Lighting exposure question

Hi

I seek some understanding. If photographing a bird that has white areas and also very dark feathers and the light level is moderate ( whatever that means), if you expose to bring out detail in the dark areas then the whites are too hot. If you expose the whites correctly then the darks are without detail. Now my question. If you expose such that the whites to the max but not blown and add a flash in backfiil manner will that bring in the details in the dark areas without blowing the whites? If that is correct, why does it not the blow out the whites?. Is it because the percent increase the light areas is less than the percent increase dark areas?

Thanks Ray:confused:

If you are using your flash to fill in the shadows, you usually have it set at -1 or -2(stops). So the flash is providing 1 or 2 stops under 18% gray, so think of it as gray light rather than white. If you add gray to white(with light) you still have white.

Hopefully that helps, otherwise, someone with a better explanation will come along.
Dug

The flash is adding light at a different angle filling in areas that were in shadow thus reducing detailess areas while increasing details.

Thanks for your answers. I was not asking about shadows but about birds that have white feathers and very dark feathers and how to properly expose both and how a flash helps, specifiaclly how a flash helps bring out detail in the black feathers w/o blowing the whites that nearly overexposed already.

Thanks Ray

Exposure is cumulative and reciprocity is generally in effect. The problem as stated is one of trying to map the dynamic range that you have into the number of levels you have (bit depth). One approach would to keep the contrast settings as low as possible and the bit depth as high as possible (raw) and try to resolve in post-processing. That being said, I think Jim was spot-on, since it's typically unlikely that the flash and ambient light have the same direction. If they were, the flash would just push everything to the right, possibly blowing out some highlights.

The flash will add light to the highlights so you have to take this into consideration when you calculate your exposure.

Some clarifications. Let's first assume that the black and white feathers are both illuminated. Then the dynamic range is set by the reflectance of each. An extreme case might be ~95% for the white feathers and 3% for the dark. The dynamic range would be ~0.95/0.03 or about 32. Add in some shading, e.g. say the dark feathers are facing away from the incident light, then the apparent reflectance will be lower.

Flash adds light. So adding a small number, say 10% of the intensity of white brings the black feathers op by .1+.03 = 0.13 and the whites 0.1+0.95 = 1.05. So the dark thing were raise by .13/.03 = 4.3 or over two stops, but changed the bright white feathers 0.14 stops. So, yes you do need to compensate to no saturate the whites, but it is a small effect. The the dark feathers were in shadow, the light added by the flash would be a greater factor, but still little effect on the white feathers.

In general I expose to not saturate the whites and bring up the bottom in post processing.

Roger

Thanks to everyone for your answers.
Roger thanks very much for the detailed answer. If I understand your answer correctly it confirms what I thought, the flash effect is greater on darker areas as apercentage of increase.

Ray.

Ray,
Yes, that is correct.

Roger

Quote:

---

Originally Posted by rnclark

Some clarifications. Let's first assume that the black and white feathers are both illuminated. Then the dynamic range is set by the reflectance of each. An extreme case might be ~95% for the white feathers and 3% for the dark. The dynamic range would be ~0.95/0.03 or about 32. Add in some shading, e.g. say the dark feathers are facing away from the incident light, then the apparent reflectance will be lower.

Flash adds light. So adding a small number, say 10% of the intensity of white brings the black feathers op by .1+.03 = 0.13 and the whites 0.1+0.95 = 1.05. So the dark thing were raise by .13/.03 = 4.3 or over two stops, but changed the bright white feathers 0.14 stops. So, yes you do need to compensate to no saturate the whites, but it is a small effect. The the dark feathers were in shadow, the light added by the flash would be a greater factor, but still little effect on the white feathers.

In general I expose to not saturate the whites and bring up the bottom in post processing.

Roger

---

Thanks Roger! The math makes sense to me but my gut says if you add the same amount of light to different tones you raise them all by the same level. I think I understand why my gut is wrong and it has to do with adding light but expressing the increase as a proportion. I'll think more!

Quote:

---

Originally Posted by John Chardine

Thanks Roger! The math makes sense to me but my gut says if you add the same amount of light to different tones you raise them all by the same level. I think I understand why my gut is wrong and it has to do with adding light but expressing the increase as a proportion. I'll think more!

---

John,
Here is an analogy: two bank accounts. In account A you have $10 and in B $1000. This is like the bright and dark areas of an image. Now we add $20 to both accounts (the added flash in the image). Account A now has $30 and B $1020. We've increased account A by 3 times, but didn't make much difference in account B. It is the same with fill flash adding to the scene: makes a big difference in the dark parts and not much in the lights.

In the flash fill, the light adds linearly, not in proportion, like adding money to the bank account.

Roger

Good analogy Roger.

I am by no means a flash expert but unless I am missig something (and it wouldn't be the first time :)), I think that the above conclusions are incorrect. First, let me restate what I understand those conclusions to be :

"If you have very light and very dark areas being illuminated by the same amount of incident light and add some fill flash, the fill flash will have a greater effect on the dark area than it has on the light area, i.e. the percentage increase in "brightness" in the dark areas will be greater than the percentage increase in "brightness" of the light areas."

I believe the percentage increase will be the SAME in both the light and dark areas. Let's say that there are 1000 units of incident light illuminating both the light and the dark areas. Further, let's assume that the light areas are 90% reflective and the dark areas are 5% reflective. We (and our camera) only "see" the reflected light and it is the intensity of the reflected light that we see as the "brightness" of the various areas on the subject. Without the flash the light areas are reflecting 900 units of light and the dark areas are reflecting 50 units of light. Let's say that we now add 100 units of light as fill flash. This has the effect of increasing the amount of incident light to 1100 units. The reflectivity of the subject, i.e. its tone, is not affected by the fill flash. Therefore, with the fill flash, the light areas are now reflecting (0.90 * 1100) = 990 units of light and the dark areas are reflecting (0.05 * 1100) = 55 units of light. The "brightness" of both the lights and the darks is increased by the same percentage, i.e. 10%.

If we are instead talking about the effect of fill flash on areas of the scene that are illuminated by DIFFERENT amounts of incident light, then we get different results. Let's say that there are 1000 units of incident light hitting one object directly but another object in the scene is in a shaded area and is therefore only illuminated by 200 units of incident light. If we add 200 units of fill flash, then the brightness of the object in the shade will double but the brightness of the unshaded object will only increase by 20%. But if both objects were instead illuminated by the same 1000 units of incident light, then they would both increase in brightness by 20%.

So what do you think?

The end of my last post got dropped and now I can't edit it either so I'll just add that I would be very interested in your comments and/or corrections to the analysis presented.

Keep in mind also that the discrete values of the scene's dynamic range are going to be mapped (generally linearly for digital sensors) into one of 2^bits available buckets. Anything pushed past the edge is going to be blown out and will be nonrecoverable in post. I come from a film background so tend to think in terms of log D exposure, s-curves, etc. The percentage reflectance analysis sounds reasonable mathematically, but I don't see the relationship to the curve other that shifting things to the right by the same amount. With film we say expose for the shadows, develop for the highlights, or exposure controls density and development controls contrast. It looks like you have two controls in a digital sensor, exposure and contrast. So the question in my mind becomes "is it possible to expose the shadows while keeping the contrast under control sufficiently to not blow out the highlights?"