Capture IR

[Bigshredder]

Hi folks
What is the easiest way to make an IR from a mix of your own guitar sound? Let me clarify. I use Reverberate LE on three different channels with three different impulses and I then use the eq in Reverberate on each IR to blend them all together. It now sounds killer but I would like to mix them together and because they have different eq and vol settings I think that I need to capture the sound as a EQ match and then save? I'm I right? I haven't done it before so I thought I might ask here first
I have tried all the IR loaders but the reverberate sounds by fare the cleanest and best. So how to make one IR from your combined IR+eq?

 

[nezvers]

It shouldn't be hard - http://www.voxengo.com/product/deconvolver/

 

[deLuther]

Or http://www.ultimatemetal.com/forum/backline/957890-new-way-cabinet-impulse-extraction.html

 

[Bigshredder]

Or http://www.ultimatemetal.com/forum/backline/957890-new-way-cabinet-impulse-extraction.html

Thanks
Call me stupid but I didn't see you match any tone in that video :/. You played with a dry signal and then opened a nother program and then loaded an IR. I Don't get it ?
My sound comes from my amp and through 3 different impulses. I wanna match that sound so I don't have to load 3 different tracks with an IR loader and an eq. I'll try these methods tonight when I get home. I Hope i works

 

[rapucore]

It shouldn't be hard - http://www.voxengo.com/product/deconvolver/

This. Quite simple.

 

[deLuther]

Bigshredder, just think that instead of preamp signal and miked signal you will have your signal before reverberate (i.e. preamp signal, ampsim out), this will be your source, and after processing, which will be your target.
In videos I`m matching preamp signal to miked signal and compare results, thats all.

Another way do doing what you want: just mix your impulses with the same levels, as you are using in reverberate and on tracks. Convolution is linear process, so you can do it such way, but, in case of additional EQs, that will not work.

 

[FrankTheSmith]

Post a clip of the final tone playing a riff, and a DI playing the same riff and maybe a screenshot of yo ampsims settings, and I'll make one for you.

 

[J.Lindmann]

This. Quite simple.

My impules created with Deconvolver always sound phasey and are lacking bass.
I've never been able to fix that.

 

[rapucore]

My impules created with Deconvolver always sound phasey and are lacking bass.
I've never been able to fix that.

I'm guessing you don't have some correct option box checked. Mine always come out as I intend them.

 

[deLuther]

Deconvolver sweep test tone starts from 40 Hz, so there is some lack in lows.
Also remember that impulse length has huge impact on low end resolution, so too short impulses have overly smoothed lows.
Maybe the best way, in matching case, is ti use source signal as a test tone, process target file, and check Silence option with desired impulse length (if Deconvolver is used), MatcherCL just creates matching impulse from two files directly (length can be specified through command line, if default length is insufficient).

 

[Bigshredder]

Thanks for all the replies guys I'll try the deconvolver. The free version can capture IR right?
Cheers!

 

[rapucore]

Thanks for all the replies guys I'll try the deconvolver. The free version can capture IR right?
Cheers!

It can, absolutely! Let us know if you run into complications.

 

[J.Lindmann]

My impules created with Deconvolver always sound phasey and are lacking bass.
I've never been able to fix that.

Oh....I am sorry...I have to correct myself as I just startet another attempt to create an IR.

I was completely wrong:
My IR do not lack bass...they are overwhelmingly bass heavy.
So the complete opposite of what I stated earlier.

I now compared the frequency response of the Ozone output to the output of LeCab loaded with my IR.
These should look pretty much the same if I understood the whole IR thing correctly.
But as you can see here...these two frequency responses do have some similarities but overall these ain't looking alike at all.
(First one is the ozone output, second one is the IR output)

I watched several IR creation tutorials using ozone and deconvolver.
I copied all settings (cubase settings, devonvolver settings,lengh of the sweep, lengh of the processed sweep, ...)
But I repeatedly can't get decent results...

 

[Vihaleipä]

Oh....I am sorry...I have to correct myself as I just startet another attempt to create an IR.

I was completely wrong:
My IR do not lack bass...they are overwhelmingly bass heavy.
So the complete opposite of what I stated earlier.

I now compared the frequency response of the Ozone output to the output of LeCab loaded with my IR.
These should look pretty much the same if I understood the whole IR thing correctly.
But as you can see here...these two frequency responses do have some similarities but overall these ain't looking alike at all.
(First one is the ozone output, second one is the IR output)



I watched several IR creation tutorials using ozone and deconvolver.
I copied all settings (cubase settings, devonvolver settings,lengh of the sweep, lengh of the processed sweep, ...)
But I repeatedly can't get decent results...

If you're recording actual cabinet impulses maybe your room is the reason why your all impulses are turning very bass-heavy? Try decoupling your cab from the floor with milk crates or something similar, and make sure your cab isn't too close to the corners of your room. Assuming you are making traditional cabinet impulses instead of match eq-style IR's.

 

[deLuther]

J.Lindmann, can you post sound before supposed impulse (source) and after (target), in original quality (32 bits will be better)?
I`m will try to use MatcherCL for impulse creation.

Here my comparison of matching with Ozone5+Deconvolver against my MatcherCL:
http://forum.guitarplayer.ru/index.php?topic=319541.msg7934107#msg7934107 (full post in russian)
Just samples.
Preamp signal (as source): http://forum.guitarplayer.ru/index.php?action=dlattach;topic=319541.0;attach=1069283
Miked cab (as reference and target for matching): http://forum.guitarplayer.ru/index.php?action=dlattach;topic=319541.0;attach=1069284
Result with Ozone+Deconvolver: http://forum.guitarplayer.ru/index.php?action=dlattach;topic=319541.0;attach=1069287
Result with MatcherCL: http://forum.guitarplayer.ru/index.php?action=dlattach;topic=319541.0;attach=1069288
Impulses:
http://forum.guitarplayer.ru/index.php?action=dlattach;topic=319541.0;attach=1069285
http://forum.guitarplayer.ru/index.php?action=dlattach;topic=319541.0;attach=1069286

 

[J.Lindmann]

If you're recording actual cabinet impulses maybe your room is the reason why your all impulses are turning very bass-heavy?

No, all my attempts have been made with ampsims.
But thank you for your input!

J.Lindmann, can you post sound before supposed impulse (source) and after (target), in original quality (32 bits will be better)?

Yeah sure.
I made a little EQ-matching experiment.

This is just overloud TH2:


This is TH2 plus Ozone (the matched tone which I wanted to make an IR from):


And this is what I ended up with
TH2 plus IR:

 

[deLuther]

This is insufficient quality for impulse extraction (material should be lossless, not from mp3).
Ozone have only 8192 analysis block size, while MatcherCL uses next power of two after sample rate (i.e. 65536 for 44k), so better to use MatcherCL straight from TH2 to desired tone and, as result of processing, you will get matching impulse, no any kind of deconvolution required, so no loss of precision between intermediate stages (internal processing performed in double floating point precision).

ps. Also better to match from preamp signal (or without cabinet simulation) to desired tone, to avoid some unwanted peaks on frequency response, because of preamp signal smoothness.

 

[J.Lindmann]

Thanks for your help.
MatcherCL may be a great tool, but my results with Deconvolcer are just terrible.
Shouldn't I receive at least halfway decent results using Deconvolcer?

 

[rapucore]

Thanks for your help.
MatcherCL may be a great tool, but my results with Deconvolcer are just terrible.
Shouldn't I receive at least halfway decent results using Deconvolcer?

Could you take some screen captures of your IR creating process with Deconvolver? That might help us sort out your problems.

 

[deLuther]

Thanks for your help.
MatcherCL may be a great tool, but my results with Deconvolcer are just terrible.
Shouldn't I receive at least halfway decent results using Deconvolcer?

You can try MatcherCL yourself, it`s free
(check third post for link to dedicated thread)

I`m also have troubles with impulse extraction from plugins (in case of matching) with sweep test tone using Deconvolver, while have good results (with minor flaws, because of mentioned test tone flaws) for cabinet impulses.