raw histogramm

[StefanKeller.AC]

Do you have access to the raw picture, and can calculate a histogram based on it?
This would be great to see the potential of raw-develop in highlights.

[garry23]

For those of us who favor ETTR capture this would be a great feature. In addition, if the actual histogram presentation can not be made more readable, ie bigger, I wonder if ML could have the option to present the right most quarter only.

At the moment I use blinkies to estimate the ETTR exposure, but of course this is on a highly JPEGed image.

Bottom line: I wonder if ML could support the ETTR shooter?

Garry

[scrax]

This is interesting: http://www.rags-int-inc.com/PhotoTechStuff/ETTR/

Finally, the image in the preview window has been color rendered and re-sampled down to a small size.   This is the data shown in the histogram.   There is no practical way to map all of the raw data in an image into a histogram that you could use effectively in the preview window.   The camera can capture all colors in the spectrum, but the rendered image is limited to the gamut of an RGB color space.   So, in addition to exposure clipping the histogram will include gamut clipping.   This is also true for the blinking highlight and shadow tools.   They might imply an exposure problem when none exists.   Out of gamut colors will contain RGB values at zero and 255.

[StefanKeller.AC]

I don't favor ETTR. I exposure more or less to my subject,
but I take care that important highlights are not (to much) blown out (in raw)
A raw histogram would be helpful to check...

[garry23]

ETTR is an emotive subject. The fact our camera meters to 'midde gray' does mean that we are not fully exploiting the data capture in RAW, ie the binning of data is not linear in RAW. This is why we use exposure compensation.

I believe ETTR has its maximum value when you are capturing a dynamic range that is addressed by a single image.  What we don't want to do in this situation is clip any highlights with 'structure' in them, ie details we wish to see. I will ignore speculars.

As the histogram and blinkies are calculated from an over sampled JPEG, ie a few percent of the RAW data, it is often hit and miss that we have pushed as much data to the right as possible. My simple understanding is that with ETTR we are increasing the s/n at a given level, ie noise stays about the same, but we 'just' gather more signal.

For scenes with a wider dynamic range we have bracketing.

Bottom line: I believe a RAW based measure of highlight clipping would take ML to another high.

Garry

[a1ex]

For now, the closest approximation is UniWB. ML will reduce the green color cast, it's not perfect, but better than nothing.

[clint]

Ron Brinkmann (visual effects artist and digital imaging pioneer) mentions the value of a RAW histogram in his blog here http://digitalcomposting.wordpress.com/2011/05/20/your-camera-it-lies/

Alex you might be interested in his other entry here about his desired camera functions:  Later in the article its becomes a bit more software focused.
http://digitalcomposting.wordpress.com/2012/07/02/camera-features-i-want/
If you haven't gotten any attention from him or any of his associates yet (I'm sure you wouldn't be able to talk about any specifics or out of manors just not), he might be interested in what ML is doing these days.  Though I'm sure he and the rest of the educators & leaders in VFX are paying plenty of attention (if not for professional purposes now, at least for personal).

[Francis]

A lot of the software features he mentions has been discussed on the forums and before the forums on the google group  but I can assure you there are no secret talks with bloggers whose topics couldn't be revealed. Magic Lantern is open source and you can see the new commits nearly every day.