Is this the right way to evaluate a camera's latitude?

marvinhello

marvinhello

Well-known member
Hello fellow cinematographers, I recently attended a local workshop and the DP demonstrated a camera latitude evaluation process (with a RED ONE MX).

Here are the screen grabs from the test video he made, one with 709 gamma applied on the r3d and one with RedLogFilm gamma applied.
https://dl.dropboxusercontent.com/u/12279480/RED_TEST_709.zip

https://dl.dropboxusercontent.com/u/12279480/RED_TEST_RAW_LOG.zip

Here is what he posted after our discussion at the studio:

If you remember we measured reflectance values of properly exposed gray chart. If properly exposed neutral gray field (18%) would be in Zone V. Then we measured the reflectance value of white field which was about 1 and half stop brighter that neutral gray – this would put white field into Zone VI-VII. The black field had reflectance value two stop below neutral grey, it means it was in Zone III. The dynamic range of that gray card is than 3 and half stops.

As we know white field (in proper exposure falls into Zone VI-VII. When this field is overexposed by three stops it falls into Zone IX-X. Black field on properly exposed gray chart was in Zone III. If this field is underexposed by 5 stops then it falls into Zone -I. (minus 1). Difference between darkest and brightest value camera is able to register without loosing detail determines camera latitude. In our case it’s difference between -I and IX-X which is approximately 10 and half sops. Not 8.

This is the correct evaluation procedure. You would come to similar conclusion in evaluating in 709 REC, which would be approximately 8 stops, not 5.

One more time: Latitude of the camera is determined by difference between brightest and darkest area camera is able to record just before loosing details. It is not difference between exposures on gray chart. Latitude of Camera is not determined by how much over or underexposure is camera able to handle. This is very simplistic approach. If you had scene with dynamic range of 10 stops, you would have to be dead on with exposure in able to maintain details in highlights and shadows. If your dynamic range of scene was only 5 stops, than you can over or under expose at least by one stop without loosing detail.
Click to expand...

The way he put it is a bit confusing, could anyone take a look the files and offer some clarification please? are there any flaws about this test and the DP's evaluation?

Thanks!
 
Liam Hall said:
Looks to me like someone is confused between dynamic range and latitude.
Click to expand...

I agree. Back in the olden days of Film film, while I was using still film... I don't think Motion Picture Film film users used different terms... 'Latitude' meant essentially, 'how bad an exposure can I have, and still get a printable result'... ok... perhaps more 'technical'... 'This film has 1, 2, 3, or 'no' stops of overexposure'... For example 'slide film', aka 'reversal' had very little latitude in regard to over exposure... highlights over exposed were 'gone'... one could however 'under' expose slide film and get more saturated colors...

Film film latitude could also be for underexposure... as in 'how many stops under can I go, before development time degrades the image to just golfball sized grain'...

So, a film like Panatomic X, with ASA 64, was considered to be a B&W film with 'little' latitude, perhaps 1 stop, whereas Tri-X had possible 3 stops of 'overexposure' latitude...


What overexposure latitude meant is that if I exposed 1, 2, 3 stops over, and then 'printed down', I was still able to retain detail in the highlight values. Usually this also required 'dodging' the shadow areas unless one also compensated for the over exposure by 'under developing' the negative... or printing on a 'softer paper'... etc.

In Digital cameras Dynamic range has been a major talking point. And to find that range, the procedure listed in the original post is one such method.

But strictly speaking, 'latitude' would be something like this:

If I over expose a shot by 2 stops, can I bring the over exposed highlights to within the display's range, and still retain 'texture/detail'. This is typically done with some sort of manipulation of the 'gamma' curve, or the like.

The answer for a camera that has an 8 stop dynamic range may be 'no'... for a camera that has a 12+ stop range, the answer may be 'yes'... there may be other issues such as 'color shifts' in the high values that may lessen the ability to compensate, but the idea is that one could possible correct in some sense for the over exposure.
 
marvinhello said:
is it correct?
Click to expand...

As I have noted, the person giving this presentation, in my mind has confused 'dynamic range' with what was used most often for 'latitude' in Film film days. There was a usage such as 'give me a film that has a long (or wide) latitude', which would suggest 'dynamic' range, but it was most often with 'here, this film has 2 or 3 stops of latitude, it should be good for your needs'.

Further, since other than slide film, all Film film was destined for 'printing' to be viewed, the negative had to be developed with a print range of 7-8 stops, ranging from 'blank white' of the print paper, to 'fully devloped' silver in the dark regions. For paper prints, this was done by either development time changes in developing the negative, or print paper 'contrast range', which would now be the 'dynamic range' of print paper.

Paper and film was often designated as 'high contrast', meaning 'narrow' dynamic range, or 'low contrast', meaning wider dynamic range.
 
Holy Hell, the Zone System… It still gives me a headache every time I see it. To me, it always seemed like one of those things that someone just over-complicated for the hell of it. I get the idea(I think we all do), but when it's written out, it's like the equivalent of someone who wrote a 100 page book explaining how to tie your shoes(an act that takes about 5 seconds and you don't even think about it).
 
Latitude is how much you can push and pull an exposure during post and still get a good image.

For example:

If I want to retain highlights in an image I might underexpose the mid-tone by a couple of stops, then push them back in the grading suite. Conversely, if I want more shadow detail I might lift the exposure in camera a couple of stops and then pull that back down during grading. This is called 'pushing and pulling'. Different digital cameras and film stocks react differently to pushing and pulling. So you need to test. On one camera you might be fine pushing the mid-tone around but find you get too much noise in the shadows or your highlights start falling apart. It's not as simple as saying this camera has an 'X'stop latitude. Test, test, and test...
 
Not a great example, but it should give you the idea.

This was underexposed to retain highlight detail in the sun and flares.

RWUnder.jpgRWCorrect1.jpg
 
Last edited:
marvinhello said:
could you elaborate? what's wrong with his evaluation? thanks
Click to expand...

There's nothing wrong with the evaluation, he's just using an incorrect term. Latitude is the amount you can under/over expose and still get a good image. He's using it to describe the range between the darkest and lightest points a camera can read. The correct term for that is dynamic range.
 
You must log in or register to reply here.
Back
Top