DIY Impulse Responses

[Jake_JW]

I've been playing around with a couple of old hardware reverb's and some other bits and bobs making impulses for reverb for a while now, is this something anyone else has had a go at/is interested in? They always need a lot of tweaking (specifically taking the room size/decay time right down in reverb plugins) but they seem to sound pretty swanky.

It's got to the point now where I very rarely use the impulses/presets supplied with Cubase.
I'll post a couple of examples for people to play around with later if anyone is interested, I'd be interested in what people here think.

 

[pikingrin]

That would be cool, I've been thinking about trying to make some of my own for a long time just never gotten around to it. I got a tascam portable recorder quite a few years ago for that exact purpose but have only used it for everything else...

What kind of equipment are you using? Are you actually micing up live spaces or are you just using the old reverb units?

 

[Jake_JW]

I've tried both - micing spaces is the devils own work. Firstly, you need a measurement mic (not too expensive, but I'm not buying a mic for this one purpose). I have tried various LDC's and SDC's but always with poor results. Secondly you need a space that is actually worth recording. Nearby I have a couple of cathedrals that are worth it (Winchester cathedral especially) but I haven't seriously looked into the reality of getting permission to record there. It's on my list.
Thirdly you need a means of emitting either a sine wave sweep, or a tiny blurt of white noise, at maximum fidelity and with a flat frequency response and enough sound to actually get the room to make a decent amount of reverb without driving a preamp hard (and adding noise/saturation).
Then you have the added bounus of trying to create a stereo/surround sound recording of the thing and it gets very painful.

My own work has only really yielded results using hardware units, I have a few old ones lying around, a yamaha and a lexicon at the moment, both early digital.
Mostly, I find that the plate/early reflections (ambience) in these units absolutely trash what comes with Cubase (and Logic) - and combining that with a little eq/saturation before loading the impule into a convolution creates some really nice effects.

 

[jonny deep]

I've tried both - micing spaces is the devils own work. Firstly, you need a measurement mic (not too expensive, but I'm not buying a mic for this one purpose). I have tried various LDC's and SDC's but always with poor results. Secondly you need a space that is actually worth recording. Nearby I have a couple of cathedrals that are worth it (Winchester cathedral especially) but I haven't seriously looked into the reality of getting permission to record there. It's on my list.
Thirdly you need a means of emitting either a sine wave sweep, or a tiny blurt of white noise, at maximum fidelity and with a flat frequency response and enough sound to actually get the room to make a decent amount of reverb without driving a preamp hard (and adding noise/saturation).
Then you have the added bounus of trying to create a stereo/surround sound recording of the thing and it gets very painful.

My own work has only really yielded results using hardware units, I have a few old ones lying around, a yamaha and a lexicon at the moment, both early digital.
Mostly, I find that the plate/early reflections (ambience) in these units absolutely trash what comes with Cubase (and Logic) - and combining that with a little eq/saturation before loading the impule into a convolution creates some really nice effects.

You can find lots of good impulses for free online as well. Not suggesting you shouldn't be doing what you're doing, but for those units and spaces you can't produce yourself, there are lots out there for the taking. You could consider building a library of your own impulses up and making them available as well.

 

[Jake_JW]

You could consider building a library of your own impulses up and making them available as well.

I could...

But I find that there is a certain satisfaction to using something that isn't widely available and that I made myself (call me selfish, you wouldnt be wrong). Experimentation is fun in itself too - a very interesting area, especially when you get to sampling impulses from more esoteric gear and using it as something other than what the origianl unit was intended for.

In any case, I have a certain sound I like with reverb that I haven't been able to find anywhere else, particularly with plates most of what's out there isn't what I'm looking for.

 

[TripleM]

A while back I spent a couple of minutes trying to just mic up stuff. In listening to the wav impulse files themselves, it seemed like something you could do easily yourself. So I tried things like tapping a glass jar with a pencil and such (I like short, dry reverbs). But when I tried to use them, they sounded horribly muddy. So you need something that is incredibly crisp sounding. After 10 minutes I decided it was a heck of a lot easier to just use the impulses I DL'd.

I may try again some day. But it's harder than you might think.

 

[ShortySMC]

I've been doing this too... for a long time.... (I use Waves IR-1 to do this now... I've like the responses better than when using Altiverb)

I haven't actually used any hardware... but plenty of actual spaces.

I've thought for a long time that it would be cool to start and host a collection of responses.
I use an FTP service to transfer files to clients... I could host them there and make a public link, if anyone is interested in sharing/contributing.

I've got some new spaces I'm wanting to get out and capture this summer as well...

 

[Jake_JW]

Sure, I'd be interested.

Sampling hardware is SUPER easy - I've been using voxengo deconvolver and a hardware unit, that's it.
I'm not sure if other software to do the same thing is available, but since it worked I didn't check.

As long as its static (no modulation effects, so delay and chorus/flange etc are out of the question) sampling your hardware is very straightforwards. What i've found though is that processing the result BEFORE turning it into an impulse response leads to some pretty unique results.