@Dietz said:

Hmmmm ... I'm really pondering the idea of having  that "pop mode" institutionalized in MIR, somehow. Maybe all it needs is a Preference that moves the initial Dry/Wet values by a pre-defined amount. 

I was actually thinking of suggesting something like an "easy button" to mass edit everything. A slider or a preset wouldn't hurt. But I wonder if it wouldn't help to have more than just a dry/wet preset mode, but something that changes the profiles (pure, air, silver, etc) and maybe room EQ to fit that mode. I know that makes it a bigger request. Even if it's not a button, just some saved MIR presets might help that. I'm just not sure how that would work given that MIR doesn't know what instrument I've put into it until I've selected an instrument preset.

-Sean

@Dietz said:

...there is not one "proper" amount of mixed-in direct (unreverberated) signal: The other library you were mentioning was made especially for contemporary movie scores - which is a perfectly valid approach^*), but most certainly not the only one to orchestral recordings in general. More "classical" oriented orchestral work requires much more "room" in the final mix than the typically drier, more "artificial" movie scores with lots of acoustic elements competing with the actual orchestra...

 ha ha! Thanks! I agree, but I don't like pegging classical into one whole of recording philosophy. I don't think you necessarily meant it that way. I actually suspect we're fairly agreeable, just coming at this from different angles.

[....]

Most definitely. That's also the reason why I put the "classical" between quotation marks and wrote about "the _typically_ drier" filmscores. I just tried to get the point accross as obvious as possible.

Kind regards,

@Another User said:

The second starts with a short example fromStar Wars. I'm using VSL Trumpets, horns, and bones as wellas Spitfire horns and bones.

J.Hall,

Well not every example goes in the same order, on purpose. Sometimes VSL is first and other times it's second. But I've sent some examples to a few people lately where I didn't tell them which was which and they couldn't tell either.

The latest example highlights something very interesting to me.

See here. (download only, as it's WAV)

VSL goes first in the early longer examples. Then the ending lower notes are VSL-SA & VSL-SA again.

There are some inconsistencies I've found, but with some careful listening I'm fairly confident those are just because Teldex and Lyndhurst are obviously different inherently. I mean, they aren't even remotely the same shape. lol So in a way it's unfair to compare them. The bottom line of course is the sound though. But these examples clearly show that MIR is very capable and flexible at getting the sound where it needs to be. I only wish there were more scoring stages... or more 'non-consert-hall-size' spaces with diverse early reflections. IMO it would lend more to what MIR can do, effectively expanding it's sonic ability.

VSL actually sounds very similar to the other samples when dry. The second it touches verb they are as different as Apples and Coca Cola. Why that is I have no idea. But once I EQ up the high end of the instrument, then lower the high end on the Room EQ in MIR, I'm in a closer ballpark. And something about the 200-300 frequency area seems to matter a great deal. But after that, I'm starting to think that after your audio source, it's all realy up to the characteristics of the space you have it in. That's why I want more scoring stages. I mean, I'd ultimately love to see MIR have as vast a library like Altiverb. But I suspect with how sophisticated MIR recording sessions are, they'd have to chain Dietz to a laptop for the rest of his life. 😉

And I agree that VSL is indeed incredibly flexible. My big gripe there has always been not having enough flexibility over the staccatos in reps, length and tightness. But then again, I have far greater complaints with other libraries. Generally, VSL's instrument flexibility is more powerful and true to the instruments, legato especially. Plus, I only own the older VSL libraries, and mostly the S.E. versions. So I can't really compare any of this to the Dimension Series, which obviously looks like they would resolve such concerns. Perhaps on the next film I can justify the temptation. 😉

Cheers,

Sean

@SJSF - sounds like you used Teldex to blend with SP?  This is a very interesting thread and I'd be interested what MIR venues other people have tried for blending SP and other libraries.  Personally, I ended up using Mozartsaal because it had a similar warm and wide soundstage.  Pernegg is also interesting being a church structure like Lyndhurst, but it sounds substantially larger.

NG1,

Correct. To me the most important factor is the size and character of the early reflection. I've AB'd it too many times to admit to be proud of. I've found that I've had to radically alter the room EQ. I've actually come to wonder about the accuracy of impulse responses in general because of this. For example, we all know a Tuba and a piccolo are different in terms of "exciting the space". But there are many reasons why. And I wonder whether something as static as an impulse can really do the job. MIR + Teldex is stellar. But I've realized that the loudest instruments are where I have the biggest problems. Brass and percussion mostly. I've demoed other rooms to try to match with other spaces as well for the rest of the instruments. It's almost like there is a ringing to all of them. I just can't articulate it well enough. But I suspect it's something to do with the principle of convolution in general. I had to get back to work, so I've given this little thought for a while. But I'm still very invested in pinpointing this and doing something about it.

Cheers!

-Sean

"It's almost like there is a ringing to all of them."

Do you mean a ringing in the impulse responses or in the sound you are trying to recreate?  If you mean in the impulse responses then I understand exactly what you are saying.  I had a live recorded trumpet in MIR once with  a huge ringing resonance at a certain frequency.  I don't know if the issue was in the impulse response, or perhaps there was a resonance in my recording (not noticeable in the dry signal) that was being amplified by the IR.  I solved the problem by moving the instrument to the opposite side of the stage.

As for possible limitations of static impulse responses, I've long wondered if dynamic convolution (like the technology used in Acustica Nebula) would offer any benefit to capturing spaces.  It would certainly be interesting to hear a comparison, although the CPU load would be extreme with hundreds of IRs.

I should also point out that I use MIR almost daily and would highly recommend it, so I hope this post isn't taken as overly critical.  I just want to use it better.

NG1,

I definitely mean in the IR itself! I noticed the exact same thing. Dry it sounds fine. It's only once it touches verb that mayhem happens. It could be in part that rooms have "sweet spots" which create problems to compensate for...? I usually just adjust the room EQ to deal with it. But that tends to merely tame the obvious nasty spot(s), where the overall sound also has a similar undesireable quality to it that when dry wasn't present. So I believe it's a larger issue. I just don't know what yet.

I've looked at Nebula because of this as well. Although I'm still not sure it's the right solution. It's explorative in regard to dynamic convolution, which is great. I'm just not sure that's the core problem yet, or the only one. And I agree completely. All this is worth discussion and hopefully exploring and eventually discovering solutions.  MIR is still a wonderful product. I wouldn't want to live with out it! I'm not sure I could actually... I would likely suffocate from a lack of air. Get it? Air... ha........ ha... ;)

Cheers,

Sean

I use dynamic convolution since the days it was intruduced as a commercial product for the first time: The legendary "Replicator" by the even more legendary company Sintefex. Lateron, the same, highly impressive technology got licensed by Focusrite for their LiquidMix and LiquidChannel series.

Despite the groundbreaking technology, neither Focusrite nor Sintefex originally offered reverb for these boxes, for one main reason: You don't need dynamic convolution to "sample" a linear system like a real room. As long as there are no non-linear dynamic variables necessary to describe a system (i.e. distortion, time-variant modulation and so on), linear convolution is perfectly capable to achieve the goal we're after: To virtually re-create a hall.

Resonances, strange flutter-echoes, strong damping-effects ... all of this is what defines the "voice" of a hall. Like in case of a "real" recording, we have to find good spots for each instrument on each stage, and sometimes we will have to "change the light" with EQ's or filters, like you would do as a movie director in a room with some darker areas. 😊

Kind regards,

Dietz,

Great insight, thanks! Most of my commentary has been more suspicious rather than a difinitive opinion. So I absolutely welcome the explanation.

One note: I reduce the "ringing" (for lack of a better term) by lowering the high end. Point is, I've yet to see good results without doing that. That's why I wonder about the abilities/limits to convolution. Needless to say, even doing that, while dramatically helpful IMO, still leaves the brass with iffy results. After a ton of trial and error I've gotten some results I'm happy with for several instruments. But I've found that every instrument is just so different here. I've thought to myself many times "man, I wish I could EQ the room on a per instrument basis". Feature request? 😉

Flutter echoes! Why do I forget all these terms? So much to absorb and so little brain to do it with. 😉

So I tried telling Ernest Cholakis once about how Lyndhurst seems to have a brilliant flutter quality to it in the early reflections. I didn't use those words and who knows how poorly I described it. But it seems like the hall has both a smooth/rich and semi-abrasive  light sand paper quality to it. It's like chocolate. It's generally very smooth but has some grainy texture in it. I couldn't POSSIBLY make more sense, could I? 😉 Anyway, the shape of that early flutter (as it seems to be that) is quite literally the thing I love about the hall so much. I've wondered if it relates to the shape of the floor area being hexagonal, as I know curved surfaces can be rather interesting.

Question: is there any way of manipulating a hall's voice in verb, such as simulating flutter echos? I know, I'm asking a crazy question here. But it seems like having the means to shape existing spaces a bit more would help liven up and breathe character into samples. It's either that, or it seems I'd just have to wait until a million more spaces are captured until I find "the right one" I really like for each instrument. I'd rather shape the reflections though if possible. Seems like it would be far less work.

Please forgive the limited vocab and understanding here. I have good ears, it's the mouth & brain that have to catch up sometimes. 😉 This is all very interesting to say the least.

Cheers,

Sean

Well crap!

The more I think about this, the more I think the idea of flutter emulation would be ridiculous... that the only real solution is just to use a different space. Again, I love Teldex for my winds. Just thinking I want something different for Brass. Plus, that's a much... much simpler solution I suppose. The only problem is, I can't find a space I like. :(

More MIR packs to come? wink wink...

-Sean