meaning of instrument wet/dry?

[Dewdman42]

what is the exact function of the wet/dry slider in the instrument section of MirPro (VEP)?  Does fully dry mean no spatialization as if it never even hit the plugin at all?  Why would we ever want to have it on anything other then 100%?

The wet/dry slider in the output section is a seperate thing...what is the exact function of that?

And the non-VEP MIRPro plugin..it has a wet/dry slider on the plugin window itself, and there is no instrument wet/dry slider inside the main UI.  So what does that slider do?  is it different in function, even slightly, then the one in the VEP version under the instrument section?

I know some of this is also related to the directivity feature...so if someone could please explain how all of the above works in relation to each other, I would appreciate it.

In normal use it sorta seems to me like the normal case would be to put MIRPRO plugin on every track and the instrument slider ouught to be set to 100%, meaning spatial placement is entirely handled by MIRPRO.  Doesn't make sense to me that the default would be 50%, but maybe someone can explain all of this in a way that will make sense?

There is BONE DRY...as in no mir spatial processing before it hits MIR, then there is SPATIAL DRY, which is the spatial placement of each instrument based on the IR's and its location...(which presumably effects EQ and early reflections), and then there is the ROOM WET sound, which is what is picked up by the virtual mic arrays.  

To me it seems like the default should be to block the BONE DRY sound, and get the SPATIAL DRY sound and then various amounts of ROOM WET sound......I guess?  

Anyway, if anyone understands exactly...like Dietz...would appreciate clarification.  I already read all the manuals, it still wasn't absolutely clear to me.

[Dietz]

It's less complicated than it might seem at the first glance, although I admit that the parameter labelling might be misleading. 😊

There are only three "states" of a signal which are of interest for MIR Pro:

- "Wet" is the sound of the Venue derived from the pre-rendered IRs for an Icon/Instrument (or any other signal source) on a MIR stage. - IMPORTANT: The direct signal component (i.e. the part of the impulse that reaches microphone first, before any reflection from the wall/floor/ceiling) has been cut away from all IRs, to avoid the ugly phasing artifacts that would occur when mixing all the IRs from a chosen position.

- "Dry" is the readily positioned and pre-processed version of the input signal which is put into the place of that cut-out direct signal component mentioned above. This means that the all-important Ambisconics-processing has already taken place, plus width, rotation and all chosen ways of pre-filtering, like Instrument Profile, Air Absorption, Character and so on. - IMPORTANT: This is NOT like bypassing MIR!

- "Bypassed" would be just that. 😊

The Global Wet Dry Offset in the Output section technically controls the same feature like the individual, Icon-based crossfader, but for all Icons on the MIR stage at hand. - IMPORTANT: This offset does NOT allow for "more than 100%" or "less than 0% wet" on individual Icons, for obvious reasons. It's an offset, not a multiplicator.

... it might be worth emphasizing again that "Full Wet" is NOT the sound you would hear in the hall, as the (often very quiet) direct signal component would be missing.

If you want to dig a bit deeper, there is a little add-on to MIR Pro's manual, called "Think MIR!" available in the donwload-section of your MyVSL user area on this site.

HTH,

[Dewdman42]

K this is what I think I’m understanding from what you have said, please correct anything I still have wrong. [list][*]unless mirpro is completely bypassed, then the original totally dry signal is cut out and not available. [*]each instrument has spatial processing applied to create a spatial dry signal which contains the instrument profile eq, width, pan, etc. This would NOT have any early reflections as part of spatial location, presumably. And I guess it is not emulating a direct close mic at the instrument either.. It is the direct sound of the instrument as if it were going through the air some distance to the main microphone array before the room has added reflections. Presumably this dry signal is delayed a bit to represent time through the air to the main mics? [*]A wet signal is also generated for each instrument which contains all reflections including early and later reflections and reverb tail of the room, it’s all part of the wet signal. [*]each instrument has a wet/dry control, on the icon as well as in the instrument area of the main mir window; which adjusts the balance between the spatial dry signal mentioned above and the reflection signal of that instrument. [*]the output section has a global wet/dry which is the same as the instrument but is just a global offset, could think of it like a VCA I guess. [/list] Questions: [list=1][*]if we have more then one virtual mic array present for the room, which one is contributing to the spatialized dry signal? [*]the non vep mir plugin has a wet dry slider in the actual plugin. Is this still functioning as mentioned above or is it allowing the bypassed dry signal to be mixed in somehow? [*]I presume that if I wanted to combine the room sound of mirpro with a close mic sound, I would need to split off the totally bypassed dry sound in the daw or vep before it hits mirpro and handle it as I wish through other plugins, while mirpro would always produce a mix for the instrument that is effected with instrument profile eq, pan, width (dry) mixed with all the room reflections (wet). Note that if I were trying to do this i would probably want wet set to 100% to avoid phase problems between the close mic and the spatially dry signal of mirpro. [/list]

[Dewdman42]

Is it fair to say that a 50% dry/wet setting for each instrument and for global output is generally what we would have expected the main mic position to hear if we were actually there in the room rather then virtually based on ambisonics? and that any setting other then 50% is basically to isolate more of the spatially dry or more of the reflected room, in a way that couldn’t be done in real life but still might be useful?

[William]

"It's less complicated than it might seem at the first glance..." 

... for Dietz. His mind is wrapped around this fully (I heard he took the Ambisonics processing treatment himself).

But I am still trying to figure it out  - this is related to my incomplete thread about "pure" MIR.

Are these correct statements? 

1) The "Dry" is the input signal processed for positioning, EQ, etc. but has no audible reverberation from the hall.

2) The "Wet" is the input signal that has gone through the processing to make it resemble what the impulse did when reverberating in the hall, and is purely the reverberation of the hall with the positioning characteristics of an impulse in that specific spot (which distinguishes it from a hardware reverb that just slathers the same reverb over everything and uses the dry signal to establish positioning).

3) The mix of the two on an individual instrument is somewhat similar to a hardware reverb mix of dry/wet though each component is far more processed to begin with.

[Dewdman42]

I have some questions too about what specifically is contained in the mirpro dry signal.  What specifically is being done there to mimic the spatial location?  Is the "depth" taken into consideration in terms of delay to the virtual mic array?  Or is it just applying the profile EQ, pan and width?  The further away a mic is from a mic, the more that lows and highs are rolled off...does MirPro do this to the dry signal depending on the location of the icon in the room relative to the mic array?  is the sound delayed, per the distance?  What if you are using more then one mic array at different distances?

It seems like it would be a little advantageous to have early reflections isolated from the rest of the reflections.. in order to help with spatial positioning yet still allowing us to dial in how much of the later complex reflections we want in the mix.  But if i am understanding correctly, ER's are in the "wet" component, along with all the longer reflections.  Do I have that right?

[Dietz]

Interesting questions! It will take some time to answer them thoroughly, though, so please allow for little delay until tomorrow. Thanks for your patience!

[Dietz]

Here we go! 😊

@Another User said:

I presume that if I wanted to combine the room sound of mirpro with a close mic sound, I would need to split off the totally bypassed dry sound in the daw or vep before it hits mirpro and handle it as I wish through other plugins, while mirpro would always produce a mix for the instrument that is effected with instrument profile eq, pan, width (dry) mixed with all the room reflections (wet). Note that if I were trying to do this i would probably want wet set to 100% to avoid phase problems between the close mic and the spatially dry signal of mirpro.

This would be a screnario where you would (ab)use MIR as some kind of glorified AUX-send based reverb engine. You would have to set MIR to 100% wet. - There are rare occasions where this might make sense in an artistic way of doing things (... mixing drums for a typical pop music track comes to mind), and in these cases your approach makes sense. Don't expect anything close to the realism MIR is able to achieve when used "properly", though. 😉

[Dietz]

@Another User said:

and that any setting other then 50% is basically to isolate more of the spatially dry or more of the reflected room, in a way that couldn’t be done in real life but still might be useful?

The answer is: "Yes, as long as you don't misunderstand "Real Life" as a single value".😊  The default values are more like an approximation to a "believable" acoustic impression of a hall. Deviations from this default value can make sense to a surprisingly large degree.

HTH,

[Dietz]

Hi William!

@Another User said:

3) The mix of the two on an individual instrument is somewhat similar to a hardware reverb mix of dry/wet though each component is far more processed to begin with.

Yes, but please keep in mind that the Dry/Direct signal component in MIR Pro is already processed, too (unlike in case of your typical reverb engine, which will leave the source signal untouched (hopefully!)).

[Dietz]

@Another User said:

in order to help with spatial positioning yet still allowing us to dial in how much of the later complex reflections we want in the mix.  But if i am understanding correctly, ER's are in the "wet" component, along with all the longer reflections.  Do I have that right?

Yes, exactly.

[William]

That is a great reply, thanks for that info Dietz.  

[Dewdman42]

Thanks for the clarifications. One thing is still not clear to me and you referred me again to the document which I have already read and still do not understand exactly what is in the mirpro dry signal. Above you said the dry signal can be thought of like a spot mic but if that is the case then the dry signal does not have anything to effect perceived stage depth, which is part of spatialization. It could be panned the way a spot mic can be panned in the mixer I suppose, but I would not regard that as having been spatialized. Any clarity on what exactly is done to the dry signal? If they are truly non spatialized spot mics, then I am not sure I would want 50% dry/wet at all. The reason I am asking these questions is because I recently watched mike verta’s Masterclass about template balancing, which was quite interesting for me. He goes through an approach to using some basic plugins such as eq and reverb to accomplish 3D panning on the stage, ie spatialization. His results are impressive. I have invested in MIR Pro and intend to use it but I am struggling to see how any of the mirpro dry signal would not ruin spatialization. If MIR Pro is primarily providing spatialization through the room IR’s then shouldn’t the slider most likely be closer to wet ? Of course if we go closer to wet it usually sounds like too much tail also.

[Dewdman42]

Alright, reading through some old posts on this forum...I think I have come to the following understanding about the mir-positioned dry signal (please correct errors):

1. Its been stated that the dry signal should be thought of like a close spot mic.  In the real world, a close mic does not have distance information in the signal, nor panning, etc..  so its a bit different then a close mic in that regard I think?
2. If Air absorption is checked, then a hi shelf EQ will be applied to the dry signal to simulate the air dampening some high frequency, increasingly based on distance to the main mic.
3. If distance dependent scaling is checked, then distance from the main mic determines the level also, for sure to the dry signal, not sure if this effects the wet level also, I think maybe so, the IR's will pick up the sound from the instrument, based on the instrument profile and directivity profile and its position in the room will matter a lot in terms of what ends up at the main mic.
4. If directivity filtering is on, then the direction of the player's nose can effect many aspects of the sound (how the sound will be effected exactly is unknown to us other then through trial and error), including level and maybe EQ?  The directivity profile is always effecting the wet signal, but it only effects the dry signal if this box is checked.
5. Each instrument can have a character profile that applies some factory black-box EQ...the exact curves unknown to us.  The character profile uses the instrument profile in some way, and does lots of "magic stuff".
6. Dry/wet slider for each instrument adjusts the level of the close spot mic dry signal, which also happens to have some of the other elements mentioned above, which infer distance from the main mic, even though a real world close mic would not have that distance information.  The wet signal includes the entire room, and stuff like air absorption, directivity, etc..would inherently be in the room IR's, which the wet/dry slider can be used to suit taste.

So more questions..

1. is LR panning contained in the mir-positioned dry signal (unlike a real world close mic) or is it expected that we pan the channel in the mixer (like we would with a real world close mic)
2. In Mike Verta's masterclass on orchestral in the box mixing, he talks about needing a low-shelf also to emulate distance from the main mic.  Mike Verta in his videos did an example with and without this rolloff, feeding into a typical convolution reverb.  The one without the low shelf sounded like a PA speaker at the end of the hall and the one with the shelf sounded more natural.  His claim is that with increased distance, some low frequency is lost over distance.  So if you're trying to place some instruments back in the mix, in terms of depth or distance from the audience, then using a low-shelf seems to be his solution.  In his videos where he demonstrated with and without that low-shelf, the differerence was very clear to me.  What does MIR do in this regard?  I notice there is a black-box "Distance" character profile, which might be doing that with "magic", but its unclear.  Also does the icon distance from the main mic effect how much the low end would be rolled off (less proximity effect).  What can you say about this topic or how MirPro handles it in the dry signal?  The wet signal will presumably have that taken care of by the IR's.

That's all I have for now.  My conclusion is that the mir-postitioned dry signal is not really emulating a real world close spot mic.  I understand why you're kind of thinking about it that way, but since there is the possibility to pan it, air absorb it, distance scale it, directivity filter it...and possibly the low end is being shelved off based on the icon position too, I'm not sure..  well then its not really a close mic after all...other then the time alignment that is being done to make sure we don't have phasing problems.    Am I understanding correctly at this point?

[Dietz]

Hi Dewdman,

You keep me busy! 😄

@Another User said:

I understand why you're kind of thinking about it that way, but since there is the possibility to pan it, air absorb it, distance scale it, directivity filter it...and possibly the low end is being shelved off based on the icon position too, I'm not sure..  well then its not really a close mic after all...other then the time alignment that is being done to make sure we don't have phasing problems.    Am I understanding correctly at this point?

I think so. Enjoy MIR Pro!

Kind regards,

[Dietz]

For your convenience, the following except from legacy Vienna MIR's manual contains the Character Preset typology I mentioned above:

-------------------------------

Character Presets 

The fact that MIR "knows" a lot about the instruments on its stages makes it possible to supply very deep and customized ways of handling them. It provides for the seemingly simple but highly efficient way of changing an Instrument's timbre or "character" by applying one of MIR's built-in Character Presets.

Individual, hand-crafted Character Presets are available for every single Vienna Instrument (with the exception of a few less commonly used percussions). Provided that the suitable Instrument Profile was applied (...), there will be at least five "colors" to choose from just by clicking on the pull-down menu bar.

Common settings are:

• Pure (no Character Preset applied; default) 
• Air 
• Silk / Silver 
• Bite 
• Distant 
• Warm 

HINT: Most of the time, selecting an adequate Character Preset will show that little (if any) other processing is necessary down the line. (...). 

CAUTION: If you have the feeling that your computer is running out of CPU power, be aware that some Character Presets may use up quite a bit of it.   ... that's a non-issue in 2019. ;-)

In cases were there are no customized settings for an instrument or if a General Purpose profile is employed, more generalized presets are used. Those presets are marked with additional asterisks (*) after the name.

... it is hard to explain sound in words! So why don't you take a look at the following, generalized examples of how a specific MIR Character Preset will change the sound of an instrument.

[Dewdman42]

Dietz, those eq curves are very helpful thank you!  When I load Teldex, I see another character profile show up sometimes called MIRX Teldex.  Any clue about that one?

Ultimately I can always put my own eq in front of mirpro in order to eliminate a closely mic’d proximity warmth from a sound that is meant to be further back. I just wanted to have a clear idea what MIR Pro is doing to the dry signal based on distance. Looks like in a generic way the distance character profile would also provide a bit of a one size fits all low shelf to imply some distance that way. I will play with that.

By the way I wasn’t meaning to infer anything “bad” about calling the dry signal the "close mic". Just seeking absolute clarity about what is in that signal; and for the moment trying to be absolutely literal about what a close mic would have in the signal vs what MIR Pro potentially has in the dry signal.

So it’s a close mic signal after a mix engineer has potentially tweaked the sound in certain ways. Maybe a good way to think about it is: it’s a close mic after an engineer has panned it, eq’d it for character, adjusted the level and eq to mimic distance and air absorption (but not removing proximity effect unless distance character profile is used but still that is not variable based on icon position), and accounting for either the mic being attached to the players nose or being attached to the floor. That’s a mouthful though I get it.

Regarding proximity effect that may be baked into samples, that will always depend a lot on the samples being used. Many sample libraries have been close mic’d and are very present front and center. When they are lathered with reverb they will sound wrong. Our brain is hearing both proximity and reverb and it doesn’t compute. So that’s why there can often be a need to low shelf the low end in addition to hi end.  But it will always depend on the source samples being used, how they were produced to begin with...the mics, distances and processing on those source samples are all part of the close mic signal chain.

I don’t hear a lot of proximity warmth when I test out MIR Pro with various vsl instruments, which is probably a good thing most of the time. If vsl shelved out the proximity from the close mic samples to begin with then it’s a non issue when using them in MIR Pro unless you specifically wanted to hear the proximity. Other libraries, however might need more shelving and I am wondering about whether low shelf would need to be adjustable to account for the desired depth of the instrument, or distance from the main mic. Verta calls it z-plane.  It would have been cool if the icon on the soundstage effected this aspect of distance also in the dry signal. 

Perhaps if VSL had provided two parallel sets of IR's, one for the dry signal and one for the wet signal (rather then just removing the Dry from the IR's), then the IR's themselves I guess might have captured that.  But since dry has been removed from the wet IR's...then we have to mix certain assumptions about distance back in using conventional mixing techinques such as EQ.

[Dietz]

@Another User said:

Perhaps if VSL had provided two parallel sets of IR's, one for the dry signal and one for the wet signal (rather then just removing the Dry from the IR's) [...]

Like pointed out before: This could be done in simple stereo convolution reverbs*), but not in case of MIR, where we mix and match more than 100 individual IRs for a single stereo source. Severe phasing issues and ugly colouring would be the result.

You know the proverb: "You can't have your cake and eat it." 😉

*) ... where it still sounds bad more often than not

[Dewdman42]

I'm aware of Mirx Mode, but not the character profile by that name when I'm not even using Mirx Mode.  What is that doing?

I was not meaning to infer that we should bypass MirPro and do things the manual way, there is no point in purchasing MirPro if that is case obviously...Still trying to get confirmation that I understand what MirPro is doing to the dry signal?  It does sound like, however, I will probably want to think about using EQ and other plugins pre and post MIRPRo in order ot handle z-plane oriented spatiality which MIR Pro is currently NOT handling on the dry signal, mainly proximity effect.

My hypothetical point about parallel IR's, is not what you were understanding from my words I don't think, but we don't need to cover that I was just fantacizing out loud and it has distracted you from my core questions.  

[Dietz]

@Another User said:

It does sound like, however, I will probably want to think about using EQ and other plugins pre and post MIRPRo in order ot handle z-plane oriented spatiality which MIR Pro is currently NOT handling on the dry signal, mainly proximity effect.

If it sounds right for you, then there's nothing wrong about that! 😊 I use track- and bus-based processing in MIR centered mixes as much as I do in "conventional" ones.