There's no hater at least by my side.

I listed a few names of brands that I use daily so I answered before even reading your post.

I did blind tests when after having analyzed and replicate freq/phase/ir/harmonics/dynamics plots I couldn't distinguish Nebula programs from algo counterparts so I know for a fact that we're already in an era where, as it is for now, is simply an inferior technology given the limitations it has even when we're talking about sound quality.

Ever heard a guitar distortion/fuzz in Nebula/Acqua (with their original technology)? It's laughable. They had to develop the new "Terra" technology (Ai/machine learning) to release a guitar pedal suite.

Compression? Only in the recent months/years Nebula managed to do something that wasn't weird, and I could be mistaken but IIRC they don't use simple Volterra Kernel Synthesis but it's an hybrid approach for latest compressors, still comps like Relab 176 do it better in algo land for example.

EQs without noise/saturation: simple convolution do the exact same without the CPU hassles. With a tool like DMG Equilibrium you could recreate virtually any curve/phase plot you can think of in algo. Would you need it? Of course no. Want to have saturation/Eq interactions? Plenty of alternatives just sounding as good.

Replicate the sound of a 5150 or a Mesa dual recto (just to name 2) with simple Nebula Volterra Kernel Synthesis: good luck!

Replicate the intricate interactions that happens with an analog gate: you can't. Where's the gate section in Sand/Pink etc? There's no one, did you ever ask yourself why?

Use whatever you like but please stop calling other people haters and please stop spreading the incorrect information that Nebula can replicate analog hardware in general better than algo/neural/etc... modeling because in many cases this is simply not true at least in these days and with the computational power available at the moment for the majority of people. Hybrid/other approaches inside Nebula/Acqua are not to take in consideration of course, they are simply another technology, a better one in fact!

 

Anyone using Nebula on Apple Silicom? Is it running native or on Rosetta? (can see from the Activity Monitor's "Kind" column).

 

What nebula plugins did you tr/compared exactly ? Out of curiosity !

 

This place is nuts. The original post was 'tell me your favourites'. What's with us now having the great honour bestowed upon us with the 'I'm a little bit triggered' (I'm assuming) professional? audio engineers who just can not bring themselves to accept we know what we are hearing - the fact that Nebula gets us where other things can't. No one asked. AlexB does the SSL gate with mind-blowing precision and TimP some amazing transient control. NAM is in its infancy, unable to replicate compression but very promising. You watch - this post will be quoted and countered with the usual opinion dress as fact rhetoric. Have a 35 year recording and high end audio and studio electronics manufacture / restoration career and get back to me. Oh and also - tell us your favourite Nebula libs.....!

 

Last edited: Feb 25, 2026 at 3:14 PM

Those are all solid stuff, but i use em when i don't have CPU available. In general i think Nebula sounds better than all of that stuff.

 

Last edited: Feb 26, 2026 at 6:53 AM

I reply just for conversation, I don't want to start a flame or anything. "Mind blowing precision" as to more accurate than other SSL emulations? Have you got actual data that confirms this? If not it's subjective opinion.
EDIT: I actually just tried AlexB 4K Channel Dynamics Ultimate AI rev. 3 and these are my findings in regard to gate/compression.

The expander it's not very smooth, it's quite jumpy but somehow usable in the expected range of work. Still UAD SSL (I chose this because I use UAD so I know it well) worked smoother and with not inferior sound quality: I would say subjectivily better because of its smoothness when changing treshold, like an usual expander, where AlexB seemed to have a few discrete steps when changing treshold: no good for me but still usable somehow.
Perceived sound quality: honestly on par but difficult to judge because of the quirks

Gate section is completely unusable: it goes from nothing to a strange pumping to completely chopped up with no in between steps, with fast attack and release and 40 db of range where UAD or any other analog or digital gate that I used goes smoothly. Unless I'm missing something this gate is completely broken.

Anyway AlexB himself states in the manual that he wasn't satisfied at all with compressors in Nebula, even his ones and tried to do them from scratch with this project and it costed him 960 hours of work and he states that the result is "very close" in regard to compression. The compressor is working and it does what it says but I can't hear nothing that speaks superior quality honestly but I should have more time to listen in case I'm missing some special qualities.

For what I heard, at least in regard to gate section, it's completely unusable unless I missed some settings that I did wrong inside Nebula 4. Mine has the default settings, which AlexB says that
should be right for installing and running his programs so idk what else could be wrong, given that other programs from the suite work. EDIT: I didn't miss anything, see next post.
What was working didn't wowed me sincerely and all of this with 52 ms of latency: big no for me.
CPU was very very low though!

end of EDIT, I'll continue with the previous answer

TimP's "amazing transient control" whatever it means is not an objective measurement.

What could end the debate it's a test where a Nebula developer and non Nebula developer take a unit, the very same unit, and emulate it. Then do blind tests and measurements and see who got closer. To this day this test doesn't exist AFAIK, so we're in the field of subjective impressions and biases.

Analog gear can sound different from unit to unit and it's difficult to compare different units with slightly different component tollerances, calibrations etc... Sometimes it could be the case that one prefers an emu or another (even with the same technology) not for a quality factor but because of the differences between the units sampled.

There's no need to go personal, I'm not 35 years in "high end audio and studio electronics manufacture / restoration career" but I'm in this world since almost 30 years and I started using Nebula and libraries from AlexB, CdSoundmaster, AITB etc... 15 years ago.
The technology has advanced but still there are limitations and difficulties for Volterra Kernel Synthesis in regard to things like compression/distortion etc... There are many papers free to read about the subject. It's not like straight sampling or even NAM captures where you sample, without having to worry about anything and you're done. There are computational and technical issues that must be taken into account when working with Nebula (from the developer side) when doing compression and other things. Even developers you mentioned talk about these things, it's not my subjective opinion.

I want to be clear again that I'm talking about straight Volterra Kernel Synthesis and not about hybrid approaches that include algorithmic, machine learning, AI, neural networks etc... that is another subject. Not all recent AA products use straight NAT sampling (Volterra Kernel Synthesis)
I'm not on a mission to convince anyone, nor I'm "triggered" or anything, I just expressed my view on this subject, not opinions dressed as facts, but when talking about Acustica Audio there's always this heated and overly defensive aura from its users.

Again, just use whatever you like!

I respect your opinion, in the end all that counts is that you can obtain what you have in your sonic vision. I simply obtain it with a combination of other tools and I'm very happy with it. I'm "maniac" about sound quality and believe me, this is an understatement.
It's just that, at least for me, is so much convenient to use other tools and obtain better results that what I could obtain using other technologies.

The freedom to not have to think about freezing, bugs, sluggish meters, high latencies or high CPU (often both) etc... make me focus on the sound only and leave my CPU open to add touches here and there that make my mixes feel just right and sound as "analog" "warm" "silky" creamy" "airy" "with mojo" "weighty" etc... as I could ever want and I'm enjoying the process in doing it instead of struggling behind it.

If I had to always freeze etc... I would just buy some real analog and call it a day!

If you obtain this with an hybrid approach good for you! It's the result that counts! I'm open to incorporate AA products in my workflow if it meets my subjective requirements, not hating or anything as I said earlier.

2 that come to mind are Magenta (Massive Passive) and Taupe.
VNXT's EMT 140 converted to simple IRs gave me the very same sound.

 

Last edited: Feb 27, 2026 at 4:02 PM

Beside all hassle and through loops jumping I find the sound very static and average. Sorry ... but for that lot of processing power ... nah, snake oil.

 

I wouldn't call it snake oil, for certain things it's very good although for me not worth the hassle.
What I can't understand is how can someone say that "AlexB 4K Channel Dynamics Ultimate AI rev. 3" "does the SSL gate with mind-blowing precision" when it's COMPLETELY BROKEN.

I took 2 hours today to make sure that I wasn't missing something in settings or other things that could lead to errors in the working of the program but there are none. All other programs from the suite work as intended. The expander section in the very same program works less or more as expected, even if it's not smooth as it should be (and I'm being kind) but the gate section is completely broken: read my previous post about it.

I don't think it's a coincidence that Acustica Audio doesn't offer any gate programs in their lineup, maybe they know that it will not produce good results because of how this technology works: it has, like all technologies its limitations.

Even the compressor side wasn't convincing me 100% but it works and in some settings I could make some UAD and AlexB sound pretty much the same but with 50 milliseconds less latency from UAD. At least for now, I'll continue to pass.

These are facts, on contrary the phrase "does the SSL gate with mind-blowing precision" is not even an opinion dressed as a fact, it's a completely false and opposite to reality information. Please don't spread false information guys!

 

Last edited: Feb 27, 2026 at 3:18 PM

Let's just not miss the point: Nebula is a Hardware color box. It's not a utility plugin, it's not a mixing suite, none of that. We got sampling of filters, EQ and compression, but the sauce of Nebula is the interaction between gain and a really dynamic distorion, where every stage sounds in a different way.

Lets take for example, any AlexB EQ. If you read the manual you will see that the HQ version (the one with harmonic distortion for EQs) will come last on the chain. TimP, you put the "clip" module after the EQs (which are COMPLETLY linear).
Why? Because boost and cuts will distort and undistort with the exact harmonics of the sampled hardware at different gain stages. You can use you DAW stock eq with the TimP Pultec, for example. Just put the clip module after the EQ. It will not be a 1:1 Pultec like the lib, but you will be free to experiment with clean, digital filters and may find really cool presets. Sometimes, more than one band of the lib EQ is not needed at all. Nothing is really obvious, so that's why a lot of myths and misconceptions will surround Nebula thing. But we got more experienced users who are not me and they can help people understand better whats going under the hood.

 

Ok but this is not something exclusive of Nebula. Algorithmics/Neural etc... technologies do the same, it's not that they add static harmonics on top of eq curves. I'll stop it here anyway, good music to all!

 

Algo plugins are always a guess game from the dev, even if it's highly trusted dev. They will listen to the hardware and replicate it's behaviour with DSP, from scratch. If there's a huge team with great ears (take UAD as example) they can come real close to the behaviour just with code.
Profiling is a new and interesting approach, although few controls are available and some devs sell flawed profiles and neural plugins (hey Tone Empire i'm talking to you!).

Nebula takes different snapshots from the hardware at different stages, capturing both harmonic distortion, dynamic behaviour and frequency response, all at once and no guess. They are there.

I'm just saying that because there are a lot of "analog" algo emulations totally flat, and with even harmonics at -80db. Not the case of the devs you mentioned. But still, famous devs.

 

Many algo models use circuit modeling and then devs fine tune with ear listening/analyzing.

In the case of Nebula their approach works great for capturing equalizers and light saturation. When going for compressors/tapes etc…things get more complicated.

Devs have to fine tune even in Nebula, especially attack/release behaviour in the case of compressors, as I said AlexB in his manual says that it took him 960 Hours to get things close to the hardware for the SSL: it’s far from an automated process and very similar if not equal to the algo modeling you described (ear Tuning), plus they have to circumnavigate NAT (Nebula sampling processor) flaws in regard to attack phase. Also they have to take care of truncation of Irs that can cause bugs like the infamous “echo bug” plus others that affect even simple equalizers programs in Nebula, including AlexB’ SSL EQs.

Tim P and Cupwise programs seem to take care of mentioned issues though but still they have to handcraft their models in regard to compression action.
If someone is interested I can link many papers about Nebula technology or discussions, mainly on Gearspace where devs discussed these things.

Also it has to be said that the majority of the equalizers modeled by Nebula are high end stuff that exhibit very low THD so that it’s not that relevant to use them as color boxes.

I hope that we’ll get to have 100% accurate emus in the future but as for now I’m very happy with the state of algo modeling: they basically nailed it for the most part of what it’s important and audible so for me the advantages far surpass the cons and in the case of dynamics there’s simply no contest as for as it is now, IMHO.

Discussion could be very long and passionate about this subject, I’ll try to stop here as I don’t want to seem like an AA basher or anything but if something interesting should comes we could continue the discussion in a new thread. Have a nice weekend!

 

Last edited: Feb 27, 2026 at 10:05 PM

Really digging AlexB Neo Console, seems it's modded Neotek Series 1, overall sound reminds me the most to Neve 8014 (Neev 14), but low end is just the ticket, think I found my favorite one for EDM projects.

 

Is there a installation guide to these(requirements)? I tried years ago, but my CPU could not handle it and the GPU version did not work for me..

 

Secret weapon of mine for virtual instruments. If you're doing a full ITB mix you can benefit a lot from this console.

 

I've not been able to find much info for using the GPU to offload the workload for Nebula profiles. Is it supposed to work like Analogx models, which works very well. I know the latter uses new technology in GPU Audio but I remember some sort of use of Cuda years ago with Nebula.

 

@ptpatty I think it was possible in earlier versions, though it did not work well, this was really years ago, maybe over 10 years, can´t remember.
What is the latest you guys use for these profiles or are these standalones like some of the R2R releases?
I never really dived into these nebulas, due to stated reasons in previous post..

 

You think that's bad ? I was wrong about every single choice I made in life.

 

if they were as easy, simple and low demand as slapping on a stock parametric eq I would have them on absolutely everywhere

but as such you have to be much more conservative. the way I've used them over the years has been to swap them in at the end over their 'will do for now' counterparts.

a bit like treating your mix session you made at home as your demo, before taking it to Abbey Road studio to get in all properly recorded and pristinely created.

they are as good as real analogue, and you should treat them that way. if you was gonna get a song cooking in the 90s you wouldn't pull out $15k of gear. but you bet your ass if you create something ready to go out, you want that gear eventually on the final production.

find your workflow of what works, but seriously if you're still sleeping on nebula libs at the mixing stage or mastering (and DON'T have said $15k+ worth of analogue gear) you're missing out.

Tim p. AlexB. Cupwise. incredible creators and we are thankful to be able to access these *real* analogue emulations for so cheap, all on the comfort of our laptops and home studios.

kudos to all involved, if you can support the creators' work you like and use so we can keep enjoying these tools