I would just not expect this behavior in a nondestructive raw developer. And honestly, I see no reason why they could not fix this. Just rotate all the local adjustments…
Usually I only keep my raw files and export different crops when I require them. Now I have to undo all my crops, save dozens of full size tiffs, and redo the crops with proper alignments. And keep those tiffs if I don’t want to repeat that procedure again. I don’t see any adavantage of a raw converter anymore.
Yes I agree with you both. This should be fixed. I was just giving a tip for new users to remember until it gets fixed(if ever).
It is not as simple as just rotating the local adjustments when geometry changes are applied. Take for example the application of a perspective change after local adjustments are applied. Not only would the location of the local adjustment mask have to be moved (not rotated), but in addition the shape of the mask would have to be altered to take the change in perspective in to account. The same would apply for lens distortion adjustments made after local adjustments were applied.
It would be great if all it took was a rotation of the masks, but its way more complicated than that. Those of us who have been using PhotoLab for some time know that changes to geometry should be applied prior to to local adjustment masks, and in fact those masks should be among the last things applied to an image.
Mark
Yes you are right, but the pixels are also all somehow transformed when you apply those lens corrections. The exact transformation can be applied to the coordinates of the local adjustments. It is simply an oversight, that had not been considered during development. To add it afterwards is a bit more work than if they had considered it right from the beginning. But at one time or another, they need to fix it.
Even if you do it first, there are tons of situations when you want to change it afterwards.
When I review old files I almost always make some modifications on them, and geometry could be one of them.
Having to redo all your local adjustments if you make a change is something that should not happen in an raw processing software.
The whole spirit of non destructive photo editing is that you can change some of the settings later without having to start from scratch again.
What you are asking for might be an appropriate expectation, but it is not an aspect of non-destructive editing.
Non-destructive image editing simply means that your original images, raw and RGB, are not actually modified during the editing session and those edits are instead burned into an output file when editing is complete. This is accomplished by writing the edits to a database or separate sidecar files during the editing session.
In some cases, proprietary software like Canon’s DPP, or Nikon’s NX Studio will write the edits directly into a specific area of their raw files which is separate from the raw shooting data. In that case those edits are only visible when using the proprietary software and also must be burned into an output file to become permanent.
Mark
@mwsilvers you cannot be serious, you are nitpicking about some formalities. Are you telling me that it would be acceptable for a nondestructive raw converter, that you after you had applied an exposure correction, you are not able to apply white balance corrections anymore? No it is not. What the software does internally and in what order, I don’t care, its up to the software developers. But as a user, I should not have to care.
This is exactly the advantage of nondestructive editing. We only save the instructions of what should be edited. The software will apply these instructions in it’s appropriate order. As soon as this is not possible anymore, we are discarding one of the big advantages of a nondestructive workflow.
Maybe… however for my use this is clearly a show stopper…
I won’t spend any time on local adjustments if any change on a geometry settings is going to ruin everything.
Image processing isn’t - for me - a linear process.
I am not nit-picking. Non destructive editing has a very specific definition. It was used incorrectly and I commented on it.
Mark
I understand how you feel and I can’t argue that PhotoLab may not meet your requirements in its current state. However, this issue is brought up so rarely that I suspect it is not a serious problem for most users. I have been a PhotoLab user for five years and use local adjustments and geometry changes extensively. Understanding those limitations from very early on, I haven’t really consciously thought about it in years.
Mark
So maybe it’s not the good term used… clearly what I want in an image editing software is to be able to adapt any settings whenever i want.
I need flexibility, my process is not linear but has to be very quick.
(that’s why also miss snapshots - that are not the same as virtual copies - and a before/after comparison where you can choose an intermediate state for your “before”).
We all have requirements which may differ from each other. Prior to moving to PhotoLab, I was a long term Lightroom user. While there are a few features in Lightroom that I would like to see implemented in PhotoLab, I am very happy I made the move and have never looked back in regret. I never found snapshots as useful as virtual copies and therefore don’t miss that feature. Am I wrong to believe that you prefer Lightroom? Perhaps that software would be a better fit for you.
Mark
If I can just chip in. If you used Photoshop, you have to first use the ACR module to convert from RAW into a bitmap that can be edited. Once you have created the bitmap, then you can no longer go back into ACR to adjust anything on the RAW level. This is a linear process and a limitation of Photoshop.
Because local adjustments are made in relation to the scaled and rotated image as it has been edited so far, further scaling and rotating would mean re-calculating all the geometry of the local adjustment masks so far created.
This is extremely complicated and I can perfectly understand why DxO might be reticent in allowing it
Whichever editor you use, things are never going to be perfect for everyone. If you don’t like this aspect of PhotoLab, maybe you should try another editor?
[quote=“mwsilvers, post:32, topic:11898, full:true”]
Prior to moving to PhotoLab, I was a long term Lightroom user. While there are a few features in Lightroom that I would like to see implemented in PhotoLab, I am very happy I made the move and have never looked back in regret. I never found snapshots as useful as virtual copies and therefore don’t miss that feature. Am I wrong to believe that you prefer Lightroom? Perhaps that software would be a better fit for you. [/quote]
I use both, Lightroom is my main software, and I use Photolab when needed as a pre-processor (with DNG output) as Lightroom is really bad to handle noisy images (it kills the color in areas with lots of noise).
There are lots of good software available, and it’s very a good thing that there are different.
On this topic, my remark is not a complaint (as there are several other blocking points for me to use Photolab as a Ligthroom replacement), just a mark of surprise… as I would consider this feature request as a basic requirement for any raw software with local adjustements.
That’s an interesting point in fact, cause that’s not something I use to try when I evaluate a software and that’s something I’ll do now… I don’t have a clue how the other software deal with this (except Lightroom of course).
I’ll also try it on the repair/clone tool…
I used photoshop (or Gimp at the time) on maybe 5 photos in 20 years 
Even in the “old days” without local adjustments I mostly sticked to what I was able to do with my RAW software.
I used other software of course, like Hugin for pano stiching - but yes it was a pain.
When you needed to change something on one step, all the processing was to be redone…
So you end up not updating cause it’s too complex.
Yes, however for a rotation I would say it’s pretty basic stuff.
For distorsion correction or perspective, this is probably more difficult as for instance even a basic radial mask would not be a perfect circle after the transformation…
So yes it’s a bit complex, but how complex is that compared to DeepPrime algorithms ?
I’m going to have to come down on both sides of this argument. On one hand I agree with @SojiOkita that the software should work as he stated above. Further, I recall a “ghosting” issue when the ReShape tool was first introduced in beta6. This was very similar in nature to the issue discussed here(a ghost image remained after using ReShape in conjunction with LA) and DxO fixed this issue promptly and without problems. I’m not sure what the cost in man-hours would be for them to fix the whole issue discussed here, it may be prohibitive. However, it could be done.
On the other hand I also agree with @mwsilvers that I have used DxO software for photo editing since Optics Pro 8 and have never run into a problem as long as geometry adjustments come first and LA/Retouch comes last.
I also revisit old photos from time to time to see if improvements can be made. I usually do this after each upgrade of PL. I’m never trying to keep them the same, I want them to be better. Each upgrade of PL usually has many new features that can make improvements over older versions. I mostly find that I can make better improvements by starting all over.
I agree that fixing this issue would be preferable, however, I don’t see it as a show stopper.
Mark
Definitely not a show stopper or even a show slower IMO.
yeah I agree rotating the mask is essentially more of an option than needing much effort. While many argued one should work on the geometry first, noone should rule out the possibility to revisit the geometry adjustment after other tuning - such feedback optimization is rather common to finalize a process.
When you have forgotten to work on geometry first, there is a workaround: first export as JPG, then make geometry corrections in VP.