Welcome to our community
Be a part of something great, join today!
15 to 20 stop range with two RED ONE (tm)
- Thread starter Dan Hudgins
- Start date
Andrew M.
Well-known member
Read this article.
What do you think?
http://www.cis.rit.edu/class/simg232/lab2-dispersion.pdf
We need a bit of more reading to make sure we will not get in to the CA trouble.
- Thread starter
- #63
Dan Hudgins
Well-known member
- Joined
- Feb 16, 2007
- Messages
- 1,369
- Reaction score
- 0
- Points
- 0
- Location
- San Francisco, CA USA
- Website
- www.dancad3d.com
Almost NONE, not an issue.
Almost NONE, not an issue.
The sides of the Pellicle Beamsplitters are parallel and when used in front of two cameras the 2 micron thickness makes the tiny amount of Chromatic Aberation too small to be visable. Put a plate of glass in a beam of light, and tilt it, the rainbow produced is less than the thickness of the plate.
We are not talking about wedge prisms, those also come in achromatic type if you need them, if the sides are parallel and the light is parallel the issue is not something that would cause a problem. Please look at the drawings I uploaded above for a way to put the system together.
Pellicle Beamsplitter used behind the lenses there would be a very small lateral color shift smaller than a pixel or so and less than the OPLF can cause in loss of resolution.
You can use a cube prism (with parallel sides) in front of two cameras if the lenses are focused at INF (parallel rays) and they do not vignette.
The problem you get into is if you use a thick plate and have the rays converging. You cannot use a cube prism behind lenses made to have air behind them, but you can use a cube prism in front of the lenses if they are focused at INF and the face of the prism is flat on to the lenses and the lenses are of long focal length.
With the Pellicle Beamsplitters you mainly need to worry about the angle of total internal reflection, which should not be a problem except with vary wide angle lenses, which I offered a solution to above. Flatness is an issue, you need to be carful about bending the frame that holds the Pellicle Beamsplitter or having wind or vibration. They may also be hard to clean, so it is best to put the optics in a dust tight box with thin coated windows.
Almost NONE, not an issue.
The sides of the Pellicle Beamsplitters are parallel and when used in front of two cameras the 2 micron thickness makes the tiny amount of Chromatic Aberation too small to be visable. Put a plate of glass in a beam of light, and tilt it, the rainbow produced is less than the thickness of the plate.
We are not talking about wedge prisms, those also come in achromatic type if you need them, if the sides are parallel and the light is parallel the issue is not something that would cause a problem. Please look at the drawings I uploaded above for a way to put the system together.
Pellicle Beamsplitter used behind the lenses there would be a very small lateral color shift smaller than a pixel or so and less than the OPLF can cause in loss of resolution.
You can use a cube prism (with parallel sides) in front of two cameras if the lenses are focused at INF (parallel rays) and they do not vignette.
The problem you get into is if you use a thick plate and have the rays converging. You cannot use a cube prism behind lenses made to have air behind them, but you can use a cube prism in front of the lenses if they are focused at INF and the face of the prism is flat on to the lenses and the lenses are of long focal length.
With the Pellicle Beamsplitters you mainly need to worry about the angle of total internal reflection, which should not be a problem except with vary wide angle lenses, which I offered a solution to above. Flatness is an issue, you need to be carful about bending the frame that holds the Pellicle Beamsplitter or having wind or vibration. They may also be hard to clean, so it is best to put the optics in a dust tight box with thin coated windows.
Jason Sinclair
Well-known member
I'd like to see some tests on the blur that you have done because others have tested on other formats and the ghosting or blurring is in fact significant and playing with hdr images myself in still domain its a problem there too.. Not because of the blurring necessarily but because there is less exposure value in the blur area because you only have one frame information compared to the two the rest of the image enjoys, so you also get a darker space. This may be subtle but it theoretically will be noticeable and more to the point unnatural. and if this is present in shot after shot methinks it will be noticeable but if your testing proves that wrong can either post some shots with camera pans and moving subjects???
Andrew M.
Well-known member
I am waiting for permit to film in the place where such test would show walking people and glass dome, with high-rise building above the dome outside, on the backdrop of the blue sky with some clouds. There is exhibition taking place there right now thus the delay.
We could do just pan outside of the office with people walking on the street but I want some HDR in it as well.
You are wrong, when you do 60fps and then you combine two frames in to one and you play it back at 30 fps there is exactly the same blur as you would be filming at 30fps and 180 deg shutter.
Remember we take odd frames from camera A and even frames from camera B, we do merge two such frame in to one and then we use 30fps base, so the action will not be played back two times faster.
Two 60fps frames combined in to one and played back at 30fps.
Here is cut-out from the two frames combined.
See person on the left running in the background.
Nice blur that is all. If you want continues blur on his head and body you just add motion blur in AE, but I found it not needed for 30fps. Maybe for 24 you would add just a touch.
But let's wait for the real test, so far the material is not good for public display because of 3D parallax and close proximity of the door and person running, that I remind you, will not be there if we use one camera with HDR bracketing or one lenses with the beam splitter
Jason Sinclair
Well-known member
Yes lets... You are doing good work. Cant wait to see your results.
I understand using a beam splitter will cure this problem. I really think for a production system this is where thoughts should go. Music videos could use method a) but for drama i think it is limited. But again the tests, especially the panning will show how much is usable.
If somehow you could use the sensor to be split into two frames (top and bottom) and nd the top frame then you would cure this problem. Slightly less resolution but very usable... No one has gone down this track yet... In this case you only need one camera. Its worth thinking about as well.
Can you explain your methodology so we know also the post requirements and wether they fit with real life post production scenarios. Did you work on this frame individually? Or is it a grab from something that has been put through a script or automated process?
For those that use ps you can use actions simply to get results. See help files.
I understand using a beam splitter will cure this problem. I really think for a production system this is where thoughts should go. Music videos could use method a) but for drama i think it is limited. But again the tests, especially the panning will show how much is usable.
If somehow you could use the sensor to be split into two frames (top and bottom) and nd the top frame then you would cure this problem. Slightly less resolution but very usable... No one has gone down this track yet... In this case you only need one camera. Its worth thinking about as well.
Can you explain your methodology so we know also the post requirements and wether they fit with real life post production scenarios. Did you work on this frame individually? Or is it a grab from something that has been put through a script or automated process?
For those that use ps you can use actions simply to get results. See help files.