Hi Macker,

MIR is a bit different than a traditional reverb. I fear the way of using it has to follow a different approach. An easier one, probably, since there is a single stage to care about.

I don't know if the technique you are describing is the same I had been using before switching to MIR. What I did was to create some layers of early-reflections, to create the illusion of close, middle, and far groups of instruments. Instruments were bussed to these layers, to give an approximate idea of distance from the listener and the back wall. Maybe this is the equivalent of what you call stage-reverb. Then, everything was sent to the main reverb halo, that may be what you describe as the main-space reverb.

With MIR you deal with a virtual space, where each instrument is sent to its final position. From there, MIR creates all the components of the reverb, giving an idea of position on stage and enveloping the instrument with the reverb's halo. What you can do is to set the length of the reverb (as if you were actioning absorbers), mix the dry and wet signals (as if you were mixing the spot and the room mics), add a second set of microphones (for surround), and equalize the instrument's and room sound.

The main character of the reverb is therefore due to the chosen room. In my case I've chosen the Teldex, the classic recording studio very often used to record classical music. It's an average-sized stage, with a very transparent sound. Not exactly the huge space that is the RAH, or a bigger film scoring like the Sony/MGM or the Synchron.

So, if I want a lusher reverb, I guess the best option is to select a different room, with a lusher sound. I'm trying now with the Steinhofkirche, a huge building modeled after Hagia Sophia, with 8 seconds of reverb. I've lowered it to 3 seconds, to make it more similar to the RAH. Different spaces, but maybe they can be in a similar class of reverberant spaces.

Paolo

Good hunting Paolo!

My professional background is design engineering, which is probably why I tend to avoid 'big black box solutions' that offer instant results by pushing a few buttons. So I can't see me buying MIR any time soon - and that's certainly not to degrade its usefulness and convenience for others.

I've further reduced the wet signal in the Teldex version.

Just for fun, I've also done a version in the Steinhofkirche, with the reverb time reduced to about 3s (similar to the Royal Albert Hall). The sound is totally different from the Londoner hall, and is of a more "impalpable" quality.

Holst - Mercury (Steinhofkirche)

Paolo

Both versions now have a night-and-day difference to the original. Much improved. Good going Paolo!

But ... (well you just know I have more to say, lol) ...... ready?

Teldex

In your latest Teldex version I'm still hearing what is for me a somewhat irksome, strong, definite and unnecessary confinement of 3D acoustic space. The sound (to my ear) too obviously highlights the walls and ceiling of Teldex and nothing beyond. This has almost nothing to do with the reverberation time (RT60) of Teldex and everything to do with how the ear perceives 3D space. It doesn't matter if a sound bounces back and forth between two walls 5 metres apart for 1 second or for 10 seconds, our ears still get the same intuitive sense of how far apart the walls are.

On its own, MIR's Teldex convolution reverb truthfully informs our ears: "there's no escape past the Teldex walls and ceiling". Synchron Stage would sound a bit larger but would still faithfully let us know we're in a confined space much smaller than a typical concert auditorium. (Btw no, I'm not a claustrophobic, lol.)

BUT .... I've just tried out a very simple 'Band-Aid' fix. I've added to your latest Teldex mix a small amount of convolution reverb of the RAH. Now the Teldex confinement, although still there, is no longer so annoyingly pre-eminent and final. (This RAH reverb is in the venerable old Waves IR1 plugin, named coyly as "The Big Hall" - perhaps because using "Royal Albert" in product names is not permitted in the UK market.)

The RT60 of this RAH reverb is only 2.2 seconds and yet the sound clearly reaches out in 3D space well beyond the confines of any scoring stage. However, adding this RAH reverb can't completely moderate or mediate the very forceful imprint of Teldex in your mix.

Therefore I'd suggest (and now I'm hearing groans from poor Paulo!) a more ideal mix would have even less Teldex, but would include some large-hall reverb, added perhaps to the master bus (or even to a print as I've just done), such that the sound can stretch out beyond Teldex confinement - and without incurring a big RT60 penalty.

From what you said earlier, I must say I'm very puzzled as to why MIR doesn't seem amenable to the perfectly proper practice of treating scoring stage reverb as no more than a component of overall reverb.

Kirche am Steinhof

This is a very strange mix, yet not without its appeal. My first reaction was that orchestral works typically don't really suit such huge and highly reverberant spaces. But then I remembered a performance of the famously beautiful Tallis Fantasia by Vaughan Williams (which is said to have been an influence in Holst's Planet Suite), recorded in Gloucester Cathedral, England. Here it is for comparison (also hoping Dietz might appreciate this):

I'd say Mercury is way too fast-paced and dynamic for the large and stately Kirche am Steinhof, despite the amplitude envelope being reduced to 3 seconds. But at least stage-confinement is gone. And as you say, it's for fun.

BTW, I'm wondering what's up with the Celesta part. In some places it seem almost to disappear and the pitches are strange, as if it's played an octave or a 5th higher, or something. It certainly doesn't sound like the Celesta part in the Prom performance - which is what I've been accustomed to forever.

@Macker said:

BTW, I'm wondering what's up with the Celesta part. In some places it seem almost to disappear and the pitches are strange, as if it's played an octave or a 5th higher, or something.

Does this also happen in the Teldex version? I'll check it again against the score.

What you suggest about the reverb is to make it more artificial than using a convolution reverb. I'm not convinced of this solution, but I’ll do an experiment in shortening the reverb time in MIR, and increasing the importance of the artificial reverb I use only for smoothening.

While artificial reverbs tend to sound good, I find them generally less convincing than MIR. A compromise could be making the presence of the smoothening artificial reverb stronger, but then again it will sound as an artificial reverb and not a spatialization. It's in any case one of the possible techniques to be explored.

Paolo

Celesta part issues are the same in both versions. Worst case is at 1:18 when the Celesta should easily be heard above the quiet Harp ostinato but seems totally absent. Another case is at 0:28 when Celesta should be there but there seems to be something very high-pitched instead, perhaps the Glock. Is the Celesta there at all, and is it perhaps just the Glock part I'm hearing where the Celesta is missing? I've just pulled out the B&H score to have a look through.

One thing about combining reverbs -  certain venues in MIR can be used with the MIRacle close orchestral enhance, as an overall addition to the sound, and it works beautifully.  Though to use it, the convolution reverb  Wet must be turned down overall.   The advantage of that system is it has a richness but also more clarity than just convolution with more Wet.  For a while I was routing the Vienna Konzerthaus  Close Orchestral Stage, somewhat more Dry than default middle setting, into a Lexicon hardware reverb which had a Large Hall at around 40% Wet. This sounded really great and was used on a couple entire music library albums of mine. However, I discovered that MIRacle does the exact same thing without the external hardware and so have been using that on several mixes and love it.        

There was some talk about the celesta - it is a unique sound and I've had trouble with it. It doesn't cut through the orchestra like glockenspiel, but if you turn it up in volume the low/mid frequencies are exaggerated. This of course because of the close up recording as opposed to where it is placed in an orchestral setting.  I think I used an individual dry adjustment, and some EQ to tone down the low/mid.

Anyway I think overall it's sounding great and can just use some little adjustments to increase the clarity here and there.  The amazing thing about MIR is default settings are close to perfect! So if one does not stray too far from those, and just tweak things a little, it works wonderfully.    

William and Macker, thank you for your support!

Before thinking to the reverb, maybe I solved the issue of the Celesta. I had it sounding one octave too high. Maybe a case of double transposition? Dorico is for sure transposing. But the Vienna Player seems to be in the real pitch.

The VSL manuals says that the instrument is mapped one octave lower than written so it's probably already transposed, despite the shown octave transposition is zero. 

The same happens with the Glockenspiel.

Macker: congratulation for catching it! Not easy, with these pure-harmonics sounds!

At this point, I don't know if I should change the octave in the VSL sounds, or remove octave transposition in Dorico. I think I will do the former, to always deal with the real pitch. It's the same with the double basses.

Paolo

I've updated the "old" version with the due changes to the mallets (and several other small things fixed). The link in the initial post is now updated.

I've also tried a version with more MIRacle, and a reduced mix/tail of MIR. I've probably not done a great job, since I find it not very realistic, and without that "gloss", making the sound seem to be immersed in the space, that I find in the MIR/less MIRacle version.

Holst - Mercury (VI) (with less MIR and more MIRacle)

I must be the biggest fan of pure MIR!

Paolo

EDIT: Added a version for the Synchron libraries (with Synchronized Woodwinds and VI Harps).

Holst - Mercury (Synchron)

I've added some more espressivo-style details to the score.

Not being undecided enough with the reverb matter, I'm now undecided about limiting or not. The loudest part seems to ask for it.

I've tried with two limiters: Steinberg's own Limiter included in Dorico, with the Protection preset. I think this introduced some harshness that was not what I like. The, I used a modeled Vari µ in Limiting mode, and with parallel compression on to let much of the dry signal pass untouched.

I thing the Vari µ is doing a good job with the loudest passages. However, it also seems to do something to the softer dynamics, compressing the ppp-mp range as well. It shouldn't, but by emulating a tube circuitry, it is probably always compressing a little. As a consequence, I find the non-compressed version slightly more transparent.

Which one do you prefer? The difference is subtle, but it is there.

Holst - Mercury (with no limiting)

Holst - Mercury (limiting)

Paolo

I updated it again. I got a chance to fix some mistakes, and work a bit more on the dynamics. This is a piece that goes from tiny, cristalline 'pings', to the most massive burst of sound. Not easy to keep it under control.

Paolo

Hi Paolo.

Just listened to your work. What a start for me in the new year!

All the best

Fabio, thank you for listening!

Happy New Year!

Paolo

Hi,

I've finally been able to make a version for the Synchron libraries (with Synchronized Woodwinds and VI Harps).

Obviously, with second violins on the right! 😊

Holst - Mercury (Synchron)

Paolo