I'm sorry if I don't get how it works immediately, but this is really a complicate matter.

Two things I noticed:

- Mixed various orchestral instruments, with just the Main mics turned on and set to 0dB. No additional reverb. All MIDI volume controls set at the same level. The resulting image sounds realistic, if one knows how a raw recording sounds.

- The presets supplied by VSL have different MIDI Volume settings. They also have different Main mics settings, as well as different values for the other mics. I wonder why there isn't a reference mic pair set at the same level in all the presets. Would it be possible? Would it make sense? Would it allow to set the MIDI Volume at the same level in all the presets?

If setting some reference mics at the same level (be them the Main or the Surround ones), I would probably adjust the level of all the mics proportionally, so that the reference mics are aligned between presets, and the other mics preserve the same balance as in the original presets.

As a result, I hope to get the same starting MIDI Volume in all the presets, and the same level for all the Room mics, as if they were the same pair getting the same coherent image of the orchestra.

Paolo

Another discover: Horn 1 vs Horn 2. The Classic Room Mix uses the same values for both instruments. With the MIDI Volume set at the same value, they are nearly perfectly balanced, according to my ears and the meter.

However, the presets have MIDI Volume = 67 for Horn 1, and = 80 for Horn 2, that becomes stronger.

I can't understand if this is just a matter of personal taste of the developer who made the presets, or there is a reason to keep this difference between them.

I'm missing a lot of the logic behind the rules of mixing a (sampled) orchestra. "Trusting my ears" is something that has to come later, after a bit of acquired awareness.

Paolo

Going on exploring the presets, to try to understand how they work.

- Synchron Strings Pro, Vni 1, Classic Decca Tree. Main L/R -6dB, Main C -11dB, MIDI Volume 101.

- Synchron Strings Pro, Vle, Classic Decca Tree. Main L/R -6dB, Main C -9dB, MIDI Volume 95.

- Synchron Strings Pro, Vc, Classic Decca Tree. Main L/R -6dB, Main C -10dB, MIDI Volume 95.

The common reference point might be the Main L/R mics. But then, the MIDI Volume changes it. And why the Main C channel is different in the various presets?

Paolo

An experiment: an all-brass ensemble, with all MIDI Volumes set to 100, Expression starting at 110, the Main L/R mics at 0dB, the Main C mic at -3dB.

The balance seems perfectly natural to me. Still a raw sound, with no emphasis on any of the instruments, but the honest image of what you would listen onstage.

Williams, Olympic Fanfare (VSL Synchron Brass, Decca Tree only)

Paolo

An interesting investigation which becomes even more complicated if you consider the differences in let us call it starting volume of different articulations of the same instrument. When working on a classical score of e.g. a Beethoven symphony with rapid changes between long notes, legato phrases and staccato runs maybe with a tenuto note from time to time and lots of Sfz bars it is a big challenge to balance. The players balance the volume of these articulations naturally but in e.g, VSL Synchron libraries but also in libraries of other providers you have to reduce the volume (note velocity and in most cases also CC1or CC2 Modulation and CC11 Expression as I personally seldom touch CC7 in the plugin) of all short but often also long notes to balance them with the much quieter legato patches to get a natural sounding phrases. I actually also noticed huge difference in volume of the VSL Pro Strings Legato patches. Try getting Legato Espressivo to balance with standard legato in a phrase You need huge manual changes in volume parameters.

The differences in the strings example presets might also have something to do with the distances from the decca tree and other microphones of the instrument sections and the way VSL libraries handle such distances. On top of this the instrument panning also plays a role. If I understand it correctly with VSL Synchron libraries, the instruments are panned as per their location in the orchestra setup although the panning button in your DAW is in the center. 

The two Horns in SY Brass probably have not been registered at the same time and the final balancing was done in the preset. Horn 2 also has a different timbre I noticed. 

Thank you for your considerations, Mavros!

@Another User said:

The differences in the strings example presets might also have something to do with the distances from the decca tree and other microphones of the instrument sections and the way VSL libraries handle such distances.

According to my first experiments, the Decca Tree seems to give a coherent image between different instruments. It seems to me that, if the whole orchestra is listened through those mics, all set at the same level, you should get a realistic image of the orchestra, with the correct panning and depth. So, I would expect this coherence to be preserved inside the presets – but it is not this way. The same Main or Surround mics can be set differently in different presets. I think only the RoomMix is always set to 0dB everywhere (but I've not examined all the presets).

Natural panning should be already in the various room mics. But then, as you note, there is the artificial panning of the close/mid mics. But shouldn't the overall volume of the stereo/surround channels be preserved, when panning?

As for the different timbre of the two horns (but, I would add, of the other winds): this doesn't seem to impact on the volume, if not marginally. If you can check with the naked samples, maybe you can give me your impression and help me understand if I'm actually listening what I believe I'm listening to!

Paolo

Still experimenting and this is my general impression on the mics volume:

- The Decca Tree and the Mid microphones are all balanced in a natural way, and faithfully reproduce the position and natural volume of the instruments.

- The Close mic can vary very much in level. Probably, it is not possible to do a 'calibrated' positioning of the close mic, due to the quick variation with just minor movements of the microphone or the player.

Therefore, while the Mid and Room microphones are working well together by sharing the same level values, the Close mic depend on the individual instrument. The Close mics can't be set visually, but need careful listening to balance them.

Paolo

I've done a test on some music I had. In this Wagner excerpt, I can test how the Synchron Strings pro are balanced between them.

I leveled all the MIDI volumes between sections (CC7 = 100, CC11 = 110).

Then I selected the Ambience SurToStereo MixPresets for all the instruments, and moved the Main L/R mics to 0dB, with all the other mics adjusted accordingly. I preserved the original balance between mics as in the VSL presets (including the room mics, that in a real-world recording situation should have been al put at the same level).

I turned the presets' reverb send off, and turned off any reverb and dynamic processing on the master.

What I decided to leave on, as another exception, are the EQ on the separate mics. These EQ (and their settings) are simply fantastic, making the sound jump out of the speakers without losing naturalness.

Pure, with individual mic's EQs on

With the EQs turned off the sound is clear, natural and very pleasant, but maybe lacking that 'pop-up' quality added by the EQ. It's really 'pure', but it might be more desirable to get advantage of to the clever programming from VSL. (Unless you want a totally natural sound).

Pure, with individual mic's EQs turned off

Paolo

I like the clarity of the second version and the nearly floating high strings in the last measures which is less the case with the EQed version.

Have you ever tried experimenting  with VSL libraries in Dorico. It lacks a bit on the mixing functionalities but that actually brings out exactly what the plugin and samples really sound like . You also have the possibility to have a note sounding longer (or shorter) than officially written in the score which gives a very natural real player sound, difficult to obtain in a DAW if you do nit want to modify the official score. e.g. at the end of phrases before the start of the next. 

@Mavros said:

I like the clarity of the second version and the nearly floating high strings in the last measures which is less the case with the EQed version.

Thank you for listening, Mavros. I'm more and more appreciating the open sound with no EQ. I have to experiment with only adding EQ in moderate amounts, if ever. Do we really need corrective EQ in sampled sounds? And, is EQ doing more bad or good, for sound shaping?

The example is indeed made with Dorico. Note attacks are modified, so that the sound doesn't start too late, shortening the shorter notes too much. But I know there is a lot of chisel work to do, both on note attacks, articulations and envelopes, to make it finally sound goo.

Paolo

This thread seems to be the newest on the subject so I would like to add another observation which came up working on volume balancing in my orchestral template which has a combination of BBCSO and VSL Synchron for WW, Brass and Strings.

I have seen several threads including this one where CC7 of all Synchron (or Synchronised) instruments was set to 100 (or all at the same value). As per VSL input in the is forum and also in a direct interaction with them, CC7 volume is used by the VSL engineers to balance instruments relative to each other together with the microphone presets for each room in the Mixer window. So changing the preset CC7 values ruin the VSL library engineer's (good or bad?) idea of orchestral instrument balance.

For this balancing act I have for now just used long articulation only for all instruments. 9 bars with a mix of half, quarter and eights notes all in the middle of the playable range for the instruments. I have left VSL at default CC7, CC11 (127) and CC1 (80/127=63%). I have used the Classical Room Mix preset. In BBCSO I set CC11 and CC1 also to 127 (100%) and 63%. Zero reverb in BBCSO plugin and no reverb or other effects on the tracks. Velocities for both VSL and BBCSO are set to the Logic default 98/127.

To compare I have listed peak dBFS and LUFS of the whole 9 bars. Attached a pdf with my findings which might be interesting to discuss.

To get a volume balance, I only adapted the volume in the BBCSO plugin for WW and Brass instruments (mostly downwards) to more or less match the VSL values. The strings I left untouched for now as the number of instruments are not the same. So arbitrarily I took VSL Synchron WW and Brass as "correctly" balanced however this might be questionable if we look at the peak and LUFS values I found for the VSL libraries .

There is the issue of pitch at which the instruments are playing of course but as an example I would think that a piccolo playing one octave higher than a flute would be louder than that flute. In BBCSO it is the case. In Synchron WW it is not. Same for a Trumpet and a Horn. Again BBCSO shows what I would expect Synchron Brass not. Also Cimbasso and Tuba seem very quiet relative to he other brass in the VSL library.

Attached a pdf with my findings which might be interesting to discuss.

Everyone seems to be in full balance. No comments to my post. In fact I recently saw two Youtube video‘s by Berklee discussing balancing different libraries. Their method is similar.

In view of the announced contest I was looking at the Hello Free instruments. I wonder if they are ballanced by VSL and if so what the reference point is. The Big Bang Orchestra is the oldest free instrument but it is massive even compared to the Fujara Flute. Not easy to combine in a natural way with the new ones.