Hi gang!
I am working on some enhancements to Aeolus integration with Zynthian which should drop in the next stable release. Aeolus is a pretty good pipe organ simulator. It uses generative algorithms to emulate the various sounds of pipe organs, not samples and is very configurable. Zynthian uses the default Aeolus configuration which presents a Grand manual, two Swell manuals and a pedal board. (Excuse me if I get things wrong. I am not a pipe organ specialist.) Each of these four keyboards / divisions has a set of stops controlling the sounds. There are also swell controls (although I donāt think they work in the current stable release). Each division is presented as a chain.
The current implementation has some constraints which I am removing, e.g.
- Any quantity of keyboards / division (1ā¦4)
- Any MIDI channel (not just 1ā¦4)
- Save, rename, delete presets
So why am I telling you this? I want your collective wisdom to shape some development, particularly around presets. I plan to provide two banks of presets:
- General
- Local
The presets in both banks are actually the same but selecting a preset from the General bank will restore stop values for all divisions whilst selecting the preset from the Local bank will restore stop values for the currently selected chain / division. This is similar behaviour to a real pipe organ that may have general pistons and local pistons. The preset names would appear the same in both banks (because they are the same - just a subset is restored if selected from the Local bank.) What is your opinion on this behaviour? I wanted something that aligned well with a real organ whilst providing the advantages / flexibility of a digital implementation.
The swell controls include:
- Swell / expression
- Tremolo frequency
- Tremolo depth
- Sustain
(Sustain isnāt really part of swell and isnāt really part of a real pipe organ.)
I had originally thought to include these in the presets (as we might do with other synth engines) but it occurred to me that might be undesirable, e.g. recalling a local preset for a division whilst playing another may have undesirable effect if the swell parameters changed. Maybe we only recall swell parameters from the General bank or not at all. What do you think?
The presets currently provided were dreamt up by our wonderful @jofemodo but at his own admission may not be particularly useful. I am struggling to find examples of presets (piston settings) that we might use as a default / factory set. It seems that most organ music scores seem to leave the orchestration to the organist with just a few suggestions, e.g. āf but sweetlyā. I have found a few examples of common settings from some organ lessons but I wondered if our lovely community of organ players would like to contribute some presets with their names and stop settings (and swell settings if appropriate). This could be in the form of lists of stops per division or as Aeolus preset files.
Zynthian presents 4 keyboards driving 4 divisions. Aeolus supports up to 6 (I think). Would there be an advantage to implementing a bigger organ with the ability to constrain the quantity of divisions when adding an Aeolus instance? I have used the default config with 4 divisions:
- Manual I
- Manual II
- Manual III
- Pedals
and allowed creation of 6 different layouts:
- Manual I+II+II+Pedals
- Manual I+II+Pedals
- Manual I+Pedals
- Manual I+II+II
- Manual I+II
- Manual I
on the basis that you will pretty much always want Manual I and addition of pedals is likely to be required for any combination of manuals. The list could grow by adding:
- Manual I+II+III+IV+Pedals
- Manual I+II+III+IV+V+Pedals
- Manual I+II+III+IV
- Manual I+II+III+IV+V
which is still a manageable list of combinations. My geeky, engineering side says that the flexibility would be good and there is likely to be someone who appreciates the extra keyboards / divisions. My project manager side says to avoid adding features that few (if any) users want and who is going to design those extra divisions? There are mega (physical) organs out there with even more manuals so it may be nice to have them but in reality, most of us only use at most two hands and two feet to play and with instant recall of local pistons, the extra manuals may not be needed, i.e. you can recall a preset for one of your three manuals rather than physically move the a fourth or fifth manual. My preference is to stick with the 4 manuals / divisions. Please let me know your opinion, preferably supported with real-life scenarios / workflows.
And finally, stop sniggering @wyleu.
Most players are totally ok with 2 manual (think home players and office), while adding the third one offers more pipes for flexibility for enthusiasts, curious, and concertists.