I see. I didn’t want to come in as impolite, I just recalled this a few days ago as something I would have loved to see completed. I apologize.
Everything related to this feature is here, you could say that this is really dead and unless someone wants to continue it I doubt that the original developer would continue it.
I recently merged this into the latest release branch of Krita and did a local build using the docker build instructions, spat it out as an AppImage to run on Linux. Everything went fine, but the comments above are right that it adds a huge number of blend modes, the drop down is a little unwieldy >w<
However.. this is also the most excited I’ve been about a Krita feature in a while. These blend modes are incredible, as an artist that makes heavy use of the “Greater” blend mode, as soon as my build was done I spent about 6 hours streaming with these and I’m never going back, I will merge this patch and self-build every new Krita version that gets released. This is making me faster at doing soft/sloppy coloring for commission sketches, and the results just look better. I’m using fewer layers. It’s just a joy to work with. We are missing out by not having this in Krita proper.
So, all of the comments above are valid. This is too many blend modes to throw in at once without separating out the alpha channels.. BUT.. if you’re an artist and are capable of running your own build for your system, then sit down and use this thing for just 3 minutes and you will immediately see what I mean. It is incredible, this makes Krita a more capable art program. There are some amazing features coming up in Krita 5.3 soon that I am very excited about (I cannot wait to start using the new comic panel tools), and I bring that up only to emphasize how good this feels to paint with - that I think this is a more exciting feature than most of the stuff I’m looking forward to.
Now, this has been sitting as an open pull request for 2 years at this point (or close to it); it is pretty obvious that the full version of this is not going to get finished any time soon. X3
But this doesn’t have to be all or nothing. I would really seriously advocate for merging in one of these modes. They’re all neat, but just having “Wet Normal” would enable a lot of the same stuff in a round-about way - because you can still use layer blend modes to mimic many of these effects (and in some cases you still might want to because once you reach 100% alpha, the color mixing will still require you to do some smoothing to avoid ghosting effects and there’s no real way to mimic the greater effect for color pigments on a single layer). :3
Perfect is the enemy of the good; obviously I would LOVE to have the full separated blend modes that were being talked about up-thread. But.. take the existing pull request as is, fix the very, very minor merge conflict on one line (I just went with the head), and then delete or comment out all of the registrations except for “Wet Normal”, and you’ll end up with one of the most useful brush blending modes in this entire program. That can be merged. The other blend modes can get uncommented or come back later when somepony who’s good with C++ and Qt and has the energy has time to handle the alpha separation.
”Wet Normal” on its own (even if it’s the only blending mode added) is a major differentiating feature for Krita. I am not aware of another drawing program that has an equivalent to this.
I’m speaking as somepony who draws inside Krita as my full-time job: using this feels amazing. The “Wet Normal” mode (as it exists today with zero changes other than trimming down the other registrations) is an easy way to get a killer feature that allows for coloring/blending workflows that are not otherwise possible to do to the same level of quality. I don’t know enough about Krita’s processes to know what else is required for this - if it needs more tests, if there’s additional cleanup that didn’t happen.
But if it’s only the UI that’s the only blocker, then.. I dunno, this should be merged.
A possible solution would be to allow the user to select the blend modes they want, similar to how brush bundles can be disabled.
Definitely, but I think even that requires making a UI, and it doesn’t seem like anypony is available to do that. What I’m getting at is that - if anypony feels comfortable looking at tests or anything else that needs to be signed off on - the extra blend modes could be commented out and this could be merged down as is, with no changes and no additional UI work.
And definitely there can be additional UI work to make it better in the future, but Krita is losing out on an amazing feature by waiting for somepony to have the time to do that.
The smallest version of this is have one more blend mode in the dropdown menu, have that be the only one, and then it could be shipped. If we think about methods to actually solve the problem of having too many blend modes in the dropdown, then this isn’t going to get merged probably. So I want to caution against actually trying to solve the problem - I think the solution is that we don’t need a solution right now, one more blend mode won’t make or break the UI, and while Krita shouldn’t get into the habit of adding more of them until the problem is solved, this one is worth shipping earlier.
Personally think that Deevad’s idea could work best, specially considering how digital painting-focused is Krita, maybe they could be the default for the watercolors and/or the mix brushes, and assuming that you can’t select more than one blending option, the drop-down menu could just be reserved for the wet mode(s) (assuming it’s not a hassle to code in that is)
And who knows, could also be useful to set terrain for a realistic blending mode (or something close to it)
I love Deevad’s idea, and I think it makes a ton of sense, and would minimize maintenance work as well <3
BUT.. is anypony available to build it? Because this pull request has been open for two years. :3
If somepony here is willing to submit a pull request with the new UI, amazing. I’ll support it 100%.
If talking about the new UI is going to mean this doesn’t get merged, then the UI is a distraction. I think this should get merged into 5.3. I don’t know enough C++ to literally build a UI for it and create a new pull request, but if there aren’t additional test requirements, I could create a pull request or.. I don’t know the process, try to update the existing one? That’s not me saying nopony should do a UI if they have the time.
It’s me asking, is there somepony here who can commit to taking the time to build a UI in Krita’s existing codebase and submit a pull request? I can’t commit to that, I don’t have the time or expertise. And if nopony can step up and say “yes, I could do this and I want to”, my opinion is that we should merge one of the blend modes into 5.3, and leave it open for whenever somepony does have the time.
I don’t want aspiration to get in the way of one of the most exciting features I’ve seen in Krita in a while.
Having played with the the Wet Normal blend mode for a little over a week with commissions, this is straight up changing my “sloppy” coloring process in a way that’s a lot faster, uses fewer layers, and (imo) looks better than what I was doing.
I can’t promise when I’ll have time to do this, it might still be a few weeks at least - I need to find a period where I’m not focused so heavily on commissions. But when I have time I’m going to clean this pull request up, fix the merge error, and re-submit it with only the Wet Normal blend mode added.
I don’t really know C++ super well (I don’t know entirely how the code here works), and I don’t have a ton of experience with Krita’s codebase. If anypony feels they are more situated for this and they have the time to look at it, I would welcome somepony else jumping in - I certainly don’t know enough to add tests, and importantly don’t have the time to learn this stuff.
But, I can fix a merge error and delete some extra lines of code, so if that’s all that needs to happen then I’m happy to make a new pull request, and I’ll link it here once I have time to do so. The pull request I submit will have zero UI changes. It will add a single blend mode (“Wet Normal”) to the dropdown and nothing else. If anypony knows anything else that was blocking with this other than the need for a UI to handle multiple blend modes, let me know so I can avoid wasting my time on it lol X3
It would be amazing if a Krita dev could help you out. This sounds like such a good feature.
Could you please share some examples (even simple ones) showing how this improved your work? It might help getting more people on board.
Looking at this from afar, could it be that “Wet Normal” is the same or sufficiently similar to the “Marker” from Paint Tool SAI? I’ve implemented that in Krita, see this thread for an implementation, instructions and downloads for testing: Feedback Wanted: Marker Mode - #10 by Drawpile
(tl;dr instructions: the Marker mode is in the Mix category, use it with a brush that uses Build-Up painting mode for it to work like in SAI.)
The majority of my artwork is NSFW and can’t be shared here, but I’ll try to look for some SFW examples or break down more what I’m doing when I have time.
It does look similar, yes.
I like having this as a blend mode because I regularly toggle it on and off for individual pens - I don’t like the idea of needing to create new brushes for each mode, or needing to open up the brush setting each time I switch.
Having said that, in the same spirit as what I said above, this mode is SO useful that it just literally does not matter how it gets implemented or where. If you have momentum to get your change merged, do it. I have no loyalty to any placement or UI for this, I don’t care how it’s implemented, this is a feature that Krita should have. <3
At this point, I’m just not willing to debate (or contribute to a debate) about whether the feature should be in brush settings or in blend modes, or have its own separate sub-menu. If it gets added, I think artists will love it no matter what, so I just don’t think it’s worth looking at anything other than what approach currently has the most momentum to get merged.
Well, if Marker mode and Wet Normal are the same feature, then we obviously don’t need both in Krita. And if one of them gets merged, the other merge request will need to be adjusted.
It also probably makes more sense to use existing terminology. The term “Marker” already exists in other software, so we might as well call it that in Krita as well, rather than coming up with a new term “Wet” for it.
Like I said, I don’t really care which request gets merged, I just want one of them to get merged.
Any debate about where the feature should go or what terminology to use I think can happen after the feature ships - I don’t think we’re locked in stone with any of this.
It’s a really great feature for artists - we can polish things up and debate that stuff after its in their paws/hands.
I think the two focus on different goals. “Marker mode” may be similar to “Wet Normal”, but @SirPigeonz has also created a series of other things such as “Wet Parallel” and “Wet Multiply”.
Its core logic should revolve around a series of alpha-based blends, such as Wet/Wet-smooth/Wet-smooth Lite/Wet-dry/Wet-dry Lite.Subsequently, he combined these with ordinary color blending modes; however, this resulted in a somewhat large number of combinations.Therefore, in the end, he decided to split the blending of color and alpha into two separate options（Later, he left the forum）.
The mixing formula for alpha should be here:
I’m aware that the original concept adds a bunch of other stuff as well. I did read through the thread, code and merge request before asking.
I’ve tested this in a build myself now: yes, Marker and Wet Normal serve the same function, namely not building up opacity on the layer you’re drawing on while still blending colors. Both work best in build-up paint mode.
I do want to note, the pull request I’m thinking of making here would only add the “wet normal” mode - so afaik there’s almost no difference at all between what @Drawpile is proposing and what I’m proposing other than implementation details. It may be worthwhile to go with the original pull request here so that the option to add more blending modes is still there in the future (I think that @SirPigeonz set up some templating stuff to allow new blend modes to be built on top of this which look like they could be pretty useful - I’m not the best C++ coder but it doesn’t look like the Marker mode has those templates).
But in terms of placement/naming.. looking into it a bit more I think there’s really no difference at all for the user between the two pull requests being proposed here other than what section the mode gets put in and what it gets called.
I just checked, and the Wet Normal option shows up in the brush settings where Marker will show up. And Marker should (I assume) also show up in the blending drop down so that addresses my concern about quicky switching between them.
The big difference to the user will be that one of them will be called “Wet Normal” and one of them will be called “Marker”. I don’t think this is worth fighting about, it just seems.. so easy to change the name later if we add the other blend modes back in or continue to expand on @SirPigeonz’s ideas about a separate alpha mode.
I have not looked at @Drawpile’s code in more detail to compare the exact algorithms (I have glanced at it, but.. again, my C++ is very very rusty); I would like to leave the work @SirPigeonz has done to make this support being templated into other color blending modes - I think there’s something really cool here that hasn’t been explored where we could start to experiment with a lot of different alpha modes. So if that isn’t something the other pull request supports, I’d vote for merging this one (or probably better, having @Drawpile pull the relevant code changes to support templating into their pull request and merge that one, since I can’t stress enough I’m not really qualified to make a pull request for this).
BUT.. in terms of the name.. like.. yeah.. let’s just call it Marker and put it in the Mix section. <3
I’ve run into more than a few CSP users who don’t know the Marker blend mode exists there, so using the same name might even provide more attention for the feature in CSP. X3
I’ve got no issue with any of this, I’m down for anything, I just think we should get this into Krita. I don’t think it makes any sense to debate the name, “Marker” is fine.