Vienna Symphonic Library Forum
Forum Statistics

181,616 users have contributed to 42,178 threads and 254,543 posts.

In the past 24 hours, we have 3 new thread(s), 11 new post(s) and 38 new user(s).

  • Philosophy of wet / dry with MIR

    Hi everyone.

    I did a search on wet / dry threads in here so apologies if this has been covered.  (incidentally can't search for just wet, or just dry due to 4 character minimum limit on search.  D'oh!).

    OK.  So, I'm thinking that since MIR is fundamentally a convolution reverb that is designed to emulate / simulate a real room, then the whole idea of dry is not really comprehensible.

    In a real room, you don't get to adjust the mix of signal that is subject to the room vs signal that isn't.

    I know every reverb since time immemorial had a wet/dry knob.  So is the wet/dry fader in MIR just a sop to appease audio engineers?

    The dry signal is such a low volume compared to the wet one that it's basically drowned by it anyway.  I suspect the dry signal is actually just a detraction.

    Is it therefore valid to mix with 100% wet?  That then sounds like a room?

    Or is the idea one of room vs close mics?  Problem with that is that depending on your instruments you may or may not have this from the source.

    Cheers all.

  • last edited
    last edited

    Welcome 169127,

    An interesting an important question. 😊 The good news is that the answer is quite logical and easy to understand.

    Of course an impulse response from a real room will almost always contain at least subtle remains of the direct signal - i.e. the part of the source signal that reaches the ear / the microphone first, without being reflected by any surface. Early convolution reverbs indeed presented this part of the signal directly in the IR, so the convolution would "sound like the room". MIR took this concept several steps further, using multiple directional IRs to represent the interaction on an instrument with a room in a previously unachievable complexity.

    The downside of this concept: There is no "ideal" impulse generator in our Real World, only good loudspeakers with little, but still evident phase errors. If MIR would simply mix the IRs including the direct signal component, ugly phasing artifacts would be the result. Not good.

    But audio engineers use spot microphones close to individual instruments or ensembles since a long time, to tighten up the sound they get from their room mics. MIR makes good use of this typical workflow: We removed all remains of the direct signal from all our IRs, which gives the possibility to mix in the "perfect" spot mic to any amount, in form of the original input signal: It gets ENcoded to Ambisonics first, to match the original position of the IRs, and it gets DEcoded with the pre-processed IRs according to the chosen Output Format. Like that, no phasing will occur, and the virtual spot mic always matches the timing of the Main Mic (... which is a big issue in Real Life).

    .... long story short: Full wet is _never_ the "real room" in case of MIR. The default dry/wet setting of the Icon on a Venue's stage tries to mimic the original ratio to a certain extent, as long as the option for "Distance Dependent Scaling" is ON for the dry signal's volume handling.

    If you want to know more about it, there's a little primer covering MIR's ideas and concepts available on this site: -> Think MIR!


    /Dietz - Vienna Symphonic Library
  • OK that's very helpful.  And it explains why the dry signal is so low - it's the portion of the original IR that was the direct signal?

    I guess this means in any instrument we should try and get the equivalent of the close mic as the source.  Many libraries of course have a multitude of more distant mics, we should basically take these down and choose only close mics, because Mir emulates the remote mics in the room that MIR is emulating.

  • Exactly. There is no direct signal component present in MIR Pro's IRs, although sometimes the 1st reflection may give you the impression, especially for those IRs that were directed towards the floor. MIR replaces that part with the readily positioned, phasing-free version of the dry input signal - the "perfect virtual close mic".

    Hint: As soon as you activate the option to use the chosen Instrument Profile's directivity filtering on the dry signal, too, you can indeed seemingly "turn away" the instrument from that virtual close mic. :-) Just be aware that this might change the sound of the dry signal component considerably.

    And yes, it's best practice to use the least reverberant mics of any recording for placement and spatialisation within MIR.

    /Dietz - Vienna Symphonic Library