A Basic Procedure for Template Balancing

As a neophyte I’m at least aware of numerous tutorials/articles on the subject. But I’d like to benefit from the collective wisdom and different approaches/tips from you pros to initial, basic balancing of your new template for volume and pan in Cubase. Reverb is another subject but I’m assuming it comes after basic volume/pan balancing (for newbs at least).

I’m looking for a basic sequence of steps in a first pass to general balancing of volume and pan. Nothing fancy yet with spatial positioners like VSS or VSL’s (makes my head hurt). Its confusing to me for openers because there is a large number of volume and pan sliders everywhere, some duplicates for convenience some not e.g. in the Inspector, the Master header of Kontakt, individual instrument panels in Kontakt, in the Fader, … And panning affects volume (DOH).

So… when you build a new template what are your steps to achieving basic overall balance ?

I have an orchestral palette of around 50 instruments organized into Group Tracks (STR, WOODS, BRASS, PERC).

Here’s an example of a sequence I think of and would like to know yours:

  1. Set every single Volume slider you can find (Inspector, Kontakt Main and its individual instruments, Cubase Fader…) initially to the same midrange number (70 ?) to get in the ballpark.

  2. Set every Pan you can find for every instrument in the palette to your best guess of conventional placement (e.g. harp hard left, Violins 20-40 to left, Cellos 20-40 to right, Basses hard right, Woods center, etc.). Ditto re Group pan settings. Don’t worry about depth yet (that’s another Post). Or maybe also adjust volumes a little for far away instruments in the back down to 60 or whatever.

  3. Make some individual Group test MIDI file(s): sequential sustain notes midrange from every instrument in the Group followed by some chords of instruments in the Group.

  4. Play back a loop of this MIDI test file for the first Group generated in Step 3, and adjust relative volume of all the instruments in the Group to balance.

  5. Repeat 4. for the next Group and its instruments. Wash, rinse, repeat for all Groups until clean.

  6. Then make a MIDI test file(s) of various chords from various instruments in various Group combinations and different subsets of their instruments in each Group.

  7. Play back these multiple Groups test files(s) of instruments playing from different Group combinations, now adjusting relative Group volume and pan faders of each Group for overall balance of each Group relative to the other Group(s). Maybe, for example, a “Groups” file with some instruments in Woods and some in Brass, another “Groups” file with some Strings and Percussion, etc. Maybe diddle relative pan settings also at this time of various Groups combinations playing back with various numbers of their Group’s instruments to set the relative pan positions of the Groups.

  8. Take up golf instead.

Sorry for the prolix post, maybe I’m making this too complicated. But I’m floundering and bewildered here. Your specific steps and approaches to basic balancing of a new template would be incredibly helpful to me. I wonder if there’d be a DAW-specific market for these test files covering various genres- light classical, epic, small chamber, stage band, …

Hoping for some real basic help here (and BTW my next mockup assignment is due in a week).

Thanks all,


Not a helpful answer, I guess, but I spent a fair amount of time on the subject, reading all I could find on template balancing, trying many configurations, three separate reverbs for simulating distance from listener etc. Almost took up golf. I ended up buying MIR24 from VSL and I’ve stopped worrying about such things. If you’re using a lot of VSL instruments this could be the way to go for you. Just my 2c.

Thanks for responding. I’ve heard alot of good things about MIR. I don’t own any VSL yet but plan on it in the not too distant future, probably when I get my first slave. Maybe VSL special edition. BTW a respondent in another forum (vi-control.net) basically said he thinks I’m overthinking it, but had some good suggestions. I’m not at all sure all you have to do is listen to some real orchestras and try to mock them up. Great for other purposes, but I just don’t see how that’s an efficient way to balance.


Listening to real orchestras is certainly not everything but it is very important. On a more technical level, if you want a starting point, I would just play a fortissimo note on every instrument of your ensemble simultaneously, including timpani, horns, everything, and make sure the output bus tickles the 0 db mark. You have to make sure that your full sound at the highest level doesn’t cause you trouble with dynamics down the road. But don’t be too cerebral about it and if you need to cheat, cheat. Nobody cares about your template, the only thing that matters is your mixdown.

If you’re just starting then your first arrangements probably aren’t going to sound wonderful and it is very, very normal. You’ll get better by doing and listening, and this takes time. So just do and listen :slight_smile: As your work is refined so will your template, it’s an incremental and (possibly) endless process: learn from previous mistakes, modify, move on.

BTW you don’t need a slave computer to use MIR.

For my templates I’ll get the score of something that has a good dynamic range and the instruments I have access too. Then I’ll input the score into my DAW as MIDI data. That gives me some music as a reference point to start calibrating my template against.

For VSL I started with MIRx for reverb, but since the listening position is that of the conductor, the sound is still a bit off for my ears. I now route everything through one of Cubase’s IRs at about 60-70% wet and that seems to do the trick, blending the instruments much more effectively. The natural volume, EQ and pan presets in MIRx have also been hugely useful to achieving balance between instruments.

Getting accurate dynamics is proving a little tricker, as one setting for the dynamic scaler doesn’t seem to apply to everything. I suspect that this is because sample layers are different depending on what patch you use.