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.

Jannard

Jannard

Red Leader
Joined
Dec 28, 2006
Messages
8,248
Reaction score
7
Points
0
Problem: Noisy footage in tungsten.

Reason: Sensors are weakest in the blue channel. There is very little blue light in tungsten and incandescent lighting. Potential trouble spoken here.

Also, many are setting exposure under tungsten lighting using their monitors set to 5600K. This gives a false brightness that often results in setting exposure too low. Under-exposure in a blue staved sensor = noise.

Solutions.

1. Don't under-expose. Make sure to check exposure with tungsten lighting by some other means than a 5600K monitor. Or set ISO 250 (which will force more exposure) rather than ISO 320 or 500. Either open up a stop from what you see with a 5600K monitor (test this 1st), use a hand held exposure meter or some other method. Share your other method here.
2. Use daylight balanced HMI, LED or CFL lights instead of tungsten or incandescent lights.
3. Add blue gels or filters somewhere in the chain (which has the consequence of lowering light levels).
4. Test. Find a method that works for you by testing before you shoot something important.

We have seen some pretty nasty tungsten lit footage. And we have also seen some incredibly great low light tungsten footage (see the new reel). The difference can be as little as one stop under-exposure. There is not much forgiveness in a blue starved light situation. A little testing can do wonders.

Jim
 
Hey Jim,

Any indication of greater forgiveness in noise under tungsten light with the new Mysterium X chip? I would like to think/hope so. If so, is it marginal, substantial or oh my god good?
 
Honestly, I have had no issues shooting as I normally would with f900, Varicam or film. In fact the Red is noticeably less noisy than varicam (red 640 asa, varicam 0 db gain). I'll post examples when the shows I've shot have cleared to do so.

I think you are being very diplomatic Mr. Jannard, but I think the jist of it is, operator error. Underexposed footage is just that, don't under exposed. Duh.

Cheers,

Nick
 
Hey Mike- I just gonna ask that 8)
I am wondering too- is there any hope for a different sort of debayer or some other solution?
(if that is even possible- but Graeme can do anything, right?)

not a big deal though- i,ve been shooting with as much daylight balanced lights as possible.
(but in a pinch- it's good to know how to make tungsten work)
 
Jannard said:
Problem: Noisy footage in tungsten.

Reason: Sensors are weakest in the blue channel. There is very little blue light in tungsten and incandescent lighting. Potential trouble spoken here.

Also, many are setting exposure under tungsten lighting using their monitors set to 5600K. This gives a false brightness that often results in setting exposure too low. Under-exposure in a blue staved sensor = noise.

Solutions.

1. Don't under-expose. Make sure to check exposure with tungsten lighting by some other means than a 5600K monitor. Or set ISO 250 (which will force more exposure) rather than ISO 320 or 500. Either open up a stop from what you see with a 5600K monitor (test this 1st), use a hand held exposure meter or some other method. Share your other method here.
2. Use HMI, LED or CFL lights instead of tungsten or incandescent lights.
3. Add blue gels or filters somewhere in the chain (which has the consequence of lowering light levels).
4. Test. Find a method that works for you by testing before you shoot something important.

We have seen some pretty nasty tungsten lit footage. And we have also seen some incredibly great low light tungsten footage (see the new reel). The difference can be as little as one stop under-exposure. There is not much forgiveness in a blue starved light situation. A little testing can do wonders.

Jim
Click to expand...

wise words.. amen
 
Something as basic as color monitor set-up is good info. for those us prone to make a mistake. Good thing I don't captain commercial aircraft.
 
Here ya go, keyed with a 55W bulb and edged with a litepanel...
ISO 400 with a Schneider 1/4 CTB filter and T2.1 (The Jpeg compression make the noise much worse than it is in motion)
Mamet.jpg
 
Each new sensor we will release will improve some in every aspect (as expected). But you still need to be aware of certain sensor characteristics. Protect the highlights in bright daylight because the shadows are very forgiving. Protect from under-exposure in low light/tungsten lighting because noise hides there.

Properly exposed images can be incredible... as you know.

Evin... tips for the troops?

Jim
 
The sensor doesn't like low levels of tungsten light, so the solution is... don't give it low levels of tungsten light! Give it more light or bluer light or add a blue filter (which requires more light).

OK... but will the later cameras & sensors handle tungsten light better at higher ASA ratings? Ideally we'd be able to shoot at least at 500 ASA in tungsten light and get the same blue channel noise as we currently get at 320 ASA in tungsten light. Is that a probability?
 
Thanks

Thanks

Great common sense tips, Jim, that novices like me would miss! I love tips of how to shoot higher quality images.

Thanks for the useful answers. Not a Red owner...yet!
 
David Mullen ASC said:
The sensor doesn't like low levels of tungsten light, so the solution is... don't give it low levels of tungsten light! Give it more light or bluer light or add a blue filter (which requires more light).

OK... but will the later cameras & sensors handle tungsten light better at higher ASA ratings? Ideally we'd be able to shoot at least at 500 ASA in tungsten light and get the same blue channel noise as we currently get at 320 ASA in tungsten light. Is that a probability?
Click to expand...

That is probable... :-)

Jim
 
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).
 
I would also suggest (only in certain controlled lighting situations) to use tungsten on your subject, and fill in the background with daylight balanced (or at least somewhat cooler) lighting.

If you can keep spill to a minimum and have some good grading options in post, this could help... again in certain controlled situations.
 
On the Postside:

If you experience blue noise in Lowlight Tungsten shots due to the 3200°K setting,
try to push the slider up towards an more ideal 4500-5000°K in Redalert.

Then render a 4K Full Debayer to get the best out of it (DPX or Uncompressed), even if you finish in 2K or less.

Grade it back to the liked temperature and by scaling down you can achive real nice textures.

Jeff
 
I always recommend either Daylight balanced light or wearing an 80C or 80D. If you watch the histogram in View raw, you'll get the best results when the channels start to line up, in other words, if you feed the camera the closest light to 5000k that you can, you'll get the best results.
 
What about available light noise?

Because I'm mostly working in a very "rough" conditions of everyday life and available light only (no time for a light-meter, additional lighting, etc...).

I use and set RED1 mostly at "native" 320 ISO and 5000 K.

For available light I always try to get the fastest lens as I can afford (f/1.0, f/1.2 or T.3 if I can get or if not then f/1.4 or T1.8 or T1.9).

The rule of the game here is to bring enough light to the sensor, more light then less noise you get.

Then later R3D RAW always give me more opportunities to deal and "tweak" a footage.

If I have some noise then I use AE with Grain reduction and play with it.

Here are the examples shot with RED1 in available light, lens Canon 50mm FD L f/1.2 @ f1.2, ISO 320, 4900 K, shutter 1/50, 25 FPS, post in AE: a bit of Saturation and Grain reduction...

AL_018.JPG

Shot hand held with RED1 in available light (indoor), lens Canon 50mm FD L f/1.2 @ f1.2, ISO 320, 4800 K, shutter 1/50, 25 FPS, post in AE: a bit of Saturation and Grain reduction>>>Noise Reduction Settings...

AL_017.JPG

Shot hand held with RED1 in available light (indoor), lens Canon 50mm FD L f/1.2 @ f1.2, ISO 320, 4900 K, shutter 1/50, 25 FPS, post in AE: a bit of Saturation and Grain reduction>>>Noise Reduction Settings...

AL_019a.jpg

Shot hand held with RED1 in available light (indoor), lens Canon 50mm FD L f/1.2 @ f1.2, ISO 320, 4900 K, shutter 1/50, 25 FPS, post in AE: a bit of Saturation and Grain reduction>>>Noise Reduction Settings...

The pictures processed in RA 320 IS0, about 5000 K, Colour Space: Camera RGB, Output: RED Log.

(Noise at the pictures above is due to JPEG web compression).
 
camera fill

camera fill

when i started shooting stills with digital in 2002 with a medium format system, the sensor (cmos) behaved similarly to what jim is describing. i ended up using alot of subtle "camera fill" specifically to fill the shadows. I would vary the fill depending on the contrast required, but the aim was to maintain the lighting characteristics while producing a slightly flattened file so that we could crunch it in post.

over the years the stills sensors have improved so much that i no longer need this technique (35mm sensors seem to be better)

i am happy now with current stills sensors, because i don't have to compromise the way i work (ie: light how i like, with what ever lights I want, including tungsten)
 
David Mullen ASC said:
... will the later cameras & sensors handle tungsten light better at higher ASA ratings? Ideally we'd be able to shoot at least at 500 ASA in tungsten light and get the same blue channel noise as we currently get at 320 ASA in tungsten light. Is that a probability?
Click to expand...

Jannard said:
That is probable... :-)

Jim
Click to expand...

great news!!

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