hi again, knitting some dropped postings together...

peter sikking wrote:

yahvuu wrote:

actually a question for peter (yahvuu): how complete is this overview?

most notably, the porter-duff modes are not listed. I'll have a look to make the overview as complete as possible.

I am interested in that. modes are like a box of chocolates, you'll never know what you gonna get.

the overview has been updated, new URL is: http://yahvuu.files.wordpress.com/2009/09/table-brightness-1600b.png

turns out the porter-duff modes mostly differ by alpha handling. Ignoring alpha they match one of 'normal', 'nochange' or 'black' at 100% opacity.

However, a few other new modes were collected. We should now have everybody and their brother's sister aboard.

Alchemie foto\grafiche wrote:

between mode used in PS and not (yet ?) included one of the most request would be Exclusion

to second that:
"One such 'killer app' is its ability to capture midtones in images, and its ability to isolate the edge regions of soft masks." http://www.digitalartform.com/archives/2009/06/exclusion_photo.html

additionaly, 'exclusion' is quite cumbersome to substitue by building from other modes.

peter sikking wrote:

peter (yahvuu) wrote:

yes, the evil plan which spun off my posting is to get rid of the layer mode concept.

funny enough I had realised before your mail rolled in that that is impossible, for the following reason:

the layer mode controls the mathematical operation used compositing of that layer. there is no _reasonable_ way (right now) to achieve the same results in a different way. here the fundamentals of what it does prove it to be unique.

yep, that's absolutely true on the technical side. The goal is to hide that from the UI: blending could be presented as an adjustment of about ten parameters. Which is much more fine-grained than choosing from a limited set of blend modes. Furthermore, it seems feasible to provide parameters with artistical meaning, similar to brightness, contrast, saturation, etc..

peter sikking wrote:

maybe there are simply zero arguments to add modes...

each new mode adds one finer step for adjustment of blending characteristics.

If i want to darken a layer by itself, the curves tool allows nearly infinite
different characteristics of darkening. If i want to darken a layer by another one, there are a total of 3(!) characteristics available: 'darken only', 'multiply' and 'burn'. Of course, there won't ever be enough layer modes, which is one reason why they have to die long-term...

no, your analogy is wrong.

layer mode sets the mathematical operation.

- if you want control over the strength, that is what layer opacity is for
- if you want fine control over the characteristics curve, use curves on the layer
- if you want to mask where what is used, use the layer mask

that's an interesting question wether all layer modes can be constructed from a few basic blend modes and suitable curves.

There are several answers: - strictly mathematically, in general: no, the layer mode determines a different quality of blending than any combination of curves on the single layers can control. Even if applying all three possible curves: one each for the base layer, blend layer and the result. A counter-example is softlight_ps, i think.

- strictly mathematically, within the set of common blend modes: yes, mostly, but the required curves are very technical in nature, so this approach is anything but user-friendly. Two examples:

glow(a,b) = b*b/(1-a) is the same as applying a curve g1(x) = 1/(1-x) on the base layer, and a curce g2(x) = x*x on the blend layer, and finally using the 'multiply' mode, f(a,b) = a*b.

Using three curves, vividlight can be constructed via hardlight: - apply g1(x) = (1-x)/2 on the base layer - apply g2(x) = 1/(2x) on the blend layer - apply g3(x) = 1-x on the result (adjustment layer)

So i don't think that clever use of curves can make up for lack of control over blending characteristics. The layer modes repertoire should be as rich as the UI can bear: to provide many nuances of blending, not so much for collector's ambitions.

Interesting insight gained: the curves tool can be an even sharper weapon if values outside the 0..1 range are allowed. IIRC, that is already on the radar. With GEGL under the hood, early clipping is not an issue, so a simple curves offset is sufficient to seemlessly explore the range between 'additive': a + b
'grain merge': a + b - 0.5
'dark addition': a + b - 1
and possibly beyond these three blend modes.

greetings, yahvuu