Similar sequential captures (time lapse) behave differently to edit controls.

Forum|Forum|2 years ago
February 26, 2023
3 replies
2913 views

Camera: Fujifilm X-H2s

Camera Raw: 15.2 version on Win 10

Bug: please compare the two files and adjust the high lights, they are not on the same range

I’ve been using Fujifilm X-H2s for a few months, I did a few times of time lapse, I found the raw files sometimes were not on the same exposure, I thought it was the camera’s problem or my wrong operating, until then, I picked the two files and sent to my friend who is using Mac system, and he showed me they were on the same range. Even I asked another friend to try on Win System, he got the same situation, so now I am reporting this to you, any problem please send me Email, <Removed for Privacy>, thanks.

The system didn’t allow me to upload raw files, please check this link: https://drive.google.com/drive/folders/1Vmw345gNFjmF2urbsUqL2lC7e5usRwMT

BUGXH2S.zip

Windows

This topic has been closed for replies.

Rikk Flohr_Photography

Community Manager

The Camera Raw team has reviewed and determined this is not a bug.

"Certain controls are sensitive to the image content because they automatically set their ranges and behaviors based on the range of tones in the picture. While these images look quite similar, their distribution of shadow content is different because of the position of the water. This causes the Shadows control to behave quite differently at higher values like close to +60.

Unfortunately, this image-adaptive behavior makes Highlights, Shadows, Clarity, and Dehaze problematic when used on time lapse sequences.

The recommendation is to avoid these controls in this situation and instead use Parametric or Point Curves, which will be “stable” across the sequence."

I hope that helps.

Rikk Flohr: Adobe Photography Org

F

FSt0p

Inspiring

Well, that's basically what I've replied in my first message.

But @MATTHEW28594348e9jk claims that results with the same files are different on Mac system compared to Windows.

I can't test by myself as I don't have access to Mac.

Maybe someone else can test on the Mac and post the results here...
If indeed behaviour on different OS's differs - this should be a bug.

F

FSt0p

Inspiring

This is not a bug - it is by design.

It may not be intuitive and what you would expect, but it is what it is.
In short - tools in Lightroom are content-dependent.

What is considered shadows, highlights, blacks and whites depends on actual image - tonal ranges are not fixed.
Because of this you get this effect.

I can't tell for sure why it acts so differently on visually very similar images thought...

For how to deal with this you could watch this video: Stop Lightroom from messing up your timelapses
Or you could switch your images to the older Process Version (v1 or v2) - this will also help, but the tools will be different to what you used to.

M

MATTHEW28594348e9jkAuthor

Participating Frequently

Hi @FSt0p It’s hard to make people believe it’s not a bug, because as I described they are working well on Mac system. If you tried the two files in Win system and Mac system, you will find the difference. If it’s a bug, it should be as same as in Mac.

So the solution is kind of pushing users go for Mac or C1? That’s upset T_T

F

FSt0p

Inspiring

I missed your point that it's ok on Mac, sorry.

Then yes - it should be a bug.

Tested on Win 10 22H2 and it's the same behaviour you've showed.

Things that I wrote still applies: same development settings will result in different visual appearance if image content changes.
For example when the clouds are moving this will lead to changes in the sky brightness if extreme highlight recovery is used, and will lead to flickering.

But it's not that extreme brigtness changes as in your case.

Also, I've tested before posting - changing ProcessVersion to v2 may be used as a workaround, but in my opinion that's a bad solution to the problem that should not exist on the first place...

@Rikk Flohr: Photography - could you please verify this and possibly convert to bug report?
For me, it's 100% reproducible on Windows with the provided images.