Impulse Response editing - Do it the right way: IRs for Space Designer

Question

Dear fellow board members,
after doing several research on this topic and many attempts to match some Impulse Responses I finally have to admit, that a little help or insight would be much appreciated. Most of the (Beamsonic) IRs, especially the Studer and Ampeg IRs, seem to have overlength (≈ 6 sec.) and some do have "silence" at the very beginning (see picture below).
I would be thankful for a helping hand explaining me, which part of this IR is of relevant importance: Do I need the "silence" tail at the end? What should I take care of when editing such IRs as the pictured one? Would only the marked yellow area be of importance? Get rid of the red areas (beginning & end)?
As many of the IRs I do have correspond to the image above, I have to go through heavy manual editing and before doing so, your help would be much appreciated.
Does anybody have gone through this before and would like to exchange some of his material for the Space Designer (IRs + corresponding Presets)? I am looking for: EMT 250, Studer A80 Tape, other Tape Sats as well, Ampeg, Westlake Studio D, other useful top quality IRs as well.
For now, many thanks in advance for sharing interest
Vic.
  Mac OS X (10.3.8)  

Answer 1

1. You maybe want to comment on what is considered to be the
"technical right way" to achieve DAW Tape Saturation via IRs, i.e.
what makes sense and what not?
Really, there is no technical right way - all we are doing is altering our recorded audio in a way that we like - it will vary depending on your tools, your tastes, the material, the artistic intent etc...
a - Insert the Convolution Reverb right on the Audio Channel
itself with Pre/Post-Compression/EQing, etc.
I will do this if I want to process an individual channel for colour. Remember to set the effect to 100% wet.
b - Send audio to a Convolution Reverb Bus (and share w/multiple
audio channels), loss of quality, etc?
You're just sending a bunch of channels - in convolution terms, the results as far as I know will be the same if you process 10 channels indivdually and sum them, compared to if you sum the 10 channels then convolute them. Remember, convolution is a static process in most tools, including Space Designer.
How would you judge the value of using "Impulse Responses"
in comparison to "Plug-Ins"?
Well, I don't think anyone's done a tape saturation plugin properly yet. I'm not an expert in this field, but had some interesting discussions with one of the Sonalksis developers over on KVR who is heavily researching into tape saturation (he's doing a doctorate in the subject) and he thinks that it is possible to do a decent tape saturation plugin - not easy by a long stretch, but possible, so I'm sure we will eventually see decent plugs appearing.
Having said that, I'm not actually a huge fan of the tape sound anyway - I like the impulses for that colour if I need it, but I certainly don't run everything through it. I don't really like the smeared transients, at least on my "less natural" pieces. Rock music (as an example) may well work better.
The plugins that do saturation in varying amounts are a mixed bag in my opinion. A lot of people love VintageWarmer, but I'm not that keen on it (again, I'm not an "expert" by a long stretch, these are just my personal opinions).
I've had some great results using Guitar Amp Pro for saturation and warming (we're talking subtle stuff here) on drums and guitars.
There are a bunch of other plugs too, but none that I have used to tape saturation. DUY's tape does saturation, but not really accurate tape simulation *from what I have heard* - I haven't used it personally myself. Charles Dye seems to like it, anyway!
. Finally: Did anybody experiance differences in quality when
comparing Convolution Reverbs (i.e. SD vs Altiverb vs Waves) in
processing the same IR?
In general, no. Mathmatical convolution is afaiu a mathmatically "perfect" process, much like the adding of audio signals in the digital realm. As long as you know how to do the maths correctly, there should be no difference, given the same IR (it's the IR that really determines the quality of the output).
Of course, different plugs implement things differently, and have different features, but by and large, the convolution engines these things use should be generally the same.
Again, I'm no expert and haven't done exhaustive testing or anything - this info comes from discussions I have and my understanding of these processes from a DSP point of view. Other may well be able to back up or debunk what I've said, so I'll leave the floor to them now...