SB 910 over exposing?

[Andrew Garrard]

A speedlight has a finitely large head; if it's quite close to the subject, it does have a softening effect, especially if you flip down the diffuser panel. The effective size shrinks with distance, so it's true that you'd get a harder shadow from a more distant speedlight. But the head isn't that big, so it's going to be fairly hard anyway unless you're at macro range. (So I think I get what Dave's talking about, but I think the effect would usually be subtle.)

 

Another aside: most subjects aren't a concave curve of constant distance from the flashlight. The farther the flashlight is, the more even the lighting across the scene. Additionally, the more distant, the more parallel the light rays from it - so light will "wrap around" a subject more at a distance, in the same way that a telephoto lens shows more of the subject's ears than a wide angle. And there's the more general concern that the ratio of distance between foreground and background affects the amount of light each receives (in a way a bit like the relationship between background blur, subject distance, and depth of field). So moving the flash to adjust its strength is a bit like having a prime lens and "zooming with your feet": it's not quite the same thing. But it'll certainly do in a pinch.

 

(This post mostly shows that I know enough about shooting with flash to scare myself with all the variables, and not enough to have good control over them.)

[rodeo_joe1]

especially if you flip down the diffuser panel.

A common misconception.

Sticking a diffuser directly in front of a tiny light source makes it no bigger, and no softer. The flip-down diffuser is there simply to spread the light coverage for super-wide lenses.

 

Light rays from the extremities of the source - typically no more than 6cm apart for a speedlight - do not get any more distant from each other. So, even with a frame-filling headshot at an in-your-face 60cm distance; a nose shadow is only going to show a few millimetres 'softening' of its edges at most. Worse with the usual central hotspot of an unfocussed flash at that distance.

 

Even if you do diffuse the flash into something of a size that can fit on top of a camera, the light remains near co-axial with the lens and those dreadful specular reflections from nose, cheekbones, chin and forehead remain.

 

There's no such thing as a compact and soft light source. Get used to it!

[Andrew Garrard]

A common misconception.

Sticking a diffuser directly in front of a tiny light source makes it no bigger, and no softer. The flip-down diffuser is there simply to spread the light coverage for super-wide lenses.

 

I was under the impression (from looking at the front of the flash when it wasn't firing, because I'm paranoid) that the flash itself was relatively central in the fresnel flash area, rather than appearing to cover it in its entirety - like a wide angle lens, for which from any given view the apparent entrance aperture doesn't cover the front element. I assume it depends where I look from. I could be wrong, but I was thinking that it's not the entire front area of the flash that normally contributes to illumination at any given point. This would be different with a diffuser in the way - all the light getting from the flash tube to the front of the flash gun is sent over a wider angle. This could be complete gibberish, and I should experiment if I ever have free time.

 

Light rays from the extremities of the source - typically no more than 6cm apart for a speedlight - do not get any more distant from each other. So, even with a frame-filling headshot at an in-your-face 60cm distance; a nose shadow is only going to show a few millimetres 'softening' of its edges at most. Worse with the usual central hotspot of an unfocussed flash at that distance.

 

Well, the "central hot spot" was what I was trying to get at with the diffuser. If we have a 1:10 ratio between the head diameter and shooting distance, that would mean the shadow cast by my nose is diffused by the order of 3mm. That's not huge, but it's not unnoticeable - it's the diameter of a constricted pupil. For portraits it wouldn't matter quite as much, because there's some sub-surface scattering of the skin, but certainly some things I shoot would make for a visible difference between a 3mm shadow penumbra and none at all.

 

Even if you do diffuse the flash into something of a size that can fit on top of a camera, the light remains near co-axial with the lens and those dreadful specular reflections from nose, cheekbones, chin and forehead remain.

 

I was wondering about that. I had a theory that if you put a (linear) polariser in front of the flash and a crossed polariser in front of the lens, you might get rid of the specular reflections and keep diffuse ones. I've never tried it, so it could go on the "gibberish spouted by Andrew" list.

 

There's no such thing as a compact and soft light source. Get used to it!

 

One compact (by some definitions) thing I wondered about - I got as far as getting one but not trying it - was a fresnel lens sheet, of the size offered as a reading aid; that is, about the size of a sheet of paper, or a smallish diffuser. In theory, it shouldn't cut out light significantly, and it should make the flash appear larger relative to the subject - so you should get softer shadows. The benefit would be that it would be flat and not lose as much light as foldable reflector/diffusers.

 

Despite sounding argumentative about a subject I've already admitted I know little about, I completely agree with Joe - at conventional distances the effect of shadow softening should be subtle, and to have a significant softening of light you really need a big diffuser of some sort. Just throwing out some ideas, though. Assuming subtle effects were irrelevant is how we ended up being confused about "equivalence" for years.

[rodeo_joe1]

I had a theory that if you put a (linear) polariser in front of the flash and a crossed polariser in front of the lens, you might get rid of the specular reflections and keep diffuse ones.

Crossed-polarised light has been used for decades in tricky copying situations, such as objects behind glass, or with a specular surface texture that needs to be subdued. Or simply to get rid of a surface sheen to improve colour saturation.

 

The big drawback is that the reflections can never be totally suppressed by X-polarisation. They still usually show up as a dark blue reflection or cast that looks totally unnatural.

 

I must admit that I haven't tried it with human skin, but a shiny dark blue area might well look a lot worse than a shiny white area. Or the reflective properties of skin might not reflect polarised light back without being scrambled.

 

Damn it! Now another experiment has been added to the list.

 

Sorry, Andrew, but I'm not even going to respond to strange theories about fresnel focussing. It's just a type of lens, no more. All I can do is repeat: There's no such thing as a compact and soft light source.

[Andrew Garrard]

Thanks, Joe. Sounds like I should at least give the portrait thing a go. If I can find a linear polariser that's not attached to glasses frames, anyway. (I actually have been known to resort to this when I've forgotten to take a polariser...)

 

And yes. I've never been terribly convinced by the theory of the tiny diffusers (omni-bounce, etc.) - I assume they rely on the environment, and on the occasions I'm using flash, I prefer not to calibrate for paint tones.