Nebula Programs and best impulses yet have arrived! (for real this time :P)
- Thread starter Metaltastic
- Start date
57 1 was with the center of the mic a tad on the inside of the dustcap edge, 57 2 was the opposite (center of the mic a tad on the outside of the dustcap edge), for the more presence ones I just changed the knob on the amp! And thanks to Ryan's tip, the mic was about 1 cm away from the grille cloth!
And Kev, I can't listen now but I'm looking forward to and am glad you got it to work!
Yeah, it peaked at like -9 db, I didn't wanna push the mic preamp and introduce more distortion
I'm really not so sure about Nebula being more accurate than impulses though dude, though obviously the only solution is to do a comparison between Nebula, impulse, and a recording all done at the same time, which I will get to in the next couple of days!
Ermz
¯\(°_o)/¯
^ Agreed, but the thing is that AE has already done this and shown me the results. Going into it blindly I actually preferred the Nebula run over the real amp and impulse, and that's no shit honesty from me right there. For some reason his last set of runs didn't have this fizz.
Could it be what giancarlo was talking about, AE? That thing about needing to do a loopback and making sure your I/O isn't doing anything weird? Also, did you deconvolve yours offline or online? All these little things could be causing issues.
Could it be what giancarlo was talking about, AE? That thing about needing to do a loopback and making sure your I/O isn't doing anything weird? Also, did you deconvolve yours offline or online? All these little things could be causing issues.
AeternusEternus
Member
- Oct 8, 2008
- 601
- 0
- 16
About Giancarlo's tests: Yup. About the deconvolution: Both. Hundreds of times. Haha.
Isn't that missing the point? If your cab hums in the room (with no signal), then surely the Nebula program should in the same situation? I remember you saying before Nebula wasn't reproducing the hum you could hear.
I haven't read up about Nebula or anything, so I don't know the specifics of how it works but the first thing that strikes me is this: it seems like everyone is trying to get an exact copy of a complex, harmonically rich sound without seeing if they can even match "silence" - i.e. amp-> cab-> microphone -> preamp with no signal. If you cab hums but your Nebula program doesn't, it's already failed. Also, if I understand it right the volume of the sweep shouldn't make a difference like that - surely it should just make for a Nebula program that reacts like a cab when the volume is turned down on the amp.
From what I know, the elements involved at this stage are:
1 - the amp/poweramp/etc.
2 - the cab
3 - the mic
4 - the deconvolver
5 - the sweep/tone
6 - the program/vector
7 - Nebula
If Ermz made impulses at the same time that don't have have a high-end/low-end issue, that rules out 1, 2 and 3. The deconvolver is supposed to be the same as used in something else right? So it's fairly safe to rule out number 4 too.
If Nebula is doing something to the sound, you'd assume there's settings to control it somewhere - which no one seems to have mentioned. Granted, it's got a dog of a front-end, so everyone could be missing something, but chances are it's not that.
Which doesn't leave much. Like I said, I have no idea how this all works, so I'm unclear on a few points:
Firstly, what does the sweep sound like? Is it just like a regular sine-wave sweep like you'd use for an impulse? It it just a repeating wave at different volumes or something? Or does it do other funky stuff? I'm assuming it's different, otherwise I don't see how the Nebula program could sound any different to an impulse. Whatever it is, the actual process of recording it is just a mechanical one, so it's probably safe to rule that out - especially as I think Gian Carlo said you could make an impulse from a 1 kernel program thing, which suggests the sweep is just a fancier sweep than normal.
Secondly, what exactly is the 'vector' that Nebula creates? Presumably if it's designed to make dynamically-reactive convolutions, it's essentially like a 3D impulse - frequency along X and amplitude up the Y (like an impulse), but with source volume along the Z. Gian Carlo said that a 1 'kernel' program is effectively an impulse right? So presumably the kernel defines how many divisions there are along the Z axis - effectively how many layers of detail there are.
Assuming any of ^that^ is right, then having the kernel setting at any value between 1 and 'Lots' (whatever the max. is), there has to be some kind of averaging/aliasing going on to make the program dynamic - otherwise you'd just have a series of impulses with obvious discrete steps between them. To me, that seems like somewhere that the output sound could be affected. Surely there's some potential for wave cancellation there?
Does a 1 kernel program give you a different sweep to, say, a 5 kernel one? If so, all you need to do is make programs for 1/middle/maximum kernels etc., run the sweeps through the same set up, then run a normal impulse-making-type sine sweep through Nebula on each of the programs and look at the results in a spectral analyzer.
Also, if you run a sine sweep through Ermz's new impulse and the corresponding Nebula program, you'll be able to accurately see what the difference is - and if the program is dynamic, then try it the sweep at different volumes, as that should make a difference in the results.
Steve
And all this is too much for me to want to think about, so someone else come up with tests I should do, and I'll do 'em!
have you tried deconvulving Metaltastic's tones without distortion kernals?
it was mentioned harsh highend could be caused by these in some way or other. maybe an a-b test with and without distortion kernals could be of benefit?
i am 100% behind nebula, impulses are yesterday's tech and sound lifeless and flat.
i think giancaro can answer all questions, but someone needs to ask in a very specific way with clear sound examples. there is a lot of confusion about everything, we need some clarity i think
it was mentioned harsh highend could be caused by these in some way or other. maybe an a-b test with and without distortion kernals could be of benefit?
i am 100% behind nebula, impulses are yesterday's tech and sound lifeless and flat.
i think giancaro can answer all questions, but someone needs to ask in a very specific way with clear sound examples. there is a lot of confusion about everything, we need some clarity i think
LeSedna
Mat or Mateo
Playing the devil's advocate :
- Does Nebula REALLY sound any good for other things than cabs ? I mean, really accurate ?
- Also, if yes, don't we have problems for distorted sounds because, we all know this, it's the most messed sound we can have and the one that has the most extreme characteristics, that could maybe make the nebula's algorithms fail ?
personal question : why do you say "impulses vs nebula", aren't nebula "programs" 2D or 3D "impulses" ?
- Does Nebula REALLY sound any good for other things than cabs ? I mean, really accurate ?
- Also, if yes, don't we have problems for distorted sounds because, we all know this, it's the most messed sound we can have and the one that has the most extreme characteristics, that could maybe make the nebula's algorithms fail ?
personal question : why do you say "impulses vs nebula", aren't nebula "programs" 2D or 3D "impulses" ?
I`ve heard an incredible drum room program that totally convinced me of Nebula's capabilities.
Also, AE's test was pretty dead accurate when comparing miked cab with nebula program. I`m still convinced that the template he got is really good and that his programs are not quite IT, just because of his mic positioning and setup (though they are still usable and show nebula's potential for cab emulation).
I think more people that own a cab, and have some recording experience should give a go at the offline deconvolsion thingie, and let AE make the programs out of them. It would give us a variety of programs, and I believe it would be a faster way to get a feel for real Nebula posibilities
Also, AE's test was pretty dead accurate when comparing miked cab with nebula program. I`m still convinced that the template he got is really good and that his programs are not quite IT, just because of his mic positioning and setup (though they are still usable and show nebula's potential for cab emulation).
I think more people that own a cab, and have some recording experience should give a go at the offline deconvolsion thingie, and let AE make the programs out of them. It would give us a variety of programs, and I believe it would be a faster way to get a feel for real Nebula posibilities
grywolf627
Member
Exactly. I'm REALLY grateful for the work AE has done the contributions Marcus has made. But up until now, the ONLY Nebula cab programs have been from them...even in the previous go around. We really need to get some others involved to help paint a broader, and hopefully clearer, picture of how Nebula sounds.
AeternusEternus
Member
- Oct 8, 2008
- 601
- 0
- 16
I dunno where the hell Guitarhack is, I'd think he'd be down to run a 1 minute 36 second series of sweeps through his rig!
noarin
Producer / Engineer
- Apr 16, 2006
- 569
- 0
- 16
anyone up for helping me with this?
http://www.ultimatemetal.com/forum/...l-like-helping-me-sample-my-od808-nebula.html
http://www.ultimatemetal.com/forum/...l-like-helping-me-sample-my-od808-nebula.html
Rex Rocker
Call me Hugo!
You're right! I really like these impulses.Nooooooooo, ignore them, cuz I used the Orange channel rather than Orange channel cloned to modern! (you don't have to understand what that means, just know it made a big difference for the better in the sound of the power amp
Pretty cool dude, could use a bit more mids and maybe a bit less treble, but overall I dig it! And Nebula is a program that loads "samples" which are similar to impulses in the way they're captured but a lot more dynamic; to give you an idea, the file that I ran through my power amp/cab to capture it for Nebula was a series of sweeps at progressively lower and lower volumes, so it captures and models how the thing reacts at different levels, thus making it more dynamic! Though there seems to be an issue with accuracy, at least on my end...
Oh, and did you use 57 1 or 2? For reference, as I mentioned, 2 is a bit farther from the dustcap, in case you were wondering what to listen for as far as changes between the two!
Ermz
¯\(°_o)/¯
What was running through my mind is... when does Nebula know when to switch from a 'low volume' impulse to a 'high volume' one as we play through it. If for instance the signal levels we are feeding nebula are not what it wants to be seeing, we could be constantly using the 'low volume' impulses, and given that those would have a relatively higher noise floor, it could explain the fizz (in some contrived, arse-backwards ermz logic way).
I still can't reconcile the idea that AE had such a similar result to his real cab with Marcus having an audibly different one. What are the factors at play here that could be affecting the outcome? Impulse offset? Converters? Amp hum? Test tone level? Methinks the developers are welcome to chip in on that one.
I still can't reconcile the idea that AE had such a similar result to his real cab with Marcus having an audibly different one. What are the factors at play here that could be affecting the outcome? Impulse offset? Converters? Amp hum? Test tone level? Methinks the developers are welcome to chip in on that one.
I'm not so sure nebula would give a shit about hum. It's there when its not sampling, as soon as the sweeps start, the 60hz idle hum would not be present. Nebula is looking for the sweeps it has assigned, and comparing them to the recorded sweeps to create the program. Add a bunch of silence before and aft the sweep and I'm sure it'd be present, but largely ignored due to how low a signal it would be compared to the blaring of the cab. AE has proved nebula works, I can't understand the doubt.
Also, guitar cabs are really fucking fizzy, as is what the sm57 see's. I see impulses as a filter, thats all they are. Adds a bunch of notches and cuts from the signal. Nebula appears to take the entire spectrums dynamics individually to get the feel it has. It's better, we know this as fact.
Also, guitar cabs are really fucking fizzy, as is what the sm57 see's. I see impulses as a filter, thats all they are. Adds a bunch of notches and cuts from the signal. Nebula appears to take the entire spectrums dynamics individually to get the feel it has. It's better, we know this as fact.
Once again, I'm not willing to agree to this, cuz I've heard a few comparisons between a mic'ed cab and impulses made with the same mic position through the same power amp (and then applied to the preamp signal), most notably by GuitarHack, and the only differences were in the lack of dynamics, especially on palm mutes - tomorrow I'm gonna do my own 3-way comparison between mic'ed cab, impulse, and Nebula program to see once and for all!
