3DLuts, Perceptual rendering intents, and gamma

[Avery]

It appears that when creating a 3D LUT using the “Perceptual” rendering intent (as well as related intents such as Saturation, Perceptual Appearance, etc), absolute and relative gamma curves produce identical results– ie, a Perceptual intent 3DLUT with Bt.1886 gamma set appears identical to a Perceptual LUT with a pure power gamma of 2.4, and far closer to a Relative Colorimetric LUT with gamma of 2.4 than a Relative Colorimetric LUT with Bt.1886 gamma.

This is consistent across multiple profiles for two different displays for me.

Is this an intended feature or implicit effect of the intent, or a bug? Is there any way to create Perceptual-intent LUTs with absolute gamma curves at all?

(I believe I’ve checked DisplayCAL’s and Argyll’s documentation pretty thoroughly, so my apologies if there’s something about it I’ve managed to miss.)

[p.dada]

I confirm. I just created some SDR Rec. 709 3DLUTs for madVR.

BT.1886 with a Perceptual rendering intent produces a WAY darker image than either BT.1886 Relative colorimetric or Absolute with white scaling intents. It looks like a Gamma 2.4 TRC to me. I’m also wondering if this is a bug.

[S Simeonov]

Same here, it produces a straight pure power curve of 2.4 gamma.

[p.dada]

So, should we be doing something about it?

[S Simeonov]

So, should we be doing something about it?

Like what?

[p.dada]

I don’t know!? Can we debug it somehow? Has Florian stopped working on the software?

[Vincent]

Perceptual intent is “reversible”, hence I’d not expect any behavior that limuts that. For me it is working as intened.

[Author]

Posts