For a while now, I’ve been using virtual copies to compare different edits of a single RAW image. Most often, I want to compare different “Color rendering” settings - film emulations, DCP profiles, adjustments with the generic rendering… So, I make some adjustments, including setting a color rendering; then I make a virtual copy, select it, and change the adjustments a bit; then, I make another copy from this one and make other changes. Then I start switching between the Master and the various Virtual Copies. More often than not, at least one of these switches back-and-forth won’t be smooth: the Viewer will first show the image with no adjustments applied, then abruptly shift the image significantly to the right and down, then return the image to its proper position in the viewer window with the adjustments applied. The problem is repeatable; but then if you stop for a minute and return to this, the problem goes away. Create another virtual copy and make some adjustments, and the problem returns for a while, then goes away again. Keep switching and eventually PhotoLab crashes.
I just opened support ticket #196900 to have this investigated, but also want to know if others are experiencing this and see if the computer hardware could be making a difference. I have 16GB RAM, a GeForce GTX 1060 3GB video card, an older Core i7 CPU, and a 1 TB SSD drive. Windows 10 x64. Thanks for your kind help!
I think this is the first time I’ve ever seen PhotoLab crash. The jerky switching among virtual copies is nothing new for me, though. Actually, it’s the eventual smoothing out of the switching back-and-forth that is new to me with PL3.
Thank you for trying that, Svetlana! I suppose the jerkiness might be connected with particular adjustments, too? Did you happen to see that the transitions from one virtual copy to another weren’t smooth and quick enough to make a good comparison?
If I may provide some additional feedback related to crashes: When PhotoLab crashes, a window pops up saying that the program has stopped working. The message in the window asks that I click the send button to provide DxO with a report, open a support ticket, and paste an alphanumeric code into that. I did that, and still support is asking me to provide a tremendous amount of information, including crash logs. I see no indication that what I submitted according to the instructions in PhotoLab was looked at or considered. This means I’ll be doing a lot of work on this when I find the time and makes me think it was right to open this thread as well. The info I was asked for isn’t unreasonable - it’s quite thorough; I just don’t like having to duplicate my own efforts if that isn’t necessary - in which case, DxO should change something about how crashes are handled.
I haven’t been asked for a test case by the support agent, but will be providing it,
Yep, it happens from the very beginning while cache file is not created yet, after that the switching is smooth on my PC.
As for the crash report itself, can you, please, remove the logs from here %UserProfile%\Documents\DxO PhotoLab 3 logs and then try to reproduce your crash. As soon as you gen the crash window, please, press “Do not send” button (in this case the crash report is stored on your PC here %UserProfile%\Documents\DxO PhotoLab 3 crashes).
Then attach it to this post and I’ll create an issue and combine it with your support ticket. In this case we will be able to start investigation faster.
P.S. I’m testing exactly with your image+VCs - no crash (so the crash report is a must have to find the cause).
Thank you for confirming that. Yesterday brought new updates to Windows and nVidia software - I’ve installed those and will see if I can crash PL. But I still want to know what goes to DxO when I press Send, if not the crash report. If that operation isn’t useful, and we should instead “Do not send,” why ask customers to Send?
I can reply, it’s exactly the crash report that is sent to our server. But there is one thing - they are sorted in folders by ExceptionPattern and not the number of the report which makes it rather tricky to find exactly yours only by the report ID (it’s done so because the crashes are analyzed by the patterns). That is why we sometimes ask you to provide the report or logs to speed up the investigation.
Well, sometimes crash log is actually not enough to find the reason of the crash: we can also need configs, presets, images, sidecars etc. And it all depends on the exact situation and case. So, please, provide the data the support asked you to make investigation faster and efficiently.