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 :)

Is ISO a Sensor adjustment or Image Processor adjustment?

Status
Not open for further replies.
So let's close the thread as this has been answered about 27 different ways.

Why am I even typing this and wasting more of my energy on something as retarded as symantics.

Conversations are actually allowed to progress throughout the thread, it does actually happend sometimes.

David
 
David Battistella said:
Another way of looking at it is if you are familiar with GAIN settings on video cameras.

IE:
ISO 250 -3db GAIN
ISO 320 -0db GAIN
ISO 400 +3db GAIN
ISO 500 +6db GAIN

David
Click to expand...

+/- 3 dB gain corresponds to half a stop increment/decrement (6 dB being 1 stop). +3 dB doesn't exactly correspond to ISO 400.

All this is of course not really relevant when recording with the Red
 
Last edited:
Hey folks,

You know. I actually understand all of this stuff, realy really well.

I make this analogy for people who do not understand film and come from a purely video background as a way of explaining that the ISO in RED is more like a GAIN feature on a video camera and IT DOES NOT MAKE THE CAMERA'S SENSOR ANY MORE SENSITIVE THE WAY A FILM STOCK RATED AT 100 ISO IS LESS SENSITIVE THAN ONE RATED AT 400ISO.

This is getting really exhausting.

It's and example, NOT A TRUTH.

Fuck.


David



ISO in RED
 
Stephen Williams said:
Sorry but that was never really clear, from the information you posted.
Click to expand...

You know. If this isn't clear, then maybe you don't really fully understand the R1 System all that well.

I'm not being baited into this bullshit any further.

David

David Battistella said:
Think of it (ISO) as a post tool rather that a aquisition tool.

If you shot a dark scene wide open on a t1.8 lens, you could set the camera at 320 or 4000 ISO but niether of those settings would affect how much light is hitting the sensor plane.

Do you get what I mean?

It just means pushing the r3d in post will make for a better image than previously. The m-x sensor is probably faster (more light sensitive) but it is a combination of things (including new post tools) that will allow you to extract better image quality from the r3d.

David


David
Click to expand...
 
David Battistella said:
You know. If this isn't clear, then maybe you don't really fully understand the R1 System all that well.

I'm not being baited into this bullshit any further.

David
Click to expand...

Hi David,

Not sure if the first or second paragraph made me laugh more.

Best

Stephen
 
How is this advancing the discussion and what is it that you are aiming to prove?

It's a very particular sickness.

david
 
David Battistella said:
How is this advancing the discussion and what is it that you are aiming to prove?

It's a very particular sickness.

david
Click to expand...

Hi David,

I am not aiming to prove anything, other than I have a sense of humor.
I prefer to read accurate information on any forum that I take part in.

Sounds like FUD to me.

Best,

Stephen
 
I came upon this site in a state of confusion after googling for the difference between 'gain' and 'ISO' and at the time of posting this comment find myself in good company, my state of confusion has been bolstered.


I am left with my original and undesired conclusion that the REAL difference between these two units of measure is that ISO increases the GAIN in the till.


Fried reds for breakfast anyone?


May 2010 be good for you all.
 
Joel Kaye said:
A higher ISO just "tricks" people into lower exposure therefore protecting highlights. A raw viewing mode with real overexposure indicators would be best. Knowing the ISO is helpful when matching light meters, other types of cameras etc.
Click to expand...

Ok these thread feels little tense right now so dont jump me know=). Iso is metadata that only manipilate the gamma curve and therefore dont touch the highlights right? If this is correct i think I understand iso./Joel
 
ISO rating - theory

ISO rating - theory

So let's summarize (and yes there some nuances across firmware builds omitted, so this is the basics)

Adjusting the "ISO rating" parameter on a RED camera is not the same as adjusting "ISO" on a DSLR as the latter is directly affecting your recorded data.


On a RED camera the "ISO rating" is used by the RGB monitoring path to "gain up" the digitized sensor RAW data with a gamma based Tonal Response Curve which lets you visualize the Bayer pattern sensor RAW data. In this sense, the ISO rating is "just" metadata - it does not alter the recorded RAW data.

EXCEPT THAT -

By interacting as an operator to what you see in the RGB domain monitoring path, you can use the ISO rating to help you bias exposure i.e. increase your protection for highlights or for shadows. It is this reaction to ISO rating by the camera operator that can lead to an alteration of the recorded RAW value.


In post, the ISO rating that is recorded with the RAW data becomes the default setting used to "gain up" the RAW video during the RGB conversion. As the RGB data is our deliverable, it's common to use use a different ISO value and a user-adjustable Tonal Response Curve to get the exact "LOOK" that we desire.


David Mullen explains this concept very clearly earlier in this thread.

David Mullen ASC said:
Metadata is just a set of instructions or settings attached to the R3D file to be applied to the RAW signal when it is time to process it in post. The metadata settings also are applied to the video-out that goes to the eyepiece and monitor.

Of course, since they manipulate the image seen on the set, the metadata settings will affect how you interpret and adjust the image. For example, if you set the camera to 1000 ASA and 3200K color balance, then a scene lit and exposed at 1000 ASA in 3200K lighting will look correctly exposed and neutral in color (though it may look noisier). The actual R3D frame would be rather dim and orangey since the camera is closer to 320 ASA and 5000K, but the metadata settings are basically telling an image processor (either in camera for the monitor out, or in post) to boost the overall underexposed signal and boost the blue channel to compensate. You can choose to ignore the metadata settings if you want to work with something closer to the way it really looks in the original recording.

However, shutter speed is not metadata, it actually affects the recording, just like the f-stop of the lens, the frame rate, and what colors you use in lighting, how you focused the shot, etc.
Click to expand...
 
I prefer to think of it as:

During recording ISO setting amplifies/attenuates the MONITORING PATH and NOT the recorded data. Those monitoring path adjustments are reorded along side the untouched sensor data as metadata.

During post, that same ISO metadata is applied and amplifies/attenuates the DISPLAY/RENDERING output. Post tools, however, allow you to override the existing recorded metadata, and apply your own.

ISO thus never changes anything recorded DIRECTLY. It does change the monitor. That may mean that YOU change the exposure via aperature, shutter-speed, etc...

-sc
 
Therein lies the problem with this whole ISO metadata feature. I think it should be left off (in-camera) by default, but can be enabled by changing a camera setting. Considering how many people are getting confused by this (since it doesn't act like an analog gain), I think people should just expose normally and then adjust the image in post by eye. The metadata may help some people get started, but I think it's simply causing too much trouble for a lot of people.
 
Chuck Z said:
Therein lies the problem with this whole ISO metadata feature. I think it should be left off (in-camera) by default, but can be enabled by changing a camera setting. Considering how many people are getting confused by this (since it doesn't act like an analog gain), I think people should just expose normally and then adjust the image in post by eye. The metadata may help some people get started, but I think it's simply causing too much trouble for a lot of people.
Click to expand...

I think that means you are suggesting that the monitor shows only the raw output? I tried that for a while, and while I was comfortable with the knowledge that I would make it look different in post, my clients were not nearly so keen on the whole idea. And how do you expose "normally" for all the various abnormal scenes we all want to shoot? You can light everything strictly by meter, but without monitoring that has some relation to your eventual post path, you don't have a very good way of judging the results. You can't look "through the lens" as you could on an Arriflex.

The real benefits of the ISO-based processing path for monitoring are that you get a pretty good idea of what the shot will allow you to do in post, considering your output target, and that you get to talk with DPs and gaffers in a common language that has been developing for their entire lives.
 
ISO rating

ISO rating

Chuck Z said:
Therein lies the problem with this whole ISO metadata feature. I think it should be left off (in-camera) by default, but can be enabled by changing a camera setting. Considering how many people are getting confused by this (since it doesn't act like an analog gain), I think people should just expose normally and then adjust the image in post by eye. The metadata may help some people get started, but I think it's simply causing too much trouble for a lot of people.
Click to expand...

Well the appropriate solution to not understanding something is to leave it at the default setting - which is ISO 320.
 
Status
Not open for further replies.
Back
Top