 4000 on that duty lately (I go light on buss compression, for me it's about transient shaping and glue more than level control).

The limitations of Nebula compressors have never really been a problem for me, probably just because of my workflow and how I typically like to use compression in the first place. That's certainly not going to be true for everyone. There are still a number of classic compressors I'm waiting to see a nice Nebula version of (how's that A*I 2500 coming, Alex? =) would surely love to see a detailed, deep Fairchild library as well, I think we just have a couple flavors of the circuit without envelope followers right now).

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 1:33 am
by lipa
I think I could write a little on this topic as I have almost all nebula compressors.. It's two or three topics here:

1. Nebula programs as color/preamp only - every compressor library can be used only this way (mostly, there are special programs with zero compression solely for doing this) and Nebula is the best at achieving this color.. (better than any algo).. My favourite flavours are fenix (2 set of tubes to choose from), the drum compressor (all of them, although they impart a LOT of color, especially vca and tube ones), 76D (3 models to choose), kultube (veeeeery nice, pretty subtle but present tube tone). I also love Henry Olonga 670 mojo - very nice thick color. Then of course, after this, algo comp can be used to easily dial the actual compression..

2. Nebula behaviour - yes there are limitations, as You all know, nebula can't do extreme slamming and ultra fast attack.. but, somehow, the programs are getting better and better.. the drum compressor library is the biggest step ahead in this regard.. but there are others that work really well (talking about: german mastering dynamics - quick and clean, 76D - vibey but no real 1176 instant and distorted attack, fenix - nice mojo, 4KD - used to love bus compressor, but lately I prefer the algo version because it's cleaner - funny :) )..

:?: one big question here: I'm embarrased but my interface won't allow 96 or 192 khz.. :oops: can anyone chime in and say if these compressors can get any more aggressive and fast in higher res?? I'll be switching converters within a week so I will report back on this too.. (I've heard that nebula engine really works twice (or 4x) as fast in 96/192khz..) :?:

3. You can tweak Nebula to get faster or cleaner response.. (look in edit/glob smooth parameter and change it to pow8 if You get artifacts with bass heavy sounds, there's also ahead control allowing to get 4 or 5 ms of look ahead) there are others in master page but I forgot the names of parameters so I won't write now..

Generally I strongly recommend trying the drum compressor library, it's vast (I didn't have time to check even half of the presets..) and most of the presets sound great and don't require any tweaking..(in fact You get 4 totally different compressors instead of one - vca, fet, opto and tube) :P

Now, if any dev is reading this pleeeasse cover that itchin A*I 2500, and be my hero! also, everyone is waiting for a fairchild, N**e 33609, C******r zener/germanium (there are those new great little devil compressors..).. sooo.. :roll: :mrgreen:

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 2:00 am
by SWAN
thanks all for the replies esp ngarjuna and Lipa-it is great to get some user experience...I will have to get the FATSO, FENIX, TDC and prob GMC also... 8-)

My concern is - because there is no metering or proper GR meters - Im still not clear on whether some of these programs are actually compressing in the way one would expect. All this talk of using it for colour and with an algo compressor...it makes me wonder - 'why-whats wrong with its compression'?

maybe we can clear this up...
  • As I understand the limitation is no fast attack - correct?

    And there is sometimes a limitation in the 'amount' of GR?

    Can anyone clarify why these limitations exist?

    Can we conclude then - if you are not asking a very fast attack, and not huge amount of compression - then these programs are generally quite accurate to the hardware...?
I know it is something to ask - but I think it would be helpful for the developer to be clear on each release what the limitations are per program - so we know what to expect. Saying 'we know the Nebula limitations - Im not sure is enough'. It would also help in explaining how to best use the programs. I remember reading some backlash following Alex B release of his 1176 because there were people disappointed that it did not do some of the 1176 style effects...

I think CDSoundmaster has been quite clever with his TDC release. To release a product that leverages the Nebula strong points only. Despite not always liking his approach to other programs - for compressors - this seems like the best way forward...I look fwd to bass, keyboards and other program packs....

I know Henry only samples compressor colour - and he is clear about that which is great.

I know we are supposed to use ears - but its quite a lot of trust to load these programs on the audio when not clear on certain aspects...

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 2:29 am
by lipa
well.. I'm no nebula engine expert.. I only use ears to evaluate these limitations.. I actually think it was a very brave to put together such library (the drum comp).. but yes, there is no heavily distorted and fast attack in nebula, it kinda always "click" on transients.. I guess it should be much better in 96/192 khz. That's about it! The rest of limitations is standard nebula stuff (no ratio knob, or no attack knob etc..).
I understand You, I'm biased towards nebula compression. I really love some of the libraries but there are times I choose algo comp because they're more predictable and ultra fast to dial-in.. and most have proper metering.. :roll:
(lately it's been 50/50 in my mixes)

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 5:08 am
by biomuse
SWAN - see:

http://www.acustica-audio.com/forum/ind ... =viewtopic

You've got the general idea - Nebula isn't very good for substantial, predictable compression, and it does not handle transients well at all.

Ideally you want to use a clean, fast algo compressor, limiter or saturator in front of Neb, and the Neb compressors can then provide the color and personality.

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 7:58 am
by mpodrug
Thanks guys for the useful informations. I'm just wondering if anyone else favoring Signaltonoize's compressor libraries? I've got his 76 stripe and find it to be very very good. My real grief is the lack of GR meter, because my ears are not trained enough to do it by ears.

When you guys say nebulas compressors can't handle heavy slamming style, how much of gain reduction can it really take before it behaves weird?

Also, i remember i have read a monthly tip where you can edit a secret parameter (forgot the name), which allows nebula handle attacks down to 1ms. Shouldn't it make the fast attack limitation obsolete?

I've read somewhere about the weird behavior at the very first transient. What's that all about?

I hope those questions are not too noobish :)

A monthly tip dedicated to compressors, the libraries and how to make them work best, and which areas are still in work would be really nice.

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 8:18 am
by biomuse
The answers to all your questions are in Enrique's Understanding Compressors Tip of the Month thread, the link to which is itself in the thread I link to above.

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 11:11 am
by SWAN
Thanks Biomuse

some exerpts from Tips of the Month:
  • In software compressors each incoming sample has a gain changement so they are accurate for all audio events including transients.

    Nebula processes blocks of samples because it tries to be accurate on frequency and harmonic content and switching kernels too much often will lead to other unwanted effects like ringing noise or fake sounds.

    How many blocks of samples? This is the PROG RATE value of the emulation preset shown in KERN page as RTE. The emulation preset rate is a parameter fixed by the library developer.

    Fast emulation preset rate means that the compressor will react in a faster way. If a emulation preset is loaded with rate of a single sample you have exactly the speedness of a common software compressor, but you can't have this working correctly in Nebula unless kernels are very short and as result Nebula will add artifacts, noise and aliasing.

    Slow emulation preset rate means that compressors will react in a slower way and as result less clipping issues, artifacts, troubles and aliasing but Nebula could start processing sound later, and transient could not be processed correctly.

    Latest compressor emulation program are based on RAWFUNS, attack and release times are sampled but with limited gain reduction range, around 30 dB are sampled, this limitation is related to the noise floor from the sampled gear.

    In order to get a similar behavior like software compressors you should try moderate threshold values, from -10 dB to -30 dB and if you need more gain reduction you can increase the gain in and decrease the gain out (depending the gain reduction and the make up values)...

Re: Nebula Compressors

Posted: Sat Sep 08, 2012 11:18 am
by SWAN
There is a few things I still dont understand:

From Nebula tip of the month:
  • When a compressor is loaded with an emulation preset rate with a value more than 2 milliseconds, for example 20 milliseconds, Nebula will be late about 18.5 milliseconds due the look ahead setting and as result the transient could not be processed correctly (you hear something loud at the beginning because compression is not engaged yet).
The preset rate is 20ms...however there is a look ahead function - I assume is 1.5ms (?)...hence Nebula will be late at 18.5 ms?

The compressor program can only make a gain change every 20ms-that is quite a long time between gain change samples?