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 :)

Rating the RED

B

bobbystone

Well-known member
Joined
Mar 30, 2008
Messages
210
Reaction score
0
Points
0
Can anyone help me with this? I've been trying to establish an ASA rating for the camera so that I can use my light meter. I consistently get a much lower rating than the 320 established by RED. Here's my methadology:

Put the camera up shooting a grey card with white and black patches. The card is evenly lit in tungsten. I set the view to RAW, turn on False color and adjust the aperture till the grey card is green and the white card is between orange and red. To get the corresponding stop on my light meter(after reading the grey card with the spot meter) I have to adjust the asa on the meter to 80. This is 2 stops LOWER than the advertised rating. I'm not bitching about the rating, but I'd like to establish what it truly is.

Comments, thoughts, critiques on my method?
 
bobbystone said:
The card is evenly lit in tungsten. I set the view to RAW, turn on False color and adjust the aperture till the grey card is green and the white card is between orange and red.
Comments, thoughts, critiques on my method?
Click to expand...

Could you try the same test with DAYLIGHT ?
 
I have been confounded by this as well, though having only tried it a couple of times.
In daylight, by my 3 incident meters, and various gaffers' meters, I've gotten ISO 125 or 160. That's by trying to make 18% grey green with false color (I had been setting exposure in Rec 709, however)
Lighting a scene with the meter set at 320 looks fine, but then opening a stop (depending on the scene) from there almost always looks a bit better. So I've taken to just using false color, with a RAW check to make sure the highs are not too high.
The false color tool is very good; however, keeping consistency in contrast
(fill light), scouting, and prelighting would all be made more efficient with accurate metering. It's puzzling to me why this seems elusive.
Maybe it's me...
Cheers,
Harry
 
Daylight tomorrow

Daylight tomorrow

I'll redo the test tomorrow in daylight and post results. But 2 stops is a pretty big discrepancy
 
ISO is totally dependent on what LUT is used to map which code value to mid grey. Also, definition of ISO has an absurdly low % for white at 107% compared to mid grey at 18%, which means it offers very little highlight protection.

That's why the ISO of RED is natively set at 320 so you don't clip the highlights like some video camera, and it's also dependent on the precise curve you use to develop the RAW image with.

Graeme
 
harryclark said:
I have been confounded by this as well, though having only tried it a couple of times.
In daylight, by my 3 incident meters, and various gaffers' meters, I've gotten ISO 125 or 160. That's by trying to make 18% grey green with false color (I had been setting exposure in Rec 709, however)
Lighting a scene with the meter set at 320 looks fine, but then opening a stop (depending on the scene) from there almost always looks a bit better. So I've taken to just using false color, with a RAW check to make sure the highs are not too high.
The false color tool is very good; however, keeping consistency in contrast
(fill light), scouting, and prelighting would all be made more efficient with accurate metering. It's puzzling to me why this seems elusive.
Maybe it's me...
Cheers,
Harry
Click to expand...

I concur.

I just did a big greenscreen job and metered at 160 (cause of my 80C filter).

Same thing as Harry, I found inevitably I would open up a stop from what my meter said, based on reviewing the histogram and false color.
 
Measuring RAW view with the false color and trying to get the grey card in the green zone will not give you any good measuremente for your lightmeter, since the RAW view is a linear 1.0 gamma curve (well, it´s not actually a curve but a line ;D), and you will mostly finish your works with a 2.2 gamma, so it´s with the redspace LUT (or 709) where you want to do that test.
 
Discrepancy? This is a shift in understanding digital - not video. When the Arri D-20 first came out, it too was rated at 320, but by all my testing indicated that it was really 160. The guys at Arri put me on the phone with one of their techs in Germany ( I think we got him out of bed). He clarified the situation. Yes, by following the waveform it would appear to be 160, but by under exposing the grayscale (to around 35 to 37 IRE), the dynamic range was extended or optimized the grayscale , and the "lifting" the image later. This was confirmed by myself after shooting dozens of commercials with the camera.

Remember that the video output is only an interpretation of the RAW image being recorded. Like Graeme said - "ISO is totally dependent on what LUT is used to map which code value to mid grey".

It's kind of like a restaurant that slightly undercooks their take out menu, because they know that people are going to microwave the meal at home - yeilding the perfectly cooked meal.

We know we will need the over-exposure protection later, so we under expose in camera for an extended range later - with no loss in quality. This undercooking is built in by using the 320 ISO rating.


BTW Arri published a slide wheel or disc that showed the relation of ASA to dynamic range. Does anyone have it? This would be helpful to people trying to understand the charecteristics of the RED. I'll dig around to see if I can find mine.
 
I've found that in most cases, metering at 160 will get you a pretty solid exposure in REDspace. Stuff I've metered at 320 is usually underexposed unless it's a high-contrast scene with some large-area highlights I'm trying to save.

That said, I don't use a light meter.
 
bjonsson,
I have a couple of those wheels. It really clarifies how the camera "is what it is", and changing ISO from 100 to 500 merely changes whether you clip whites or crush blacks.
Glad to have other people agreeing with my observation of ISO 160. Maybe I'm not nuts.
MacGregor, so you think I'm doing the right thing by false color in Rec 709?
Harry
 
I set my Red camera @ 400asa and I set my meter at 250. It seems to work out.

I have found the same to be true with Daylight and Tungsten.

Prost,
Rich
 
My approach has been to set up a grey card and light, lens set at T4, until the card is covered with an in-camera zebra set at 44-47 IRE, using RAW view. Using several light meters and adjusting ASA until the meter reads T4 has given me a consistent 200ASA of the camera.

I have only done this test under tungsten, and will redo it under 5000K asap. My assumption is that this will give me 160ASA as others have found.

macgregor said:
...the RAW view is a linear 1.0 gamma curve
Click to expand...

I don't think this is true. If you look at the footage from the test above in true linear in RedCine (ie with the 'Color' button at the top switched off), the grey is a lot darker than it was on the monitor while shooting.

I agree with the idea of the highlight protection that RedSpace / 320ASA gives you, but I've often found this mapping results in noisy shadow areas that are 1 to 2 stops under mid grey. I need to be able to give a DP good advice about the balance between clipped highlights and noisy shadows for a particular situation.

Can Graeme maybe answer the question of what exactly we are seeing when we choose 'view RAW'?
Or better yet tell us what RedCine settings equate to 'view RAW'.
 
Just did a quick test and discovered that the following settings in RedCine give exactly the same histogram:

1. Linear (ie Color button switched off)
2. Gamma 1.0 / 320ASA / 5000K / everything else default

So 'view RAW' in camera equates to 160ASA, but a linear view in RedCine equates to 320ASA.
 
As far as I remember, view raw has no matrix, so no white balance correction or colour correction of any kind, no gain, and a full range REC709 curve to vaguely perceptually linearize the image - because seeing and shooting with a linear light image would be very misleading and not as useful.

Graeme
 
Graeme Nattress said:
As far as I remember, view raw has no matrix, so no white balance correction or colour correction of any kind, no gain, and a full range REC709 curve to vaguely perceptually linearize the image
Click to expand...
So to re-create view RAW in RedCine (for testing or production reasons - there are others here who have asked for this) you would use REC709 gamma, at 160ASA, 5000K, no other adjustments?

Graeme Nattress said:
seeing and shooting with a linear light image would be very misleading and not as useful.
Click to expand...
I agree absolutely. It's useful to know the difference between view RAW and linear light.
 
ISO360 is unity gain. Yes, REC709 Gamma, CameraRGB colourspace + ~5000k for colour temp will get you very close.

Linear light is useful in a post environment to see what, precisely, it is you have, but next to useless to shoot with as it will make you really over-expose.

Graeme
 
Graeme Nattress said:
ISO360 is unity gain. Yes, REC709 Gamma, CameraRGB colourspace + ~5000k for colour temp will get you very close.
Click to expand...
360ASA? Interesting that if you set the camera up at 360ASA / 5000K / REC709 / no matrix or other adjustment and then flick to view RAW, you appear to lower the exposure by about a stop. I'd love to provide specific IRE values of a grey card to illustrate, but I'm not near a camera for a while. Brings to mind -3dB gain on video cameras...
 
I will check here, but 360 should be unity...

Graeme
 
Hi everyone,

I did some test yesterday about rating the RedOne.

RedOne @ 25fps, 1/50, 320 ASA, tungsten:
Grey chart at 50% on a waveform gave me 80 ASA on my Sekonic.

Maybe you do your test at 1/25 so you get one stop more: 160 ASA.

There is still this 1 stop between 160 to 320 ASA that i don't understand.

I'm preparing a movie and during the prep, I only have my sekonic with me and I do need the correct rating of RedOne....

You're telling us that we should do our test with Rec709.. but in that case, we loose our raw data dynamic range, don't we?

It would be cool to have a "raw view" in RedCine in order to know exactly what we have recorded. Something close to that seems to be: gamma custom, RedSpace Gamut... is it correct?

Thank you for helping me.
Julien
 
You must log in or register to reply here.
Back
Top