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 :)

Problem... solution.

It is very important to understand the tech you use. If you have a idea of how the sensor and the camera works, you can avoid the biggest problems. The idea of "protect the highlight" i daylight shooting, and "protect the shadows" in tungsten should be in the REDONE shooting manual at page 1.

The idea that the sensor don't have a fixed ISO, rather then using the bandwidth - dynamic range, is very important IMHO.
 
We are one guilty party of making noisy images in the beginning when shooting indoors. We're getting good results lately though.
On interviews for example we only use a blue gel on the backlight and background light. This is usually enough to feed the blue channel, while still keeping the same amount of light from our key and fill light, making it still possible to bounce these or use a softbox. This way it is also possible to get a lot more room to play with the colors.
 
One thing that comes to my mind when talking about noise in the blue channel: Trying to WB the picture by setting Kelvin to 3000K in RA or RC leads to more noise than processing the footage with the antive 5000K and WB the pictures later in post with a dedicated CC tool.

Hans
 
Hans von Sonntag said:
One thing that comes to my mind when talking about noise in the blue channel: Trying to WB the picture by setting Kelvin to 3000K in RA or RC leads to more noise than processing the footage with the antive 5000K and WB the pictures later in post with a dedicated CC tool.

Hans
Click to expand...

This is exactly right from what I've seen. And yet I would think it shouldn't be so. Gregor Hagey's test ages ago come to mind.

http://www.reduser.net/forum/showthread.php?t=18042

Frankly, this should be in the manual! Everybody's advice, 5600K light sources, ETTR (low contrast situations), 80D filters (half ritalin!), CTB Gels, all work one way or another to cram more blue light down the pipeline. But Post is the lost stop (pun intended) for noise.

I've had to light a fair bit of studio greenscreen for 2K 120 frame tungsten with lowish budgets. The last job was for ML-Baseball. If noise is evident, always process cooler then the light source (4.5-5K) and CC later. It just seems to work. Maybe Graeme can explain! It just seems to introduce less noise for any given blue light level. Perhaps it has a subtle curve/+desat/soften? effect as well. Perhaps its not completely "neutral" either. Hopefully somebody technical should look at the numbers quantitatively and shout-out one day.

As for the ML-baseball shoot, the footage had to mix with period footage that probably had a 3 stop range at best and was grainy as hell and B+W so blue noise was a non-issue. It's nice to deliver maximum quality to their post department that were RED virgins though.
 
I have had no problems with noise and tungsten light as long as there is sufficient exposure. I come from a film background. Meter with an incident meter and make sure there is enough light. That has included night exteriors with no noise problems. Daylight sources are definitely better for the Red sensor, but on small indie budgets that doesn't often happen.
 
Would it make more sense for a cinema camera to be natively balanced for tungsten? Usually when shooting with daylight we have much more light and can afford the light loss of an 85 filter, rather than needing an 80A under tungsten conditions.
 
Of course, if you have control of a situation you can expose correctly, but I'm still wondering how many people are getting vertical lines in their grain when shooting at ASA 500 and below... (even in 5600 lighting.) You can crush it out under many circumstances, but in Doc or Reality work -- not always the case. I don't mind grain if I have to push the hell out of a shot (it's expected), but I cannot use footage with vertical lines that look like scratches on film. Very weird looking. As many places as I posted this, I never get an answer as to if this is a problem all the sensors have/ if future firmware will fix it/ if future chips will be less effected.

Jim?
 
JanneJansson said:
The idea that the sensor don't have a fixed ISO, rather then using the bandwidth - dynamic range, is very important IMHO.
Click to expand...

Could you elaborate on this a little, please? What does this mean, practically? I don't quite follow. :embarassed:

-Tim
 
Evin Grant said:
Don't be afraid to push the camera just understand what it wants. You can get a very nicely textured image by just adding a little blue back in and evening out the channels. It's when the blue gets severely under exposed that the noise punches through.

Also Lightbulbs unlimited has fantastic 5500K bulbs that screw into normal fixtures. That and a few Kino flos and you can light an entire interior at 5500K and no one will be the wiser (Except David Mullen).
Click to expand...

Forgive a somewhat novice question, I'm just checking my understanding. If you're looking for a very warm tone, it's okay if the blues are rather low, right? The problem arises if you try to raise those underexposed blues to make the image more neutral, because the information that used to be all close together at the bottom end of the blue channel gets spread out unnaturally, resulting in noise. Is that correct?
 
Brent J. Craig said:
Would it make more sense for a cinema camera to be natively balanced for tungsten? Usually when shooting with daylight we have much more light and can afford the light loss of an 85 filter, rather than needing an 80A under tungsten conditions.
Click to expand...

It has some issues. Tungsten balance infers dB gain to blue channel off the bat. Most B4 systems do this using analogue gain I believe. No real advantage vs post colour correct. Its true, of course, regarding your second point of course, except when shooting massive night exteriors using HMI's to light buildings/crowds/runways. Shooting tungsten, you have to correct the lens or the lights with resulting stop loss along the way. Ultimately it means more generators and crew when it gets really really big. So now you have 6-12 dB analogue blue channel gain and you have lost 1-2 stops of light from the sources. Or you can use tungsten lights which are less efficient candle/watt....more gennies....more crew etc.
 
James Derek Holloway said:
It has some issues. Tungsten balance infers dB gain to blue channel off the bat.
Click to expand...

Only if the sensor is already daylight-native. If you were designing your own sensor from a blank sheet of paper, couldn't you design it to be tungsten balanced? I think one of the other digital cinema companies did this - Dalsa maybe?
 
Tim Nolte said:
Forgive a somewhat novice question, I'm just checking my understanding. If you're looking for a very warm tone, it's okay if the blues are rather low, right? The problem arises if you try to raise those underexposed blues to make the image more neutral
Click to expand...

Yes, the sensor is naturally balanced to around 5000K, so in 3200K light, there is insufficient blue if you are trying to get a neutral image with no noise in the blue channel. However, if you are OK with the 3200K light being rendered as warmer than neutral (let's say, by setting your balance to 4000K), then you are pushing the blue channel less heavily to make it look neutral.

As for why they don't make the camera tungsten balanced, all silicon-based digital sensors are less sensitive to blue wavelengths; the only issue is whether you can get away with biasing them towards tungsten-balance with minimal noise in the blue channel.

I was testing the Genesis camera, which is set-up to make 3200K the base color temp, like a lot of Sony cameras -- and it is fairly clean at 3200K in terms of noise -- but when I switched the camera electronically to higher color temp settings like 4300K and then 5600K, rather than use filters... guess what? It got even less noisy.

So basically reducing the noise of the system allows you to get away with a 3200K balance for a camera with fewer visible noise artifacts, but no matter what, you are still dealing with the fact that sensors prefer or need more blue wavelengths.
 
It's the way silicon works. All silicon sensors are relatively insensitive to blue light, and tungsten is blue deficient. Two ways you can change this - analogue gain up on the blue, or filtration to reduce the red and green to match the blue's native level. Both are compromises, either premature clipping on blue, or reduced camera sensitivity.

Graeme
 
High hope for Black Silicon in the future. It could really throw a spanner in the works because it has a pretty flat response curve across visible spectrum. It has other issues though and may need heavy IR filtration. But its almost 100x more sensitive! 6+ stops has advantages and complications.

It could make for an amazing Low light Epic brain for nocturnal wildlife shooting...or military!
You could light night exteriors with LEDs and Diving shots would be hugely improved.
There is very little Red spectum light (or any light) at depth.
 
From the charts I've seen it still looks blue deficient compared to red and green. Guess we'll have to wait and see until someone makes a sensor for how it performs.

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