Solved Making impulse responses

  • Thread starter Thread starter danny.guitar
  • Start date Start date
D

danny.guitar

Guest
If I want to record my own impulse responses, how do I do it? I've been searching for awhile but can't find any simple instructions.

I wound up at a noisevault forum and then ended up finding & downloading Voxengo Deconvolver demo but I'm not exactly sure what to do next...:confused:
 
Thanks Glen. I also missed the Help button on the software which explained it briefly but still wasn't sure I was doing it right.

I gave it a shot using the test tone generator that comes with the software and it's hard to tell if it worked. Of course, my room isn't very big but I recorded a direct guitar and loaded up the impulse file into SIR and it did sound more like it was in a room.

The links you gave me were a good read and very helpful. According to the emusician article, I guess I did the "Sweep" method using that test tone generator in the software.

Also, I didn't really think about it at the time, but I should probably have the mic on the stand with it all the way extended and up as high as possible to get the most ambience out of the room, is that correct? Or possibly even pointing away from my speakers/the source of sound to capture mostly reflections?

Contrary to what I had thought, there wasn't really much noise in the resulting impulse response file I made, even though I'm using a kind of noisy mic (Behringer ECM8000) with the gain cranked up.
 
Also, I didn't really think about it at the time, but I should probably have the mic on the stand with it all the way extended and up as high as possible to get the most ambience out of the room, is that correct? Or possibly even pointing away from my speakers/the source of sound to capture mostly reflections?
The way I figure it is you want to stick the mic where you have the sound you want to capture in the IR. Where the ambience you want IS is where you stick the mic.

I have never actually created an impulse, but I know when recording natural verb I often have the mic pointing away from the sound source; after all it's not the source I'm recording, it's the verb.

Do you want the side reflections, the rear reflections, the corner reflections, etc? Do you want them the way they sound in the corner or the way they sound in the middle of the room, etc? Set the niic accordingly, i would think. Or do you want the room in total (from a certian position), in which case an omni could be in order.

G.
 
Yeah I was kind of going for capturing the whole room.

So maybe as far back as possible? Have the speakers on one end of the room, and the mic on the other? Not sure, if for this case, I should point the mic away from the source or not.

I may do a few different recordings from different parts of the room.

Oh and one more quick question which I think might be a cool idea.

Since I have those in-ear binaural mics, do you think a binaural room impulse would be cool? :D It's worth a shot at least.
 
I may do a few different recordings from different parts of the room.
I would think that if one recorded in the room before, and took advantage of the room's verb, that the idea would be to capture IRs at the mic positions that they had already found to work best for certain recording effects (up there for a nice slap, over here for a warm glow, back there for a thick fill, etc.)

If the room is new to you in that regard, then, yeah, I'd probably try a whole bunch of mic positions (and mics) with real music pumping (doesn't necessarily have to be a live band) and see what works for what.
continued... said:
Oh and one more quick question which I think might be a cool idea.

Since I have those in-ear binaural mics, do you think a binaural room impulse would be cool? :D It's worth a shot at least.
I'd think - someone correct this if it's wrong - that you'd have to make sure that the L and R impulses are used on seperate tracks (or panned seperately on a stereo track, if your editor will do that) to basically provide a stereo reverb result.

Again, I'm no expert at this topic, that's only an edumakated guess on my part. There's got to be someone else here who has more experience with making IRs that could speak more authoratively on that.

G.
 
I would think that if one recorded in the room before, and took advantage of the room's verb, that the idea would be to capture IRs at the mic positions that they had already found to work best for certain recording effects (up there for a nice slap, over here for a warm glow, back there for a thick fill, etc.)

If the room is new to you in that regard, then, yeah, I'd probably try a whole bunch of mic positions (and mics) with real music pumping (doesn't necessarily have to be a live band) and see what works for what.I'd think - someone correct this if it's wrong - that you'd have to make sure that the L and R impulses are used on seperate tracks (or panned seperately on a stereo track, if your editor will do that) to basically provide a stereo reverb result.

Again, I'm no expert at this topic, that's only an edumakated guess on my part. There's got to be someone else here who has more experience with making IRs that could speak more authoratively on that.

G.

Thanks, Glenn. You're right about knowing which spots have the right verb for different situations. I'm gonna try a bunch.

This software has a stereo option and I think it will be the same as a stereo reverb once processed and then loaded into the SIR plugin.
 
I'd think - someone correct this if it's wrong - that you'd have to make sure that the L and R impulses are used on seperate tracks (or panned seperately on a stereo track, if your editor will do that) to basically provide a stereo reverb result.

Again, I'm no expert at this topic, that's only an edumakated guess on my part. There's got to be someone else here who has more experience with making IRs that could speak more authoratively on that.

Not speaking from an incredibly experienced standpoint on binaural technology, but having worked sporatically with binuaral heads partly thanks to my time at Full Sail and a couple of others outside of that...

I'm pretty sure most binuaral assemblies like this should be split already with two outputs for both mics respectively.

From what I remember, the binuaral heads I had worked with where essentially two mics inside a fancy molded foam designed to mimic the human head. At the bottom, you had whatever adjustment controls you needed and two XLR outputs.


My only thought about capturing impluses with binuaral setups is to expect it to sound somewhat different than your standard stereo micing technique.

I would expect alot more seperation between your left and right channels with a bigger hole down the center of that stereo image. That could be good for the mix, and it may not. It may give the illusion of having a wider stereo field since your imitating the way the human ear picks up sound.

Again, that just depends on how you go about micing. I would expect it to sound loosely similar to spread pair micing with a sound treated divider between the both mics. It's just different phasing, imaging, seperation, etc.

I've never sampled impluses, but I would expect the EQ response of those mics to affect the quality of your impulses. So if I had the means to capture a room exactly how it is, then I would make the effort to rent reference grade mics.

I still say it's worth a shot. A little more experimentation dosn't cost you anything.
 
Thanks, Lee.

As of now I'm mainly doing it just for fun, but I do plan on using these verbs for some of my recordings. I may even sample a few other rooms in my house like my living room or something.

The 2 binaural mics each have their own channel so I can record each one separately or as a real stereo file.

I think I may try a stereo mic setup using my other microphones and compare.
 
I think that, according to the theory (which seems like a great idea) you want to have the sound source (either the sweep or the impulse) at the location of the imagined musical source. The microphones should be placed at the intended "optimum listening position". The mikes should also be flat-response, omnis (like the Behringer or earthworks)

If you do stereo, just separate the mikes the width of your head.
A binaural head, or similar, might be nice, since it will capture the response as experienced by a room+head system. The whole scheme will only be really accurate if you then listen to music that was recorded in an anechoic chamber, and you listen to the recording+reverb on isolation headphones.

But reverb is the sort of thing that still sounds "good" even when it is pretty corrupted from "realistic" so it would probably sound good with small imperfections to the method.
 
Back
Top