June 2010 Archives

Why don't people who like on-camera ring lights like on camera flash?

Are they really that different? (other than the symmetrical outline shadow on the ringlight rather than the simple slightly offset shadow)

or put it another way -

Isn't every criticism people lodge against against camera flash (too much on-axis light flattens the subject) also applicable to ring lighting? So at what point does that 'bug' become a 'feature?'

Here is a Maxwell Render simulation comparing a ring light with on-camera flash. The shadows are quite different, as I expected, but I'm surprised the quality of light on the skin itself seems so similar.

Or I guess I'm not surprised. That I suspected they might be quite similar was the reason for the test.

UPDATE

Another comparison between a ring flash and on camera flash using the Maxwell Renderer. This time I changed the angle, made the skin a bit glossier, and the flash bulb a bit smaller.

You can see that the ring flash is gentler on pores (or would be, if this CG mannequin had any)

While investigating Focus Magic-style deconvolution (which I like to use in tiny amounts to sharpen images) I came across 4N6site.com and the concept of color deconvolution.

When I saw the fine distinctions color deconvolution could make between different colors of ballpoint pen ink, I thought it might be a useful tool for hair masking. Verdict: so far, so good.

Granted with fuchsia hair against pale cyan I chose an easy example, but so far I'm pretty impressed with the plug-in's ability to drive one color to white and the other color to black.

Since hair and backgrounds contain many hues, the user can sweep out areas of interest and obtain the average color. Even with that ability, multiple mattes may need to be pulled and combined to do a precise job.

Maybe color deconvolution is already at the heart of other popular hair masking algorithms. I don't know. But I like the simplicity of this free (donation requested) plug-in.

More exploration to come.