Hmmm, yeah so I guess it's just ideally you want to be as large as practical on the globe, one of those all-things-being-equal sort of things. Weighing in the amount of power on hand changes things some.

Still, it wasn't so long ago even the "dirt cheap" level of product would be considered quite expensive. Hopefully the backlash against CFL for the home further drives down the cost of LED lighting and the power and purity up.

@vicharris yeah, that is based on the assumption you've got enough bundled in there to put out some fairly strong footcandles. I wonder about the relationship to the interior source size relative to the outer diameter and how much power output influences the result.

+1 on the larger china balls. The extra space also lets you put more watts inside without risk of burning. (Not that you probably need to worry too much about burning with these LEDs.)

@andyharris

My conclusions are that I need to use a smaller ball so that I can get the drop-off greater (the ball would be closer to the subject). I'll be ordering a 20" ball when I get a moment.

No, no, don't do that. You are creating a soft source. This is affected by the paper scattering the light but it's the size of the source that more dramatically effects the quality of light hitting your subject. You want to go as large as is still practical.

Smaller sources, especially moved closer to subject, create "source-y", harder light. You become aware of the spread angle and light hitting your subject at different angles. Larger sources further away, or very large sources closer but dimmed down, give you softer, better looking light, of a more consistent direction, if you're not wanting to draw attention.

This is why they were able to move the China Ball with camera and subjects in American Hustle without looking like something on TMZ. The larger, soft source didn't create obvious shadows (which come from harder, source-y light).

Similarly, when a production is only equipped with fresnel and par-can type lights, and not shooting in the old style of harsh, hard light, they will fire the lights into very large frames of diffusion material, turning the entire surface area into the light source. That or fire the light into a very large piece of ultra-bounce or other types of cards, frames, etc. It's to create a larger source with mostly parallel rays contributing to shadows.

edit: I'd raise the overall stop of the room first with a second light fired into the ceiling perhaps (if it's a light color), so that you're not so dependent on the one light doing everything perhaps. Then bring down the effect of your China Ball perhaps by walking it further back if you don't have a dimmer on it. The dimmer would eventually be pretty helpful.

@andyharris I don't know man, that video looks pretty good to me. Did you light her on camera by chance? Also was there a backlight used at all?

Thanks for all that @andyharris! If you have time one of thoses evening, a test in the street to "boost" the practical street lighting on the talent face would be awesome (or in a car at night). But I'll try to do some test myself when I'll found thoses tools. Thanks again. Great ideas.

Ok, so far with my brand of lights, which I think Andys might be much better, here's the adjustments to pull it back from any color cast on a 18% gray card. This is with 1/4-green on them. I've included the Vectorscope before the adjustment in Resolve as well.

@vicharris On those tests I had to remove a fraction of green --- This is based on doing a 'conform' on a known grey in Resolve - I made the spot on the vectorscope round and in the middle (Is this the right way?). There might be a delay in my next test due to Talent having a nasty cold. The next test may contain a bit of a shoot out between the BMCC and the Canon DC1 which a friend of mine just bought.

@andyharris I have a few LiPos, but didn't know there are problems if you drain them. Thanks for the tip!

I bought some LED strip today, and will be starting my own experiments.

@andyharris More than likely you'll need to add some 1/4-green to em. That was what I found a little difficult in the lantern. I'll dive back into it next week and make it less ghetto :)

@ahbleza If you check out one of my pics above, you can see how these can easily be ran off of battery power. It's the main reason I started the project. So you can take it anywhere and have a little soft light in a min.

So the 2M strip of LEDs have arrived, I've put them on a bench power supply and they drawn about 2.6A at 12.6V which is 32W (was expecting 28.8W).

You wouldn't want to look at them for long.

I wrapped them around a baked bean tin and this seems a perfect fit -- know what I'm having for breakfast now. Since the strip gets quite warm I'm going to clean the can in the dishwasher then rub it down to roughen the surface a bit. That should help the self-adhesive strip.

By the week-end I should have a few photos and by next week end I hope to have done a mini project with the experimental light.

@vicharris

I ordered: http://www.ultraleds.co.uk/90-cri-120-led-1200-lumen-14-4-w-warm-3200k-led-tape-1-metre.html (2 Metre length)

Plus I have some RGB Tape left over from another project.

The Lantern arrived today, at 1Metre (36") I may have over done it a bit - its huge!

For testing I have some nice DC bench power supplies that are very stable with CV and CC controls. I have a little project in mind that would make a sensible test.

I also considered:

http://www.energybulbs.co.uk/s.a.d+t8+15w+18%5c+natural+daylight+pack+of+6%22/3269855943

I have some Dimmable HF Ballasts in the workshop, but I reckon this would take more than a weekend to construct and test.

I looked at the Litegear tape that I think you are going test, the prices are good but the UK shipping was more than the tape.

@andyharris Awesome man. What lights did you order? If you provided a link that would help out here too.

P.S. Like the name ;)