|
Post by gouge on Dec 10, 2016 19:05:26 GMT -6
I've been updating my plug library and am very short on 96k IR files.
specifically chasing freeware. any leads appreciated.
|
|
|
Post by Guitar on Dec 13, 2016 16:12:12 GMT -6
It is quite the opposite of freeware, unfortunately, but Altiverb 7 XL prides itself on just this feat!
|
|
|
Post by tasteliketape on Dec 13, 2016 20:26:46 GMT -6
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 14, 2016 12:24:01 GMT -6
|
|
|
Post by ChaseUTB on Dec 14, 2016 15:59:49 GMT -6
noob hat on for full effect: Not one of my strong suits knowing all the types and kinds. I know what sounds good though 😀 and you guys know I like to learn!
What is an Impulse response? And why would I use and impulse response reverb instead of a D-Verb or Valhalla Vintage Verb or RVerb? How does an impulse response reverb sound compared to an algorithmic reverb ( Im assuming Dverb and RVerb are algorithmic.. )
Is an impulse response reverb different from a convolution reverb?
When do you choose impulse response reverbs versus RVerb?
How does the workflow differ from impulse response reverbs to RVerb/ Dverb?
|
|
|
Post by gouge on Dec 14, 2016 16:30:14 GMT -6
In simple terms the convolution program is the software that effects the audio with the impulse response.
The impulse response is the measured response from an impulse passed through a piece of gear or into an acoustic space/instrument.
Other than that it's over my head.
|
|
|
Post by bradd on Dec 14, 2016 16:41:24 GMT -6
I've always viewed impulse responses as similar to samples. With an impulse response, you are taking a "sample" of the room versus an algorithmic recreation of spaces that you have with D-Verb or Valhalla.
|
|
|
Post by gouge on Dec 14, 2016 16:48:21 GMT -6
Also in general terms to me an ir reverb sounds a lot better than an algo.
|
|
|
Post by Guitar on Dec 14, 2016 17:03:04 GMT -6
After listening to reverb for a long time, you can pick up that "digital reverb sound" of a lot of typical hardware boxes and algo plugins. It's a very common sound that does have its place in music, even if it is artificial.
A convolution reverb using a room sample (impulse) sounds a lot more like recording in actual space. It sounds more "realistic" like real ambience. Since that's what it is modeling.
There are other types also like spring and plate, that have their own specific tonalities. These can be made into impulse responses also.
My favorite use of IR reverb lately, which feels like cheating, is to take tracks recorded in small heavily treated rooms (dead acoustics), and try to fool the listener into thinking they were done in a larger studio setting. Blend the instrument sends into this "room" and get a good balance and the mix can sound a lot larger. Sounds more like you set up a second mic further back in a better sounding room. Can also be good to get virtual instruments sounding more realistic. Or crazy amped sounding drum machines.
My dead, dry tracks can make a sort of claustrophobic rough mix sometimes, especially when everything is close miked. Using IR feels like a cheat or a band aid but it's one I sometimes am desperate for.
In a fantasy life, I'd have a bigger more lively room and it would be carefully treated, but right now that's not what's on my plate, so I have to use other tricks to get "bigger" sound.
Of course this trick is probably useful for a lot of other types of recordings, I'm sure you can use impulse response reverbs wherever you would use any reverb, it's just another flavor in the palette.
I'm sort of a late adopter of convolution verbs, but I've become a big fan.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 14, 2016 18:37:58 GMT -6
The impulse response file is a kind of extract of a signal processing, that can be mapped on a new signal that is mathematically prepared from a defined input signal and a processed signal sample. Most of the time, the defined source sample is a dirac function impulse. It is sent to a room or signal processing and the resulting signal is sampled. Then you let a deconvolution program calculate a kind of transfer function file that can be used to map to a new signal. Then it can be fed to a convolution reverb which then "maps" this characteristic to any new signal. Convolution is used in many fields of signal processing and got interesting for music the time computers got enough power for realtime processing. The comvolution and de-convolution processs requires mathematical operations that quite some computation power, involving calculations like integrals and fourier transforms etc.. There are different methods of convolution and deconvolution with different computational effort. I do not want to go deeper in the math, it is not trivial and the interested people could google up convolution and will get plenty of mathematical explanation.
The convolution reverb takes your signal and maps the reverb/echo from a room or machine in form of the impulse response file on it. This costs considerably more calculation power than most algorithmic reverbs, but you need the room or reverb unit only one time, make your dirac sample and after preparing the file with deconvolution you can use it on any signal with the convo-reverb. Results can be pretty impressing. Reverb/echo is a good application for convolution.
But still, the number of aspects of the room or unit, that can be taken by a convolution reverb are limited. You actually take the information from one sample at one volume level, several aspects like nonlinearities with different dynamic behaviour can not be replicated with this "simple" kind of trick. There are more advanced methods that take series of impulses so signals with different dynamics can get a different mapping. The Acustica Audio Nebula engine plugins use such a method and can produce even more realistic replications of rooms and signal processings. At the cost if even more processing power demands. But for some applications the results are not only impressing, they can be undistinguishable. Up to the point where they are identical. But also, not all aspects can be replicated in some cases, depending on the kind of processing and it's complexity.
For the replication of room reverb of real rooms, results are most of the time better than algorithmic reverbs. But in many cases you might not want a realistic reverb, but one, that gives a distinct color or smear or characteristic that suits the music better than a real room does. So a convolution reverb does not render algorithmic reverbs obsolete. Using the different kind of reverbs on the right occasion can be the key to a great mix. Say, you give the lead vox a lush old Lexicon algo verb that makes exactly the imperfections (in comparison to a real room) that we want to hear on the voice to make it mighty, and use a real room for the drums in a convolution reverb and maybe you sent all signals except maybe bass and kick to a reverb bus and build up a stereo image and room depth impression that you like...
|
|
|
Post by ChaseUTB on Dec 15, 2016 18:22:49 GMT -6
Thanks for the in depth resonse. Definitely more detailed and contains lots of good info to soak up. Any opinion on liquidsonics Reverberate 2? I have been trying to find some opinions regarding convolution reverbs and didn't really know another besides altiverb I can't afford.
|
|
|
Post by tasteliketape on Dec 15, 2016 18:42:27 GMT -6
I bought Reverbarate2 on Black Friday haven't used it a lot yet but first impressions is I think it's quite good , a lot of perimeters to tweak . The Bricasti samples sound better to me than the simplicity Bricasti samples I had in TLSpace. Jmo Edit: I demoed the Hofa reverb and it was pretty killer ,but they didn't have a Black Friday sale .
|
|
|
Post by gouge on Dec 16, 2016 18:33:28 GMT -6
|
|
|
Post by gouge on Dec 16, 2016 18:34:51 GMT -6
|
|
|
Post by Guitar on Dec 17, 2016 10:40:26 GMT -6
nice links! I'm gonna grab 'em
|
|
|
Post by gouge on Dec 18, 2016 1:34:03 GMT -6
|
|