well, thisone is also somehow perverted, sorry.McLilith wrote:What a perversely twisted statement that is.AndrewSimon wrote:Well I don't want to start an argument cause it's pointless but Impulse Reverbs are a "work around" not a "solution" to reverb quest on PC's.
Given the choice between the two I am sure everybody would choose the real thing not the "snap shot taker".
The real thing comes from recording a sound in a real environment, not from using a piece of hardware or a plugin. Guess what? The convolution processors actually make use of recordings taken in real spaces (if you want them to.)
What you are calling the "real thing", is simply an engineer's attempt to synthesize his concept of the real thing. It's not even something taken directly from a real space. That said, you might well prefer the non-convolution reverb, but I certainly wouldn't think of calling it the "real thing."
take care,
McLilith
they are both synthesized.
both techniques use their own algorithms.
the algorithmic reverbs are made of the simulation of natural behaviours. it strongly depends on how the coder is knowledged about physics, accoustic phenomenons and math to set it up.
the convo dev`s have to know about how to "interpret" the ir`s, easy said. doesn`t really matter if they know about physics and how a real room is buildt.
nevertheless, all native convolution reverbs have to use heavily applied reduction maths in their code, to actually make it work on an up to date pc, as "real convolution" would by far exeed the available power of todays cpu`s. that causes a big data loss, and allways lacks of quality.
imo a good algorithmic reverb has by far more advantages vs. a convolution reverb.
that has several reasons:
i nature there is allmost _no_ static room behaviours. in nature mostly all of the relevant facts change constantly, such as temperature, air pressure, moving of objects, etc.
so a reverb sounds natural _if_ the reverb modulates.
it only depends, how it does.
this behaviour simply cannot be done with convolution. only theoretically, but that would cause even more cpu power, than an ordinary "real convolution process", described above.
also there is absolutely no way of doing real stereo behaviour with a 2 channel convolutin reverb.
you at least need 2 stereo or 4 mono ir`s that are, of course, recorded in relation one to another.
this doubles the handling, the cpu and the ram usage.
saying this, i have to admit, that some ir´s are quite good. but that`s "blending" imo, as the production has to be adjusted to the reverb, not the reverb to the production, and imo this would be more useful.
etc,etc,etc ...
