Page 1 of 1

Wind supply model preferences

PostPosted: Wed Aug 02, 2017 3:27 pm
by tf11972
On creating our sampleset, we have the issue that there is a "pumping" effect, when we are playing with most of the stops pulled out. It sounds like the wind pressure is going down temporarily.
Disabling the wind supply model solves this issue, but the sound becomes slightly static.
In the CODM it's possible to modify the size of the windchests. Does this help? Or what else preferences in the CODM you recommend to get a stable, but vital sound?

Thanks in advance for your help.

Re: Wind supply model preferences

PostPosted: Thu Aug 03, 2017 3:43 am
by mdyde
Hello Thomas,

Yes -- the values of the CODM Division.WindModel_... attributes indirectly determine the magnitude, frequency and damping of the oscillations that are likely as pressures fluctuate. If you haven't done so already, I'd suggest having a thorough read of the descriptions of those attributes in the CODM user's guide -- they describe exactly the effects that each parameter would be expected to have, including formulae.

For example, increasing WindModel_ResrvrBellowsPositiveDampingCoeff will make oscillations die away faster.

Re: Wind supply model preferences

PostPosted: Thu Aug 03, 2017 3:46 am
by tf11972
Thanks a lot for your quick reply, Martin.

I will discuss this with Chris.

Re: Wind supply model preferences

PostPosted: Thu Aug 03, 2017 4:10 am
by mdyde
Thanks, Thomas.

Re: Wind supply model preferences

PostPosted: Thu Aug 03, 2017 12:14 pm
by chr.schmitz
I have the actual measures of the original organ reservoirs. There are 2 reservoirs for 5 divisions.

Should I calculate volumes of reservoirs and sizes of reservoir tables, sum these values for both reservoirs and use them for all 5 divisions?

Chris

Re: Wind supply model preferences

PostPosted: Thu Aug 03, 2017 12:32 pm
by mdyde
Hello Chris,

The CODM uses a separate reservoir/regulator for each division (although that doesn't necessarily apply in the full ODF format).

Within the CODM: assuming each of the real organ's two reservoirs is used for some pipes on each division (e.g. flues and reeds) then I would be inclined to start by summing their sizes and volumes, then splitting that sum into five according to the approximate proportions of the total number of ranks that are allocated to each division. However, you will probably still find you need to tweak the values a bit to get fluctuation/oscillatory characteristics that approximate those observed from the real organ.