Some crazy idea about MIR: Virtual IR Sampling

[YWT]

Hi, I have been using MIR for quite a long time now. Also have updated to MIR Pro 3D immediately after its release. I now have some strange and crazy idea about using MIR Pro...

I have to admit that although MIR is really great and the workflow is simply amazing, there are some strange stereo width problem in the output. I almost exclusively use MIR in mockup situation and the most difficult problem for me is that I cannot determine the correct stereo width of a sampled instrument. And the most problematic instrument is string ensembles(or simply a section of a string ensemble). If you listen closely, you can even hear the strange width problems/phase problems in the strings in VSL's own MIR audio demos. For me this is a very big problem and I have always tried to solve it. Until today it occurred to me that... Can I do virtual IR sampling in MIR Pro?

In my theory, the best usage of IR is that having the exact IR as the exact sampling microphone setup, and apply them to the sampled instruments dry output. Cause it will just match everything and get the correct result. So if I can get the microphone setup for a sampled instrument, then I use this microphone setup sample the IRs, this should be the most accurate way to get the sampled instruments, especially the ensembles, sounds like in real life in the sampled environment. But for me, there are always different sampled instruments and different microphone setup. So I am thinking that can I make virtual IR sampling sessions in MIR Pro. If I can, then I actually have the room in my hands and can get any microphone setup in MIR(although as sampled IRs, not in the software itself).

Also, I know maybe this is a far fetched idea... but can we get more complex microphone setups in MIR? like wider surround microphone setup, decca tree setup and others. Just like recording in the real life. Even something like direct single/multiple microphone setup and output connections.

I have already done a virtual sampling session here. It's still using the MIR main microphone setup (although a custom one). I have to say the result is very interesting. I will continue try doing some other experiments like doing a wider microphone setup by sample two different main microphone position...

Also I want to ask the team that do they think this is possible for doing virtual IR sampling in MIR Pro.

[Dietz]

Thanks for your interest in MIR Pro 3D, and thanks a lot for your suggestions!

It sounds easy to make something like MIR with other microphone setups than the all-dominant Ambisonics-based concept. As a matter of fact we have tried several other approaches, some of them several times even - but truth to be told: In the end all of them are either inflexible and as such much too rigid for real audio work with all its constantly changing demands (e.g. output formats!), much too demanding regarding the engines's CPU-consumption, much too complicated for the average user - or they simply sound bad. 8-]

Long story short: There are no plans to change the underlying concept of MIR 3D. :-) 

Thanks for your understanding. Enjoy MIR 3D!

Kind regards,

[Beat Kaufmann]

Hello YWT

I don't know what libraries you are working with. But as part of your experiments, maybe you can try a mix without MIR again. If you work with the new SYNCHRON libraries, you might find a preset you like. If you want to be completely free, you can use the "Default" preset with the SYNCHRON-Player to position your violins at any depth in space, for example, using various microphone signal ratios. If you want them to play in the front on the virtual stage, I recommend panning them to the left or right with the pan pot. The further you move them to the left, for example, the more you lose the channel information of the Right channel but, unlike the PowerPanner, the stereo information is not monoized, which - as we know and you observe - leads or can certainly lead to phase cancellations. When instruments play further away, the Power Panner is of course a good choice to narrow the stereo field a bit, which corresponds to real physics.

Such mixes sound different than those with MIR, but certainly no less appealing.
Here is an example showing the TIN WHISTLE from the new World library. All instruments were positioned in the corresponding SYNCHRON player of the instruments, for panning the front instruments (strings) I used the pan-pot - for gluing all things together I used some reverb tail over all in the output channel.
More info here and other VSL demos,

As always, there is often not only one solution to a problem. My approach to solving the problem is completely different from the way you have worked so far. Even if you don't want to try my way of mixing, maybe it will lead to new ideas on your side.


Good luck and a lot of success
Beat