Dynamics of all the instruments in VSE+

[denis]

After a few unsuccessful attempts at making a decent mockup straight out of Sibelius using VE Pro with VSE+, I've decided to analyse step by step what was going wrong. So I played a single mezzoforte melody line over a few bars and copied it onto all my instruments, organised by register (so I can compare what is comparable). I made sure that the settings were identical for all instruments (no fx at all, velocity x-fade on at the same value, exp at the same value, volume at the same value, no pan).

Now it seems that the strings are MUCH lower in volume that the brass, which are much quieter than the woodwinds. Also I noticed some inconsistency within instruments families: for example flute 2 seems louder than flute 1. The worst comes when experimenting with the timpani: they're so quiet compare to the real thing!

So... there must be something that I don't understand. What do I need to do to make sure that the instruments sound like in real stage conditions, so that my celli are not totally hidden behind my bassoons playing the same register, so that the balance between violas and clarinets is natural, so that flute 1 can be heard when playing with flute 2 and oboe, so that the brass are ... well ... brassy, loud and clear?

Obviously I've done some panning and depth with the VSL Suite plugins, which helps positioning the instruments, but doesn't solve the problem of dynamics at all. There must be a way to play the same note on different instruments and have them all play as loud as they would in real life.

Please, HELP!!!

[stevesolum]

I'm a relative newbie here, and look forward to comments by others more experienced, but this is how I've tackled the same problem in Vienna Instruments: I control the three basic volume levels separately (VOlUME level of the particular Vienna preset, MIDI volume,and MIDI expression). For each VI preset, I go to Map Control in the perform window and assign the volume level (fader at lower right of window) to a separate MIDI controller #. I arbitrarily use #18. I then use this fader to set the basic maximum level of the instrument in relation to others in the orchestra. Trombone ensemble and timpani louder, alto clarinet one of the softest, and so on. I'm still experimenting with this, but as I get the hang of it I'm addressing your issue of balancing. By assigning this fader to something other than the standard MIDI volume controller 7, I can use MIDI volume 7 as usual to control faders on my mixer only; Pro Tools, Cubase, or Vienna Ensemble etc. Because the bassoon, alto clarinet and others have been set to a lower level on their presets, my mix faders can be in the middle of their range and still reflect a reasonable relative level for the respective instruments. I then assign the expression fader in the perform window of V.I. (exp, far right middle) to breath controller or a fader on my MIDI controller depending upon the situation. So: Preset Volume = relative instrument level, MIDI Volume = mix level, Expression (+velocity, filter, articulation type etc.)= expressive phrasing. Hope this helps, waiting to hear others solutions. Steve

[jasensmith]

I think this is one of Vienna Ensemble's strongest attributes.  What I do is find out which VI is playing the lowest volume in my VE template. Then, in the bus view, I adjust the volumes of the other VI's to match the soft VI. 

[denis]

Alright, so the difference of level between instruments is something that everybody experiences. In a way, I'm reassured as I thought I was doing something wrong.

Is there a template or something that can be used as a starting point? Why isn't it automatically done within the library?

[Dietz]

[quote=denis]Alright, so the difference of level between instruments is something that everybody experiences. In a way, I'm reassured as I thought I was doing something wrong.

This option is available as optional feature within Vienna MIR, called "Natural Volume".

Technically, the reason for the seemingly "wrong" volumes is very simple: In digital audio, there is an absolute maximum a signal can achieve volume-wise, called 0 dB FS (for "Full Scale"). If you go over this value, a A/D-converter will distort heavily during recording.

Now you might argue that we should adjust our preamps and converters to the loudest possible signal and leave these setting alone. The problem is: As soon as you leave the area of optimal gain settings, the recording will degrade in the analogue domain as well as the digital domain. So we head for the optimal resolution of (more or less) each sample.

You might say that these differences should be reflected by the mapping and programming within a Vienna Instrument. Although this would be possible, it would lead to the situation that the violin would have to be set to -20 dB (or even less) to achieve the necessary headroom for (say) a TamTam. Quite understandably, most users would complain why the violin is sooooo quiet (... and as a matter of fact this is the only complaint we get on a regular basis regarding Vienna MIR ;-) ... how good that "Natural Volume" can be switched off).

The good thing is that you are able to write a concerto for flautando violins and can rest assured that the recored samples are pristine. 8-)

HTH,

[denis]

Hi Dietz,

Thanks for the enlightenment!

Now... MIR isn't available on Mac. That would solve all my problems (and probably bring new ones). So for the time being, I need to work with VEPro alone.

As you seem very aware of the differences of volume and the recording techniques, could you give me some hints to achieve this "Natural volume" feature with VEPro and VSL Suite? I'm sure (actually, I hope) that it would be fairly straightforward to do with the recording/production notes of the VSL instruments.

Best,
Denis

[inwinterhesleeps]

Yes, a table of volume compensations to get a good balance (such as using the MIR natural setting) would be greatly appreciated. Then we can incorporate this into our VEPro templates. I have already been doing this, but feel I haven't got it right yet.

Cheers,

Tom

[denis]

A volume compensation table would indeed be an outstanding addition to our arsenal!
So would a pre-eq table be.

[Dietz]

@denis said:

[...] So would a pre-eq table be.

 Err ... "pre-eq" in the sense of ...?

[denis]

In the sense that to my ears (and to my FFT analyser), flute 2 sounds richer than flute 1. So even if I adjust the volume so that they play at the same level, flute 1 is overwhelmed by flute 2. So either I need to "pre-eq" one of the flutes to more or less match the other one, or I need to be very careful when using both instruments together.

[Dietz]

Hm ... I don't think that there's an EQ-setting that will make Flute 1 sound like Flute 2 (or vice versa). Actually I'm sure that they were made with the intention to sound different.

[denis]

Oops... Sorry, I wasn't very clear with my explanation. Yes, these two instruments are different and that is surely their purposes!

However I found at many occasions that it proved difficult to make them blend when both played together (to me, they sound more like two different takes rather than two similar instruments playing together, flute 2 being always very dominant).

Anyway, I hope that the team who developed the "Natural volume" in MIR would be kind enough to help with a volume compensation table.

[Animus]

There's something called volume and faders.  Use them.  ;-)

[denis]

Thanks Animus, that was incredibly useful.

[Animus]

You are welcome.  :-)  Just something I learned back in school; if something wasn't loud enough just turn it up.

[jasensmith]

All smart assness aside [:P] some instruments, especially brass, are sensitive timbre wise to volume changes. These changes in timbre have an effect on how natural the instrument sounds when trying to emulate a real orchestra.  For example, if you're trying to get horns playing in mp to sound like ff by simply turning the volume up you not going to have much success.  Instead of getting something that sounds like ff horns, you'll get something that sounds like mp horns with the volume inappropriately cranked way up. 

Just something I learned back in school. 

[Animus]

oooh snap!

[:P]

[denis]

Dietz, anyone else at VSL,
any chance to get your input and help to achieve Natural Volume without using MIR?

[Dietz]

[quote=denis]Dietz, anyone else at VSL,

Right now, we have no plans to separate this feature from MIR.

Kind regards,

[denis]

Should I understand that VSL don't want to help their customer achieve the best result possible with their products, in the same vein as they have denied their customers some depth presets with Vienna Suite convolution or powerpan presets for usual ensembles?
I really look forward to hearing from you.