Impulse vs. a Real Cabinet with clips!

[aortizjr]

Well after playing around with "The Ultimate Modeling Challenge" by Lasse, and reading all the complaints, I decided to experiment. I use a lot of impulses from ReCabinet and Catharsis and Alu, but haven't really put them head to head with my own gear.

So I made some DI's, mic'd up a cab, made an impulse. Then reamped and A/B'd. Then I wrote way too many words and put it up on my blog. I have only seen one other comparison done like this and it was just lead guitar.

Each clip is 3 riffs double tracked: chuggy, thrashy, and chordy, once through alone and then with bass and drums. SSD 3.5 and a Lakland import bass with a limiter.

http://ortizaudio.blogspot.com/2010/06/impulse-vs-real-cabinet.html

Essentially this isolates all the variables and compares a mic'd cab and an impulse of that mic'd cab. All the details are there and I don't want to type them again.

I might do some blind tests depending on the responses I get on this and other forums.

I would really like to A/B modelers with their real inspirations. Or compare modelers head to head. Ugh.. so much work though and I don't have all the amps in question. If I did, I wouldn't need a modeler.

Discuss...

 

[dcb]

pretty cool, although your miced version is very muddy and i personally wouldnt want my guitars to sound like that (no offense! just matter of taste i guess)

so i wouldnt say there is any better (real vs impulse) - but igotta admit they sound close enough for untrained ears not to make a hell of a difference.

 

[Megin]

Sounds pretty close with the impulses. Which impulses did you use? Your own?

 

[P-E]

I wonder what you get if you flip the phase. How much different stuff there is.

Just sayin'.

 

[Fox Mulder]

I wonder what you get if you flip the phase. How much different stuff there is.

Just sayin'.

Aha!

 

[aortizjr]

pretty cool, although your miced version is very muddy and i personally wouldnt want my guitars to sound like that (no offense! just matter of taste i guess)

so i wouldnt say there is any better (real vs impulse) - but igotta admit they sound close enough for untrained ears not to make a hell of a difference.

Well there lies the problem with the ENGL Powerball aka. Mudball. So loved around here.

Right now I have a band tracking with it and using a cheapy Behringer Ultra-G cab sim thingy which is super bright. So the amp is dialed for that. The plan is to re-amp but maybe mess with impulses, so I can't redial.

Also in general with the Powerball, if you go more towards the cap, you get too much fizz. So I multi-mic, SM57 right around here and then an LD condenser to get more high end.

But for this test, I didn't want to play the multi-mic game. And it sits in the mix fine, especially with a high shelf. With some more EQ and compression, I think it would really come down to a taste thing.

Anyway I mention the mud tone in the Arguments, definitely worth noting.

Sounds pretty close with the impulses. Which impulses did you use? Your own?

If you read the page, I made an impulse of the same real mic and cab setup. So my own which matches the setup.

I wonder what you get if you flip the phase. How much different stuff there is.

Just sayin'.

I will edit the page to talk about this, since I forgot to add it. Tried that, there are a couple of problems.

1. The time delay between the cab and the mic. So you have to line it up by hand, not a huge deal, but the waveforms don't perfectly line up either. Interestingly, the time delay makes the signal quite a bit out of phase.

2. Even though the signals are the same, the recording times are different so there is inherent clock skew. So you end up with an inflated amount of stuff left over that actually shifts in and out over time.

You are welcome to test it since you get the first set with just the guitars and they are panned hard left and right. But you will have to align the wavforms manually then use a frequency dependent phase shifter to get them in phase, then flip... or just perfectly out of phase. But you will still have the clock skew.

Also on my Dynaudio BM5A and Yamaha HSM80's in my studio I can hear so clearly hear difference between the Impulse and the Cab that it really isn't worth it. On my Roland MA-8's in my untreated office I think I hear differences, but I don't think I could do it blind.

Oh I also tried doing matching EQ with Ozone 4. That matching was weird and added an overbearing amount of mud to the Impulse and an way too much sizzle to the Cab and Mic. But I was learning as I go with the Matching EQ. I tried different parts with different riffs, all sounded bad. Even just looking at the snapshots, there are big (~.5-2db) differences in the low mids, upper mids, and highs.

I was thinking maybe if I sent white noise and then matched, but curves and applied it to the guitar tones, but I don't know. Isn't that what an impulse is? A sampling of all frequencies? Then just a 1024+ band EQ or whatever? So I didn't bother, this gets the point across.

Personally I think if you can hear the difference now, take note of them. Then go into a blind test, you would do pretty well at picking them out. Especially if the same mic setup and sampling were done.

Adding a room impulse to me helps with a lot of that too. Probably the same would be true with just a touch of algorithmic reverb designed to sound like a room and not an effect. Hmm... I wonder how hard it is to sample my room. I think I just need a starter pistol and measurement mics.

 

[Brandon E.]

I had ALWAYS wanted to do something like this, thanks for doing it for me

They sound very similar but the real cab sounds more 3d and definitely fills in the mix a little better, but the version with the room reverb sounds really not very far off...A/Bing them I still think the real cab sounds a little more full though...

 

[P-E]

Well there lies the problem with the ENGL Powerball aka. Mudball. So loved around here.

Right now I have a band tracking with it and using a cheapy Behringer Ultra-G cab sim thingy which is super bright. So the amp is dialed for that. The plan is to re-amp but maybe mess with impulses, so I can't redial.

Also in general with the Powerball, if you go more towards the cap, you get too much fizz. So I multi-mic, SM57 right around here and then an LD condenser to get more high end.

But for this test, I didn't want to play the multi-mic game. And it sits in the mix fine, especially with a high shelf. With some more EQ and compression, I think it would really come down to a taste thing.

Anyway I mention the mud tone in the Arguments, definitely worth noting.



If you read the page, I made an impulse of the same real mic and cab setup. So my own which matches the setup.



I will edit the page to talk about this, since I forgot to add it. Tried that, there are a couple of problems.

1. The time delay between the cab and the mic. So you have to line it up by hand, not a huge deal, but the waveforms don't perfectly line up either. Interestingly, the time delay makes the signal quite a bit out of phase.

2. Even though the signals are the same, the recording times are different so there is inherent clock skew. So you end up with an inflated amount of stuff left over that actually shifts in and out over time.

You are welcome to test it since you get the first set with just the guitars and they are panned hard left and right. But you will have to align the wavforms manually then use a frequency dependent phase shifter to get them in phase, then flip... or just perfectly out of phase. But you will still have the clock skew.

Also on my Dynaudio BM5A and Yamaha HSM80's in my studio I can hear so clearly hear difference between the Impulse and the Cab that it really isn't worth it. On my Roland MA-8's in my untreated office I think I hear differences, but I don't think I could do it blind.

Oh I also tried doing matching EQ with Ozone 4. That matching was weird and added an overbearing amount of mud to the Impulse and an way too much sizzle to the Cab and Mic. But I was learning as I go with the Matching EQ. I tried different parts with different riffs, all sounded bad. Even just looking at the snapshots, there are big (~.5-2db) differences in the low mids, upper mids, and highs.

I was thinking maybe if I sent white noise and then matched, but curves and applied it to the guitar tones, but I don't know. Isn't that what an impulse is? A sampling of all frequencies? Then just a 1024+ band EQ or whatever? So I didn't bother, this gets the point across.

Personally I think if you can hear the difference now, take note of them. Then go into a blind test, you would do pretty well at picking them out. Especially if the same mic setup and sampling were done.

Adding a room impulse to me helps with a lot of that too. Probably the same would be true with just a touch of algorithmic reverb designed to sound like a room and not an effect. Hmm... I wonder how hard it is to sample my room. I think I just need a starter pistol and measurement mics.

Really interesting, thank you for that!