- This topic has 1 reply, 2 voices, and was last updated 4 years, 1 month ago by
.
-
Posts
-
Can you guys help me with understand the intent of the measurement report Gamma tracking graph? I can’t figure out how to contextualize what that graph says. (Attached)
Just in case I’m missing something obvious, here’s how I am taking it:
I see the X axis is luminance as a percent, and the Y axis is the a pure power law response that corresponds to the measured luminance at that level. And I can hover over each plot point and get the dE. That’s neat.
Yey I don’t know why I want to know the Y dimension in that unit?
It’s a bit funny because there’s nothing wrong with the plot, if what I see amounts to is a list of dEs from reference expressed as exponents
But I expect an XY plot to help me understand a Y trend of the measured data over the X range, but what is the trend when Y refers to yet another curve? I’m not looking for a fit to a lot of gammas, I’m looking for how well does my display fit a specific gamma.
For example, a trend of conformance to a nominal exponent makes sense to me—say a best fit. Or given that some standard TRCs are linear / power law hybrids, where the splice is finer point for a nominal power law response, it would be cool if you could see that kind of detail, although I think the point of that detail is supposed to be lost in the nominal response 🙂
I guess what I want to see is a graph which reveals how well measurements fit the nominal response.
Everything in this program is so well thought through that there must be a use for this graph!
Why is it helpful to see that there’s an undulation at the top range of responses where power ranges from this-to-that exponent? What does it mean that at 70% brightness there’s a slight decrease in the exponent? How does that relate to desired or expected performance?
Danku!
The unfilled circles are the measured response at profile creation time (profile is a model, so it’s best fit, but for all intents and purposes it doesn’t make sense to differentiate between model and actual measured response when it comes to tone response/neutral axis in this context). The filled circles are the current measured response. As this is a log/log graph, the Y unit is gamma. Personally I find such graphs to be a bit useless because gamma deviations have little correspondence to visual perception, the grayscale delta E will tell you the same information in a more useful form.
