Bern Caughey
Veteran
Interesting conversation going on over at CML. Added the original post below, plus a couple of the responses that stood out, but suggest reading the entire thread.
https://cml.news/g/cml-lighting/topic/how_much_colour_control_do/13271819
- - -
“Talking to a manufacturer who is looking for input…
Bi-color is the rage, magenta/green & RGB is the talk BUT what do gaffers & DP’s really want?
Now OK, in an ideal world I want DMX control of RGB but in a real world, bearing in mind speed of operation and cost, I tend to feel that bi-colour plus a G/M axis shift is all I really need. In fact it’s simple and easily controllable.
What do you want? Bear in mind weight, speed of use and cost
”
Cheers
Geoff Boyle NSC FBKS
Cinematographer
Zoetermeer
- - -
“I'll put the science in here.
Bi-colour generally gives the best colour rendering, because the two or three phosphors used are broad-spectrum and don't leave any serious gaps. Adding a G/M shift is a real problem because it means that the basic white has to be made with a proportion of the G/M control, so the broad-spectrum bi-colours have to have a hole in green to make way for it, which isn't a natural thing to have (none of the luminaires of this type that I've measured score high in TLCI).
RGB etc is an effects lamp unless the R G and B are broad-spectrum: none of those I've measured score at all well when used to produce white.
So, we have a dilemma: good colour rendering means little control other than CCT, and good coloured effects means poor white performance. Any imrovement in this area is going to mean a significant hike in the cost of the units.”
Alan Roberts
- - -
“Okay time to wade in here...
So I would like to just try and blow up this entire conversation for a second. There is no such thing as white light. There is only limited spectrum and full spectrum light sources, they are all colored. Secondly there is no longer any reason to use any single color temperature as the measurement point for valuing a light's performance, that was tied to limited white light photographic sensitivity for accurate color reproduction... color film stocks. Which is my way of saying you are all asking the wrong questions.
You should be asking for two things and two things only: spectrum and control. Forget the wide variety of numbers, letters and hyphenates we use to describe light sources, they are almost all marketing schemes from technologies and industries that have moved on. 3200K and 5600K are anachronisms of Kodak film stocks, green and magenta are camera filters you no longer use since the advent of DSLRs. They no longer accurately describe the majority of modern light sources nor camera sensitivities. "Daylight" color temp is a product of altitude, latitude, atmosphere and time of day. Moonlight on camera is an attempt to represent the relative experience of the human eye under low light conditions at night based on scotopic vision or how the rods in our eyes are more sensitive to certain wavelengths , and has no relationship to any scientific measurement of the spectral properties of daylight reflected off a giant dirt bounce card in the sky.
Which brings me back again to spectrum and control. As a photographer for motion or stills you want the light source with the greatest overall portion of the visible spectrum produced and with the ability to control aka limit that spectrum as needed for your emotional, aesthetic or technical goals. For a very long time the best way to do this was to take a incandescent light source, and have a large library of color gels. I have a shocking surprise for you. If you have ever lit something well using a gelled tungsten light, you have effectively lit for camera with a very low CRI source. For that matter there are plenty of bad RGB lights, bad bi-color lights, you can use garbage phosphors and you can have ugly but high CRI lights, etc. etc. If you buy a light because it renders colors at 3200K perfectly and use it in space with an average ambient color temp of 4500K for a camera with a sensor with a native sensitivity of 6500K, did that test matter and benefit you at all?
Spectrum and control, spectrum and control, spectrum and control after that economics and physics will always apply, its the manufactures job to give you as much of this as possible, and try to make a viable product by considering $ and size, but the rest of the number and letter soup is mostly noise.
Now get of my lighting lawn... “
Grumpy as ever
Jon Edward Miller
Chief Product Officer / Founder
@hivelighting
Los Angeles, CA USA
www.hivelighting.com
https://cml.news/g/cml-lighting/topic/how_much_colour_control_do/13271819
- - -
“Talking to a manufacturer who is looking for input…
Bi-color is the rage, magenta/green & RGB is the talk BUT what do gaffers & DP’s really want?
Now OK, in an ideal world I want DMX control of RGB but in a real world, bearing in mind speed of operation and cost, I tend to feel that bi-colour plus a G/M axis shift is all I really need. In fact it’s simple and easily controllable.
What do you want? Bear in mind weight, speed of use and cost
”Cheers
Geoff Boyle NSC FBKS
Cinematographer
Zoetermeer
- - -
“I'll put the science in here.
Bi-colour generally gives the best colour rendering, because the two or three phosphors used are broad-spectrum and don't leave any serious gaps. Adding a G/M shift is a real problem because it means that the basic white has to be made with a proportion of the G/M control, so the broad-spectrum bi-colours have to have a hole in green to make way for it, which isn't a natural thing to have (none of the luminaires of this type that I've measured score high in TLCI).
RGB etc is an effects lamp unless the R G and B are broad-spectrum: none of those I've measured score at all well when used to produce white.
So, we have a dilemma: good colour rendering means little control other than CCT, and good coloured effects means poor white performance. Any imrovement in this area is going to mean a significant hike in the cost of the units.”
Alan Roberts
- - -
“Okay time to wade in here...
So I would like to just try and blow up this entire conversation for a second. There is no such thing as white light. There is only limited spectrum and full spectrum light sources, they are all colored. Secondly there is no longer any reason to use any single color temperature as the measurement point for valuing a light's performance, that was tied to limited white light photographic sensitivity for accurate color reproduction... color film stocks. Which is my way of saying you are all asking the wrong questions.
You should be asking for two things and two things only: spectrum and control. Forget the wide variety of numbers, letters and hyphenates we use to describe light sources, they are almost all marketing schemes from technologies and industries that have moved on. 3200K and 5600K are anachronisms of Kodak film stocks, green and magenta are camera filters you no longer use since the advent of DSLRs. They no longer accurately describe the majority of modern light sources nor camera sensitivities. "Daylight" color temp is a product of altitude, latitude, atmosphere and time of day. Moonlight on camera is an attempt to represent the relative experience of the human eye under low light conditions at night based on scotopic vision or how the rods in our eyes are more sensitive to certain wavelengths , and has no relationship to any scientific measurement of the spectral properties of daylight reflected off a giant dirt bounce card in the sky.
Which brings me back again to spectrum and control. As a photographer for motion or stills you want the light source with the greatest overall portion of the visible spectrum produced and with the ability to control aka limit that spectrum as needed for your emotional, aesthetic or technical goals. For a very long time the best way to do this was to take a incandescent light source, and have a large library of color gels. I have a shocking surprise for you. If you have ever lit something well using a gelled tungsten light, you have effectively lit for camera with a very low CRI source. For that matter there are plenty of bad RGB lights, bad bi-color lights, you can use garbage phosphors and you can have ugly but high CRI lights, etc. etc. If you buy a light because it renders colors at 3200K perfectly and use it in space with an average ambient color temp of 4500K for a camera with a sensor with a native sensitivity of 6500K, did that test matter and benefit you at all?
Spectrum and control, spectrum and control, spectrum and control after that economics and physics will always apply, its the manufactures job to give you as much of this as possible, and try to make a viable product by considering $ and size, but the rest of the number and letter soup is mostly noise.
Now get of my lighting lawn... “
Grumpy as ever
Jon Edward Miller
Chief Product Officer / Founder
@hivelighting
Los Angeles, CA USA
www.hivelighting.com
Last edited:
