Harmonizing contrast/luminosity between pictures (jpg)

[Eildosa]

Hi,

I was experimenting with photogrammetry and I now have 410 sets of 21 pictures.
Each of those 21 pictures being someone doing a specific face.

I'd like to be able to "equalise/even out/hamonize" the difference of contrast/luminosity.
Being French I lack the vocabulary to explain it better so here is a picture :



As you can see the side pictures are darker than the central picture, is there a way to batch process this?
Give the 21 one pictures to a software and let it "harmonize" every pictures?

Thanks.

[Dave Humphries]

Hi and welcome to the CiC forums from me.

I think you've done quite well explaining the problem with the example shots.
Unfortunately I can't think of an immediate answer, but hopefully someone here might.

Am I correct to guess that the person's face was captured on 21 separate (different?) cameras in an instant?
or did they, or the camera, rotate around taking the 21 exposures?

Either way, the problem might have been lessened if the lighting had been more even - but you know that by now!

I'll be interested to see if someone has a solution for you.

Good luck,
Dave

PS
Could you do me a favour please?
Could you click Settings (right at the top),
then Edit Profile (on left)
and put your first name in the Real Name field
then click the Save Changes button below and to right,
this helps everyone give you more personal and relevant answers - thanks in advance.

[TonyW]

It looks to me to be a job for Photoshop scripting. However, I have to declare that I have never used Photoshop scripting and it would take me a long time to investigate. I read that if you are used to JavaScript, it is easier.

[Dave Humphries]

I think the biggest problem is going to be determining which areas in each to make the same levels; e.g. a bit of face skin and a bit of the dark background - no idea how that can be automated - given that the face isn't in the same place on all frames.

With those lamps visible in some shots, averaging won't work - and indeed this is possibly why the camera's (auto?) exposure was fooled in first place.

(so perhaps the lighting was even enough on the subject from different directions, but the fact that the tubes are visible in only some shots, and the camera(s) weren't on fixed manual exposure, is what's caused the problem)

It might be tedious, but I'm thinking that a refined human driven process (in ACR or LR) might have to be the answer - it may be quicker than time expended trying to automate a batch process.

Dave

[Manfred M]

It looks to me to be a job for Photoshop scripting. However, I have to declare that I have never used Photoshop scripting and it would take me a long time to investigate. I read that if you are used to JavaScript, it is easier.

I assume you mean Photoshop Actions? I do use Actions fairly regularly, but they are really little more than recording steps and replaying those steps from image to image. There is no decision tree / what if type functionality. Places were input is required, the script stops running and waits for user input. I think your idea has some merit, but the process should be looked at as very basic automation. I don't think Photoshop is going to provide the solution that is being looked for here.

[Cogito]

Hi,

I was experimenting with photogrammetry and I now have 410 sets of 21 pictures.
Each of those 21 pictures being someone doing a specific face.

I'd like to be able to "equalise/even out/hamonize" the difference of contrast/luminosity.
Being French I lack the vocabulary to explain it better so here is a picture :



As you can see the side pictures are darker than the central picture, is there a way to batch process this?
Give the 21 one pictures to a software and let it "harmonize" every pictures?

Thanks.

Hmmm! 21 images! I don't know of any batch processing action that could manage this but perhaps your computer could manage a similar action as you have performed here? If you paste all 21 (!!) images into a single image (but remember to leave clear space between each image) then you COULD adjust each image's brightness and contrast to some equality. Additionally you could adjust the apparel to equalize the colour. And also remove any background distractions..... A very exhausting process! After processing you then cut each modified image as a single image - hence leave a space.
this with your example......



I made no effort to adjust the background! Bon chance!

[TonyW]

I assume you mean Photoshop Actions? I do use Actions fairly regularly, but they are really little more than recording steps and replaying those steps from image to image. There is no decision tree / what if type functionality. . . . .

No, I did not mean Photoshop Actions. See, for example, http://wwwimages.adobe.com/content/d...guide-2015.pdf

This does allow conditional statements for example.

However, as I said before, I have not used this myself and implementing it would depend a lot on programming skills and would certainly take me a lot of time to do. It may not be worth the time but the OP had over 8,000 images to deal with.

[revi]

Assuming that the lighting setup and intensity of all images within a series is the same,
could you use the exposition information in the metadata to correct the brightness in post?
I mean, if exif tells you the center image is shot @ 1/125, F5.6, and the outer edges @ 1/125 F16,
could you apply a correction of +3EV to the edge image to at least get close?

[Eildosa]

I think you've done quite well explaining the problem with the example shots.

Well thank you

yup the images were captured with 21 separate camera at the same time.

There is 2 problems :
- I used cheap point and shoot cameras, they choose their own ISO/shutter speed/aperture so I get different exposures
- the sides of the face get less light because it's a demi-sphere

After hours of google searches I finally found how to do it, lightroom "match total exposure" trick :
https://petapixel.com/2014/12/15/tip...oom-quick-fix/

It works quite well, I just have to select the central pic as a reference and it will then manipulate the contrast and luminosity of each jpg image in order to match the exposure of the reference pic.