Has anyone managed to use Voxengo Deconvolver w/o MP?

[jerotas]

I have not been able to make an IR without using the Minimum Phase option. They're all invalid according to Jay Mitchell. But then other people are saying that using minimum phase destroys a lot of what I was trying to capture. The minimum phased I made IR's sound good, mind you, but I would like to hear what they sound like without MP.

I'm open to using a different program than Voxengo if I have to.

Thanks!

 

[Jay Mitchell]

I have not been able to make an IR without using the Minimum Phase option. They're all invalid according to Jay Mitchell. But then other people are saying that using minimum phase destroys a lot of what I was trying to capture.

Given that Cliff applies a minimum phase transform to the factory IRs, I wouldn't say that it necessarily destroys very much.

As a test, I deconvolved a 12 second log sweep with itself using Voxengo and got the expected ideal IR. The thing you may be missing is that the deconvolution process Voxengo employs adds quite a bit of leading silence. You need to remove all this with a wav editor before converting the IR to sysex.

 

[jerotas]

Given that Cliff applies a minimum phase transform to the factory IRs, I wouldn't say that it necessarily destroys very much.

As a test, I deconvolved a 12 second log sweep with itself using Voxengo and got the expected ideal IR. The thing you may be missing is that the deconvolution process Voxengo employs adds quite a bit of leading silence. You need to remove all this with a wav editor before converting the IR to sysex.

Thanks for the reply Jay. I wasn't doing any trimming. You're saying I should open the deconvolved (with minimum phase) wav file and trim the silence at the beginning before using the Axe-Fx IR Convertor on it, right? There's about a 1/2 second of silence in my deconvolved wav files.

 

[AlbertA]

I have not been able to make an IR without using the Minimum Phase option. They're all invalid according to Jay Mitchell. But then other people are saying that using minimum phase destroys a lot of what I was trying to capture. The minimum phased I made IR's sound good, mind you, but I would like to hear what they sound like without MP.

I'm open to using a different program than Voxengo if I have to.

Thanks!

On a minimum phase system you can derive the phase response from the frequency response as they are hilbert transforms of one another.
As far as I understand, and Jay may provide more insight into this, speakers are largely a minimum phase system. Hence the phase response is derived when applying the minimum phase transform to an impulse response, will deviate from the actual measurements, but won't completey destroy it.

Jerotas,
You have to zoom very close in the waveform and trim as close to the big peak as you can (try to choose the closest point to 0 that you can find close to the peak).

 

[jerotas]

On a minimum phase system you can derive the phase response from the frequency response as they are hilbert transforms of one another.
As far as I understand, and Jay may provide more insight into this, speakers are largely a minimum phase system. Hence the phase response is derived when applying the minimum phase transform to an impulse response, will deviate from the actual measurements, but won't completey destroy it.

Jerotas,
You have to zoom very close in the waveform and trim as close to the big peak as you can (try to choose the closest point to 0 that you can find close to the peak).

Ok, I did what you're describing last night to a wav file that had been deconvolved with minimum phase turned off. When I opened that wav file in Albert's Axe-Fx IR Convertor, it was obviously invalid. The phase graph looked like a 500bpm heartbeat. If I don't trim the silence, it's still invalid. I have to use Minimum Phase unless there's some other trick.

So like I said, has anyone been able to make IR's without Minimum Phase? Or should I just forget about it and use MP and not case about the deviation from actual measurements?

 

[AlbertA]

On a minimum phase system you can derive the phase response from the frequency response as they are hilbert transforms of one another.
As far as I understand, and Jay may provide more insight into this, speakers are largely a minimum phase system. Hence the phase response is derived when applying the minimum phase transform to an impulse response, will deviate from the actual measurements, but won't completey destroy it.

Jerotas,
You have to zoom very close in the waveform and trim as close to the big peak as you can (try to choose the closest point to 0 that you can find close to the peak).

Ok, I did what you're describing last night to a wav file that had been deconvolved with minimum phase turned off. When I opened that wav file in Albert's Axe-Fx IR Convertor, it was obviously invalid. The phase graph looked like a 500bpm heartbeat. If I don't trim the silence, it's still invalid. I have to use Minimum Phase unless there's some other trick.

So like I said, has anyone been able to make IR's without Minimum Phase? Or should I just forget about it and use MP and not case about the deviation from actual measurements?

Jerotas, I'll work on the set that you sent me, and returned the trimmed versions to you.

 

[jerotas]

Jerotas, I'll work on the set that you sent me, and returned the trimmed versions to you.

Maybe your deconvolving program is different (actually can product valid files without MP turned on). Or...are you using Voxengo? If you are, I don't have any reason to believe that your trimmed versions will be an improvement...unless I'm missing something. Why don't you just send me 1 so I can hopefully become a believer and ask more questions to become enlightened

The ones I sent out have a 1/2 second of silence at beginning and end of the sweep recording, FYI.

 

[AlbertA]

Jerotas, I'll work on the set that you sent me, and returned the trimmed versions to you.

Maybe your deconvolving program is different (actually can product valid files without MP turned on). Or...are you using Voxengo? If you are, I don't have any reason to believe that your trimmed versions will be an improvement...unless I'm missing something. Why don't you just send me 1 so I can hopefully become a believer and ask more questions to become enlightened

The ones I sent out have a 1/2 second of silence at beginning and end of the sweep recording, FYI.

I'm using voxengo demo. You have to trim the silence after deconvolution, getting as close as you can to the peak, this requires your wave editor to be able to zoom into the waveform (I usually zoom until I get to the sample points).

I'll send some that I've trimmed.

 

[jerotas]

I'm using voxengo demo. You have to trim the silence after deconvolution, getting as close as you can to the peak, this requires your wave editor to be able to zoom into the waveform (I usually zoom until I get to the sample points).

I'll send some that I've trimmed.

That's what I tried last night. Opening it after trimming shows a quite invalid response (phase). Aaarrrrrrgh!