It is currently Wed Dec 01, 2021 1:20 pm


Approximating a commercial ODF's wind model in the CODM

Using the CODM to create your own organ definitions, exchange CODM organ definitions, ...
  • Author
  • Message
Offline

larason2

Member

  • Posts: 403
  • Joined: Thu Feb 04, 2016 9:32 pm

Approximating a commercial ODF's wind model in the CODM

PostWed Sep 08, 2021 1:05 pm

I have been working on a new CODM file, which is working well, however, the sound of the pipes doesn't have the same quality or volume as the original ODF file (the commercial file). I know how to increase the volume and the wind model volume of a rank, and how to adjust the volume of the wind model for each division, as well as adjusting the harmonic levels, etc. however I'm wondering if anyone has any tips for duplicating the settings found in the ODF? When I just increase the volume or wind model level, it doesn't approximate what I hear in the commercial ODF very well.

When I go through the ODF of my custom organ, the relevant section appears to be the "<Pipe_SoundEngine01_Layer>" I am considering duplicating the relevant settings of each pipe in the commercial ODF into the same or a similar rank of my CODM ODF. I know that editing ODF's isn't officially supported, but any advice about how to recreate a given commercial ODF's sound as closely as possible using the CODM would save me a lot of time! I could also do more experimentation with the CODM, but there are so many parameters to the wind model that experimenting would take a lot of time as well. I have found that adjusting the wind model in the CODM is complicated, and that changes that I make don't always translate into an obvious improvement in the resulting custom organ. I also don't want to over wind or under wind the pipes, so deciding on a division air volume setting or how much to adjust the wind model settings for a particular rank is difficult to do.

Thanks for any help that could be provided!
Offline
User avatar

mdyde

Moderator

  • Posts: 13719
  • Joined: Fri Mar 14, 2003 1:19 pm
  • Location: UK

Re: Approximating a commercial ODF's wind model in the CODM

PostWed Sep 08, 2021 3:44 pm

Hello larason2,

If the differences you're referring to apply even to sustained notes a few seconds after any notes have been pressed, or released (i.e. if the differences are 'static' rather than temporary changes as notes start/stop playing), even when only a few pipes are sounding at once, and if the sample sets are modern-ish (19th/20th century) organs then the differences you're referring to are probably simply differences in voicing, unrelated to the wind model. If so, you could try to replicate the sound by adjusting values via the voicing screen for your CODM organ (and/or via the equivalents in Rank attributes in the CODM ODF).

Usually organs' wind supplies (especially 'modern' 19th/20th century ones) mainly impart 'wobble' to pipes as pipes start or stop sounding. Some sample sets of much older organs do also intentionally impart more variation even when notes are sustained (e.g. modelling hand-pumped bellows, which produce irregular 'wobbles', or by modelling wind supplies that are barely sufficient, resulting in some noticeable loss of pressure when a reasonable number of pipes are sounding at once).

The CODM's wind supply model is intentionally simple, being defined only by attributes on the Division and Rank tables. The CODM documentation covers the effects of those attributes fairly extensively (especially for the Division table). For more complex wind models than that you'd have to use the 'full' ODF format, as covered in the sample set creator's documentation, but that's complex, and I'm afraid I'll have to leave you to experiment with that if you want to go down that route. Hope that's of at least some help!
Best regards, Martin.
Hauptwerk software designer/developer, Milan Digital Audio.
Offline

larason2

Member

  • Posts: 403
  • Joined: Thu Feb 04, 2016 9:32 pm

Re: Approximating a commercial ODF's wind model in the CODM

PostWed Sep 08, 2021 5:42 pm

Thanks Martin,

That's very helpful. Given your answer, in the division table, what is a rule of thumb for the volume of the wind supply per rank for a 16 foot, 8 foot, etc. that will create a good wobble, but not sound artificial, for each division.

Also, for a given rank on the rank table, what is the difference between increasing the rank volume and increasing the wind model volume? Is it possible to create artificial effects by increasing the wind model volume too much? Should you increase them both at the same time, and by how much relative to each other if you are increasing volume generally?

Thanks!
Offline

larason2

Member

  • Posts: 403
  • Joined: Thu Feb 04, 2016 9:32 pm

Re: Approximating a commercial ODF's wind model in the CODM

PostWed Sep 08, 2021 5:44 pm

Another thought I had - many organs are winded such that if you are holding one pipe open, and open another key, the wind supply in the pipe that is open will wobble, and this contributes to articulation in the pipe that is open. Is this modelled in the wind model? Is there any way to replicate this in the CODM by adjusting wind model settings? Thanks!
Offline
User avatar

mdyde

Moderator

  • Posts: 13719
  • Joined: Fri Mar 14, 2003 1:19 pm
  • Location: UK

Re: Approximating a commercial ODF's wind model in the CODM

PostThu Sep 09, 2021 6:00 am

Hello larason2,

I'm afraid I don't really have time to help with this myself -- very sorry, but I'll try to give some very brief pointers:

larason2 wrote:Given your answer, in the division table, what is a rule of thumb for the volume of the wind supply per rank for a 16 foot, 8 foot, etc. that will create a good wobble, but not sound artificial, for each division.


I can't give a quick answer to that, but have a read of the documentation for the following attributes:

Division.WindModel_TotalResrvrAndWndchstVolMetresCubed
Division.WindModel_AreaOfReservoirTableMetresSquared
Division.WindModel_ResrvrBellowsPositiveDampingCoeff
Division.WindModel_WindchestPressureDropPctAtMaxLoad

... and perhaps try some of the values from some example CODM ODFs, then adjust them to taste, based on how the documentation for those attributes says they will affect the result. Also see the sample set creator's guide for guidelines on configuring wind models.

The CODM generates one parallel-rise weighted (not sprung) wind regulator (reservoir) and one wind-chest per division, with the two connected together, and all of the division's pipes on the wind-chest. It automatically calculates appropriate attributes for the generated ODF objects (WindComparment, WindCompartmentLinkage, pipes, etc.) to ensure that the regulator will maintain a pressure of 6" of wind to the pipes under static load, so as to try to make the CODM simple to use and understand.

larason2 wrote:Also, for a given rank on the rank table, what is the difference between increasing the rank volume and increasing the wind model volume? Is it possible to create artificial effects by increasing the wind model volume too much? Should you increase them both at the same time, and by how much relative to each other if you are increasing volume generally?


As with the equivalent voicing parameters,

Rank.Amp_PipeMIDINoteNum036/96_WindModelModDepthAdjustDecibels
Rank.Pitch_PipeMIDINoteNum036/96_WindModelModDepthAdjustPercent
Rank.HarmonicShaping_PipeMIDINoteNum036/96_WindModelModDepthAdjDb

... would not normally affect pipes' amplitudes/pitches/brightnesses under normal static loads -- they primarily affect (turn up/down) the magnitudes of the any temporary/dynamic 'wind wobbles' imparted by the wind system.

Likewise for any of the Division wind model attributes in the CODM ODF. The wind model mainly affects dynamic behaviour (as pipes start/stop sounding), not the normal static sound of sustained pipes.

If you're referring to Division.WindModel_TotalResrvrAndWndchstVolMetresCubed, if a division has more/lower-pitched ranks then it will draw/need more wind in general, so you would normally want that attribute's value to be bigger for it than for a lower-wind-demand (fewer/higher ranks) division, unless you're specifically trying to make it unruly.

larason2 wrote:Another thought I had - many organs are winded such that if you are holding one pipe open, and open another key, the wind supply in the pipe that is open will wobble, and this contributes to articulation in the pipe that is open. Is this modelled in the wind model?


Yes -- that's the main effect of a real organ's wind supply, and of Hauptwerk's wind supply model. Since the CODM generates one regulator/reservoir and wind-chest per division (rather than having multiple divisions' wind-chests sharing the same regulator), in the CODM pipes from one division won't normally interact with those from another. However, in the full ODF format, the topology of the wind system can be whatever the ODF programmer wants. E.g. if I recall correctly the full St. Anne's ODF has the Great and Pedal divisions' wind-chests fed from the same same regulator, so pipes on those divisions will interact with each other.

larason2 wrote:s there any way to replicate this in the CODM by adjusting wind model settings?


For pipes on the same division it will happen inherently anyway, but not between divisions in the CODM (as above). Adjusting the CODM's wind model attributes (especially on the Division table) will affect how the 'wobbles' behave (magnitudes, frequencies, durations, etc.).

For more detailed information, I'd suggest studying the sample set creator's documentation for the wind model thoroughly (as well as the CODM Division attributes' documentation), and looking at some example ODFs (e.g. by exporting them to SQLite).

I hope that helps, but I'm afraid I really will have to leave you to your own devices now -- sorry again. Or perhaps other people on this forum with CODM/ODF programming experience may be able to offer advice if needed.
Best regards, Martin.
Hauptwerk software designer/developer, Milan Digital Audio.
Offline

josq

Member

  • Posts: 870
  • Joined: Tue Jan 24, 2012 12:11 pm

Re: Approximating a commercial ODF's wind model in the CODM

PostThu Sep 09, 2021 6:54 am

Perhaps oversimplifying:

CODM wind model gives wobbles/oscillations like a tremulant, but the amplitude of the oscillations is (gradually) decreasing over time

Division.WindModel_TotalResrvrAndWndchstVolMetresCubed - large value: you need to play more pipes to hear the oscillations. Small value: large oscillations even with low wind demand

Division.WindModel_AreaOfReservoirTableMetresSquared - large value: fast oscillations.

Division.WindModel_ResrvrBellowsPositiveDampingCoeff - large value: only a single oscillation upon increase of wind demand. Small values: many oscillations, more tremulant-like.

Division.WindModel_WindchestPressureDropPctAtMaxLoad - large value: a lot of detuning when wind demand is high.

A tremulant-like effect can be realistic for historical organs. You can cleary hear the effect at the end of this video recording in Groningen: https://www.youtube.com/watch?v=HObZ2P5towk. Unfortunately, this effect is not very accurately reproduced in the sample set.
Offline

larason2

Member

  • Posts: 403
  • Joined: Thu Feb 04, 2016 9:32 pm

Re: Approximating a commercial ODF's wind model in the CODM

PostThu Sep 09, 2021 8:29 am

Martin and Josq: that’s very helpful, thanks so much! I will spend more time with the sample set creators guide and the CODM documentation, but this has helped a lot already.
Offline
User avatar

mdyde

Moderator

  • Posts: 13719
  • Joined: Fri Mar 14, 2003 1:19 pm
  • Location: UK

Re: Approximating a commercial ODF's wind model in the CODM

PostThu Sep 09, 2021 8:52 am

Thanks. You're very welcome. Hope you manage to tweak it to your taste.
Best regards, Martin.
Hauptwerk software designer/developer, Milan Digital Audio.

Return to Custom Organ Design Module (CODM)

Who is online

Users browsing this forum: No registered users and 0 guests