Looking at my dof images in post #13 and #15 I noticed that the Poser built-in DoF doesn't have the transparency (hair) problem that shows up if you use a Z Depth render and do the DoF in postwork. To get a depth-map for postwork that respects the transparency I think (and this is just a guess - I'll try it out later) that two additional renders would be required - one would be a transparency mask (a normal render with all the lights off and any distant background behind the transparency switched off) and the other would be a Z Depth render with the transparent object switched off.
Anyway, I was surprised that the Poser in-built DoF is not as slow as I recall it being. Maybe that's because I'm using PP2014, and the last time I tried it I was probably using Poser... 6 ? I'd noticed that the helicopter was very grainy (I usually go to Render Settings > Automatic, slide the slider to halfway between Draft and Final, switch to Manual and select 'Acquire From Auto'). I read something since this challenge started that said to increase 'Pixel Samples' and/or decrease 'Min Shading Rate' on the render settings to reduce the graininess, so I increased Pixel Samples from 6 to 12 and halved Min Shading Rate from 1.0 to 0.5.
Here's the latest - it's the version with Vicky, so not suitable for this challenge, but it's another step in the process of understanding DoF. :
View attachment 12407
Another small addition - I also used the Poser Atmosphere Depth Cue with bagginsbill's distance falloff correction (I should post that over in the Poser Shaders forum - I've added a
placeholder thread that I'll update after posting this) and a large half-distance to provide a subtle atmospheric effect.