Author Topic: Landscape Focus Stacking: A simple tweak that will make ML even better  (Read 3778 times)

garry23

  • Contributor
  • Hero Member
  • *****
  • Posts: 2217
As developers tweak and add to the nightlies I was hoping one kind developer would tweak the LV DOF display.

First of all why?

As a landscape photographer I often undertake focus stacking. This is not of course like macro stacking, where the focus steps are linear and small. Landscape focus stacking is non-linear, and change frame to frame.

For the occasions when the ML returns focus info, eg my 24-105 F/4L, ML provides a realtime aid for the landscape focus stacker.

I focus on the nearest feature I wish to see in focus, note the depth of focus either side, and move my focus into the field until the second focus point overlaps with the first. And repeat until the last focus reports infinity.

At the moment it’s a little tricky for several reasons.

First, you need to do some mental arithmetic as the focus distance is in absolute units, ie ft, and the focus distance either side is relative, ie to the focus distance. What would be great if the focus info reported the absolute distances for all three values. This could be user selectable, ie relative or absolute distance reporting.

Second, the text is really difficult to read, unless you have young eyes. It would be good if the user could select the size of the text, ie normal, double or 4x, say. BTW this would be great for other ML feedback, eg the spotmeter, which is really small and difficult to read.

Finally, I’m not sure what equations are being used, but I assume they use a pupil ratio of 1. So my final ‘request’ is can we use the full equation, including the pupil ratio and have that ratio and the CoC as two user variables, ie see www.toothwalker.org/optics/dofderivation.html. Other variables in the equations are known to ML, ie aperture and FL. 

As I say, I hope some kind developer could do these tweaks and make the DOF ML feature a real tool for landscape photographers.

PLEASE IGNORE SOME OF THE ABOVE: AS I NOW REALIZE THAT ML DOES REPORT ABSOLUTE DISTANCES :-)

Licaon_Kter

  • Hero Member
  • *****
  • Posts: 519
  • M
As you can see in the commit: https://bitbucket.org/hudson/magic-lantern/diff/src/lens.c?diff1=f713bdd0b63e6e8d0b874d84fd94f068c5830c86&diff2=fca7f4b00653&at=unified
The distance is as reported by the camera, according to your lens.

garry23

  • Contributor
  • Hero Member
  • *****
  • Posts: 2217
@Licaon_Kter

Thanks for the link. I know the lens is providing the distance. My request is more about the calculation of the depth of field and how this is presented to the user.

I just posted this on landscape focus stacking. http://photography.grayheron.net/2015/04/magic-lantern-shows-way-yet-again.html

As I say, I hope one of the developers can find time to tweak things.