Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Circle of Confusion

Does it make sens to do it before Build 15? I don't know if there is an objective and empirical way of determining the outmust resolution on files that are software based (r3d up to Scratch output)?

See you.

Pat
 
I understand your point software will change it somehow but.... what do you say to the AC´s. Hello mate you just have to pray and somehow you will have everythng needed in focus ? :ranting2: <--- His reaction
It has sense doing it before 15 because you want have it until somewhere in march and i am shooting before that time.
And yes there are empirical ways of doing it. I will work on it during this week
Hopefully I will come to a conclusion.
Jon Corcuera
 
I now actually work like if it would be 35mm... But is it enough... or too much?
Testing the whole pipeline until 4k screening would for sure tell us...

BTW, I love your 1rst AC :-)

See you.

Pat
 
For something like this, I believe we need different CofC's depending on the final product. For example, something going to 1920x1080 HD will be much more forgiving that a 4K filmout.

I can tell you that looking at the footage in 4K on an Apple Cinema Display the Circle of Confusion is much smaller than we are used to with 35mm!

A good starting point might be to see what DSLR photographers have come up with.

And if anyone knows David Eubank, we will have to get him to do a Red version of Pcam .
 
CrewPix.com said:
For something like this, I believe we need different CofC's depending on the final product. For example, something going to 1920x1080 HD will be much more forgiving that a 4K filmout.
Click to expand...

It could be (should be) ouput format independant and we should take "best resolution possible" achieved with REDONE to determine CofC's.

You would'nt argue about TV series shot on 35mm negative. Why should we change that because of the variety of possibilities of output formats RED gives us?

my 2 cents...

Pat
 
Tresch said:
Would you accept to shoot 35mm tv series with soft focus just because it will end on a TV screen?
Click to expand...

No, but I would calculate the focus split for a product shot based on how it will be used. Why ask the DP to light to an 11 when 5.6 will give you what you need for the maximum resolution the shot will ever be output? Asking someone for four times more light isn't a good career move.
 
Well, since RED's sensor is basically S35mm sized, (4mm shorter in height) you can use a 35mm guide on a kelly wheel or depth of field calculator to figure everything out, could you not?
 
The cc is directly proporcional to the distance that is wachted and the enlargement. There for I think that Theaters and TV will have diferent CC.
I pretend to do both at least.
Th softer Idea Tresch is not completely like that they will be in focus. Then you can decide if it is valid for your sho or not. The extra info wan´t heart ...
Jon Corcuera
 
Circe of Confusion determination

Circe of Confusion determination

It's likely that 35mm circle of confusion (CofC) figures would be correct, but it would be good to know for sure. A medium consisting of orderly pixels rather than random grains may make a difference. (Here I'll reference Dan Burkholder's book Making Digital Negatives for Contact Printing especially the parts on stochastic printing. Burkholder's website is (http://www.danburkholder.com/Pages/main_pages/book_info_main_page1.htm) if you're interested.)

As part of learning 4x5 photography I had to determine my own CofC by shooting tests and examining full-sized prints at proper viewing distance, etc. Among the valuable things I learned: Just because a subject is inside the DOF range doesn't mean it's sharp (only one plane is actually in focus) and that that what is in 'good enough' focus is a subjective thing - hence a personal circle of confusion.

I'd thought to do the exercise again with our Red, but what has stymied me is that I have no way to view the image at 'finished quality and viewing distance' - not without a 4K projector. I could do a test for a 1080 'print' - and that makes sense for an HD out - but was wondering if anyone had done something like this at 4K.

The pCam software on my Palm uses a CofC value of 25.4 microns, which I believe is .0254mm, for 35mm aperture (not S35). Any other values being used?

Nelson
 
CrewPix.com said:
No, but I would calculate the focus split for a product shot based on how it will be used. Why ask the DP to light to an 11 when 5.6 will give you what you need for the maximum resolution the shot will ever be output? Asking someone for four times more light isn't a good career move.
Click to expand...

Agree 100%

35mm for TV = 1/500 inch
35mm for Theatre = 1/1000 inch
S16 blowup for Theatre = 1/2000 inch
RED 4K = ? 1/2000 inch ?

my 2 cents,
antoine.
 
Corcuera said:
Has anybody calculated de cc (circle of confusion) for the red at 4k, 3k and 2K?
It is about time to make some kellys...
Will have to do my own anyway but just asking for having a preinfo.
Jon Corcuera
Click to expand...

CoC for 4k, 3k and 2k are the same for the same sensor! It is about 7 micro meters.
 
Pawel,

How did you find out the 7 micro meters?

IMO CoC vary in relation to the size of the screening, hence different CoC for TV or theatrical release, although it is the same size of image capture: 35mm film

If your 7 micro ( I assume for theatrical release) is right it means it is 3.6x smaller than the CoC for 35mm film (0.0254 vs 0.007 milimeter) ....

thanks,
antoine.
 
laboprod said:
Pawel,

How did you find out the 7 micro meters?

IMO CoC vary in relation to the size of the screening, hence different CoC for TV or theatrical release, although it is the same size of image capture: 35mm film

If your 7 micro ( I assume for theatrical release) is right it means it is 3.6x smaller than the CoC for 35mm film (0.0254 vs 0.007 milimeter) ....

thanks,
antoine.
Click to expand...

CoC is a property of the sensor or film limit of resolution. Assuming photo sites are about 5 micro meters, after Debayer pattern, you end up with approximately 7-10 micro meters of CoC. This is the number you use to estimate critical focus and DOF at the acquired resolution (the same for 4k, 3k or 2k). If you want to downsample the acquired resolution because of your output (say TV display) by, say half the acquired resolution, then you can double the CoC. Hope this makes sense :nerd:
 
Calculating CoC

Calculating CoC

Sharper lenses + "sharp" digital sensor = less apparent depth of field. The old standards (like CoC 1/1000th inch) will let you down big time.

Here's how I would work it out:

Micron = micrometer = µm = 0.001mm

29 sq µm pixels = 5.4 µm pixel pitch = 92 lp/mm maximum theoretical resolution

92 lp/mm x OLPF factor 0.8 = 74 lp/mm maximum captured resolution (as per Graeme Nattress)

74 lp/mm = 6 µm airy disc size; So there would be no point is a CoC smaller than .006 mm for any reason.

Now, determine the maximum resolution of the output format. Let's say it's 2k:

92 lp/mm (full 4k) /2 = 46 lp/mm = 10 µm CoC [0.01mm, 0.0004 in]. Anything within that CoC would look sharp in 2k projection.

0.0005 in / 12 µm would probably be good for 1080p final product.

Most good modern lenses have decent response at 40 lp/mm, so the MTF at that spatial frequency would be high enough to be captured effectively by the sensor.

I should add that if your limit for acceptable focus is an airy disc diameter of 12 µm, then your focus calibration (accuracy of lens setup and camera's lens mount flange focal distance) will have to be at least within the same tolerance, a total of 12 µm, so +/- .0005 mm. This is the same tolerance I use to set up film cameras, which can be set up by direct measurement.

I can't see this happening with old beat up lenses bought off shady sellers on eBay, or with frequent lens mount swaps on the RED ONE without checking the FFD with proper test equipment (and that means an accurately calibrated sensitive collimator and an accurate wide, fast test lens).
 
CinemaTechnic said:
Sharper lenses + "sharp" digital sensor = less apparent depth of field. The old standards (like CoC 1/1000th inch) will let you down big time.

Here's how I would work it out:

Micron = micrometer = µm = 0.001mm

29 sq µm pixels = 5.4 µm pixel pitch = 92 lp/mm maximum theoretical resolution

92 lp/mm x OLPF factor 0.8 = 74 lp/mm maximum captured resolution (as per Graeme Nattress)

74 lp/mm = 6 µm airy disc size; So there would be no point is a CoC smaller than .006 mm for any reason.

Now, determine the maximum resolution of the output format. Let's say it's 2k:

92 lp/mm (full 4k) /2 = 46 lp/mm = 10 µm CoC [0.01mm, 0.0004 in]. Anything within that CoC would look sharp in 2k projection.

0.0005 in / 12 µm would probably be good for 1080p final product.

Most good modern lenses have decent response at 40 lp/mm, so the MTF at that spatial frequency would be high enough to be captured effectively by the sensor.

I should add that if your limit for acceptable focus is an airy disc diameter of 12 µm, then your focus calibration (accuracy of lens setup and camera's lens mount flange focal distance) will have to be at least within the same tolerance, a total of 12 µm, so +/- .0005 mm. This is the same tolerance I use to set up film cameras, which can be set up by direct measurement.

I can't see this happening with old beat up lenses bought off shady sellers on eBay, or with frequent lens mount swaps on the RED ONE without checking the FFD with proper test equipment (and that means an accurately calibrated sensitive collimator and an accurate wide, fast test lens).
Click to expand...

Very nicely described, Jorge.
 
I have found 35mm CoC to be very unsatisfactory. I add at least 1 stop to any calculations made using the 35mm Kelly wheel or Pcam.

The fill factor (gap between pixels) on the chip is the main cause for this discrepancy. Film is covered with randomized sensitized particle, and the sensor is not, things appear to fall off much quicker as a result.
 
You must log in or register to reply here.
Back
Top