1

I am eagerly looking for a camera in which I can adjust the gamma correction. The cameras that I have, do this correction by default (which are not adjustable) and I need to remove this correction. Any suggestions?

Thanks

2
  • 4
    Any camera which is capable of shooting RAW and a decent image processing tool can apply any tone curve you like. What are your cameras and what's your workflow for getting from the camera to a final image? – Philip Kendall Jan 30 '18 at 12:18
  • 1
    Without some amount of gamma correction, 8 bits is not enough for decent representation of an image. Since JPEG is only 8 bits it wouldn't make sense to offer that option. – Mark Ransom Jan 31 '18 at 18:11
5

Get a camera that can save its raw data. With the raw data, you can apply whatever brightness curve you want in post-processing.

There are many cameras out there that can save raw data. Look around. Pretty much most cameras above the point-and-shoot level can do this. All pro and prosumer cameras can do this.

0

You don't want to adjust the transfer curve of JPEGs — sRGB is a standard for a reason. 8bit JPEGs are intended for low-data density storage, and using a transfer curve is an important part of that encoding.

8bit JPEGs are wholly inappropriate for linearized image data.

If you want linear, "no gamma" (which actually means gamma 1.0), then you want to use the RAW images from your camera, not the 8bit JPEG. All competent cameras offer a RAW option. RAW is linear (no gamma curve, or more precisely gamma 1.0) straight from the sensor, and before debayering.

From this RAW you can debayer into a linear-gamma1.0 image and save to a floating-point format such as 16bit half-float .EXR which is the current preferred standard for storing linear image data.

0

You don't have to undo or change the OECF in the camera, you could always do it in post. Just let the camera do a decent OECF (for perceptually uniform coding), then in post apply the inverse EOCF back to linear-light (like Raw), followed by your preferred OECF. Or combine the EOCF and OECF into a single EECF, and call it "tone mapping".

It is more important to consider how a 1D conversion curve is applied to 3D color data. Is it applied 3 times to R and G and B individually, or is it applied once to the Luma value and then the same gain applied to the 2 Chroma values ? And is that Luma+Chroma constant-Luminance or not ? All results will be different, at least for saturated colors. Tone mapping in color is not trivial.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.