Playing with light modifiers and strobes. Kelvin correction.

[Some Guy]

Bought a new face cloth to go on my old soft boxes.  I noticed the white material was really white, and a rip-stop nylon too.

My Sekonic C-500 color temp meter read 5,250 on the strobe without the new cover.  With the new cover, 5,950 Kelvin.  Lots of blue or maybe ultra brighteners in the cloth maybe?

Left outside in the sun to fade.  No difference.

Washed in ammonia bath for 3 hours that was recommended to get rid of the UVB stuff.  No difference in Kelvin.

So I went and bought some rip-stop white at the fabric shop that looks like the diffusion cover.  Some "Orange" fabric dye stuff too (Tulip One-Step Dye.  www.ilovetocreate.com).

Interesting part is once I mixed the orange dye in the little 4 ounce bottle that comes inside the box, if I took 1cc of that mixed dye to 1,000cc of water and about 3cc of white vinegar, the dye rip-stop nylon cloth dropped the Kelvin by 1,000 degrees off my Nikon SB-900 flash that normally runs at 6,300K and with the dyed cloth it took it down to 5,300K.  Fwiw, I left stirred it in the dye bath for 5 minutes at 140 degrees, then washed it and dried it before testing.  The dye bath solution looks like a Skylight or 81A filter almost - a very faint orange.

Need to play some more before committing to the large softbox diffusion cloth.  Maybe 1cc dye to 2,000cc water will drop it maybe 500K.  I'm not to crazy with the heads being that blue at 5,950K so we shall see.

SG

[Paul Williamson]

Interesting!

I'd suggest you shoot a ColorChecker under the modified lights and see what it does to the color rendering. The color temperature meter is making some assumptions about the spectral characteristics of the light source, and not telling you the whole story. The orange dye might be exactly what you want, or it might not.

It'd also be interesting to compare results between the orange dye trick and software-based camera profiling. Unless you're mixing lights of different colors, you should be able to calibrate out the differences.

[Ken Bennett]

This is interesting, and Paul makes a good point about testing beyond just the color temp in degrees K.

I usually fix this sort of thing with CC filters over the flash head. It's pretty easy with a speedlight.

[Some Guy]

With the studio heads, the color temp changes about 500 degrees from dim to max power.  They are not Einsteins by any means!  Studios seem to hover around 5,650K on average.  The Nikons speed-lights are really blue at around 6,500K.  Sort of odd that Nikon calls their Daylight WB as 5,200K and then they say Flash at 5,600K, but their own speed-lights are way above that.

More interesting is the dual pane softboxes are over 6,000K as well.  Maybe the two fabrics add the blue and it goes way over where I'd like it at.  Single layer are warmer at maybe 5,800K average, but still too blue for me if I try and balance them outdoors with daylight.

Back to dyeing the fabric...

So far a dilute mix of the concentrated dye at around 1:3,500 to 1:,4000 seems close.  The nylon takes the dye pretty well (add a bit of white vinegar to help set it in the mix).  The fabric can take on a skylight filter look, and reading it with the ColorMunki Photo produces interesting results.

The plain white diffusion fabric used and read with the CM Photo reads as a "Bluish White" in their software (RGB=227, 233, 231).  Dyeing it produces other colors like Vivid Yellowish Pink, Light Orange, Pale Orange Yellow, Pinkish White, and even White.  The "White" read as RGB=240, 243, 239.  So it appears you can dilute-dye out the blue of the fabric covers.

I got them (dyed) down into the 5,100 to 5,400K range with the studios.  They now look like a pale skin pink color, much like some face powder in color.  No longer a bright white.  The dye bath looked like a dilute apple juice, or orange Jello, when mixed.  You can fine tune the dye out with a dilute mix of "Synthrapol Sizing & Dye Remover" which is some excess dye remover from better fabric shops, or maybe RIT Dye Remover if you go overboard.  I'm removing one where it went under 5,000K to match it better with the others.

Even with some aluminum beauty dishes I have, the Kelvin is different out of them on the same head.  Some look like bright polished aluminum, others are a duller, maybe oxidized, and have a different color cast out of them than the bright aluminum reflectors.  Strange stuff, but I'm dyeing the color slip-on diffusion covers to balance the entire lot.

Add:
I've attached an image of the large beauty dish with the dyed cover and a smaller one.  Looks sort of pinkish to me.  The white fabric patch (Original piece of diffusion fabric I found.) demonstrates the tint difference made by dyeing.  The smaller one has some binding that doesn't accept the dye, maybe some poly- something as poly fabrics seem to not like dye.  The large one was about 5,950K and is now around 5,400K.  The smaller on reads as 5,100K so I may strip some dye out of it.

Add:
With some work I managed to get all the fabrics dyed to within 5,200-5,300K with the same head and setting.  What is interesting is the faint pinkish tint of the fabric is very close at a Lee #506 Marlene color (So named for Marlene Dietrich here: http://www.leefilters.com/lighting/colour-details.html#506 ).  Now time to do some tests with the ColorChecker.

Add:
Still refining.

Seems "Golden Yellow" Rit Dye in powder diluted to 1:4,000 with water (About 1/4 teaspoon powder) along with a couple of ounces of white vinegar works to approach what Vivitar had as a filter over their speedlight flash tubes in the 285 with the zoom snout.  About 12 minutes at 160 degrees is a good starting point.  Very faint yellow in the 4 quart stirring pot.  Might take 2-3 trips to the dye bath to dial it in and best to start from weak side too.  The Rit "Golden Yellow" seems to drop the blue of the flash tube pretty quick, much more than the above which is more a pale pinkish skin color.  A second bath using the Rit Dye Remover works to prepare the nylon, and to fade out if overdone too.  Has to be really hot (170F or so) and maybe 10 minutes with a good wash following (Drops it about 200-300K).

SG