Where would those pixels live?
Since DxO is a non-destructive tool, it would presumably have to store them in the dop sidecar, which could soon become very bloated, especially if it held multiple virtual copies. It certainly can’t put them into a RAW file.
PhotoLab is very definitely not a pixel editor.
Mark
Endre,
This has been discussed before by @A77user. You also posted in that thread.
After you’ve used the perspective tool or the horizon tool and created those black slivers on your raw files you can export them to DNG, Tiff or JPEG. Those black slivers will then become part of the image itself and the Clone/Repair tool can be used to fill them in.
Additionally, regardless of the file type you are editing, after creating slivers, It must be exported to make those slivers part of the image.
In pixel editors like Photoshop you do not need to take this interim step. In parametric raw editors like PhotoLab and others, you do
See my example of accomplishing this in the thread below. I believe my example is the third post down.
https://forum.dxo.com/t/repair-tool-refuses-work-on-crops/19190
Mark
I’m sorry, but your “problem” statement is inaccurate. As I demonstrated to you in another thread this task can also be accomplished in PhotoLab after exporting the image to a DNG. Tiff or JPEG. There is no need to use other software like PhotoShop to accomplish this unless you prefer using their clone and repair tools more than PhotoLab’s. As a result I can’t vote for what I believe is an unnecessary enhancement.
https://forum.dxo.com/t/repair-tool-refuses-work-on-crops/19190
Mark
Hm… If it is not a pixel editor, is it a vector editor like CorelDraw or Illustrator ?
I think, if it would not be pixel editor, the export time must be almost independent from small or big file size RAW images.
Endre
Mark, you pointed to the solution. Let make PL a temporary file with all visible pixels on the preview ( similar to Nik Collection editing process ). It not necessary to be visible by user, just allow to add colors to the pixels - not existed before. Of course, you are right I can save the image with black parts and start to substitute the black pixels. Actually, it is still unperfect. At the borders of the image PL still unable to fill perfectly the territories.
I do not not want to fight for better cloning features, just confirm the existence of the need.
Endre
Agree. But the other corrections also are not stored in the raw file, especially not the masks from the repair/clone tool.
Endre,
Vector editors are primarily used for drawing or illustration software. As I said earlier, PhotoLab is a parametric image editor like Lightroom and Capture One Pro.
Mark
Sorry, I haven’t seen this statement. The parametric image editor as a class is new for me.
OK, I think all image editor should play role as parametric image editor, if the saving ( export ) format is different from the original. Even MS Paint is able to open and save 9 different image file format. If we treat all RAW file processing software as parametric image editor, it is based on the opening file format, rather than the software. Any image editor is parametric, if able open RAW files ? There are a plenty of image viewer ( XnView, IrfanView, ACDsee, NX Studio, Canon DPP, Portrait Professional, HDR software, etc.) works with RAW files. Especially the Panorama software always offer a stiching error fill, to make the image as big as possible.
Endre
Endre,
Opening and reading a raw file is not the same as applying nondestructive edits in the form of overlay instructions to it. I have lots of software that will allow me to open and read raw files. They are not parametric editors.
DPP, and probably NX Studio as well, don’t use a database or create sidecar files. Since they are both proprietary software owned by Canon and Nikon respectively.they are parametric editors because they store edits non-destructively in a specific area of their companies raw files. As a result those edits can only be read and applied to those images from within their own software. That is why you cannot see the DPP edits to a Canon raw file in other software. Parametric editors will work similarly on DNG, Tiff and JPEG files. They don’t actually update those file formats either. Edited files need to have the edits burned in to a new copy during export. .
Parametric editors, as a result of all this, generally have export commands to create a final image rather than the file save command.you will see in pixel editors like PhotoShop or Affinity.
Mark
Hi Mark,
It is a fact, the parametric editor is new description for me, therefore I accept its determination means they add more parameter to RAW files or just generate independent sidecar files. However, I do not think it would be impossible to determine the black parts color info. But, I agree with XMP sidecar use only, it is a far more complicated task.
I tested the NX studio now, how they treat the internal editing parameters. It is generating also a sidecar file at the same path in a NKSC_PARAM subfolder.
Endre
To me there’re two kind of image editors: pixel based and vector based. Pixel based editors change the values of the pixels in a predefined size. And so do the raw converters. Vector based images don’t have a size: they use the universe.
Another difference would be parametric or not. Vector based editors and raw converters are parametric editors: they save there edits in a list of used functions with their parameters. PL does this in the dop files or database, other converters have their own editlists.
The base function of a converter is the development of the raw info to an image. Adding info to it is more the world of image manipulating software.
Beside that a problem with a photo should lead to a learning of how to avoid that problem in the future. Been said before in this context: take more care while shooting and/or take care of more space to correct some faults.
George
Yeah ! It would be great to have an automated proposition for filling gaps 
K.
I’m astonished that anyone finds the cloning and repair tools good enough in PhotoLab 4 (haven’t tested them hard in PhotoLab 5 yet) to bother trying to fill in black edge of frame off-canvas areas. For fill, I almost always have to pass via Affinity Photo which does a beautiful job, difficult to detect. PhotoLab 4 leaves primitive blobs for me.
On the other hand, the PhotoLab repair tool is great for fixing skin blemishes, dust, etc. Just not for fill at this point.
I’d vote for better fill/clone/repair functionality first. If there were better fill/clone/repair functionality I’d vote for extending it to the black patches left over after horizon levelling and crop.
The ability to fill in those segments depends in part, on the image. Extending green grass is easy, while extending a segment of a wall is doable but not so easy.
It also takes a lot of practice and experience to use Photolab’s repair/clone tool effectively. I use it extensively, and over a period of time I’ve learned to use it successfully in a number of situations where other people seem to struggle.
This is not to say that there are not better and easier to use repair/clone tools available. There certainly are. It’s just that with time, patience and experimentation it can be a very effective tool in a lot of situations.
Mark