I was working on my composition today and was wondering how you guys create your final mix! I know everyone has it's own way, but i was just curious.
The way i do it, is to export every midi track to an audio file, store and organize it neatly in my project map, so i can find it back easily. Then i create a new project where i load all my dry audio files and build my composition out of it.
Then when my composition is ready, i load up four instances of Altiverb (one for every section of the orchestra), load up an preset, and play with the wet/dry slider! That creates nice depth and a real sense of space.
I know exporting every midi track to an audio file isn't the most convenient way of working, but it really helps reducing the amount of CPU power since four instances of Altiverb are quite demanding!
Anyhow, i was just curious how you guys create your final mix!
I most always render sections in stereo. Reason being I've created the continuity in the section during the performance and I want that to be written in stone.
Example: All strings ( other than basses ) rendered to 1 stereo track.
I consider the basses a section so they get their own stereo track.
Occasionally I've had to split the cellos out to their own track just to get a bit more control.
Using Sonar 4:
So I the usually have: ( all stereo tracks )
Harp ( not all of my pieces use a harp )
Piano ( not all of my pieces use a harp )
If I (sometimes) have ethnic or electronic instruments involved they also go to a buss of their own.
Sounds redundant, (1 stereo track to it's own buss ) but in Sonar I find it easier to work with the buss envelopes because of the way I can view them.
So all get sent to their own buss and are summed to a buss I call "The Conductor."
The conductor is then sent to the master buss as are the reverbs which are on FX busses.
So the conductor manages all of the sections ... and the master buss "A" manages the conductor and the hall(reverbs).
I use volume envelopes on the busses to fine tune the final mix, as most of it's done already in midi with CC:11 or just the velocity and occasionally CC:07 if 11 isn't giving me enough control.
The conductor buss gets a volume envelope also and is used to enhance ( or manage ) long term musical dynamics of the whole orchestra, most of which is pretty much also finished in midi, but can be made more radical, or calmed down a bit here, but the conductor.
The master buss "A" has a volume envelope to manage any last minute problems with the whole mix but usually has very little activity on it ... only the occasional tweak to stop an over at a peak or a slight rise or fall in volume to make a transition work better.
Occasionally I'll use a 6 band parametric on the master buss to boost a very slight bit of fundamentals or to roll of a tiny bit of highs or lows.
All of the section busses start out at 0dB as the mix is usually pretty close coming from the rendering since I render everything at consistent levels, never boosting an individual section to get more volume.
Mixing the sections is very easy. Most often only very slight changes in volume are necessary, but sometimes I find interesting alternatives not thought of during the tracking.
I usually spend more time on the conductors buss than anything else, other than tracking or rendering.
Steve's detailed response describes something in the ball park of how I work too. I Must bounce MIDI tracks to audio in order to hear a complete piece--my computer isn't able to play all the tracks I use (average between 30 and 60) in real time.
You use a notation program, correct? Finale I think you said? I'm not familiar with the detailed differences between using notation instead of an audio/MIDI program like Sonar, so maybe some of what you described is necessary because it's in notation? --Because I didn't understand:
"...The way i do it, is to export every midi track to an audio file, store and organize it neatly in my project map, so i can find it back easily. Then i create a new project where i load all my dry audio files and build my composition out of it..."
I guess you mean that in a notation program you have to export the audio instead of keeping it in the project by just bouncing down to a track?--It's all the exporting and then having to start a new project, importing the files--all that seems so round-about.
In Sonar, as Steve was talking about, the audio renderings of the MIDI tracks are kept right there in the project, and there's no re-assembling needed--I'm wondering why you need to "store and organize it neatly in my project map"--why aren't they automatically organized, even if you need to re-import?
In any case, I'm often working on new MIDI tracks while some finished sections have already been rendered to audio. Once I have all my audio, I often keep each instrument on its own track for most flexibility in mixing, but each section of the orchestra is going through a group bus. FX buses for each section's reverbs, as per Steve.
I use volume envelopes extensively, bringing the volume down to zero during rests, so that each section of audio on a track is wrapped with a glove-like envelope.
My main goal is to stay as flexible as possible. While I've attempted to do all the volume dynamics possible in the MIDI realm, there's more group balancing, soloing etc that can be made more effective while putting together the audio mix, using a combination of envelopes and console automation of faders.
I use spectrum anlayzers to determine what "frequency notching" needs to be made, boosting and scooping out tracks so they theoretically are fitting more nicely together like a sonic jigsaw puzzle. I'll use quite a few EQ plugins, but am sometimes forced to apply an EQ setting to a track, when I'm running out of CPU power for driving so many live effects.
The master bus will often have a limiter on it, sometimes a compressor, always an EQ.
Once I've made my 2 track mix down, then I go into Sound Forge to work on mastering the mix. That always involves maximizing the volume, but it's also one more chance to improve volume dynamics if needed, making crescendos even more dramatic, for instance. Huge spikes are toned down, and if normalizing is needed, I do that in sections. I'll sometimes find that my quiet sections really are just Too quiet-I need to bring a whole area of the mix up, and do that with a combination of volume tools in Forge and normalizing.
AND so forth--That's a roughed in sketch of what I do. I've found that each project has its own unique needs. I can't template everything, forcing material into being dealt with only in ways I have before.
I don't use a notation program, rbowser-! I use FL Studio 7. It's not the best program out there but i certainly like the workflow the best of the sequencers i tried.
What i mean by store and organize it neatly in my project map is like this:
This is the data map of FL Studio! When you click on Projects, you'll see this:
I created a map for my current project: A Hymn for You!
You see the project itself, and a map called Garritan Personal Orchestra. I always create maps of the sample libraries i use so i can store the audio files in them! This is what the Garritan Personal Orchestra map looks like:
As you can see, a neatly organized map by instrument. In these maps are audio files called Introduction Phrase, Mysterious Introduction Phrase.. it allows me to find the right audio file easily.
I've never worked with FL but in my setup I don't need to ~find~ anything since everything is organized by the Sonar project file.
I just open Sonar and do ALT+F+1 and I'm back to where I left off last work session.
I have 1 directory per song and the rendered audio and the project files for the song go in there. It's really pretty simple.
Thanks for the pictorial explanation of what you meant. But, Oh my!--You have to assemble all this on your own for each project--?? And in those Garritan folders, you don't mean that you actually copy the instrument .nki files into them each time for each project, do you? Your hard drive would become filled with duplicate sounds in no time. You must mean you link to the actual Garritan folder which is in your Program Files--?
Steve responded with a brief explanation of how it works in Sonar, and every DAW program I'm aware of--All the organizing of audio files, MIDI files etc is done automatically. We set up a "per project folder" scheme so that the song we're working on has a master folder with its name, and inside that are project files and a separate folder for the song's audio. All done automatically. Load a project in progress and, as Steve said, everything comes up in place where it should be--And there's no difficulty knowing how to find files for each instrument--they're all labelled in the project.
So it's looking to us like you're working in an extremely complicated way, having to do things which would be automatic in other programs and on which you'd spend exactly Zero time away from actually writing/performing music. Are you sure this is the way the programmers of FL Studio intended you to work? It's seeming---very primitive.
I don't copy the .nki files into every map for every project. The folders you see in the last picture contain .wav files of every midi track of my project.
In Sonar you bounce a track to a audio file that Sonar stores IN your project, but in FL Studio you can export a midi track to a .wav file that you can store where you want! These .wav files are stored in those maps for easy acces.
Interesting thread. So ultimately you have a final 2 track stereo .wav file with everything mixed down. What do you all do about mastering? I use a tool called T-RackS 24 - Stand-alone Analog Mastering Suite to create a final master. Link here;
Mastering is I think just fancy EQ of the final mix (correct me if I'm wrong Dan), but is something of a black art. Tweaking the presets can give me a really nice warm produced sound that I don't get with Sonar.
Oh Ok! I was worried you were thinking you had to copy the .nki files each time! lol.
"...In Sonar you bounce a track to a audio file that Sonar stores IN your project, but in FL Studio you can export a midi track to a .wav file that you can store where you want!..."
Yes I see, but you can do that in Sonar, Cubase, Pro Tools et al also. While working on a project, it's convenient to have the associated .wav files in one folder, the project's master folder. But exporting any .wav to anywhere you want is always an option. For me, exporting the 2 track mix down .wav file is important, putting that where I want.
It just seems like you're creating a lot of extra work for yourself--But you know, if something works for us, we tend to stick with that way of working until more efficient solutions come up.
I almost never use EQ. GPO sound great out of the box, but sometimes i boost a bit of the lows of GPO and maybe add a TAD more warmth if the piece calls for it.
But for anything else i just add Altiverb and export it to 192kbps .mp3!
My PC isn't the fastest machine in the world anymore and i can usually load 2 instances of GPO without loading the Steinway Piano! And i have more freedom to process the wav files if i need to do some ambient work.
As you said, everyone has his/her own way of working!