Personal View site logo
Make sure to join PV on Telegram or Facebook! Perfect to keep up with community on your smartphone.
Please, support PV!
It allows to keep PV going, with more focus towards AI, but keeping be one of the few truly independent places.
The Myth of colour science
  • 6 Replies sorted by
  • hmmm... seems to me this guy is primarily a photographer and not a videographer or he would have known you don't just dial in Canon colors on a Sony camera and call it a day or the A7S would have had a bit more life to image with these simple in camera "Canon" color tweaks.

    That said when you work hard enough like Blackmagic did to get close to the color science of the Alexia, you can get close enough (excluding dynamic range) as they did as demonstrated in this results video of this blind camera test. I think Red came in last do to their color science... so IMHM color science clearly does make a difference.

  • @majoraxis

    Thing is, companies do everything to make "color science" looks like magic. While it is tuning of on sensor filters transmittance curves (very limited!) and after this only small things they do in algorithms (can be emulated with LUTs quite good usually).

    Alexa main advantage is their color filters. You can find some previous topic where I posted curves.

  • When IPP2 came out for RED, you could apply it to the much older RED cameras and their 'look' transformed into something completely different. There was no mechanical change, just software updates for color science. Also, color science does matter as each CMOS sensor has their own sensitivities to UV, IR, and spectral locus points. The color science gives both hue and saturation points across a wide variety of gamut mapping. If you want to match up cameras perfectly, you have to align their hues in a nonlinear fashion(which is harder on cheaper cams cause they warp over exposure) If you're curious what this looks like, shoot a color chart while changing the exposure in real-time. The hues will warp in the vectorscope and won't be a straight line. That's called 'cheap color science'.

  • Also, color science does matter as each CMOS sensor has their own sensitivities to UV, IR, and spectral locus points.

    It shorten it - it is all defined by transmittance spectrum. UV and IR part of filter are usually on cover glass atop the sensor.

  • I'm having hard time matching the 4.6k sensor of ursa to Alexa mini . Tried the color space transformer on resolve ,helped make log even ,but colors seem off. Alexa always seems to have vibrant and pleasing colors ,while ursa seems like its over saturating things to get same result but not as pleasing, if anyone here have matched them, would love to get a better idea of ideal settings etc