Eh no!
I will be in holidays tomorrow. In case I have no internet there I now told the results to Ermz who can reveal it then.
Part 2 of "The ultimate IR, NEBULA, REAL DEAL test"
- Thread starter Skaldir
- Start date
Eh no!
I will be in holidays tomorrow. In case I have no internet there I now told the results to Ermz who can reveal it then.
Wisheraser
Member
Just to be different, I'll guess:
A - Amp
B - Impulse
C - Nebula
That said, I liked B the best, A was a VERY close second, and I liked C the least. If my guess (and it is a complete guess) is right, then you've managed to create the "Scarlett Johansson's pussy" of impulses. Bravo!
A - Amp
B - Impulse
C - Nebula
That said, I liked B the best, A was a VERY close second, and I liked C the least. If my guess (and it is a complete guess) is right, then you've managed to create the "Scarlett Johansson's pussy" of impulses. Bravo!
TonyS.
www.SoundCollab.com | www.risingstorm-music.com
For me, the interesting thing in this test is not to find the right answer, but rather the track the most people like.
I think its the real amp vote, mostly?!
I think its the real amp vote, mostly?!
Nimvi
Member
Agree, I'm secretly hoping that I prefered the new nebula cab over the real one...that would be interesting.
Btw, I've just noticed that most of us have been calling it "real amp, ir and nebula". It's all the real amp, it's the CAB that is either real or impulse. Maybe that's why it hasen't been guessed correctly yet? That would be awkward
Ermz
¯\(°_o)/¯
Nailed it. This is the actual result. Unicorn told me last night.
I'm somewhat disappointed, personally, as I expected Nebula to do better. Though I think this does speak volumes of Unicorn's ability to make a good IR. It seems most of us couldn't confidently pick whether A or B was the real amp. So if all else fails hopefully we'll be graced with an awesome IR library.
Not giving up on Nebula just yet though, as I'm almost positive we've had better results with it in the past. It does have a tendency to somewhat deviate from the frequency response of the cab more than the IR, despite the fact that it has dynamic layers. It's that harshness in the highs that I've wanted to eliminate from it since we started. I'm keen to try one of Ola's songs next time around and see whether our results are consistent.
Nimvi
Member
I think that's the best attitude you can have towards it Ermz. What you guys are doing is pioneeringwork after all, and Rome wasn't built in a day. I mean...the first static impulses weren't stellar by any means, but nowadays there is quite a library of very usable ones available. Stay at it and you WILL get rewarding results in the end.
Ermz,
how do unicorn's impulses behave when eq'ed?
I remember that especially you and jefftd mentioned that impulses don't really stand eq'ing very well. does this statement still hold true?
how do unicorn's impulses behave when eq'ed?
I remember that especially you and jefftd mentioned that impulses don't really stand eq'ing very well. does this statement still hold true?
Correct!
I wanted to add that the normal impulse is mostly very close in terms of frequency response. But there is a thing, that is more noticable with more bassy signals that f.e with palm mutes it seems the bass comes a bit later than with the real cab. Nebula handles that way better to me.
But its hard to hear with this example.
When I made the cancellation test with the real mic'ed cab (you can do it yourself also) the Nebula programs also canceled out more signal as the impulse. So it IS closer to the real deal.
Well, I will look further into it. We still have a few ideas left.
Here is the impulse and the Nebula program for all.
Kalthallen-IMW-Impulse_2.zip
Notice it might not be the brightest mic position, but it also sounds quite good with amp sims.
//Markus
Mago
Austrian Blech Machine
azistenostora
Member
~BURNY~
Member
- Apr 20, 2005
- 5,091
- 67
- 48
SwambA
Processed Chicken
Correct!
I wanted to add that the normal impulse is mostly very close in terms of frequency response. But there is a thing, that is more noticable with more bassy signals that f.e with palm mutes it seems the bass comes a bit later than with the real cab. Nebula handles that way better to me.
But its hard to hear with this example.
When I made the cancellation test with the real mic'ed cab (you can do it yourself also) the Nebula programs also canceled out more signal as the impulse. So it IS closer to the real deal.
Well, I will look further into it. We still have a few ideas left.
Here is the impulse and the Nebula program for all.
Kalthallen-IMW-Impulse_2.zip
Notice it might not be the brightest mic position, but it also sounds quite good with amp sims.
//Markus
Thank you!
Kazrog
Kazrog, Inc.
Great stuff. I couldn't hear much of a difference between the 3 versions, really a toss up, and I had no clear preference. Thanks for posting!
Its a self build 4x12 cab with dimensions like a Marshall 1960a. v30 speakers
Pipotron3000
Member
- Aug 5, 2010
- 32
- 0
- 6
Thanks for the test !
I ABXed the files...and ended thinking it was :
A real cab
B Nebula
C IR
I like B better than C...and i'm a little disappointed by the final answer
Because in your previous test, Nebula was nearly on par with the real cab for me.
May be what Ermz pointed at since the first experience with AE impulses, some high freqs troubles ?
On the other side, it is better than the first attempt by AE, and i don't blame him, he was the first to attempt this
Some ppl still use the AE Nebula impulses, blended with IRs.
I wonder how your C - Nebula version could react to an auto EQ matching with the B - IR version...
PS :For ppl who doesn't know it, OwnHammer do commercial Nebula cab programs...and it rocks :
http://www.ownhammer.com/media/speakercabinets/412-GTR_Mesa_V31/
Listen to the Mills 412-GTR Acoustics Afterburner 4x12 Celestion Vintage 30 #1 - Metal Full mix...epic
A good motivation to do better than him (even if it is difficult)
I ABXed the files...and ended thinking it was :
A real cab
B Nebula
C IR
I like B better than C...and i'm a little disappointed by the final answer
Because in your previous test, Nebula was nearly on par with the real cab for me.
May be what Ermz pointed at since the first experience with AE impulses, some high freqs troubles ?
On the other side, it is better than the first attempt by AE, and i don't blame him, he was the first to attempt this
Some ppl still use the AE Nebula impulses, blended with IRs.
I wonder how your C - Nebula version could react to an auto EQ matching with the B - IR version...
PS :For ppl who doesn't know it, OwnHammer do commercial Nebula cab programs...and it rocks :
http://www.ownhammer.com/media/speakercabinets/412-GTR_Mesa_V31/
Listen to the Mills 412-GTR Acoustics Afterburner 4x12 Celestion Vintage 30 #1 - Metal Full mix...epic
A good motivation to do better than him (even if it is difficult)
Pipotron3000
Member
- Aug 5, 2010
- 32
- 0
- 6
I think i found WHY Nebula impulses can sound strange in the highs.
When you do a linear impulse IR, you "just" capture the final spectrum.When you apply it to your amp sound, it is just a "mask", it doesn't add any harmonic, saturation,whatever to the signal.It is just a frequency mask (even if it is more subtle than that).
If there is nothing at 1000Hz input, there will still nothing output.
When you capture a Nebula program with NAT, it analyse witch harmonics are added in the chain with multiples tests (ie NAT send a 500Hz sine and see there are 1000Hz, 2000Hz... added harmonics in the final result).
When you apply the program to your chain, it add the harmonics NAT captured !
Like with any usual saturation Nebula program.
You get the added harmonics of any pedal, preamp, power amp and mike preamp.
If there is nothing at 1000Hz input, but something at 500Hz, you will get something at 1000Hz output, unlike the IRs
So you get too much harmonics, because you get the ones from your guitar sound chain without cab AND the harmonics from ALL the Nebula captured chain.
This is like putting an already processed chain without cab in another processed one with cab.Like chaining one amp directly into another and capturing the final cab sound.
I think the only solution is to get a minimal harmonics path.
Witch is the opposite of most IRs captures i've seen, pushing amps and even using pedals
Like using a transistor guitar power amp alone, at medium level.
Or at least using a tube power amp alone, at medium level.
This is related to the inerrant very different way IRs and Volterra kernels works
IR : linear, no added harmonics to the input sound
VK : non linear, add captured harmonics to the input sound.
It combine cab spectrum capture (like IR) AND added harmonics from signal path (like a saturation plug).
I wish it will help getting better results
I don't want to start another shit storm war, this is just the way i see things, using Nebula EQ, saturation,reverbs... for years
I think i really got the trouble.
There is a simple test to do : creating a 1 kernel program.
No more saturation capture with 1 kernel.
If it sounds good, we got our culprit : saturation capture
It was so easy for Nebula devs i don't understand why they didn't answered this simple (for them) question...
When you do a linear impulse IR, you "just" capture the final spectrum.When you apply it to your amp sound, it is just a "mask", it doesn't add any harmonic, saturation,whatever to the signal.It is just a frequency mask (even if it is more subtle than that).
If there is nothing at 1000Hz input, there will still nothing output.
When you capture a Nebula program with NAT, it analyse witch harmonics are added in the chain with multiples tests (ie NAT send a 500Hz sine and see there are 1000Hz, 2000Hz... added harmonics in the final result).
When you apply the program to your chain, it add the harmonics NAT captured !
Like with any usual saturation Nebula program.
You get the added harmonics of any pedal, preamp, power amp and mike preamp.
If there is nothing at 1000Hz input, but something at 500Hz, you will get something at 1000Hz output, unlike the IRs
So you get too much harmonics, because you get the ones from your guitar sound chain without cab AND the harmonics from ALL the Nebula captured chain.
This is like putting an already processed chain without cab in another processed one with cab.Like chaining one amp directly into another and capturing the final cab sound.
I think the only solution is to get a minimal harmonics path.
Witch is the opposite of most IRs captures i've seen, pushing amps and even using pedals
Like using a transistor guitar power amp alone, at medium level.
Or at least using a tube power amp alone, at medium level.
This is related to the inerrant very different way IRs and Volterra kernels works
IR : linear, no added harmonics to the input sound
VK : non linear, add captured harmonics to the input sound.
It combine cab spectrum capture (like IR) AND added harmonics from signal path (like a saturation plug).
I wish it will help getting better results
I don't want to start another shit storm war, this is just the way i see things, using Nebula EQ, saturation,reverbs... for years
I think i really got the trouble.
There is a simple test to do : creating a 1 kernel program.
No more saturation capture with 1 kernel.
If it sounds good, we got our culprit : saturation capture
It was so easy for Nebula devs i don't understand why they didn't answered this simple (for them) question...
Omega_Void
Your Eventual Destroyer
@Pipotron3000
Most of the IRs I've seen used only the poweramp section to drive the cab.
I'd assume that's what's been done here, so the problem you describe of adding harmonics twice over shouldn't occur.
So (provided the same poweramp settings were used in making the impulses as on the cab recording), Nebula should have captured poweramp and speaker harmonic distortion not present in the preamp signal but present in the recording of the mic'd-up cab.
Nebula therefore ought to sound more like the mic'd-up cab than the impulse does. In practice though, it DFW.
If I'm understanding Nebula correctly (and I've paid near-zero attention to it), it's based on a sort of "multiple snapshots at different input levels" type of thing?
Loudspeaker behaviour isn't all about what the input is now, it's a also about what's the difference between the input now and what it was just beforehand. Think "physical momentum".
My guess is capturing speaker distortions improperly may be the culprit, and worse than not capturing them at all.
Most of the IRs I've seen used only the poweramp section to drive the cab.
I'd assume that's what's been done here, so the problem you describe of adding harmonics twice over shouldn't occur.
So (provided the same poweramp settings were used in making the impulses as on the cab recording), Nebula should have captured poweramp and speaker harmonic distortion not present in the preamp signal but present in the recording of the mic'd-up cab.
Nebula therefore ought to sound more like the mic'd-up cab than the impulse does. In practice though, it DFW.
If I'm understanding Nebula correctly (and I've paid near-zero attention to it), it's based on a sort of "multiple snapshots at different input levels" type of thing?
Loudspeaker behaviour isn't all about what the input is now, it's a also about what's the difference between the input now and what it was just beforehand. Think "physical momentum".
My guess is capturing speaker distortions improperly may be the culprit, and worse than not capturing them at all.
Pipotron3000
Member
- Aug 5, 2010
- 32
- 0
- 6
I'm just saying capturing Nebula programs is not like capturing IRs
Take an IR of a very pushed out distortion pedal (it is easy using a loopback).Load it in a cab impulse loader.Apply it to a dry clean DI guitar track.
What do you obtain ?
A perfectly clean sound.
Do the same with Nebula...you get a big distortion at the output
Because Nebula capture the saturation of the path too, whereas IR just capture a spectrum mask.
If you push your power amp with Nebula, you will create so much saturation that it will create too much harmonics when you apply this program to an ALREADY saturated sound (the output of your virtual amp).
This will never happen with IRs, because they are linear.
And this is the main reason why ppl got troubles with Nebula.
They know how to capture IR, but Nebula need a clean chain unlike IRs.
This is simple to prove : take any massive IR made with big pushed out amp and apply it to a clean DI track -> it still clean.
An IR is just a sort of "subtle auto EQ".
Nebula do the same BUT add saturation/compression too
I can't try the Unicorn impulses (no Nebula commercial version), but i'm pretty sure if you apply them on a perfectly clean DI track, you will get a subtle saturated/compressed guitar sound at the output (if they got more than 1 kernel, of course).
If someone can try it and post the result here
When i speak about saturation, it include dynamic too.
If you push your power amp to get a typical PA compression, it will be applied back by Nebula
This will never happen with IR (linear, again).
It can lead to BIG differences between an IR and a Nebula program done with the same rig.
More you push your power amp/mike preamp, more the differences will be obvious
I wish my explanation is cleaner now
Take an IR of a very pushed out distortion pedal (it is easy using a loopback).Load it in a cab impulse loader.Apply it to a dry clean DI guitar track.
What do you obtain ?
A perfectly clean sound.
Do the same with Nebula...you get a big distortion at the output
Because Nebula capture the saturation of the path too, whereas IR just capture a spectrum mask.
If you push your power amp with Nebula, you will create so much saturation that it will create too much harmonics when you apply this program to an ALREADY saturated sound (the output of your virtual amp).
This will never happen with IRs, because they are linear.
And this is the main reason why ppl got troubles with Nebula.
They know how to capture IR, but Nebula need a clean chain unlike IRs.
This is simple to prove : take any massive IR made with big pushed out amp and apply it to a clean DI track -> it still clean.
An IR is just a sort of "subtle auto EQ".
Nebula do the same BUT add saturation/compression too
I can't try the Unicorn impulses (no Nebula commercial version), but i'm pretty sure if you apply them on a perfectly clean DI track, you will get a subtle saturated/compressed guitar sound at the output (if they got more than 1 kernel, of course).
If someone can try it and post the result here
When i speak about saturation, it include dynamic too.
If you push your power amp to get a typical PA compression, it will be applied back by Nebula
This will never happen with IR (linear, again).
It can lead to BIG differences between an IR and a Nebula program done with the same rig.
More you push your power amp/mike preamp, more the differences will be obvious
I wish my explanation is cleaner now
Similar threads
