I've a scene file for the Panasonic Varicam and would like to put together
a graph showing IRE values per overexposure and underexposure.
The settings are optimised for post color grading -- along the lines of
the Ben Allan Scene File #1. As well, I'd like to share them with the
rest of the CML Community.
The settings use the CineGamma Film Rec curve as a framework with slight
modifications to the black levels to allow for even more manoeuvrability
in post color grading.
For this test, I know I need to set the 18% grey card to a certain IRE
value. I've seen information that a properly exposed 18% Grey Card creates
an IRE value of anywhere from 42 to 56 IRE. (Which is really .42v to .56v?)
What is the correct response value for the 18% grey card at proper exposure?
That in place, I can accurately graph my response curve.
Look forward to responses!
Jason Scott Cohen
Dallas, Texas USA
I may get flamed for this but I do not think that there is a correct answer
to your question. Where you place middle grey determines where you want
the middle to be. The lower the IRE value you assign to it allows more
space for what is exposed above middle grey. If you assign a "high"
IRE value to 18% grey, then there is more room for everything exposed
beneath it. This translates visually into smoother transitions from one
degree of exposure to the next. It all depends on what is in front of
the lens and where you want to allow room for exposure given the fixed
number of units on the IRE scale and a video camera's latitude.
A Kodak Standard Grey Card has a reflectance of 18% which renders as a
"normal" mid grey exposure density on film. In video terms this
equates to 50% or 0.5 volts IRE.
This is a purely theoretical measurement and, I for example, peg the normal
exposure at 45% to gain a little extra highlight detail. Some people measure
the grey card exposure to be in the mid 50%'s by checking it's level after
setting a "normal" video exposure of skin tones at 70%.
I hope this helps.
Ben Allan ACS
Director of Photography
Jason Scott Cohen wrote
>For this test, I know I need
to set the 18% Grey Card to a certain IRE value.
Okay, I'll jump in on this one.
The answer is, "It depends". I know, I know. You hate it when
they say that.
The 50% or .5v volt value for 18% optical grey is a good starting point.
(For folks looking at NTSC with a 7.5 IRE setup it's 53-54-ish IRE). There
are dependencies. The actual gamma grey point is also affected by the
other settings for the shoulder and toe performance of the transfer curve.
If you start the shoulder transition sooner and reduce shoulder gain in
Zone 8-9 a bit more aggressively then there is often an associated slightly
higher grey point (55-ish?) for 18%.
Conversely, if you keep the shadow settings deep - "crush blacks"
- use a lot of roll-up - call it what you will, but it's the opposite
effect - with lower gain set in Zones 1-2 and a slight reduction (45-ish?)
in the mid grey point 18% point.
The uber issue is about translating emotion to photometric principles.
Do you want the scene to be sullen or mysterious? Do you want a cheery,
All those tweaks along the transfer curve play their part in mapping the
light across the Zone system as you would have the camera interpret it.
Then there's the light and the exposure. Contre jour? Dawn/dusk? What
to keep - what to clip…you have some freedom here.
Thus, the 50% point for 18% optical grey is a "norm" - with
creative latitude. A classic, fairly linearised "video gamma"
setting gets you to 48-52%.
FYI: Another useful "rule of thumb" - with a typical camera
gamma setting the linearized transfer curve response across the centre
range of the signal will be about 16 to 18% per stop.
Fellow, Advance Development
Co-founder, Avid Technology
Ben Allan writes
>A Kodak Standard Grey Card has
a reflectance of 18% which renders BA as a "normal"
mid grey exposure density on film.
Ben, 0,5 volts isn't 50 IRE or 50%. 50 IRE (or 50%) s equal 0,65 volts,
in reference to sync level, or 0,35 volts, in reference to video black.
Image is between 0,3 and 1 volts, the rest is for sync (0 to 0,3 volts).
Adriano S. Barbuto
Sao Paulo / Brasil
I did some test in this way ( a relation IRE-T Stop-RGB value-Negative
density) with F900.
You can see it in a magazine called Shooting from Spain.