What's new

Ways of mixing the orchestra

Anevis

Member
Hi guys.
I suppose this question has appeared on the forums a few times, but I´m gonna ask it again, since I haven´t seen it recently.

How do you approach orchestra mixing? Do you mix seperately each part of a section or only the whole section? For example with strings, do you mix everything together or each part individually, 1st/2nd strings, violas etc..?
The way I approach is that I only EQ the unecessary frequencies of the sound for the 1st strings then for the 2nd strings and so on and put a limiter on them. Then I mix the whole section together with some compression, reverb, saturation, you know.
 
I mainly work with "dry" samples...

1a. Strings, Woodwinds, Brass, Percussion all together on one group track.
1b. Solos on separate group track.
2. EQ and compression mainly in the group tracks.
3. adjust volume ratios within the groups.
4. create different acoustic room depths.
5. route instrument groups to the corresponding room depth.
6. now panning the instruments.
7. now coordinate the different instrument groups (mostly automation).
8. read out final track with enough headroom.
9. mastering

Beat
 
I mainly work with "dry" samples...

1a. Strings, Woodwinds, Brass, Percussion all together on one group track.
1b. Solos on separate group track.
2. EQ and compression mainly in the group tracks.
3. adjust volume ratios within the groups.
4. create different acoustic room depths.
5. route instrument groups to the corresponding room depth.
6. now panning the instruments.
7. now coordinate the different instrument groups (mostly automation).
8. read out final track with enough headroom.
9. mastering

Beat
4. create different acoustic room depths.
any tips how to do something like that?
 
Everything. Individual processing on busses, individual instruments, separate reverbs for each drum library. The only thing is that I tend to send the same reverb is strings and brass but sometimes I have to send more reverb to certain tracks to compensate for libraries sounding different.

With that said, that's the approach to achieve a coherent sound when mixing different libraries. You have to do indivdual processing, say you blend century brass which is super sharp with some spitfire brass that you use for other things, it has to sound coherent together so they need to be treated a bit differently.

When mixing live orchestra it's a bit different, you're most likely focusing on EQing different mic positions that record the whole orchestra at once, so the only individual EQ control you have is on close mics, which won't even be mixed loudly in the mix in most cases, so it's about getting a good overall sound from your mic positions.
 
Most of the newer orchestral libraries are so well recorded and mixed that there rarely is a need to process them a whole lot, if at all, during mixing.

If you're not going for a "real" sound but rather the modern hyped epic sound, then processing is necessary, obviously.

I only really do a few things:
1) bus similar sections to dedicated buses, so I can ensure the spacial characteristics are "the same" when viewed in the context of the entire mix.
2) surgical EQ (rarely needed) and removal of low-end for higher-pitched instruments (a high-pass filter on e.g. violins cutting below 200 Hz).
 
I think all this processing sounds bonkers unless you're going for a hyped sound (nothing wrong with that of course).

I'm with @AllanH -- the new sample libraries sound good already and torturing them one at a time is more likely to lead to that big "accordion" sound than a natural result.
 
I have yet to mix any kind of multi-track recording that doesn't need at least _some_ EQing on more or less every track - sometimes just to tame a resonance which builds up only in certain keys, or to get rid of low frequency build-ups, sometimes aiming for massive changes to get things working with (not against) each other.

I use compression lavishly, most of the time just very soft ratios like 1.2:1 to glue individual samples, but sometimes quite aggressively, especially in case of percussion or solo instruments. If in doubt I write some automation*) moves instead.

Saturation builds the bridge between the frequency and the dynamics domain as it changes both aspects (dynamically, one should add). Sometimes I use it just for more "interesting" textures (mostly in the treble range), sometimes for more even loudness (which is a result of the louder harmonics). I _love_ distortion, but you really have to have your act together to use it fruitfully in an orchestral context. ;-D

It doesn't make sense to do all of this in solo mode, though. Listening to tracks in solo is useful for restoration work, not so much for mixing. All the beautiful bass will be too much when it appears on all tracks, for example. ... this is the reason why I don't really believe in the idea of so-called "pre-mixed" samples, because this would mean that somebody has magically anticipated all your arrangement- and performance-decisions. 8-) (... yes, I write that as the guy who created myriads of MIRx-settings for Vienna Symphonic Library :-P).
Starting points - yes. Prêt-à-porter - no.

Reverb and any kind of spatialization is something that has to be applied to individual sources in context of the whole mix, too. For me its highly connected to width- and panning decisions, as reverb is the only straight-forward way to achieve depth in a stereo setup (... and in surround, too, to a certain extent). Don't hesitate to automate*) the "wetness" of instruments or the whole mix. If done properly this will add motion and "drama" to an otherwise static mix.

*) Automation is the key to any vivid and emotionally compelling mix in any case.

Finally, don't underestimate the possibilities you get from mix bus processing. Personally I tend to "mix into" a quite elaborate bus chain all the time. Its actual setup depends a lot on the job at hand, and it is nothing that is set once and left untouched for the rest of the time (quite contrary). Still it makes live so much easier, as I can spot weak points where they really hurt (in the mix) and fix them "bottom down" in the source track(s) thereafter.

That's an example of how it sounds when I combine real (life) orchestra and some samples, using the approach outlined above:





EDITED for valid YouTube-link.


Have fun!
 
Last edited:
Hi guys.
I suppose this question has appeared on the forums a few times, but I´m gonna ask it again, since I haven´t seen it recently.

How do you approach orchestra mixing? Do you mix seperately each part of a section or only the whole section? For example with strings, do you mix everything together or each part individually, 1st/2nd strings, violas etc..?
The way I approach is that I only EQ the unecessary frequencies of the sound for the 1st strings then for the 2nd strings and so on and put a limiter on them. Then I mix the whole section together with some compression, reverb, saturation, you know.

For orchestral stuff, I generally do more cutting EQ than boosting, at least when I first start. I don't use as much compression as I would for rock, pop and hip hop....but I automate A LOT. Lots of pushing and pulling to accentuate the dynamics depending on the track.
 
I think that the number one thing before even touching an EQ or fader is to strive to achieve a clear and balanced orchestration. Sometimes some of the muddiness or lack of clarity and depth can be solved by adjusting the arrangements and orchestral parts. BTW this applies to all genres of music - a lot of "mix" issues are actually more to do with problems with the arrangement/orchestration.

That being said, once that's been addressed then things will very much depend on the different libraries being used. If the libraries allow for a lot of control of MIDI CC (Dynamics, Vibrato, Release, etc.) This can be also considered very much part of the orchestration process and can have a profound effect on the timbers and balance of the instruments.

If the libraries have a baked in wet sound and are recorded in situ. (i.e. SF libraries), you may need to address some of the low end buildup of the hall with some EQ or HPF filtering. If the samples are dry then adding some early reflections and a couple of hall reverbs for the tails along with any gentle EQ / filtering and panning to help different libraries and sections sit well together.

Additionally you could have the longer tail reverbs act as a sort of glue for all the sections adding to taste depending on the library's wetness/dryness. Compression used very sparingly and judicially can also help glue things together - especially when blending different libraries.

In general, I find that small adjustments along the way tend to be more effective than some drastic processing on the master groups or Mix bus. Just subtle tweaks that can have a strong net cumulative effect overall.
 
Last edited:
I mainly work with "dry" samples...

1a. Strings, Woodwinds, Brass, Percussion all together on one group track.
1b. Solos on separate group track.
2. EQ and compression mainly in the group tracks.
3. adjust volume ratios within the groups.
4. create different acoustic room depths.
5. route instrument groups to the corresponding room depth.
6. now panning the instruments.
7. now coordinate the different instrument groups (mostly automation).
8. read out final track with enough headroom.
9. mastering

Beat

Yes, I do too, work with "dry" samples as it´s way easier to process them in the mixing stage.
 
Most of the newer orchestral libraries are so well recorded and mixed that there rarely is a need to process them a whole lot, if at all, during mixing.

If you're not going for a "real" sound but rather the modern hyped epic sound, then processing is necessary, obviously.

I only really do a few things:
1) bus similar sections to dedicated buses, so I can ensure the spacial characteristics are "the same" when viewed in the context of the entire mix.
2) surgical EQ (rarely needed) and removal of low-end for higher-pitched instruments (a high-pass filter on e.g. violins cutting below 200 Hz).

It´s different with every track I make, sometimes I tend to go for more of a "realistic" sound sometimes for that modern sound.

2) That is what I do too, for pretty much most of the parts.
 
I think all this processing sounds bonkers unless you're going for a hyped sound (nothing wrong with that of course).

I'm with @AllanH -- the new sample libraries sound good already and torturing them one at a time is more likely to lead to that big "accordion" sound than a natural result.

I use EWHO as my main orchestra and other libraries to just fill out the spectrum.

True is that I don´t do much processing but sometimes blending different libraries together does require some extra work.
 
Top Bottom