Masking complex structures for local adjustments

Hello all,
I intend to buy PhotoLab 6.3 and am just getting to know the program by using the 30-day trial version. I have already been able to learn many image development steps from instructional videos and the online manual, and have already developed some of my images and achieved quite good results.

Regarding the local adjustments I still have some difficulties and ask for your help or tips. What is the best way to mask complex structures, for example the branches of a tree or the grid of a power pole?

I attach two example images to which my question refers. They are the unprocessed RAW files directly from the camera, which I exported to JPEG without further development, since I can’t fetch the original RAW files here.

One is a power pole in the late afternoon intentionally photographed against the clouds, through which the winter sun is only dimly shining. I would like to mask the power pole to lighten the pole grid a bit with a local adjustment. How can I mask just the grid with local adjustments to make the brightening only affect the grid? I have not had satisfactory results with either control points or a control line.

The second image is a backlit shot on a winter day. Here I want to darken the background and brighten the tree in the foreground with local adjustments. Again, I did not get satisfactory results with either control points or the control line.

Do any of you have tips or help for me on how to mask the complex meshes in such a way that a local adjustment that is precisely delimited from the rest of the image is possible?

Many thanks for any advice.


1 Like

I’d be interested in any advice here too, I have problems with trees. Judicious use of the dropper with control lines and changing the Chroma and Luma values improves the results but I don’t think they’re ever really good. I suspect the software does the best it can with a deceptively complex ask.

@Kaffeepause

Hi & welcome here!
Apart from being difficult to get a clean mask for the tree and the power pole alone, both pics are taken in counterlight. I suppose, there is not much chance to only brighten the tree or the power pole – and btw look better as silhouette.

Instead, try to brighten the whole pic and then darken the sky as long you are working on the raw-file.

1 Like

There are several approaches you can take, depending on your needs.

DXO PhotoLab is not really designed as application for compositing work. if you need to replace the sky or something like that, best to do it in an application that is more suitable for that kind of work, such as Photoshop.

If you just want to deal with tones and colors, there are plenty of sliders and features in DXO PhotoLab that will do the job with organic feel to it. DXO smart lighting or selective tone adjustments, for example.

Here is a quick was to even out the tones, but since your image is heavily compressed JPEG, it shows in the final result. RAW file would work great.

If however you need to make a mask or selection of particular object, one quick and easy was that works in most cases is to use local adjustments and control points, optionally you can combine them with so called control line. Just click with control points on the image , and use shortcut key “M” to preview your mask as black and white overlay.

Than just add or subtract parts of the image you don’t want. You add by clicking on the image with your mouse and adding more control points and you subtract by holding down the ALT key on windows, probably its OPTION key on the max.

Than go to layers panel and use chroma and luma sliders for refining selections.

its not prefect but it takes very little effort and time to get a decent selection. However I would simply use other tools for most of the images, unless you have a truly select portion of the image you want to change.

HSL controls also work selectively. And contrast controls as well. So you can combine them for some pretty powerful adjustments and fine tune control.

By the way, you can keep adding adjustments as layers and if you want to darken the part of the image back to original state, you can just do it in a second pass. So its quite powerful and easy to do.

But I would advise you get most of the work done with just regular sliders and if you need compositing to do it in a more specialized app like Photoshop. But you can get a lot done in DXO, for sure.

5 Likes

If you only want just brighten power liens or something, my advice is to use brush and paint loosely over the power lines. than simply use sliders to affect the tones like shadows, and leave the mid tones and highlights alone and it will only brighten the power lines, with no need to make complicated precise selection.

Off course with proper RAW file you would get more control than I can show with compressed JPEG’s here, but it should be quick and easy to make such an adjustment.

@Kaffeepause

The best advise you have got so far is to do this layer/masking-job in a more suitable program for this than Photolab. Precision-masking or refining masks is not one of Photolabs best branches. Sometimes it´s possible to manage but it is often far more inefficient than in more suitable softwares.

Some do these kinds of jobs in Photoshop and I do them in Capture One. Today many other RAW-converters has these kinds of tools too. To do what you want to do is very easy in Capture One with “Magic Brusch”. Even Lightroom has effective tools for this now.

We can´t expect anything at all to happen in this area in Photolab until next Christmas - best case, since DXO only makes bigger upgrades once a year and the last upgrades has had almost all focus on adding XMP-data support and an image database á la Lightroom, even to Photolab. I guess that´s the backside of their licensing model. Even Capture One has gone into the subscription business model and the upside of that is that they now release improvements when they are ready to be released instead of saving them to the big release of the next version in the end of the year.

1 Like

@Wolfgang

Hi Wolfgang, thank you very much for your welcome and for your hint.
When shooting, I slightly underexposed so that the bright areas wouldn’t burn out, and was hoping to lighten the dark areas since I shoot in RAW.

Admittedly, these images are extreme examples, but when it comes to developing images, I don’t like to be limited by the developing software.
With the RAW developer I’m currently using, there is a kind of “magic wand” that masks the grid of the power pole and the tree accurately enough to lighten both. I was hoping there was something similar in Photo Lab for complex structures like this.

Thanks again for your tip, I will try it out.

Hi Kaffeepause :slight_smile:

you can try the Auto mask, but …

maybe better use a Control Line, set the picker in the sky, use < M > to activate the mask view and adjust Chroma + Luma at the RHS (rechts).
To include the clouds just continue from the active CL and add more CLs + pickers (or Control Points instead). When it’s getting too complex, add new CLs + picker (or Control Points) …

There is no magic AI or whatever wand in PL.

1 Like

@MSmithy, thank you for your helpful tips. I’m not interested in doing image composition with PhotoLab or doing sky replacement. I would actually do that in another program. My approach was just making the gray color of the pole stand out a bit (just very slightly) and solving this with a local adjustment. Since I always shoot in RAW, the image information is available for developing the photo (unfortunately I couldn’t upload the original RAW files and so I uploaded an undeveloped JPEG version instead).

Before posting my request here, I had tried without success to create a reasonable mask with control points for the power pole.
Your hint to pressing the option key to subtract areas from the mask was very helpful.
I then experimented with the control line and the eyedropper and finally was able to create a satisfactory mask from the combination of control line and subtract control points (screen shot of the mask is attached).

@Stenis
I have been using Capture One Pro in the perpetual license for several years. Now that Capture One is moving to a subscription model, I’m looking around for a perpetual license alternative in a timely manner. Subscription software is absolutely out of the question for me and is way too expensive when I ignore the advertising bells and do the math. All software that has been changed to a subscription model, I have replaced with other programs with a permanent license.
PhotoLab is a very interesting candidate for the switch and I would like, given enough time, to convert my workflow to the alternative with the perpetual license. Since I already developed the image of the power pole in Capture One, I wanted to develop the same image in PhotoLab for comparison purposes and struggled with the masking of the pole grid.

Thank you all very much for your kind help.

1 Like

Attached is the current development status of the image as I was able to convert it to PhotoLab with my beginner skills. There is still much to learn and I look forward to learning new things.

Off topic, though you did bring it up :grinning:

DxO PL6 Elite + FP6 + VP4 >= ÂŁ400 in year one
Capture One Pro >= ÂŁ179 in year one
Adobe Photography Plan (20GB) >= ÂŁ119 in year one

Your maths only (nearly) works if you skip a complete version each year and can get a half price upgrade the next (which I admit is likely)

I don’t understand your argument on cost, though I fully support your right to avoid subscription models if that’s what you want

Nice.

One thing I noticed as a side note, is that often simple changes of sliders can produce more organic-looking results than precise masking, at least when it comes to tone and color processing. Composition is off course different operation.

Capture one and Lightroom or Adobe Camera RAW , especially on the Adobe side have the AI assisted masking tools that can at time be a good way to precisely change something, but I rarely find myself using it because I can get virtually all I need in DXO. I still sometimes use Capture One and Lightroom or rather Adobe Camera RAW,mainly because DXO does not always provide support for all cameras and lenses or is quick about releasing it. I don’t complain too much because they are often much better corrected. However, sometimes clients send me files that are shot with unsupported camera and I need to retouch the images, so I still sometimes use Adobe Camera RAW or Capture One.

Also, as I’m sure you know, if you still use for example Photoshop, Adobe Camera Raw is the same as Lightroom in much of its functionality and one can export linear DNG with all the DXO corrections and use that raw to further develop it in Adobe Camera RAW , Lr, or C1, After that one gets all the benefits of DXO plus any other DNG reading raw processing apps.

Personally, I use DXO PhotoLab for about 90% of my work and only use Capture One in rare occasion where I need a specific preset like some InfraRed simulation stuff or if a new lens or camera is not yet supported by DXO.

Anyway, welcome to DXO user base and cheers!

I have also rejected to step into subscriptions as far as it has been possible. A lot of people seem to have got upset on Lightroom and Adobe and now they are mad at Capture One too. Soon there is only DXO and Photolab left but for how long?

I haven´t got all that upset yet as long as they don´t mistreat their old customers that prefers to stay on the perpetual licenses. I upgraded to C1 23 some months ago and I will be able to live with that for many years to come. It hasn´t changed a bit. I got an update just some days ago and I will get some more maybe later this year but after that I will have to upgrade or just settle with version 23.

I had the intention to leave Photolab last year just because I was really tired of the lack of a more sophisticated layer-based editing system than the ones in Photolab. I really love all the masking centered tools in C1 and the refinements they offer but there were other things I found I couldn´t really live with. For me Photolab has an edge in efficiency and overall image quality and Deep Prime is irreplaceable but likewise important is the superior workflow and integration with Photo Mechanic and the XMP-support in Photolab 6.

So, I will continue to see Photolab as my main developer but I really need the much better precision in the C1 masking tools for a small fraction of my images and the new smart and efficient “Style Brusch”-function they released in version 22 has somewhat levelled a lot of the efficiency problems I had with C1. That function has automated and hidden a lot of the handling issues with all the layers that was cumbersome earlier. In fact, that improvement completely changed how I work with C1. I love it and will continue to use it when Photolab is to dull fot the job. No need to panic over changed license conditions.

From Capture One in February 2023

About perpetual licenses

Once you make a full one-time payment for the Capture One software version, you obtain a lifetime license. You can use this license to activate the version you purchased for, and for any older versions.

Subscription off course means you are leasing the software that can be taken away from you, spied on you, deny service, and you don’t really own the software. So the cost is far higher if you consider that. Plus as we have seen with many subscription services and Adobe is certainly one of them, you basically pay for the privilege of being a beta tester. Instead of waiting off stable release, these subscription-based companies tend to release rapidly, buggy software, which is for all intended purposes still in beta phase. And then paying subcribers report bugs, acting as beta testers and paying for the privilege. Furthermore the companies will always increase price , not lower it and in the case of Adobe for example, much of the profits is not even in the Creative Software business as it is in marketing B2B business that offers cloud services. And this can be seen in the way the treat the customers and how they are pushing rainbow politics in their apps. Just recently, Adobe bought a competitor to their UI building software that was not working. 22 Billion $ I think. I mean why, innovate when you can buy any competitor and ruin the culture of innovation.

And in some cases like Espon or Canon printers I think, you buy a printer and they rent you ink. So when you stop paying for subscriptions , even if you have ink in the printer the chip say, sorry old boy, you will own nothing and be happy.

So thanks, but no thanks. The data mining and one side power play of subscription models are far, far more expensive than the price tag you listed. Something a wise persona thinking long term, does not overlook.

I mean Adobe try to make people pay for Pantone colors, remember that? When you see subscription cost on the website, that is just the tip of the ice berg. The real cost in underneath. Consider that.

I Have to Pirate COLORS Now?? - Pantone Connect + Adobe split

Sure, we can use control lines to do some masking jobs but it makes me think of more using a hammer to hammering down screws instead of using an electric screw driver,

Since there obviously is a Control Line Tool I wonder why DXO didn’t develop a more general smart masking tool too. Most of the code must really be there already, isn’t it?? It seems that all the air just went out of that balloon before the release of version 6.

1 Like

@MSmithy
The real danger with the subscription model is that lasy fat companies will get even lasier and more fat of all the money that just poures in without needing to do any thing at all really.

I don’t know where you picked up what you wrote about printers and ink but I don’t understand it at all.

Adobe has made hostile take overs before. Around 2006 they were shaken by the fast success of RawShooter made by the danish company Pixmantec. So they bought it and just killed it. One of the strength of RS was a superior preview quality. They gave us the then mediocre Lightroom 1.0 for free and then it took more than 15 years for them to achive something near the RS previews. Despite they bought Pixmantec they didn’t seem interested at all in learning from those danes.

I guess the developers they were uninteresred in started to work for Capture One in Copenhagen instead for Adobe and there they made Capture One to what it is today - a far more sofisticated converter than Lightroom is. Still the image quality in Lightroom is beyond the one both in C1 and Photolab. Lightroom users still have to buy DXO PureRaw of that reason. Pretty remarkable if you ask me.

… but now it seems that even Capture One seems to have got lacy too … or just greedy.

How true. We see it quite often.

“Why HP Can Remotely Disable Printers If You Cancel a $4.99 Monthly Subscription, Tweet goes viral fueling criticism against HP”

That was one article headline of many similar ones, including other companies. Basically they hold users hostage by subscription model. Its insane.

And its a nasty trend of you will own nothing and be happy by WEF.

Apple Is Working on a Hardware Subscription Service for iPhones

Company wants to make owning phone like subscribing to apps
Service could tie into AppleCare and digital service bundles

https://www.bloomberg.com/news/articles/2022-03-24/apple-is-working-on-a-hardware-subscription-service-for-iphones

I don’t know what you have been doing with your software, but that simply is not correct. You must be doing something wrong. I’m pretty confident PhotoLab does a lot better job then Lightroom does in terms of detial, noise reduction, and optical corrections. To claim otherwise would be factually incorrect, regardless what you subjective opionion of it might be.

Capture One is similar to Lightroom in many ways, but they each have weakness and strengths. PhotoLab on the other hand is in a league of its own. Just fact. I can show you examples, but you can do them on your own if you like. As everyone can. I can only imagine you have been using PhotoLab wrongly if you are getting inferior results, or you had unrealistic expectations to begin with.

You can argue about subjective matters if you like, but demoiseicing, noise reduction and optical corrections are not subjective attributes. Color is.

…or desporate. Who knows.

@MSmithy
Thanks for the info about HP. I have an Epson P900 and still the old business model is used there.

Sorry, excuse me for my sometimes-imperfect English. As you can see in the part just over the one you cited I ment the opposite:

“Adobe has made hostile take overs before. Around 2006 they were shaken by the fast success of RawShooter made by the Danish company Pixmantec. So, they bought it and just killed it. One of the strengths of RS was a superior preview quality. They gave us the then mediocre Lightroom 1.0 for free and then it took more than 15 years for them to achive something near the RS previews. Despite they bought Pixmantec they didn’t seem interested at all in learning a few things from those danes.”

So of course, I mean that the image quality is better in both previews and ready-made images than the one in Lightroom.

To make myself clear on the subscription business model:

I don´t give DXO more than a year or two before they will start to use subscriptions too. The reason is “the rule of diminishing return on investments”. In a market with mature products like the one of RAW-converters, it gets harder and harder and more and more expensive to “Wow!” the users and make them voluntarily upgrading every year. Company after company will hit that wall.

For a while DXO might have got a few subscription migrants from Adobe or Capture One but I wouldn´t migrate of that reason alone since in the short run nothing changes if you sit on a perpetual license despite what these companies are doing. No need to hurry if you haven´t got a new camera without a profile support! In the long run a small company like DXO can´t afford earning a lot less than Adobe and Capture One per user and at the same time having to spend more than them on R&D in order to get their users to upgrade.

I´m pretty sure we will also see a diminishing developing speed both at Adobe, Capture One and DXO. The losers are the users. Photolab 6 had it´s obvious development focus on XMP, the implementation of a better ImageLibrary and the support for Fuji-cameras which targeted an important part of the market that earlier was completely inaccessible for DXO. I hope the next version will be a substantial lift in the layer-functions and masking tools but beside that I have really hard to see any real important changes that will persuade the users to upgrade the next November-December and the users also know by now that there will always come a Black Week in that part of the years that brings offers with 30% discounts. So less and less of the user base are really paying the regular upgrade price these days - also a reason to implement a subscription business model. So, we shall not fool ourselves - because it´s just a short matter of time before also DXO forces their customers to subscribe.

I see a close future where the interest for new features in this kind of software decreases. I remember the extreme interest we had in my country in the early nineties for new computers when 386-computers came. The same when new releases come of Word and Excel and Windows. I was a Product Manager for all Microsoft products at a big regional IT-product distributor these days. Who cares about Windows and who cares about Office today? I still use Office 2007! Yes, and there is no problem with that. That´s why Microsoft since long have switched to the subscription model for MS Office. … and who except hard core gamers care about what kind of processor there is in a computer today?? … except for people upgrading to Windows 11 maybe.

Det är möjligt, men vi bör också tänka på det i bredare termer. Att vara ett av de få företag som inte uteslutande bygger på prenumerationsmodellen kan bli en konkurrensfördel om alla andra kräver enbart prenumeration. Vi har sett detta tidigare med några mindre företag som inte kunde konkurrera med de stora företagen, men deras främsta sätt att utnyttja konkurrensfördelar var att erbjuda fantastiska innovativa funktioner och ingen prenumeration.

DXO är ett sådant företag. Affinity är ett annat. Inom videoområdet är Blackmagic ett annat exempel. Och de var alla mer innovativa och började få fler och fler användare. Lojala användare. Det finns alltså något att säga om en användarbas som i princip inte vill ha en prenumeration. Det är en möjlighet för vissa medelstora eller mindre företag. Tror du inte det?

…

It’s possible, but we should also think about it in broader terms. Being one of the few companies that are not exclusively based on the subscription model can become a competitive advantage if everyone else is mandating subscription only. We have seen this in the past with some smaller companies that could not compete with big companies, but their main way to leverage competitive advantage was to offer great innovative features and no subscription.

DXO is one. Affinity is another. In video space Blackmagic is another example. And they all were more innovative and started gaining more and more users. Loyal users. So, there is something to be said about a user base that on principle does not want to be on a subscription. It is an opportunity for some medium size or smaller companies. Don’t you think so?

DXO is one. Affinity is another. In vidoe space Blackmagic is another example. And they all were more innovative and started gaining more and more users. loyal users. So, there is something to be said about user base that on principle does not want to be on subcription. It is an opertunity for some midium size or smaller companies.

@MSmithy - thanks for your input!

We´ll see! I sincerely hope you are right but I doubt it, because these economic under toes are so strong that they will force most companies to follow the flow.

I have personally upgraded my Photolab every year for many years now. Not because I always have found the value to motivate that but just to support DXO as a splendid and so far, superior alternative to Adobe. … but the last years I also have seen some worrying signs where DXO hasn´t been all that responsive to our input. If I compare them to the even smaller Camera Bits who makes Photo Mechanic, they have a lot to learn from Camera Bits. So, the DXO support really have to step up a bit if I shall continue to voluntarily support them in the years to come. They have to improve their user dialog. It´s really not to ask for too much. If a successful one-man company like Hamrich Software (manufacturer of the market leading scanning software Vuescan) can make it, it just must be possible even for DXO to do it.

1 Like