Image software that can provide an absolute brightness value?

[TOTC]

Not sure where to put this post, but as it relates to applications in reviews, this was the first place I thought of. Apologies to the mods if this should be in a different subforum.

I am wondering if there is a simple program that could return an overall brightness value for a digital image. I realize that in any image program it is easy to click a pixel or small area and see that color value represented as RGB or various other numbers. What I am wondering, however, is if there is a way to see the overall measure of the whole image.

It would seem an easy way to take anyone's beamshots (assuming that two or more beamshots being compared were taken under identically controlled circumstances) and compare brightness between two lights. Is there something out there like this that I am missing? I'd love to use an app in which you could plug in two images and say "Oh, light X scored 60 on the overall brightness measure, while light Y scored 80."

I realize there are light readers, but I am thinking of this specifically from the angle of measuring digital images. I can think of a number of theoretical ways to do this (the simplest being an algorithm that measured RGB values of every single pixel in an image and then returned an average) but I am wondering if something already exists.

I also realize the large number of confounding variables here (not the least of which being ambient light present in the image), but I am still interested in exploring the concept.

Any thoughts, help, or suggestions are appreciated.

 

[TedTheLed]

how about taking a reflected light reading with a light meter off the computer screen?
(all other parameters; image size, screen brightness settting, meter position, etc. equal of course)

 

[TedTheLed]

probably you'd need to keep the brightest light below zone 8 to allow for brighter lights to not "white out" beyond zone 9 -- then the extra brightness wouldn't register on the screen. probably would be good to start around zone 5, that would give you some room for extra brightness.. see here for more of an explanation;

http://www.luminous-landscape.com/tutorials/zone_system.shtml

 

[BB]

Not exactly what you are looking for--but may actually allow you to do what you want--to create a high dynamic range photograph using a standard camera + software:

High-Dynamic-Range Photography

HDR provides a way to combine a range of exposures of the same scene into one image, adding significantly to the amount of data held per pixel (most digital images hold 8 bits of color information per pixel; an HDR image has 32). The result is an image with more "dynamic range"—in other words, the brights are brighter, the darks darker, and there's much more variance in between.

Basically you take 3 or more bracketing shots (from under to over exposed)... Then with software put the multiple 8 bit pictures together into one 32 bit deep picture.

There is paid (including a version of PhotoShop) and freeware software to allow you to do this function.

I would guess you could do the same thing with a beam shot and then integrate the picture.

-Bill

 

[greg_in_canada]

I think you could blur the image until there is no change from edge to edge in the image. Then just read the RGB value of any pixel.

Greg

 

[Canuke]

Those who post underexposed shots are actually doing a crude HDR implementation already; the difference is that HDR requires more shots over a greater range of exposures, and requires special tools to combine and view the results.

HDR would be great for beamshots. Unfortunately, it's a lot of work to do, and requires a number of software tool expertise to pull off.

If reviewers want to be able to objectively compare beamshots from different reviewers, then it gets even worse, with calibration and standardization of setups and equipment.

The benefit of all this work would be considerable; precise calibration would mean that the HDR image would contain accurate information about luminous flux at points on the target surface. HDR images also allow re-exposure on the fly while viewing, with the right tools, which gets around the problem of clipped/blown highlights and the lost information therein.

Here is a demonstration of how this works; in particular, check out the Hollywood Sign shot. Notice how the image re-exposes, or "auto-gains", as you pan around from darker to brighter parts of the shot. Beamshot HDR's would ideally be set up as regular shots (not panos) and would have controls to expose manually. We could then accurately compare two lights by setting their HDR's to the same exposure value (EV), and the relative brightness on our monitors would closely approximate relative brightness in the real world.

 

[Ganp]

Although this is a nice idea and a very desirable ability, IMHO there are just too many variables for this to be feasable.

The first, and most important is .... the setup for the original image, regardless of HDR or anything else.

1. The reflectivity of the illuminated surface needs to be consistant.

2. The distance between torch and surface needs to be consistant.

These are only the first considerations and without these being the same it does not matter what you do after, you will not be comparing like for like!

And even if these were the same, there are a multitude of camera settings, and positions relative to the target, that would influence the final result.

So - my apologies for being a damp squid, but unless there is a concensus on method, and a universally available target to illuminate, then IMHO, individuals own direct comparisons are the best option at the moment.



Colin.

 

[TOTC]

Appreciate the responses so far.

Just to clear up an issue, I wasn't suggesting that Individual A's photos would be comparable to Individual B's photos using my suggested method. I know that would be quite impossible. Rather, I was suggesting that a set of photos from a single individual, all taken under identical circumstances and with identical settings, would be roughly comparable.

For example, in my beamshots here, I would like to be able to say "The Fenix's beamshot scores an average value of RxxxGxxxBxxx while the CL1H's beamshot scores an average value of RxxxGxxxBxxx. If taken at face values, this suggests the Fenix sample is x% brighter but y% bluer." Or something along those lines...

I do appreciate the benefits of HDR photography, and it is something that I have successfully used in my recreational photography, but I'm not sure it gets me any closer to my end goal here.

I did a quick experiment using Greg's suggested blur method in Photoshop but found that too much information was lost, and even with a Gaussian Blur at max level, I could not produce an image that was consistent from edge to edge.

Thanks for the ideas so far