Thanks Polarbear and Rawmusic,

My question may of been a bit raw but I'll try shape it more. There's been a few things said that are already helpful, but one thing I'd like somebody to explain to me, if VSL go through all the trouble of recording as best as possible each instrument resulting in a pure sound of the instrument, including all registers, bright highs, dark lows, why would one want to alter the frequency (EQ)of that sound? Will engineers do that when recording live orchestras? You could also say an orchestra sounds great just hearing it live, with no messing around with sounds, if so, in studio. I know accoustics play an important role in sounds also. I ask this because sometimes I'm affraid to change the color through EQ of the instrument already with a natural sound.

I totally agree that quality of orchestration is into consideration. But let's assume that not the problem, and I don't think it is. An ex. I'm thinking of is thick brass playing with thick strings, à la Wagner or Tchaikovsky, I often found that the 2 sections lack contrasts in terms of timbre, they are not enough distinct, which makes me avoid using that combination in times. If this was a real orchestra I'm sure that this would not be a problem. My approach is that each section must sound great without the help of the other one, and that is fine, but when certain of them are combined I would hope for more clarity between the sections. So I'm hoping some of you experienced computer musicians will be able to give me some advice.
Thanks.

Oh yeah! I will eventually give some audio examples, especially if I still have the same problem. Unfortunetally I cannot do this right now: somebody got a hold of all my recent demos blablabla, and won't give them back, the only copy too. I'm also in the process of buying all new equipement, so it may be a few weeks...also time to do new demos.

Guy

This is what VSL tried to capture, the reality. However live and recorded doesn't compare too well due to technical reasons - recording is always trying to get near reality but can't because of the limitations in today's techniques. However you can compare a recorded score to a VSL arrangement. And those tricks are the ones done by the engineers, also on real scores. It takes a lifetime to master them, and with new tools there are new possibilities and new tricks. And some are good-kept secrets... [[;)]]

Without knowing a specific line of the given example of thick strings and brass - I think this will be an example where all the things can be used: Don't let them play in the same range to seperat them from each other, make use of positioning and narrow the stereo image of all (by the this is crucial with VSL, because everything is recorded centre, which leaves you the choice where to place things). Then listen with your golden engineer's ears to detect where the instruments are still interfering with each other, and shape that with EQ. Finally apply the right touch of reverb to each instruments and whoops, the week went by already. [[;)]]

PolarBear

Rawmusic and Polarbear - some really good advice given here [:D]


Rob

What did I post you don't already do, ROb? [;)] In fact I can learn a lot from your orchestrations and compositions not only just on that matter!

I just think your well written plan to first rely on orchestration is soooo important to balance, timbre, etc.

It is something all of us should be reminded of constantly [:D]

There was a time not too long ago, when I would think - "I'll fix it in the mix" - wrong thinking - fix it now in the writing and orchestration - 9 out of 10 this takes CARE of the mix.

Rob

I think another issue to add or at least expand on is the perception of where the instrument sits in space. Narrowing the stereo field and panning obviously help as does application of reverb to affect the listener's perception of distance. To this, I've found that there is a different EQ for instruments that are close miced, as I assume VSL is, verses something farther away. I'm not an engineer, but have noticed that highs and lows both disapate as sound sources get farther away. What works for me is to try to judge and balance the relative "size" of an instrument both in relation to it's other articulations, and it's perceived size in relation to other instruments of the orchestra, and then again it's position on stage. I did this over a period of almost a year to set up a template with these settings built in. In consequence, the clarity is there and the only EQ (and mixing) that might be needed is a minor adjustment at the mastering end. The sad part is it doesn't save any time, yet makes it possible to focus more on the music itself.

Thanks guys, so fine tuning here and there will make a considerable difference if I understood you right. Regarding panning, is it a good idea to pan or to have a fix pan according to the position of players in a real orchestra? Ex. 1st viol-left, 2nd viol-middle, cellos-left, timp-middle etc. To me that seems like a dopy idea but I'll ask anyway. I say it's a dopy idea because I would think that the idea is to make it sound lively no matter what tricks it takes rather than working too theoretically. Does this make any sense?

I'm looking forward to testing out the given advice. Thanks!

Not dopey at all. Do also remember that some instruments, even though they have been recorded in stereo, might be better loaded in as mono instruments. I am even finding that in the violin sections, particularly on the more "abrasive" (sorry - for want of a better word!) articulations, there are a couple of musicians off to the right being more enthusiastic about their sffz. Thus, when I have 2nd vlns panned 30% left, there are still major contributions from the right hand side of the stereo picture. I am experimenting with narrowing down the strings, whilst still maintaing stereo of some kind. Logic has a plugin which enables you to play around with the stereo width. Will report back with anything interesting.

The other thing to remember, is that when an ensemble is recorded all at the same time, each instruments' entire frequency ranges will be reacting with one another, interfering, masking etc. etc. To use pure instruments and mix them together in the computer can't recreate this subtle effect. When doing a pop mix, I routinely mix out the bottom end (within reason) from everything but bass drum and bass guitar. Especially electric and acoustic guitars - which sound gorgeous as solo instruments with full frequency reange - can ruin a mix with the amount of low frequency program material they provide.

I have started eq-ing individual instruments with good effect. Mostly subtractive, as it is often a part of the sound that is interfering that can stand to be attenuated. One opposite example is the trumpet sounds, where I do like a little extra sparkle in ff passages.

Have a good play, tweaking!

I concur that mono sounds often sound better. Yes they are more easy to place, but quite often you find yourself loosing a lot of the resonant body, and it feels "a bit glued in the mix" to me - in the lack of better words. Solo instruments sometimes profit from it, but I didn't like any sections in mono yet. I had good results with narrowing the stereo fields, I use the free plugin from http://www.mda-vst.com/ (Image, with M-S matrix) - it's dong its job just as fine as the more expensive Waves S1 plugin.

I also use it for panning, as Midi panning is subtractive in volume on the left or the right channel. You loose some of the beauty that's recorded. And what is a mono sample other than a hard-Midi-panned stereo sample?

PolarBear

@rawmusic said:

The other thing to remember, is that when an ensemble is recorded all at the same time, each instruments' entire frequency ranges will be reacting with one another, interfering, masking etc. etc. To use pure instruments and mix them together in the computer can't recreate this subtle effect.
Have a good play, tweaking!

That I find is HIGHLY IMPORTANT and I wish the VSL company could have 2 types of recordings: no. 1 dry and no. 2 surrounded by an orchestra in a good concert hall. This way we could benifit from those interacing harmonics, at least from the other instruments, of course it's not the same when the players are actually playing their part and there is interaction between the parts, but it would already sound more natural, I find. I mean can you imagine recording a Beethoven symphony putting each player in an isolated booth with headphones. I'm not sure the overall sound would be so hot!

Guy

@Guy said:

That I find is HIGHLY IMPORTANT and I wish the VSL company could have 2 types of recordings: no. 1 dry and no. 2 surrounded by an orchestra in a good concert hall.
Guy

Then you'd need twice as many computers to run it...!

DG

@Guy said:

[...] I wish the VSL company could have 2 types of recordings: no. 1 dry and no. 2 surrounded by an orchestra in a good concert hall. [...]

Number 1 doesn't sound good, and number 2 would restrict us to an old-school sampling library. No legato, no repetitions, no meaningful alternations, no change in the tempo of performances, no free placement on the stage, no proper tuning, and more noise.

So we decided to build our SilentStage, and make the Library sound like it does [;)]

All the best,

I'm sure you guys no what you're doing and I doubt I know what I'm saying but I still find questionable the accoustics of the instruments. especially in comparason with a real orchestra. But maybe there are ways around that which I still haven't found.

And I am well chuffed that you did. I never thought that VSL would be a tailor made orchestra that would require no additional work. A while ago, piano samples were simple note-on note-off. Now they are complicated ideas involving pseudo sympathetic resonances, several layers of release samples. Similar things may well come our way in the future, but just consider the possibilities and the variables is utterly mind boggling - impossible one might say.... Give it a couple of years I say, once processing power has multiplied a couple of hundred or thousand fold, and VSL has moved several stages on from MIR and involved in inter-instrumental acoustic relationships, calculated on the fly in real time, with no latency!

Well said Rawmusic.

Guy [[;)]]