Hi, this may not be a welcome post here, but I mean it with the utmost respect for the incredible technical achievement of VSL and their MIR product line, it is doubtless a milestone and a benchmark in the audio world and unmatched as a tool at this time for it's strengths.
However.....
I have had gripes about impulse responses for some time now. Even MIR. It sounds fantastic, there is nothing else like it out there, to be sure. But to me, at the end of the day, as much as it does place instruments impeccably and the realism is amazing, the result of impulse responses, even thousands of them as in MIR, is 2 dimensional, and flat beyond the first impression of the "moulding" of the instruments due to their virtual placement. I have considered even getting a dedicated PC (as I'm a Mac user) and MIR, but I found that although it's less accurate, mixing with a synthetic (sic) reverb is preferable to me as I feel there is more life retained from the original samples, which impulse responses, for me, just completely kill, every time. Especially in complex mixes, they just mush things together (this is where MIR pretty much avoids this problem due to the variations, but still the 2D problem is there for me). Before the engineers at VSL stop reading here... please allow me to finish my point...
It's a lot to do with preference, of course, however, an impulse response is NOT a real room, it is not really anything close to a real room, it is like a "thumbnail" image of a room, - that is stating the obvious of course - but then again, for this point, it's not.... - it is not even a high resolution photo, let alone actually being there. Furthermore, an impulse response may sound great on simple material at least at first listen there is the perception of realism, but aesthetically, it is very very flat. On complex material it suffers a lot. MIR goes a long way to correcting this, and I believe the core design and philosophy will stand the test of time, but I want to make a suggestion: simply, replace the impulse response based reverb in the engine with synthetic algorithms modelled on the real room places. To be sure: that is perhaps equally or MORE work as was required to organise and gather all the impulses for MIR as it is now, however, technical issues and present computer processing issues notwithstanding, I believe it's ultimately possible. If done to the degree of dedication and technical level as MIR has been created, I believe this could yield an aesthetic result far in excess of what MIR has achieved. Impulses responses are 2D and limited, and as a "mix tool" I believe they will never match up to a good hardware reverb. Somewhere in between lies the answer... Incorporate reverb algorithms of the level of say a Bricasti or a Quantec, or better ;) and then apply MIR engine / design philosophy together with that, I believe you have the makings of a master tool that nobody in the industry will want to be without... What I'm saying is, why not continue to have the same foresight and depth of vision as you had 10 years ago, and which you have followed through on to date. I don't know how it could be done... what kind of work would be involved, but not doubt on first thought, it seems to me it would take at least as much work as the MIR project has itself taken to date to incorporate this kind of technology into the design and function.
It may be a decade away, perhaps longer, but that didn't stop you guys having the vision and foresight to record your libraries as you have, dry and direct, with so many articulations, knowing that the software for both the instruments as well as the mixing software, would come in time and you left it open for maximum flexibility. A brave and intelligent move. It's been more than a decade since you started and look where it is. So I don't think this idea is far off of reality, any more than MIR was in 2001. I believe MIR could really be viewed as a stepping stone, as, presently, neither the software / hardware synthetic reverb platform, nor the MIR platform, with the greatest of respect and recognition of what it *can* do (which is a lot), deliver ideal results each on their own, at least to where it *could* be (for me personally, just my humble opinion). Each method has it's merits, and their drawbacks. Combine the two taking the best from each... ?!
then again perhaps the engineers will think "go to hell!!!" haha. [:)]
