Horizontal/ vertical tilt correction

This might be a bit pie-in-the-sky but as a photographer I’m not particularly skilled at judging correct horizontal/ vertical alignment whenever I take a shot where there may be certain distractions that inhibit my ability at getting this right.
Thus during editing the usual correction procedure will have to be undertaken in order to properly straighten the composition. This usually means that depending on the size of the error a certain number of pixels will have to be sacrificed as a result of the necessary cropping that must follow.
Adobe Photoshop has a useful tool called Content Aware Fill which can often help in minimizing the amount of pixel wastage.
What I think would be a really natty feature would be a special tool which can do all the necessary corner fill work completely automatically once the user has made the desired geometric correction.
Would this be possible using currently available technology?

It’s not supported in Photolab at the moment. It may be too difficult to achieve reasonably in a raw editor rather than with Photoshop which is a pixel editor. In any case it only works effectively if the area to be filled in doesn’t have a lot of recognizable detail.
Mark

I often end up with night sky or grass areas which need to cloned in for my football pictures after levelling and cropping. I do the RAW development in PhotoLab leaving the black areas and then clone and fill in Affinity Photo.

It would be lovely if PhotoLab were to include good enough clone and fill tools to be able to do this within PhotoLab but there’s no rush. Could come with v4 as the pixel tools improve.

1 Like

Actually, as it it turns out, I was able to do it in PhotoLab 3 after all. I purposely skewed a raw image, finished all the processing and then exported it to a 16 bit tiff file. I then used PhotoLab 3’s great new clone tool to fill in the black areas of the tiff file. It won’t work on a raw file because the black areas are not recognized as part of the image. The attached image was done quickly just as an example of the possibilities. It wasn’t automatic, but it only took a few minutes. Obviously, the results will vary depending on the type of content that needs to cloned. The original image was captured at the Grounds for Sculpture in Hamilton, New Jersey.

Mark

Skewed image

Black areas cloned in

8 Likes

Well done Mark.
My motivation for putting forward this suggestion was induce the user to stick with using DXO PL for as many of his processing requirements as possible.
Obviously with the ever-growing range of alternatives at the disposal of the image editor the temptation to multiple edit is quiet significant.
It would therefore make economic and therefor marketing sense for a company like DXO to encompass as much of the processing chain as is practicable.

2 Likes

I get it Mark, very clever.
Can I steal the idea for my tuto ?

Pascal

It wasn’t really my idea. The subject was brought up before in a different thread and Alec mentioned doing it in Affinity using a clone tool in this thread. Since I believe he has indicated having difficulty installing PL3 on his Macs I guess he hasn’t had the opportunity to review the new clone tool yet.

Of course you can use it though. You can use my examples if you like although I did do it quickly just as an example, and it is not as clean as I would have liked. My cloning was a little sloppier then it would have been If I had taken more time.

Ce n’était pas vraiment mon idée. Le sujet a été abordé précédemment dans un autre fil de discussion et Alec a mentionné l’avoir fait en affinité en utilisant un outil de clonage dans ce fil de discussion. Comme je crois qu’il a indiqué qu’il avait de la difficulté à installer PL3 sur ses Macs, je suppose qu’il n’a pas encore eu l’occasion de revoir le nouvel outil clone.

Bien sûr que tu peux l’utiliser. Vous pouvez utiliser mes exemples si vous le souhaitez, bien que je l’aie fait rapidement à titre d’exemple, et ce n’est pas aussi propre que je l’aurais souhaité. Mon clonage était un peu plus négligé que si j’avais pris plus de temps.

Traduit avec www.DeepL.com/Translator

Mark

2 Likes

Merci Mark pour la traduction :wink:

Of course you can use a third software.
The best argument to use your solution is try to do so in PhotoLab.
Keep in mind that is not its first usage.

Pascal

Great experiment Mark. Bottom left corner is a bit weak (leaves cut off, visible lines). This kind of work is still better done in Affinity Photo, or Photoshop (hopefully CS6 so the photographer is not stuck paying Adobe monthly fees for the rest of his or her days).

Jumping in too deep too fast ends up in existing tools and performance being neglected. A slow PhotoLab with more features is infinitely worse than a fast PhotoLab with a more limited but focused feature set. I’ve heard PhotoLab 3 improves performance as well as adding better clone and better colour tools. That’s the way to do it.

If it weren’t for the OS fiasco I would have installed PhotoLab 3 already (I bought it day of release). As it is, I’ll probably have to refund it as after weighing the options and reading about the nightmare of trying to keep DOP files working properly in PhotoLab 2 if PhotoLab 3 is installed I’m not inclined to try to have both of them installed at the same time. Fortunately PhotoLab 2 is still pretty awesome.

I missed to say: Cancel the Lens Sharpness correction on opening the TIFF file.
Pascal

Thanks. As I indicated I did this quick and dirty just to see the feasibility of using the clone tool in PL3 for this purpose and I wasn’t concerned about getting it 100% right. Additionally there is a small learning curve to get the best from this tool and this was the first time I tried using it. If I were to attempt it again, I feel confident I could end up with cleaner and less obvious results. At first I didn’t care for their implementation of the clone tool but after learning how to use it properly I now like it very much. It does require a bit of practice to get natural looking results, especially on larger areas.

Sorry you can’t install this version. After what we went through with the lack luster PL2 implementation, IMHO this one makes up for it. It’s far from perfect but I think it’s moving in the right direction.

Mark

2 Likes

But if you are using 2 or more programs in tandem to perform an image process this will quiet likely demand more in system resources than using just 1.
On the other hand you can of course complete one step using one program, shut this one down on completion of that step then start up the next one to complete the task. But don’t forget to set your stopwatch.

It’s not always just about how long something takes. Some people are just more comfortable doing things a certain way, especially if they will be doing multiple things in each program.

Mark

2 Likes

Nice trick to export as a TIFF file to be able to clone outside the original image :slight_smile: !

2 Likes

My hardware is very well kitted out in terms of memory. The laptops all have 16GB and the desktops between 32GB and 128GB. 4 core i7 on the low end and 12 core i7 Xeon on the high end. All production drives are SSD.

If you are working with image or video files, the price in terms of time very high to work on underspecced hardware. I try to avoid large sessions on my laptops, mainly triage there. Of course all this is hardware is no excuse for developers to get sloppy and to code poorly, counting on the latest hardware at all times. DxO is somewhere in the middle in these terms. PhotoLab performs acceptably well on even the laptops, as long as the source files are not from my 5DS R. But then those files are slow on my 12 core desktop as well.

In any case, running PhotoLab with Affinity Photo with both open is not really a problem. Chromium with twenty or thirty tabs open uses more memory than either of those. Shutting down one’s browsers before a photo edit session is the most important single performance step of which I know.