Excellent work, Harv. Yet again, a very well structured walk-through that gets the information across clearly. I'm now motivated to get out some lights and experiment with temperatures and my own skin tones. Thank you.
Interesting video, Harv. I think your constant referral to "tungsten" muddies the analysis. I believe that using actual tungsten lighting would produce superior skin tones than most any LED set to 2700 kelvin. Perhaps producing a video with tungsten compare/contrast to LED would reveal it so.
Great overview on the value and impact of mixing color temperatures with intent. For the newbie, you may want to explain how to do a custom white balance when working with mixed-color temperatures. For example, are you doing it with the key light only before turning on the other lights? Doing it with all the lights on, even if your white/grey card is on the key side of the face may yield different results? Perhaps there may even be motive to do one or the other depending on desired results.
This is great as its something I mess around with a lot on my videos and looks like a similar size room as well. I think around 5600k is the one I am going to go with for my next few videos.
Good info on the subjective nature of mixed light color temps. If you haven't seen already, MarkusPix just did a video that answers some of your questions (why CRI is a joke, etc.) and he has some good examples using a Sekonic color meter that clearly shows the spectral differences and why all modern LED lights are nothing like real tungsten video lights. But I noticed you didn't show or mention post-production white balance color correction (objective) or color grading (subjective) and how it affects the results you are seeing. WB settings in cameras are still not all that accurate. I manually set the camera WB to the dominant key light temp and then adjust in post.
Hmm, I am a little confused, tungsten is 3200k not 2700k? In addition, lights set to 3200k look a lot better as clean color using a Sekonic C-800 for a measurement device.
The spyder colour chip card is not a good match for rec.709 broadcast, as none of the chips hit the std broadcast saturated colours, or skin tone variations, that is why the Calibrite (Xrite) color checker cards are more widely used. The spyder was designed with print photography in mind. If you use resolve your card would be more useful as it allows you to closely compare the shot card, and attempt to auto balance.
It is also relatively common to set camera to 5600k, and use lighting temp to set the mood. Because this allows one to shape the light the way you want in a consistent fashion.
Thanks for doing this Harv - very interesting. 2,700K skin looks best to me even though (or maybe because) it’s a bit more magenta. In comparison the higher kelvin tests look a bit/green/blue but look fine in isolation. 3,200K incandescent tungsten lights were always considered the most colour accurate and flattering for skin tones back in the day.
You deserve more subscribers, sir. Thanks for all the hard work you put in for these videos.
Many thanks 🙏🏻
Nice content! Lots of take-aways.
Excellent work, Harv. Yet again, a very well structured walk-through that gets the information across clearly. I'm now motivated to get out some lights and experiment with temperatures and my own skin tones. Thank you.
Fantastic! As always, it's all about context. :D
Thanks for the shoutout! :)
Interesting video, Harv. I think your constant referral to "tungsten" muddies the analysis. I believe that using actual tungsten lighting would produce superior skin tones than most any LED set to 2700 kelvin. Perhaps producing a video with tungsten compare/contrast to LED would reveal it so.
Great overview on the value and impact of mixing color temperatures with intent.
For the newbie, you may want to explain how to do a custom white balance when working with mixed-color temperatures.
For example, are you doing it with the key light only before turning on the other lights?
Doing it with all the lights on, even if your white/grey card is on the key side of the face may yield different results?
Perhaps there may even be motive to do one or the other depending on desired results.
Questions I'm keen to know the answers to as well - hope someone helps!
This is great as its something I mess around with a lot on my videos and looks like a similar size room as well. I think around 5600k is the one I am going to go with for my next few videos.
HUGE LIKE for SIKTH shirt!) Oh, by the way, thx for the great tutorial tho!
Good info on the subjective nature of mixed light color temps.
If you haven't seen already, MarkusPix just did a video that answers some of your questions (why CRI is a joke, etc.) and he has some good examples using a Sekonic color meter that clearly shows the spectral differences and why all modern LED lights are nothing like real tungsten video lights.
But I noticed you didn't show or mention post-production white balance color correction (objective) or color grading (subjective) and how it affects the results you are seeing. WB settings in cameras are still not all that accurate. I manually set the camera WB to the dominant key light temp and then adjust in post.
Hmm, I am a little confused, tungsten is 3200k not 2700k? In addition, lights set to 3200k look a lot better as clean color using a Sekonic C-800 for a measurement device.
The spyder colour chip card is not a good match for rec.709 broadcast, as none of the chips hit the std broadcast saturated colours, or skin tone variations, that is why the Calibrite (Xrite) color checker cards are more widely used. The spyder was designed with print photography in mind.
If you use resolve your card would be more useful as it allows you to closely compare the shot card, and attempt to auto balance.
It is also relatively common to set camera to 5600k, and use lighting temp to set the mood. Because this allows one to shape the light the way you want in a consistent fashion.
Thanks for doing this Harv - very interesting. 2,700K skin looks best to me even though (or maybe because) it’s a bit more magenta. In comparison the higher kelvin tests look a bit/green/blue but look fine in isolation. 3,200K incandescent tungsten lights were always considered the most colour accurate and flattering for skin tones back in the day.
Helo House M.D!