Intelligent Masking

[JoPoV]

Yes, possible.
Everything is possible since we know nothing.

DxO has been the first one (I think) to get in the Ai thing with denoising.
And has not yet shown anything in AI masking.
And it’s software needs refinements.

Indeed it seems they’re not in game but anyway I would bet on nothing here.

But I would prefer first to get current mask system at full capacity than have “the” mask tool before :
so any relevant tool in mask system : tone tone curves, fine contrasts (the 4 parts).
All current masks at full capacity :
control point with ellipse and color picker, luminance mask in pl, way to mix any mask together.
And other things , to get what exist at full potential .

But that’s me.

[Stenis (Sten-Åke Sändh (Sony, Win 11, PL 6, CO 16, PM Plus 6, XnView))]

@Stenis

Over 50 have voted for this feature request now and that is pretty good and more than I ever have seen before for any other request, so if DXO is checking the votes they ought to understand that many really want features like AI-masking in Lightroom and Capture One even in Photolab.

Now it is 70.

[Stenis (Sten-Åke Sändh (Sony, Win 11, PL 6, CO 16, PM Plus 6, XnView))]

@JoPoV

I’d like to add to that, for all the photographers who use these adobe processes : don’t complain when you’re out of work, because agencies will replace you with AI softwares that in a matter of seconds will come up with more proposals than you, and at a ridiculously low price : it’s you who will have given them this opportunity. And those Ai softwares will use your work without any compensation. Ironic, no ?
(And it has already begun).

The proletarization of photographers will continue and the prices of photographic pictures will generally continue to be depreciated. I see few other ways for photographers to survive in a long run than to get more efficient than they are today. So, the ones that do not use the AI-support available in RAW-converters and for example also Photoshop today and refuse to do so in the future will probably soon be out of business.

I´m not talking about generative AI here, just about all smart new AI-supported features that now gets developed in most other successful converters and photographic tools of different sorts.

Just a few days ago all that like me subscribe on Capture One got a new pretty interesting upgrade (16.5.0) of some of the earlier published AI-aided features in that program. It is not a revolution but a very interesting evolution.

An overview including the new People precision Masking features including the very powerful possibility to include masks in your own presets. Up to sixteen for one picture.

Capture One Pro Tutorials | Introduction to Layers & Masks

Even earlier there was a function called Smart Edit, targeting wedding photographers. Now it has got far more powerful and versatile and got renamned to Match Look. One reviewer started to think that this is the beginning of the end of presets as we know them. The new is that it now covers not just two variables but almost anything you want to do with your pictures dynamically.

Capture One Pro Tutorials | Match Look

Even the function AI Crop that came in the previous release I think can be a massive time saver for studio photographers taking product pictures.

Capture One Tutorial | Ai Crop in Capture One Pro and Capture One Studio

I have hard to see that there are so many others that has such a focus on productivity as Capture One seem to have had the last two or three years.

… and yes, studio-photographer have been faking backgrounds for over hundred years in their studios when taking wedding photos etcetera. So why shall they not be able to do that in a modern studio too? In Capture One´s Studio-version you can.

First picture from my grandfathers and gradmothers marriage 1917. The second picture is taken when my grandfather was eighteen. Both taken with a faked background, so that is nothing new and special at all to change backgrounds. Not really anything to get upset about under the right conditions.

[Wlodek]

“AI” masking lives in RGB world. PL has perhaps the best noise reduction on the market, manages colors “well” during the edits, and it’s simple enough and ergonomic to be productive for basic photo development. For high-end fine-tuning, use PL output in PhotoShop, Affinity, DaVinci, etc. For basic work, PL output may be good enough. It’s hard to say what DxO strategy is, but they seem to be focused on the image pre-processing science, which they do quite well, with some things still to tune and add. Maybe they should join forces with, say Affinity (+C1?), to make a complete product portfolio (??).

Off-topic:

@Stenis

The proletarization of photographers will continue and the prices of photographic pictures will generally continue to be depreciated

That has already happened few years ago, when many agencies fired a lot of experienced photographers, switching to photos provided by the readers. The quality went down, but who cares? But it seems some still do – anything extreme comes usually with a backlash.

@Stenis

I see few other ways for photographers to survive

What ways do you see? Some artists have to make their living from advertisements, and that market may soon vanish due to “AI generatives”. It may be like with my friend, who came from a family of “masters of typesetting art” going back few generations, and became himself a graphics editor. So, some photographers may become consultants.

@Stenis

AI-support available in RAW-converters

That’s mostly “ML” used for denoising and avoidance of demosaicking artifacts.

@Stenis

So, the ones that do not use the AI-support available in RAW-converters and for example also Photoshop today and refuse to do so in the future will probably soon be out of business.

If you take pictures of events (e.g. weddings, sports, etc), people will hate fakes, except perhaps for a few funny add-ons, where “AI” can ADD a creative touch. War time reporters are still a different story (e.g. common loss of touch with reality).

Currently most of things marketed as “AI” are really standard algorithms using (Deep) Machine Learning to get the “right” parameters. Sometimes the results are funny, as you can see from examples of object removal and inventing the background which was not there. It is still very primitive on average.

@Stenis

Match Look. One reviewer started to think that this is the beginning of the end of presets as we know them.

Did you use that? The direction may be useful but the devil is in the details, which may not work under very variable lighting. Will customers really care? The primary goal is productivity, and maybe there’s not too much to save here?

@Stenis

focus on productivity as Capture One seem to have had the last two or three years.

Much longer, since C1 started with processing studio photos from PhaseOne MF backs about 25 years ago (?).

Somehow humanity has managed with fire, electricity, and atomic energy inventions…

[NaturalThought]

Bump. DXO if you’re reading this I also want intelligent masking features.

[ColinG]

DxO does not actively (or otherwise) frequent this forum, so far as I am aware.

[JoPoV]

Here is what I heard on a youtube “webminar” this year from DxO evangelist about AI masking :

" … par contre le jour ou on va vous sortir quelquechose a base de selection D’IA ou je ne sais quoi, je ne sais pas si nous on sera les rois du monde, mais vous en tant que photographes vous devriez le devenir. C’est ce qui s’est passé avec deepprime, qui a révolutionné et qui revolutionne toujours aujourd’hui la gestion du bruit et de la haute qualité du dématriçage."

" … on the other hand, the day they come out with something based on AI selection or whatever, I don’t know if we’ll be the kings of the world, but you as photographers should be. That’s what happened with deepprime, which revolutionized and is still revolutionizing noise management and high-quality dematrixing.
deepl translated."

So wait and see.

[FarmerPhotos]

i watched a lightroom video this morning a New Zealand landscape photographer “dont not use ai masking during edits do old fashioned way”

He went on to say only ai that should be used is object removal