Creating Spike for an IR

jonnyclueless · 14th November 2005, 12:05 AM

I want to make some IRs of an M5000 I have. From what I understand, it's best to use a digital spike rather than a sine sweep to create IRs with Digital gear (accessed digitally). I created a region in PT one sample length and drew in a spike.

Is this th correct way to do this? It seems to work, but I don't know if there is a better method or not. Anyone have much experience in this area?

js1 · 15th November 2005, 05:30 AM

It'll work. But you may want to use a sine sweep instead. For a few tips, go here:

http://noisevault.com//index.php?opt...pv=0&topic=4.0

js

js1 · 15th November 2005, 05:37 AM

Actually, I should have read your post more carefully. It's not a case of digital gear vs. analog gear. It's more a case of delay/reverb fx (use spikes for these) vs. EQs and the like.

The one sample spike is fine. You didn't mention the amplitude of the spike - I wouldn't go 0 dbfs, I'd probably shoot for -10 dbfs in case there's some gain somewhere.

js

jonnyclueless · 15th November 2005, 06:43 AM

Quote:

Originally Posted by js1

Actually, I should have read your post more carefully. It's not a case of digital gear vs. analog gear. It's more a case of delay/reverb fx (use spikes for these) vs. EQs and the like.

The one sample spike is fine. You didn't mention the amplitude of the spike - I wouldn't go 0 dbfs, I'd probably shoot for -10 dbfs in case there's some gain somewhere.

js

I tried to make the spike as high as it would go because the output of the reverb unti(s) were barely even visible even with the spike at maximum. It seems like it's a bit shy of 0dBFS, but I couldn't draw it any higher anyways. Actually, it's that the output seems so low in comparison to a full level spike that makes me wonder if I am going about it wrong. In the end, the IRs seem to sound fine to me, but I have no way of knowing if they could be better if I don't know th proper way of doing it.

I was told that when recording digitally, to always use a spike instead of a sweep. I'm not sure why, and still wonder if there are some effects where a sweep is more appropriate. If something is a multi effect that includes some kind of EQ, I imagin a sine would be better, no? I am really curious what problems I would have usin a sine, because it seems like it would be better overall to cover every type of FX.

I also wonder if there are instances where recording mono input, even if the device has stero input is better. For instance on a flange, I am likely to catch the sweep at a different part of the cycle when recording the 2nd input. At least thats what seems like a likely result from my brief understanding of how IRs work, but I may be wrong for all I know.

chester · 15th November 2005, 01:24 PM

hi there, what you mean by a spike is digitally called a DIRAC (DIRAC-DELTA-FUNCTION).

http://www.noisetime.com/impulsecreation.html

You can download such diracs in different formats from this page.

Much fun

chester · 15th November 2005, 01:27 PM

Sine-Sweep-Method will give you a better Signal-to-Noise-Ratio if you capture analog devices. In the digital domain its best to go with a dirac. For example via SPDIF or ADAT. If you want to capture analog devices like a platehall or analog eq you will go better with a long Sinesweep (longer than 10s) to get best signal to noise ratio, same for capturing real spaces like a concert hall.

jonnyclueless · 16th November 2005, 12:54 AM

Thanks for the info. So far for analog and rooms I have been using a 30 second sweep with 7 seconds of silence at the end. I'm curious what the problem is with using sweeps digitally? Is it something to do with filtering in the upper frequencies or something?

Going to check out that link now, thanks!

jonnyclueless · 16th November 2005, 11:28 PM

Doesn't seem to be anyone alive at noisevault. Is using a DIRAC ok for pitch type effects? I hught pitch shifting was done by changing the sample rate or something, but with only 1 sample, is this possible? Thanks for any insite!

chester · 17th November 2005, 12:53 AM

its nearly impossible to capture non-linearities like pitch shfifting or distortion for ex. therefore dynamic convolution is needed. but, this is patented by a firm called sintefex.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Creating his OWN Team!!	Solar	High end	8	7th August 2006 07:27 AM
Creating SX screenshots..	farjedi	Music computers	12	20th April 2006 10:46 PM
Help creating sampler instruments	Studiocat	So much gear, so little time!	2	3rd March 2006 12:38 AM
creating rex files	Renie	Music computers	2	21st April 2005 02:59 PM

14th November 2005, 12:05 AM	#1
jonnyclueless Lives for gear Join Date: Aug 2004 Location: Los Angeles Posts: 800	Creating Spike for an IR I want to make some IRs of an M5000 I have. From what I understand, it's best to use a digital spike rather than a sine sweep to create IRs with Digital gear (accessed digitally). I created a region in PT one sample length and drew in a spike. Is this th correct way to do this? It seems to work, but I don't know if there is a better method or not. Anyone have much experience in this area?

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes	Rate This Thread
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode	Rate This Thread:

15th November 2005, 05:30 AM	#2
js1 Gear maniac Join Date: Nov 2003 Posts: 245	It'll work. But you may want to use a sine sweep instead. For a few tips, go here: http://noisevault.com//index.php?opt...pv=0&topic=4.0 js

15th November 2005, 05:37 AM	#3
js1 Gear maniac Join Date: Nov 2003 Posts: 245	Actually, I should have read your post more carefully. It's not a case of digital gear vs. analog gear. It's more a case of delay/reverb fx (use spikes for these) vs. EQs and the like. The one sample spike is fine. You didn't mention the amplitude of the spike - I wouldn't go 0 dbfs, I'd probably shoot for -10 dbfs in case there's some gain somewhere. js

15th November 2005, 01:24 PM	#5
chester Gear addict Join Date: Aug 2005 Location: the whole world Posts: 368	hi there, what you mean by a spike is digitally called a DIRAC (DIRAC-DELTA-FUNCTION). http://www.noisetime.com/impulsecreation.html You can download such diracs in different formats from this page. Much fun

15th November 2005, 01:27 PM	#6
chester Gear addict Join Date: Aug 2005 Location: the whole world Posts: 368	Sine-Sweep-Method will give you a better Signal-to-Noise-Ratio if you capture analog devices. In the digital domain its best to go with a dirac. For example via SPDIF or ADAT. If you want to capture analog devices like a platehall or analog eq you will go better with a long Sinesweep (longer than 10s) to get best signal to noise ratio, same for capturing real spaces like a concert hall.

16th November 2005, 12:54 AM	#7
jonnyclueless Lives for gear Join Date: Aug 2004 Location: Los Angeles Posts: 800	Thanks for the info. So far for analog and rooms I have been using a 30 second sweep with 7 seconds of silence at the end. I'm curious what the problem is with using sweeps digitally? Is it something to do with filtering in the upper frequencies or something? Going to check out that link now, thanks!

16th November 2005, 11:28 PM	#8
jonnyclueless Lives for gear Join Date: Aug 2004 Location: Los Angeles Posts: 800	Doesn't seem to be anyone alive at noisevault. Is using a DIRAC ok for pitch type effects? I hught pitch shifting was done by changing the sample rate or something, but with only 1 sample, is this possible? Thanks for any insite!

17th November 2005, 12:53 AM	#9
chester Gear addict Join Date: Aug 2005 Location: the whole world Posts: 368	its nearly impossible to capture non-linearities like pitch shfifting or distortion for ex. therefore dynamic convolution is needed. but, this is patented by a firm called sintefex.