The effect of Exposure stacking on bug shots

[JBW]

The first shot is one of the 17 shots used in the exposure stacking. It is sharp but it suffers from a lack of tonal depth.

The next two shots, created from the stack, lack ultimate sharpness but have deep colors.

[DanK]

Did you stack images that had the same exposure?


Sent from my iPad using Tapatalk

[JBW]

Did you stack images that had the same exposure?


Sent from my iPad using Tapatalk

I stacked shots with varying exposure settings from -1.7 to -0.0

[Stagecoach]

Very good example Brian.

Pondering over this further and re reading your previous thread discussing 'exposure stacking' to increase colour/contrast/tonal range (terminologies each used loosely) I had a close look at the first two images you used here.

The two images consisted of;

a) 1 of the original 17 different 'exposures'.
b) the result of stacking the 17 different exposures together.

There are two things we do not know;

c) What you used to vary the exposures, it could have been Aperture, Speed or ISO? So let's forget anything to do with DoF/sharpness changes and noise (which is not evident)

d) If you have done any further PP to the 'result' produced by the 'exposure' stacking. So let's assume you have not.

I took the two images, resized, aligned pretty close and labelled them to make it easier to undertake comparisons;

No 1


No 2


Opening these two images in my PP package ACR I am unable with all the basic tools available make the 'Original Exposure' look like the 'Result'. It would be interesting if others could

So my conclusion is that your method has done far more than simply use the average or mid exposure of the 17 or something that is easily achieved in basic post processing on a single image. What your result shows I believe is an 'increased' dynamic range than that what was achieved in any single exposure, but if we measure at the bright orange bit where the legs meet the body the two images have almost equal values.

[dem]

I stacked shots with varying exposure settings from -1.7 to -0.0

17 shots separated by 0.1 EV? Could you get the same result by simply adjusting the curves in a single raw file?

[JBW]

17 shots separated by 0.1 EV? Could you get the same result by simply adjusting the curves in a single raw file?

Perhaps you misread my info? It is 1.7 difference not 0.1. But to answer your question I've never managed to do it by a simple curves adjustment.

[JohnRostron]

I'm confused.

You took 17 shots.
The exposures varied from -1.7 to 0.0 EV
The difference (between shots) was 1.7 EV, not 0.1 EV.
Or am I misinterpreting this?

John

[JBW]

I'm confused.

You took 17 shots.
The exposures varied from -1.7 to 0.0 EV
The difference (between shots) was 1.7 EV, not 0.1 EV.
Or am I misinterprrting this?

John

I agree with you but then tech is not my specialty

[DanK]

Not having the originals and not being familiar with that specific software, I'm in the dark. However, the exposure-blending software with which I am familiar works by selecting from each image areas of pixels that are within set exposure limits and then combining them into a composite image. That's all it does. The result is not easy to replicate with standard editing tools because it is in effect a very localized editing, although I think one could get something similar with luminosity masks if one were sufficiently skilled with them (I certainly am not) and if one knew what luminosity ranges to select. However, this process has no impact on color, other than the making colors look different because of luminosity changes, and it has no effect on softness.

Whether Fiji is something different, I don't know. My recollection is that there is someone here who is familiar with ImageJ, so perhaps that person can weigh in.

If you had 17 shots with a total difference of 1.7 EV, that is, 0.1 between shots, you should be able to use far fewer shots. Looking at the first shot in Irfanview, it appears to span less than the dynamic range of the camera, so there is no need for exposure blending to compensate for excessive dynamic range.

[Stagecoach]

17 shots separated by 0.1 EV? Could you get the same result by simply adjusting the curves in a single raw file?

I tried but could not achieve anything similar

[DanK]

I tried but could not achieve anything similar

Nor surprising, because if what Brian is doing amounts to exposure fusion, the adjustments are all local, not global.

[JBW]

Nor surprising, because if what Brian is doing amounts to exposure fusion, the adjustments are all local, not global.

Here's a thought Dan, focus stacking lines everything up and then picks the sharpest areas from each shot and merges them into one. Perhaps Exposure stacking does basically the same thing. Only instead of choosing the sharpest areas to merge the richest colours are chosen?

Or... perhaps with stacking different exposures the sharpest parts are changed in relation to the exposure resulting in a much deeper set of tones?
Brian

[Round Tuit]

Here's a thought Dan, focus stacking lines everything up and then picks the sharpest areas from each shot and merges them into one. Perhaps Exposure stacking does basically the same thing. Only instead of choosing the sharpest areas to merge the richest colours are chosen?

Or... perhaps with stacking different exposures the sharpest parts are changed in relation to the exposure resulting in a much deeper set of tones?
Brian

Can you tell us which command/plugin/option of fiji that you used to create the stack?

[DanK]

Here's a thought Dan, focus stacking lines everything up and then picks the sharpest areas from each shot and merges them into one. Perhaps Exposure stacking does basically the same thing. Only instead of choosing the sharpest areas to merge the richest colours are chosen?

Or... perhaps with stacking different exposures the sharpest parts are changed in relation to the exposure resulting in a much deeper set of tones?
Brian

It seems possible that the program is selecting on the basis of saturation (hence deep), but the programs I know of do not use color as a criterion to select. They use exposure. However, changing tonality can change the appearance of color.

[JBW]

Can you tell us which command/plugin/option of fiji that you used to create the stack?

simply 'images to stack'.

[JBW]

It seems possible that the program is selecting on the basis of saturation (hence deep), but the programs I know of do not use color as a criterion to select. They use exposure. However, changing tonality can change the appearance of color.

it is perplexing and i shall have to try again when the opportunity arises to see if the results are similar.

[Round Tuit]

simply 'images to stack'.

Sorry Brian, I meant to ask which command you used to PROCESS the stack. My bad

[DanK]

Fiji is ImageJ, right? I took a quick tour of ImageJ documentation over lunch. My reaction is:

1. This is like using an 18-wheeler when one needs a wheelbarrow. it is an enormously flexible program designed to do many things--e.g., dealing with 3-dimensional "hyperstacks"--and it is not aimed at the general population. Both of these things make both the software and the documentation complex.

2. I never did find a description of the various ways in which it creates a composite.

3. For most people, I think it would work out better to use simpler programs designed to do what one needs and documented accordingly so one knows what edits one is actually imposing. In this case, we are talking (I think) about exposure stacking and focus stacking. For the former, one option is Enfuse, which will also do focus stacking. I use this as a plugin to Lightroom, but I think you can use it separately. For the latter, I use Zerene, which is superb, but there is at least one free alternative, the Combine programs.

[xpatUSA]

Fiji is ImageJ, right? I took a quick tour of ImageJ documentation over lunch. My reaction is:

1. This is like using an 18-wheeler when one needs a wheelbarrow. it is an enormously flexible program designed to do many things--e.g., dealing with 3-dimensional "hyperstacks"--and it is not aimed at the general population. Both of these things make both the software and the documentation complex.

2. I never did find a description of the various ways in which it creates a composite.

3. For most people, I think it would work out better to use simpler programs designed to do what one needs and documented accordingly so one knows what edits one is actually imposing. In this case, we are talking (I think) about exposure stacking and focus stacking. For the former, one option is Enfuse, which will also do focus stacking. I use this as a plugin to Lightroom, but I think you can use it separately. For the latter, I use Zerene, which is superb, but there is at least one free alternative, the Combine programs.

Y'all might find this of interest:

https://www.dpreview.com/forums/1054

Personally I use align_image_stack.exe (command-line utility) then TuFuse Pro (has a GUI) which can do exposure and also focus in one shot. Been a while though, they're all quite painful to use ...

[JBW]

Sorry Brian, I meant to ask which command you used to PROCESS the stack. My bad

That is a much more complicated process. I did some global touchups in Capture 1 Sony Pro, exported to Fiji in Tiff then returned to C1 and completed things in JPEG.