Welcome to our community

Be a part of something great, join today!

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 :)

Red one dof

M

Mike N

Well-known member
Joined
Nov 4, 2008
Messages
232
Reaction score
0
Points
0
Hello,

An AC showed me today a Iphone app (can't remember the name) that shows technical things about using lenses and such amongst them a DOF calculator for RED 35mm and 16mm. When using this the dof range for the Red one was less than that for 35mm. How can this be ? I don't remember the Reds sensor being larger than 35mm format ?
He also told me that because of this focus pullers find it harder to work with the red as it is more unforgiving than 35mm film which has a larger DOF.
:ack2:

Is this true ?
 
Mike N said:
Hello,

An AC showed me today a Iphone app (can't remember the name) that shows technical things about using lenses and such amongst them a DOF calculator for RED 35mm and 16mm. When using this the dof range for the Red one was less than that for 35mm. How can this be ? I don't remember the Reds sensor being larger than 35mm format ?
He also told me that because of this focus pullers find it harder to work with the red as it is more unforgiving than 35mm film which has a larger DOF.
:ack2:

Is this true ?
Click to expand...

IMHO yes, Red is very critical for focus.
 
Not true. At least not as far as the DOF and most properties of the image go. The sensor / OLPF combination seem to soften more so than film if the lens is a bit out of focus, so in that respect, true I guess. I haven't really seen it be an issue though.

DOF is a function of aperture size in relation to focal length. Sensor/film/image size plays a role as the larger the image area, the greater the FOV/AOV. When you calculate DOF, you do so based on your lens -- focal length and aperture. Sensor or film size does nothing to determine how deep or shallow the DOF is. Larger sensors or film sizes allow for increased shallow DOF effect because longer focal lengths can be used to maintain similar framing to a shorter lens on a smaller format.

The RED One, when shooting 4K, is a bit smaller than Super35 -- it's really close in size to 35mm academy format, just a touch larger than that. RED One at 4.5K with the latest build is just a bit smaller than S35, but very close, I think it's 24.19 wide at 4.5K (4480 pixels actual).

At 4K, the RED One sensor area is 22.12mm wide. S35 is 24.89mm wide for most 3 and 4 perf variations.
 
Last edited:
Yea now I realize sensor size doesn't influence dof. So then, I remember choosing an 85mm lens at 2.8 aperture and if I recall corectly a distance of 4 meters. When I pressed the RED button it showed me a set of values and the total in focus range was let's say half a meter and after I pushed the 35mm the values changed showing the total in focus range was a wee bigger with 3 cm i think. When you are trying to keep the actors face in focus this 3cm difference really makes a difference.
So it the app mistaking because as far as phisics is concerned it's not corect. The in focus area should be bigger on the red and not on 35mm.
 
Mike N said:
So it the app mistaking because as far as phisics is concerned it's not corect. The in focus area should be bigger on the red and not on 35mm.
Click to expand...

The circle of confusion is smaller on a digital 35mm sensor than on 35mm film because the imaging plane is thinner.
 
FWIW, both the iSee4K and the Pocket DIT apps seem to be using the same functions to calculate DOF. But the numbers are incorrect in the way they change with format size. Which leads me to wonder just how they are both calculating DOF. Other than focal length and aperture size, the only other factor to influence DOF is the circle of confusion. CoC is often associated, incorrectly, with format size. It is more a function of the focal depth of the lens system. This focal depth is also different between digital sensors and film because film has a few microns of depth to it. Edit> Cail beat me to that one.

So these apps are either making errors in their calculations. Or they are making assumptions that are incorrect. There is no input on these apps to account for CoC, FFD, etc.. So if all factors are equal in that you're shooting with some 50mm prime lens with a PL mount. It makes no difference whether you mount it on a RED, F35, 35mm film camera, 16mm film camera, etc.. It's still the same lens with all things being equal right up to the imaging surface (film or sensor). So in that respect, all numbers for hyperfocal distance, near and far focus distances should remain the same. And differeing film/sensor sizes are in effect a crop within the coverage area of the lens. Just as when you shoot 2K on the RED One, it is a crop. The optical properties don't change, DOF, hyperfocal, near and far focus, don't change. Even if these apps are calculating with a different CoC for film vs. RED (CMOS), that doesn't explain the difference between 16mm and 35mm, unless they are making more assumptions somewhere.
 
Jeff Kilgroe said:
The optical properties don't change, DOF, hyperfocal, near and far focus, don't change. Even if these apps are calculating with a different CoC for film vs. RED (CMOS), that doesn't explain the difference between 16mm and 35mm, unless they are making more assumptions somewhere.
Click to expand...

From the iSee4K settings page (for custom lens crop factor and CoC)

CoC presets are: 16mm (0.0125), 35mm (0.029) and RED (0.025)

It says these values are "for 1080P display".
 
Yeah, I see that now that I looked at the settings. Interesting how when I specify a custom CoC value, then the calculator no longer works. :-/ Also wondering how they derived separate values for 16mm vs. 35mm film and what a delivery resolution of 1080p has to do with anything.

If we're going to be really pedantic about the whole thing, not all film stocks have the exact same imaging depth and it is also possible for analog film, within a mechanical system, to vary in position by a micron or two as it moves. It seems to me that the app is taking the 1080p resolution factor into account as an additional factor to CoC and that is what is giving us the wildly different numbers.

Also have to say that DOF calculations are pretty complex and open to a lot of interpretation as there are many variables from one optical configuration to another.
 
Last edited:
The calculations done by isee4k use the standard formulas for hyperfocal distance, and DOF detailed here (http://en.wikipedia.org/wiki/Depth_of_field#Derivation_of_the_DOF_formulas). And will deviate when used for macro, and super telephoto.

The CoC's used in iSee4k were determined before RED released their official sensor measurements. I have asked repeatedly for people to provide me with more accurate CoC's, if they dislike the ones currently used, so far no one has done so.

As far as I can tell the custom CoC feature works. Try re-installing, and if that fails to fix it let me know and I'll bug check it.

b.t.w The acceptable circle of confusion is influenced by visual acuity, viewing conditions, and the amount by which the image is enlarged. The original CoC values used were a compromise between my everyday usage of the RED for the small screen, and other users delivery to big screen. I added the "1080p display" disclaimer to help clarify but clearly did nothing of the sort.
 
Hi Jeff,

DOF is subjective, what is acceptable focus to the user. Many have claimed that film in the cinema is less than 1k resolution. Based on that information I would expect to use a different COC with Red than 35mm film.

It's very obvious to me that split's that should work with RED just don't.
YMMV

Stephen

Jeff Kilgroe said:
Not true. At least not as far as the DOF and most properties of the image go. The sensor / OLPF combination seem to soften more so than film if the lens is a bit out of focus, so in that respect, true I guess. I haven't really seen it be an issue though.

DOF is a function of aperture size in relation to focal length. Sensor/film/image size plays a role as the larger the image area, the greater the FOV/AOV. When you calculate DOF, you do so based on your lens -- focal length and aperture. Sensor or film size does nothing to determine how deep or shallow the DOF is. Larger sensors or film sizes allow for increased shallow DOF effect because longer focal lengths can be used to maintain similar framing to a shorter lens on a smaller format.

The RED One, when shooting 4K, is a bit smaller than Super35 -- it's really close in size to 35mm academy format, just a touch larger than that. RED One at 4.5K with the latest build is just a bit smaller than S35, but very close, I think it's 24.19 wide at 4.5K (4480 pixels actual).

At 4K, the RED One sensor area is 22.12mm wide. S35 is 24.89mm wide for most 3 and 4 perf variations.
Click to expand...
 
Expanding on what Cail said, one factor in the difference between film and a digital sensor is that because film is comprised of layers of emulsions, the different colour components do not focus in exactly the same place. This gives film a 'fudge factor'. With a digital sensor the R, G & B photo-sites are all in exactly the same plane, so there is only one point of absolute focus. Any AC who's worked on RED destined for projection will have found this out pretty quickly in dailies.
 
So what|s the solution...? -:)

A prism with mirror distance-defocusator thingy inside the optics to emulate filmplane distance?

/joking

G
 
Jeff is right that DOF is format independent. I have a full set of tables for Panavision lenses and they are depth of field for any given lens, irrespective of whether you shoot Academy 1.33, 1.85, 2.35, etc.

Also, as has been pointed out DOF is subjective. If you do a 1/3 - 2/3 split to keep foreground subject and background subject in focus, the degree of sharpness at the front and back is not the same as at the 1/3 point. But a judgement call was made on these tables as to what is "acceptably" sharp. Ultimately, this is what we all must do in working with any lens in any situation.
 
Also if you wan't to be super critical you have to take into account the "fill Factor" of the chip. Fill factor is the 'dead' space between photo sites on a sensor. Film is covered almost 100% (by surface area) by sensitized particles of varying sizes (grain). Sensors however have very evenly spaced pockets of sensitivity, surrounded by a small area of nonsensitive structure. This changes how things "roll off" in and out of focus.
 
There's a very nice article about working with the RED in a recent American Cinematographer Magazine.

Here's an excerpt from the part about DOF and the RED:


"A more innate issue, which the RED shares with other "full frame" digital cameras, was the surprising lack of DOF, which made focus harder than with 35mm. Subjects popped in and out of focus with no smooth transition.

Because digital sensors have a fixed array of uniforum pixels, the circle of confusion "jumps" from one row of pixels to the next; film, on the other hand, has randomly placed, variably sized microscopic grains, resulting in a smooth gradual focus transition. To mitigate the focus difficulty we shot masters at 2.8 and sometimes closeups at 4.0."
 
I think any focus puller will tell you that you have less forgiveness on a RED than on film.

There are many thinkable solutions to this, but even though DOF is allways the sam, how it behaves differs. Higher stops, can help a bit...
 
You must log in or register to reply here.
Back
Top