< Back to IRCAM Forum

Understanding a slight difference between oper and roomsize

Spat
#1

hi thibaut,

i m trying to understand a difference i observe in output value when i use roomsize, between what spat5.oper displays and output of the object spat5.roomsize.

In this screen you can see that for the same settings (room size and so on) if most of the ouput are the same, there are differences at early/cluster distr and modal density. That remains if you test other roomsize.

Capture d’écran 2026-05-19 à 10.50.54
Capture d’écran 2026-05-19 à 10.50.541554×1830 367 KB

thanks for the help!
-eric

#2

and there is another suspicious behavior
For room ==240, we have early distr == 0.4 as cluster distr
suddenly for room == 245 early == 0.6 and cluster distr == 0.7 … it’s a huge step, isnt it ? but maybe it’s expected ?

Capture d’écran 2026-05-19 à 16.31.49
Capture d’écran 2026-05-19 à 16.31.491018×536 65.3 KB

Capture d’écran 2026-05-19 à 16.35.01
Capture d’écran 2026-05-19 à 16.35.011012×548 66.1 KB

#3

Hi Eric,

Thanks for pointing this out.
Well, distribution of 0.4 versus 0.6 seems like “a huge step”, but since all the delay values are quite small anyway, I think the differences in actual delay settings is marginal.

We’ll check if this can be smoothed out in future versions.

Thanks,
T.

#4

  1. There’s a ‘problem’ in your screenshots :
    in spat5.oper, a reverberance of 65 actually corresponds to a decay time of 2 sec (2000 ms); however your spat5.roomsize screenshot uses 65 ms.
    This partly explains the differences you observed.

  2. That said, that might still be some minor differences between the two objects because spat5.oper further includes some ‘security clippings’ to ensure the values remain in the ‘best’ range according to the perceptual model. Some of these clippings are not included in spat5.roomsize.

Best,
T.