From OLPF 2 to OLPF 1

[Steve Freebairn]

To really get the best out of the latest available, I'd highly recommend updating RCX.

Hey Luigi, Ive been doing a lot of car stuff lately, a variety really but I do love rally. That Renault 5 is an awesome machine.

It was great witnessing Loeb's Pikes Peak record an Peugeot's triumphant return to the mountain.

Old and the new!

 

[Dane Henry]

Milan, great shots. And while image two is stunning, image one seems lacking a teensy-weensy amount of finer focus on the toggle switches. Was that a result of lower light or just a visual choice?

??? His focus plane is on the Words "Fuel Prime" and the base of the switches....if he had it in the toggle switches you would probably ask "why isnt anything in focus?"....people can't win around here

Who knows, this is just a still grab, focus may have been roving in the actual shot..

nice images Milan

-dane

 

[Milan S.]

To really get the best out of the latest available, I'd highly recommend updating RCX.

That screen grab is from June of last year, I'm running the latest and greatest.

 

[Steve Freebairn]

Haha that's funny!

That screen grab is from June of last year, I'm running the latest and greatest.

 

[Milan S.]

Milan, great shots. And while image two is stunning, image one seems lacking a teensy-weensy amount of finer focus on the toggle switches. Was that a result of lower light or just a visual choice?

They are just quick hand held shots, nothing artsy fartsy. I just wanted to show the quality of the grain you get with OLPF2 in sub-perfect conditions - available fluorescents.

But yeah, the lens was wide open, 1/48 shutter and camera is moving, not perfect for pulling stills but it works with movement.

 

[Robert Ruffo New]

It depends on how you expose it. Yes, less light hits the sensor in v2 olpf ASSUMING that you rate both sensors at the same iso. If you use a light meter like I do, and rate v2 olph at 320 iso, you are now giving v2 olpf 1.3 stops more of light than v1 therefore negating the darkness of the v2 olph and moving your 18% gray point brighter and farther away from the noise floor, thus improving the noise. The camera will now map 18% gray at a brighter value than at 800 and my guess is that the noise will be equivalent between the two sensors. The difference is that now you need 1.3 stops more light into your set. For some of us that's a lot more light to have to come up with. At t2.8 thats the difference between 12 footcandles and 32 footcandles. 12 footcandles is very doable in many situations, especially when using natural/ambient light thats already there. 36 footcnadles exponentially bumps up your lighting needs.

And it's not just cost/time. Actor comfort, neighbor disturbance in urban area and heat also come into play. When we shot film, we sued to spend a lot more time refreshing makeup between takes.

 

[Gavin Greenwalt]

Paul, this is out of context. You are showing a part of an image that's under exposed.

How can you under expose black? Black is black. ISO 2000 is ISO 2000. If he applied secondary gain so that it wasn't ISO 2000 it was actually processed at ISO 8000 I would agree with your criticism but you could shoot the back-side of a lens cap at ISO 2000 and it would be a useful illustration of the noise level of a sensor at ISO 2000. In fact it's the best way to do noise comparisons --and it's the reason you blackshade with a lens cap on. Black + ISO + No additional Gain or Gamma = Good noise test.

 

[David Battistella]

Gavin,

this logic could also carry over to display/projection no? Why would anyone try to display black? It's black, It's a zero value pixel. But if I raise that value to 20 ire and attempt to display that most likely projects or displays noise. Every medium has it's inherent noise and where that noise shows up in the dark areas.

I can't speak for bob, but perhaps he is suggesting that when you underexposed and image, crop the underexposed parts (that perhaps should have fallen into black) and display that as noisy shots a sensor produces then it could be taken out of context.

You are correct about black shading. There are also correct grading practices and gamma curves that go with each system. As an example. An ungraded log image is not an "intended" final image, it shows you the good the bad and the ugly, including underexposed shadow noise. So images need to be judged as a whole. The whole frame, where the shadows fall and the intended final look.

anyone can find noise in underexposed log images, that's easy, but log is not an intended finished look. You'd have to compare the log with the LUTS that RED provide, RG3 and RG4 as those curves read and interpret the data and send that underexposure to black (where it most likely maybe belongs)

battistella