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 :)

ISO of Red

Graeme, David's scenario is not dependent upon "scene":
If RED is, as has been claimed, to be 320 iso, then that is what
I'd expect to correspond to PLACED values, no matter what the scene:
Post aside- if I place a (320 ISO setting) sunlit face on Zone VII
(two stops above mid gray), allowing all other values to fall as
they will, and the face falls on Z VI (one stop above mid Gray)
then the ISO is off. And will be no matter what the scene
(low light tungsten response aberrations aside). If, however,
adjusting the ISO to reflect accurate placement via my ("very accurate")
spot meter meant reduced DR, then I'd be tempted to establish
an ISO based upon max DR for an average scene (let's say 320 iso)
and I'd adjust my meter's ISO until its values corresponded perfectly
with RED's output: So that ( spot meter at 160 ISO and RED on 320 ISO)
the sunlit face now fell where I placed it (with max DR) Is this accurate?
 
The definition of ISO is only for a) a corrected image, not RAW data, and b) for a certain percentage white clip (109% I think).

a) means we can't really talk about ISO with respect to the RAW data, only the developed data.

b) means using that 109% definition leads to ugly clipped highlights.

The whole ISO scheme needs completely revising with respect to how cameras now work. The ISO320 of the RED is to achieve a reasonable exposure on a scene with a reasonable dynamic range so that it looks reasonable after reasonable grading. And I find that reasonable.

Graeme
 
The whole ISO scheme needs completely revising with respect to how cameras now work. The ISO320 of the RED is to achieve a reasonable exposure on a scene with a reasonable dynamic range so that it looks reasonable after reasonable grading. And I find that reasonable.

Graeme[/QUOTE]

I was hoping someone would say this sooner or later. I'm sure many of us have been thinking the same thing, but, like me, without real solutions (outside our own quirky nomenclatures).
 
If ISO definitions are not for RAW data, then how do we describe the sensitivity of the sensor as a RAW data capturing device in a common language we can all understand?

I personally feel 320 is an aggressive ISO rating for R1 especially in a scene where the majority of tones are below key (a dark night interior), considering where this actually places normal exposure on an IRE scale (in REC709 B20).

Attached is a graph of relative exposure vs IRE levels as measured by lighting greyscales and recording the reflected spot meter readings. Spot meter assist and false color were used to measure the IRE in the viewfinder. Meter set at 250, camera metadata ISO at 320. From here, you can figure out what any other meter rating would yield as far as IRE level in any color space. All curve, brightness, contrast, and color settings were cleared before shooting this test. Tungsten light, no filter, white balance at 3200K. Minolta Spot F meter used. I am sure there are far more accurate ways of measuring this but to me this test is a decent look at where tonal values would fall in practical shooting situations with regards to light meter readings.

Two things struck me upon viewing the test results:
1) The large difference in REDSPACE values in mids and upper mids versus REC709 and RAW. REDSPACE really lifts these values.

2) Where a normal exposure actually fell on the IRE scale (far below 18%).

http://dl.dropbox.com/u/904379/Build 20 Tests Round 2.pdf
 
Graeme Nattress said:
The definition of ISO is only for a) a corrected image, not RAW data
Click to expand...

Agreed. In order to apply the idea of ISO to raw data, it's necessary to define what is meant by "ISO". The simplest definition is to state only the number of stops between mid gray and clipping in the raw data when given the same raw conversion. That ignores important issues like how much noise is in the shadows, etc., but at least it's something that is otherwise similar among raw cameras.

So I could say "When developed for 2.5 stops headroom above middle gray, RED ONE is X ISO. When developed for 4.5 stops, the ISO is X * 4".
 
Daniel Browning said:
Agreed. In order to apply the idea of ISO to raw data, it's necessary to define what is meant by "ISO". The simplest definition is to state only the number of stops between mid gray and clipping in the raw data when given the same raw conversion.
Click to expand...

As ISO is meant to give an idea of film sensitivity, in case of CMOS sensors shouldn't the corresponding number reveal when the highlights clip with respect to some nominal level.

Mid grays does not make that much of sense in case of RAW data, so the definition should neither rely on any raw conversion.
 
Lauri Kettunen said:
As ISO is meant to give an idea of film sensitivity, in case of CMOS sensors shouldn't the corresponding number reveal when the highlights clip with respect to some nominal level.
Click to expand...

Yes, and that "nominal level" is what I was talking about in my post.

Lauri Kettunen said:
Mid grays does not make that much of sense in case of RAW data, so the definition should neither rely on any raw conversion.
Click to expand...

People still need a way to tie by to it visually, though, so talking about the implications of raw conversion is necessary, but it should be defined. The simplest definition, I think, is the one I gave.

If the raw conversion puts middle gray at 2.5 stops below clipping, than the ISO is one thing (let's say ISO 125 just for illustration). If the raw conversion puts middle gray at 4.5 stops below clipping, then the ISO is totally different: ISO 500 (for illustration).

But in either case the exposure is exactly the same. All that changed in the raw conversion and therefore the ISO. Since raw conversions can vary by so many ways, the number of stops between a certain raw value and clipping is a good way to level the playing field. Essentially it works like this:
* What raw value corresponds to middle gray in the raw conversion you, personally, want to use as the ISO standard?
* How many stops are there between that raw value and raw clipping?
* Use that number to calibrate ISO.

So if you rate with 2.5 stops headroom, you will get low ISO. If you rate with 4.5 stops headroom, you will get high ISO.

So instead of discussions like "RED is ISO 160", "no, it's 320", you get "I use RED with X stops headroom, so I rate it at ISO 160. If you need X+1 headroom, you will rate it at ISO 320".
 
As Graeme has stated - 320 is a "reasonable starting point".

It builds in highlight protection ('cause you can't get that back) in favor of underexposure - which you can arguably manage.

Now that I understand a little bit about the sensor's sensitivity I adjust according to that;

If I am interested in protecting highlights in a high contrast setting (i.e. outdoors) I'll rate it at 500 ISO. If I am concerned with noise in shadow detail, I'll rate it at 200, or 160 if I can afford the lights!

This is the complete opposite of film ISO sensitivity - but that is how to manage RED's sensor. What we are accustomed to with respect to ISO and film is irrelevant with RED.
 
Now I'm confused ...

Now I'm confused ...

Okay, so now I think that I am completely confused, or maybe I'm just over thinking it. Can someone straighten me out ...

During Shawn and I's testing I also shot my Sekonic Exposure chart. After processing it it gives me values from 5 stops over to 6 stops under (That was the range of the lens and chart.) I also shot reference frames of what False Color was telling me. If you want to check out the 4k or 2k stills, or the False Color pictures you can download the file I've linked to below. I have attached the final combined chart to this post for quick viewing, but it is also in the link below.

Sample Frames and Reference Pictures: (83MB)
http://www.ryanewalters.com/downloads/ExposureTest.zip

After evaluating the results, and reading everyone responses this is what I understand so far... (Please correct me where needed.)

*** My definitions:
Rating: The process of using my meter to determine the correct exposure so that the values fall where I expect them to fall. 18% is middle grey, Caucasian skin is one stop over, etc.
***

- Rating the camera at ISO 320, and leaving the camera set to ISO 320 provides an "automatic" built in highlight protection into the image as the resulting image ends up being underexposed by one to one and a third stops.
- Rating the camera at ISO 320, and leaving the camera set to ISO 320 does not place middle grey or Caucasian skin tones where I would expect, and if I want them to fall where I expect then I need to rate the camera one to one and a third stop lower. (Somewhere around ISO 160)
- Rating the camera at ISO 320, and leaving the camera set to ISO 320 biases the exposure toward the highlights, so if there is information that I want to pull up in the shadows I could run into problems as noise could be an issue depending on my taste.
- Rating the camera at ISO 500, and leaving the camera set to ISO 320 biases the exposure even more toward the highlights and is preferred for outside, so as to protect those highlights. But it will push the image even further towards the blacks. So if I need to dig anything out of the shadows I could run into problems.
- Rating the camera at ISO 160, and leaving the camera set to ISO 320 biases the exposure toward the shadows but means that I will have to pay attention to my highlights as they can clip sooner.
- If I Rate the camera at ISO 320, leave the camera at ISO 320, and I am lighting a low contrast scene, I should be prepared to keep an eye on my shadows, as if something falls in shadow then I run the risk of introducing noise into the image if I try and bring up the exposure later in post. It would be better in this instance to rate the camera at ISO 160 and clean up my shadows by adding some fill to them.
- The camera can handle about 3 stops over middle grey and 5 stops under middle grey (Depending on your own tastes / tolerance of noise.)

Is that correct?

So when I am metering a scene, I should set the meter to the corresponding situation- something like ISO 160 for interior, and ISO 500 for exterior - however keep the camera set to ISO 320.

Additionally, when I hear / read "headroom" I am thinking that middle grey is being moved down accordingly. So with 4.5 stops of headroom, that means middle grey has moved down accordingly and the shadows have just be crunched- correct?

***Items of Note***
These are some things I have noticed on build 21 with False Color, and they worry me ... (You can see this in the reference stills in the download above.)
- In all of the color spaces the false color values fluctuate by 1/6 to 1/3 of a stop and they are not consistent fluctuations.
- In RAW mode sometimes the Straw value does not come up.
- 18% Green does NOT line up between all three views like has been stated before. (It did back in a previous build - build 20 I think?)

These issues are especially troublesome since the difference in-between yellow, orange and red are 1/3 of a stop, and if False Color is not reading correctly- especially in RAW, then it is useless as a precise exposure tool.
 

Attachments

  • ISO320-Comparison.jpg
    ISO320-Comparison.jpg
    124 KB · Views: 0
Ryan,

I agree with all of your points and have discovered the same things in my testing and real world on set experience of shooting dozens of projects on RED ONE (apart from B21 false color- still not tested). I've come to the same conclusions as well, independent of reading your post. In my above post I have a link to a graph that is similar to the data you have gathered.
 
Thanks Eric for your confirmation- I'm glad I'm not crazy ... at least about this anyway. ;)

I downloaded your chart- thanks for that, very helpful. And it backs up what I had thought too.
 
Hey Ryan
I hope it wasn't my comment that raised your doubts!

The only thing that I would clarify is that in your examples, you mention "rating the camera at" and "setting the camera to".

I don't use a light meter with RED and the ratings I suggest "rating the camera at" and "setting the camera to" are one in the same.

In other words, when I shoot outdoors in full sun, I'll set the camera's ISO to 500. This "forces" me in to a little more under exposure than even the 320 rating. All of this said - I am fully conscious of what I am doing and will keep checking the RAW to make sure I am not obliterating shadow detail.

This is not an original thought - by the way. I have Macgregor to thank for it from a post that probably goes back 18 months now!

Bye for now.
 
Dylan,

Thanks for the clarification. As for setting the camera to ISO 500- do you do the reverse for the low contrast scenes and set it to ISO 160, or 200? Then when the footage is processed is it all processed at ISO 320?
 
Ryan E. Walters said:
Dylan,

Thanks for the clarification. As for setting the camera to ISO 500- do you do the reverse for the low contrast scenes and set it to ISO 160, or 200? Then when the footage is processed is it all processed at ISO 320?
Click to expand...

Ideally everything I do will respond to the ISO setting. So if I am at 160 - I'll light the image to look good at that setting. This ensures that I have the room to grade things with some flexibility. If we need to push things up a little the noise floor has been "exposed up" a little to accommodate that.

Reality is - it isn't always practical to light a night scene to 160 ISO - but theoretically - if you can do it - you are favoring more info in the shadows. Which is generally what you want in night scenes. Highlights are less of an issue - but need to be checked (i.e. when setting practicals) knowing that you are rating the camera to "expose more".

Short answer is - everything would get processed with the in camera rating of 160 ISO because that is what I have lit for and set the stop for.
 
David Mullen ASC said:
My only question is why the False Color guidelines don't match my meter at 320 ASA. Are the False Color guidelines designed to only indicate RAW levels even when viewing in Rec 709?
Click to expand...
:hurray:
I would like to understand this better too.
 
Hi, in my test ( two days with a good DOP, gray scale chart, gray 18%, actors)

the Red sensibility is really 125 ASA in Raw mode exposition.
 
Old film references are often the cause of a bad decision.

Graeme's point should be re-read by all.

Shooting ISO 160 is not a good idea except in low contrast situations.

Shooting ISO 320 will give you the best starting place for highlight protection and low noise for a RED ONE under most conditions.

ISO Rating is a good term. "Native ISO" is a very dangerous term.

From Asia...

Jim
 
You must log in or register to reply here.
Back
Top