HDR artifacts

Forum|Forum|7 years ago
April 19, 2018
3 replies
812 views

Lightroom Classic CC version 7.3

2017 15" MacBook Pro 14,3 macOS 10.13.4

I have been using Lightroom Classic CC to process HDR images and have run into some issues which I think may be due to dynamic range being too wide per final merged image. I am not 100% sure though and wanted to see if there were any tips about what I may be able to do better.

Is there a maximum recommended number of stops between exposures to merge to HDR successfully? I was recently in a dark house (walls painted dark blue/grey) on a very sunny day. When I exposed for the exterior from inside, the rest of the walls in the the photo were black. In about half of the merges, I encountered issues with strange graininess - especially around light fixtures - or even in one instance an error message saying that Lr couldn't blend the images because there wasn't enough similar content between exposures. Each merge contains 3 exposures, one slightly over-exposed, one three stops over exposed for shadows and dark objects, and one for the outside (about 4+ stops under exposed depending on the room). Should I have taken more exposures in those conditions for Lr to be able to merge images properly?

I was using auto align and had the ghosting correction off for this batch.

This topic has been closed for replies.

J

Jao vdL

Community Expert

In general you don't want more than 2 stops difference between subsequent shots as the alignment engine will have issues

exhaustinoAuthor

Known Participant

Thanks - I will try to keep that in mind at shoots going forward. I wish there was more control of the HDR process but it seems like I'll have to buy a specific program for it if I want that type of functionality...

cmgap

Community Expert

Do you mind posting a screen shot of the bracketed images in your grid view to give us a better idea of what you were working with? While HDR is a useful technique to employ under certain conditions it is not always going to work. HDR is most often (but not limited to) landscape photography because it works well. Seeing your images may help us in making exposure recommendations.

exhaustinoAuthor

Known Participant

Hey cmgap, here's the edit, a detail of the noise, and the set of thumbnails for this particular image. Let me know your thoughts. Thanks!

J

Jao vdL

Community Expert

Yeah, they don’t give us much control at all. If the alignment fails, there is nothing you can do. If, like what happens here, it uses data from the severely underexposed image in areas where it shouldn’t, nothing you can do. It works best if you have not too much EV difference between the images