The DSLR, even in sRGB JPEG mode, holds element within the grey card at 3+ stops of overexposure on this case.
It’s because when Canon says these JPEGs are “sRGB,” that defines their right profile for show, however not essentially their precise encoding. The encoding profile — the colour changes and tone curve used to convert the linear uncooked sensor information to a viewable picture — could also be primarily based on the sRGB curve, nevertheless it has some subjectivity baked into it; possible a little bit little bit of s-curve distinction, and a few spotlight rolloff.
And that’s with the “Customary” Image Profile, sRGB, and JPEG — possible the least dynamic vary this digital camera would ever current. A uncooked file, log video, or perhaps a less-contrasty profile may provide a considerably gentler spotlight remedy.
Should you work in linear-light, you’re doing issues proper — however if you’d like your outcomes to look pleasing and/or photographed, an sRGB lookup alone shouldn’t be adequate.
sRGB and Gamma Visualized
Earlier than we skewer the sRGB “gamma” as a view rework, let’s look at what it truly is.
First, some terminology. Strictly-speaking, gamma is an influence operate. A gamma of two.2 is similar as elevating the pixel worth, on a 0.0–1.0 scale, to the facility of 1/2.2. However the time period gamma has been broadened by some to incorporate any sort of 1D tone curve utilized to, or attribute of, a picture. Life is simpler with this relaxed definition, in order that’s how I take advantage of it.
Gamma Administration is the time period I take advantage of for a workflow that makes use of 1D lookups/conversions between codecs. Magic Bullet Appears to be like 5 and Supercomp 1.5 use Gamma Administration fairly than full coloration administration.
You’ll be able to completely gamma-manage your workflow utilizing the pure gamma-2.2 and its inverse. But when your imagery is sRGB, it’s barely extra correct to make use of the sRGB curve. The sRGB tone curve is a really shut match to a pure gamma 2.2, nevertheless it has a little bit kink on the backside to unravel an previous downside.
A pure gamma curve has a slope of 1.0 or 0.0 at its base, i.e. because the values within the picture strategy zero, the gamma curve approaches a flat line. Which means calculations on the darkest pixels in your picture may very well be inaccurate, and people inaccuracies may compound by way of a number of steps of linearization and de-linearization.
sRGB has a steep, however not infinitely steep, linear slope on the very backside, after which the remainder of the curve makes use of a gamma of two.4 squished to slot in the remaining vary. The intelligent result’s that the curve is clean on the transition and strong by way of a number of generations of processing, even when the processing shouldn’t be performed in floating-point.
It’s straightforward to see how related the gamma 2.2 and sRGB curves are by graphing them:
