I always start writing at the piano (Kawai MP10) and then notate this into Sibelius, once the structure is in, I start to orchestrate it, then often the piano part is removed (unless the piano has a required part).  I only use Sibelius sounds at this stage, I have never found a successful way to get all my VSL library (VSL complete) to work with Sibelius, as the SIbelius template ends up being so huge, it takes forever to load, and that's before you have even written anything !

Once I am reasonably happy with the arrangement, I then create my "orchestra" in VI/VE Pro driven from Cubase networked from another PC. I don't load PRESET or MATRIX, I only build up my own matrices from the patches, this way I only need to load what I am actually using so that the matrices are not unnecessarily large, and it ensures that each of my pieces use what they need, and not what just happened to be in a template. The exception to this, is when I am arranging for a specific orchestra/band when I have to set the template up to match what instruments they have available, and so have to work within that.

Once I have it all ready in VE Pro, I play all the parts in realtime reading from the Sibelius screen into Cubase, then I add Expression maps to cubase and manually select all the articulations. I always use CC for this, I never attempt realtime key switching - it just doesn't work for me.

Once it all sounds good, I record the audio into Cubase track by track from VE Pro/MIR, so that I have individual tracks of audio. A little minimal processing done in Cubase (I don't want to upset what MIR is doing) and that is a  finished mix.

Thanks a lot for your answer! Much appreciated. [:)]

Marco

@marcotronic said:

...I can´t imagine playing stuff like that with so many different articulations live with keyswitches...

Hello Marco

You are right, it isn't possible to switch between articulations at any time. Sometimes you need a sfz for the first note of a fast run and the rest of it is played by the legato articulation.

You have no chance to do this fast enough by manually pressing keys.

Further: As Corte mentioned it as well if you want to simulate a true instrument you should change the articulation very often.

Here is a short example (No.21) - even if it shall show another matter here.

About my workflow

1. I always start a piece with midi-instruments. I use the Cubase-possibilities (VST-Instruments). I also try to get the tempo variations as close as it could be in the end.

2. Then I begin to exchange every single midi track by using samples (incl. X-velocity).

    If you have enough RAM you will be able to leave all the VIs till the end, so that you are able to adjust the tempo at any time.

    I'm using Basic-Presets for each instrument (Viol1, Viola, Flute etc) as a starting point.

3. I always do the mix with audiotracks.

    So that means: When the piece is ready for the final mix I produce audiotracks first.

    Then I delete all the VIs and save the peace as "Version-Audio".

    - Advantage: I have the piece for ever with audiofiles (with no effects). Who knows what systems we have in 2022... I'm independent this way.

    - Disadvantage: You need a lot of Diskspace (not a really problem today)

4. Now I'm doing the mixdown with effects, track-automation etc.

Finally two examples.

The first shall show that we should use X-Velosity with slow and long notes for keeping the dynamic in move.

The second one shall show what we can get with changing the articulations - sometimes for every single note...

Example: X-Velocity and tempo variations (Dormabain Trad. Swiss-Song) (fits to the current Christmas time [:)])

Example: a lot of articulations and variations of the tempo (Mozart: Eine kleine Nachtmusik)

Have fun

Beat

Thanks a lot, Beat.

Keeping the raw/frozen audio channels/files is a good idea ;)

I totally love your Mozart piece - the use of the different articulations is awesome. 

Marco

 Beat,

Your performance of Eine Kleine Nachtmusik is abasolutely brilliant. I assume you're using the Chamber Strings with Solo layered in?

Mahlon

@Mahlon said:

 Beat,

Your performance of Eine Kleine Nachtmusik is abasolutely brilliant. I assume you're using the Chamber Strings with Solo layered in?

Mahlon

Thank a lot and - yes. [;)]

Beat

I don't understand why keeping the unmixed audio files for future technologies is more practical than keeping the MIDI. Isn't MIDI much more pliable, and convertible to audio with a single click anyway?

@Errikos said:

I don't understand why keeping the unmixed audio files for future technologies is more practical than keeping the MIDI. Isn't MIDI much more pliable, and convertible to audio with a single click anyway?

Yes, the midifile is OK as well - for getting the tempo and note information in future...

Why saving the audio files?

Logic said "good by" to windows users with the version 5.5.1. in 2002.

Now, on the one hand I have the saved project files and on the other hand of course each final mix from that time.

Unfortunately, the "old" Logic (only 10 years) doesn't work properly with the current windows.

Sometimes I would like to use those projects for doing a remix with all the better effects of today or for extending them.

And because I used other libraries and synthesizers at that time it would be great to have a pure audiofile of all the instruments.

But - all of those projects are worthless today... after 8-10years.

So what did I learn:

I always save a midifile and also an audio file from each track of my projects.

So I'm (hopefully) more independent of all the licences, programs, computers etc in the future.

Beat

Yes that's a good idea to back up each individual audio track.  I used to do that, thought it was overkill and stopped doing it out of laziness.  Then I just recently needed the dry, un-EQ'd audio tracks and didn't have them for a new mix.  So it means going back to reconstruct the MIDI playback which is by no means instantaneous.  

It's really interesting how different people have a totally different workflow. As for me, I must say I'm not a pro like many of you and mainly compose music as a hobby just for myself. Anyway, most of the time I already have an idea in my head to start with and how I want the piece to sound like.

So I start with my basic orchestra setup in VE (strings, brass, woodwinds with the basic articulations) and put things together. Usually a melody starts to expand, give me ideas for the harmonies, then the harmonies will give me new ideas for new melodies and so forth. Most of the time a piece is growing over the time, and I work on the details and the full orchestrations and all MIR-positioning  right from the beginning, so I don't have a finished midi-song first.

Most of the time, I enter short passages with the masterkeyboard with basic articulations (not switching anything while playing) and put in the key-switches with the mouse later on, together with velocity x-fades.

When I'm done I put in my Suite-effects in VE, mostly EQs and some compressors, and I program some automation where I need it, mostly volume changes. After rendering, some mastering is applied to the finished audio file.

This may not be the best workflow, but I've come to like it and feel comfortable doing so [;)]