General > General high-speed discussion

First shots with the early bird camera!

<< < (3/3)

jasonfish:
I'm really hoping it's the lens.

tesla500:
The softness is due mainly to the lens, followed by the demosaic algorithm. The shot you picked is one of the sharper ones, and is limited more by the demosaic than the lens I believe, especially for the duck in the center. The ones at the edge are more focus limited due to the lens, and possibly motion blur. The current bilinear demosaic does introduce some softness. A much sharper AHD demosaic is on the roadmap, and of course with RAW saving the demosaic is up to the user in post. It will never be as sharp as the monochrome camera however, some loss of sharpness is intrinsic to Bayer filter cameras, but with a good demosaic it can be almost as good as mono for most shots.

Simon:
Oh yeah, the demosaic - forgot about that since I ordered the monochrome.

No rush for the sharper demosaic  ;)

Thanks.

gyppor:
The computar 12.5-75 lens is very soft at f/1.2, much better at f/2, and way way better at f/4 and above. Youtube also absolutely butchers image quality, when I saw the first images out of my own camera I was blown away by the difference.

It's also hard to get the focus right at telephoto focal lengths with close up shots because the depth of field get very very shallow. I was filming a wasp from 1m away in a 10cm diameter glass vase, and if I focused on it on the far side of the glass, it would be totally out of focus when on the near side.

For now I can definitely live with a few flaws because that one lens does almost everything I want. It's nice not to have to swap lenses.

Simon:
Just plugged those numbers (75 cm focal length, f/4, subject 1 m away) into the depth of field calculator at http://www.cambridgeincolour.com/tutorials/dof-calculator.htm , using a sensor size of 1/2.3" (closest preset I could see) and you've got just a mere 0.7 cm depth of field. That's assuming you're viewing the image on a monitor with ample resolution at 10" wide from a distance of 10".  Not 100% confident my assumptions are correct, but wow if they are.

That's a bit eye opening actually (no pun intended) - at least for me. Smaller sensors can get subject isolation from the background at much narrow apertures, optimizing the sharper middle ranges of zooms. Just need tons more light.

Navigation

[0] Message Index

[*] Previous page

Go to full version