5600k vs. 3200k

[David Mullen ASC]

Any amount of blue filtering helps, so use the amount that you can get away with. An 80C would be fine if you can afford the stop loss.

Here's a basic guide to those blue filters, 80A is the heaviest, 80D the lightest:

80 Series Color Conversion Chart
• 80A increases color temperature from 3200-5500 degrees Kelvin (commonly used with 3200K lamps)
• 80B increases color temperature from 3400-5500 degrees Kelvin (commonly used with photoflood lamps)
• 80C increases color temperature from 3800-5500 degrees Kelvin (commonly used with flashbulbs)
• 80D increases color temperature from 4100-5500 degrees Kelvin (commonly used to remove extreme orange or red)

 

[Joe G.]

Just to clarify --

You only need to get the temp near 5000K, not 5600K:

"Native color balance for the MYSTERIUM™ sensor is 5000 degrees Kelvin..."

http://www.red.com/downloads/23afb64ce6c36df30fde7ed4aff5511f24b21532/build_17_ops_guide_v3.4.1.pdf

So, some less blue should be fine (C-D).

Always consider David Mullen's advice!

 

[David Mullen ASC]

One thing to keep in mind is that blue filters are not going to help much with a subject lit with sodium vapor streetlamps -- these are partial spectrum sources, so the bluer wavelengths are missing. Filters can only subtract wavelengths -- a blue filter normally makes an image look bluer by canceling the warmer wavelengths, leaving the bluer ones more prominent. But with a sodium vapor streetlamp, the bluer wavelengths are missing so filtering isn't going to bring them out.

 

[Bryce Lansing]

One thing to keep in mind is that blue filters are not going to help much with a subject lit with sodium vapor streetlamps -- these are partial spectrum sources, so the bluer wavelengths are missing. Filters can only subtract wavelengths -- a blue filter normally makes an image look bluer by canceling the warmer wavelengths, leaving the bluer ones more prominent. But with a sodium vapor streetlamp, the bluer wavelengths are missing so filtering isn't going to bring them out.

But under regular tungsten lights, a blue filter should help right?

And if I balance the camera to 5000K, and use an 80C filter with tungsten lights, that should be a good enough compensation to make the light white right?

 

[Jeff Kilgroe]

Since RED ONE is not a "true RAW" camera, that is you cannot get the full sensor data out (as far as I have been able to tell), because the de-mosaic software does white balance clipping, this is needed because the anti-OLPF sharpen requires that the data be white clipped before it is sharpened otherwise you get off color rings on the highlights.

I wouldn't say the RED One is "not a 'true RAW' camera". That is not the case. The sensor data is there, but it's a matter of what we have access to at this point in time. RED has not opened up the R3D format for others to do their own custom debayer operations. And to start placing that much emphasis on what is or is not "true RAW" would essentially negate just about every "RAW" capable DSLR on the market today.

Interesting. Think it will be the same for Scarlet? All my lights are 3200.

Should be. CMOS sensor tech in general has an affinity for a white point near 5000K.

 

[David Mullen ASC]

But under regular tungsten lights, a blue filter should help right?

And if I balance the camera to 5000K, and use an 80C filter with tungsten lights, that should be a good enough compensation to make the light white right?

It's close enough... but to be picky, an 80C filter has a Mired shift of -81, so that means it converts 3200K to almost 4400K, not 5000K. So set the Red camera's color temp to 4400K, use the 80C filter, and 3200K will render as "white". If you set the color temp to 5000K, use the 80C filter, 3200K will render a bit warmer than white, which may be OK for what you are trying to do.

The Mired shift system is the best way to calculate changes in color temp from any starting point, assuming you know the Mired shift value of the filter or gel.

 

[Bryce Lansing]

It's close enough... but to be picky, an 80C filter has a Mired shift of -81, so that means it converts 3200K to almost 4400K, not 5000K. So set the Red camera's color temp to 4400K, use the 80C filter, and 3200K will render as "white". If you set the color temp to 5000K, use the 80C filter, 3200K will render a bit warmer than white, which may be OK for what you are trying to do.

The Mired shift system is the best way to calculate changes in color temp from any starting point, assuming you know the Mired shift value of the filter or gel.

Your right, thanks for pointing that out.

I'm shooting in a club, using their strobing lights to give the background some character, and using tungsten lights to light my subject. I'm still waiting on the producer to let me go check the location, but I'm hoping the strobing club lights are tungsten too (is it likely?)

I have a specific color palette, and my own gels.. My intention is to take off whatever color filters are on the club lights, and replace them with my own rosco gels.

The lights i'll be using are definitely 3200k. So my white light will be set to match them (well technically 4400k since I'm using an 80C filter). If the club's strobing lights are something like 5500k (which I fear they may be), I'll either have to dim them til they hit 3200k, or CTO them to 3200k, before I put my color gels on them, in order for my colors to look right, right?

Sorry, I know that was a confusing question, but I didn't know how else to word it.

 

[David Battistella]

I am not sure if this has been covered but I would strongly suggest that you shoot the new Color science of build 20 if you are going with a lot of tungsten sources. I just did a wide series of testing comparing build 17/18 and build 20 and the Tungsten is MUCH more accurate and far less noisy than we have come to expect from RED.

The sensor still loves to have the blue channel activated so keep that in mind as well.

David

 

[David Mullen ASC]

Usually it's not a problem to let the daylight-balanced strobes go blue-ish, partly because they go very bright, so it's not a deep blue color, and also because they aren't on all the time, they are very brief. Plus what's wrong with odd colors in a nightclub?

The bigger problem is the split-frame effect during a strobe flash due to the rolling shutter... now and then, it's not enough of a problem to worry about, but it may get annoying if the strobe is constantly firing.

 

[Bryce Lansing]

Usually it's not a problem to let the daylight-balanced strobes go blue-ish, partly because they go very bright, so it's not a deep blue color, and also because they aren't on all the time, they are very brief. Plus what's wrong with odd colors in a nightclub?

Thats very true, but in this certain case I want to keep all light (except the key/fill) in the club either Cyan or Purple. If the camera is balanced to the tungsten lights, and I shoot blue light through the Cyan and Purple gels, won't it add blue to the colors as well?

The bigger problem is the split-frame effect during a strobe flash due to the rolling shutter... now and then, it's not enough of a problem to worry about, but it may get annoying if the strobe is constantly firing.

I wondered about that.. I heard the Red didn't have as much of a problem with that as the Sony EX series CMOS sensors.. How often does that happen? And if it does, will I see it in the monitor? Or will I not know it happened til I get the footage?

 

[Ketch Rossi]

I also like to trow in a different shooting scenario were Arri's Tungstens Fresnels true Blue are used together with fluorescent tube lighting such as the Kino's Diva Lite 400 and the Cool Lights.

BTW, still uncertain to which tubes are best if the Coll lights rated at 5600k, but tested at 5430k or the ones used on the Diva lite series?

So here we need the best CTB to apply on the 1k True Blue and reproduce more accurately matched light to the fluorescents, for both Photography and Cinematography purposes since this is a project that we will be extracting frames for still manipulation as well.

Also what it would be the true light output with the CTB? How much would the 1k fall off?

I understand that the one one most commonly used is the 80A Full to go form 3200k to 5600k, and about the 2 Stop loss, but never did calculate the actual watt out put loss, and which is the bets one of the available ones.

Thanks

ciao