HDR Merge * Big Issues for Interior Photography
Been playing with the new HDR merge feature and there are issues.
I even made a video about it on my channel:
Worth the Upgrade? My Thoughts on Capture One 22 https://youtu.be/pJrsU7DTtCA
There are strange artifacts where highlights blend with shadows. Also, the HDR tool appears to work well for outside images (landscape & exteriors) but there are ghosting issues present… and without a way to deghost this tool is unusable.
-
John,
I think there are some aspects of your observations that are especially of concern to Fuji users if using Fuji Film Simulations. The film simulation for Fuji is specifically handled in the RAW files and here we have the concept pseudo-RAW files (not finished output files) using DNG. Presumably the DNG file type, perhaps combined with the nature of the recombined set of RAW files, does not currently allow easy support of the Fuji Profiles as currently implemented.
I have no idea how Adobe addresses the Simulations you mention but I would guess it is not the same approach.
In addition the user guide suggests using 3 images rather than the 5 you have selected. During testing I discovered that was good advice since 3 images with my Canon files gave more consistent results than more images. That was also true when processing panoramas, although of course there it was more a case of making sure one was not overloading joins with too much data. 5 or 6 images better than 8 for example.
For most things I have tried, the differences between using the new HDR merge versus to existing HDR tool especially running with a Linear Response curve have been minimal.
However, your needs would be more of a challenge for the pre-existing HDR tool alone working from a single shot. Combining that with your desire to deploy the Fuji simulations probably leaves you out on a significant limb at the moment.
From what you said in your video I will guess that the developers were originally intending (and may still intend) to create a highly controllable merge selection process but have elected (for whatever reason) to launch with an "AI" marketing proposition and simplified "time-saving" features.
Now, having spent time years ago working with some of the better thought of "manage everything" panorama and merge tools, spending hours trying to perfect the matchinig of a few jpgs (even more time with tiff files due to their size) I was delighted to discover, eventually, a much simpler solution that took away most of the need for the heavy lifting and still produced what for me were acceptable or better results most of the time.
So far I see the same thing with C1 but with some of the limitations you have mentioned being evident. On the plus side we get a "RAW" file to work with as the process output.
The learning curve, especially with the initial User Guide notes now available, should be brief.
The outputs will, in the main, either work or not work as required and people will be in a position to decide if the results are adequate or whether they will really need to go into "full control" mode which is so familiar to you in the workflow you described.
(As an aside I would guess that AI will, eventually, provide a facility that simply takes the input images as a guide to what is required, assesses what can be done with the data and, where it finds something that is too complicated to "fix" using original data, simply creates its own. A sort of CGI without user effort. My guess is that few people, if any, will care at the point of consumption of the image. Phones are probably doing that already.)
The thing is that if one has an existing workflow that involves complexity that one feels is necessary - as it probably is in your situation - it is likely to be quite difficult to accept anything less. There is nothing wrong with that, it's just human nature. In your case justifiable. Others may feel the need to have control when in reality they don't really need to spend the time they do spend pursuing details at the level they want to feel is necessary. They might be able to save considerable time for other things with very little compromise.
If I really need the control point management I still have the existing applications to use with tiffs or jpgs.
It seems reasonable to anticipate that further developments are likely to be forthcoming both from an AI direction for the simple approach currently offered and, based on your comments in the video, a more "hands-on" set of controls at some future point. All related to the creation of RAW-like intermediate files rather than TIFF or jpg.
Based on what you said in the video, after that early engagement with the development people you did not see the results until the release date. Is there any reason that you were unable to take a sneak peek during earlier testing phases? If I had been involved so early in the process I would be extremely keen to see the developments at the earliest opportunity.
0 -
Sadly I can confirm that the same ugly artefacts happen with Canon raw files. I am also an architecture and interiors photographer and honestly in this state, the HDR merge is pretty much useless. Simply because it is subpar compared to almost any competition.
0
Post is closed for comments.
Comments
2 comments