Will the new chips have more blue?

[Stokestack]

Hi all.

I haven't shot on Red yet, so don't get defensive. I'm basing this question on observations of footage over the last couple of years.

It seems that Red footage suffers from a lack of blue. I've seen what seems like tons of clips with an almost sepia-toned look; and while in some cases it's clearly exaggerated on purpose (or as an adaptive strategy), in others it just seems that blue is missing. Basically everything looks, to varying degrees, like Book of Eli.

Past reports about blue noise would seem to support this suspicion, since weak blue at acquisition time means you wind up massively boosting blue to achieve anything near a normal color balance in post processing.

Can anyone weigh in on this from a variety of shooting experience? And will the new chips provide improved blue sensitivity (if indeed that's the problem)?

 

[Tim Whitcomb]

Hi, this has become a 100% Real Name forum. you can change it here.

http://reduser.net/forum/showthread.php?t=32432

As for question, I wouldn't classify it as a "problem" but a challenge with CMOS sensors and Bayer patterns. Lots of information on here to address the blue channel.

Book of Eli was shot on Build 14 or 15... we are now on Build 21 and have ALL NEW color science... desaturating digital images is often en vogue when going for a "film look" and is not limited to the RED ONE for a coloring style choice of late.

Moreover an upgrade program is imminent for a new sensor... with much more capabilities than the current sensor.

Get your name updated and spend some time catching up on RECON and you should know your question is as almost as outdated as its answer.

 

[Ryan S]

When I was shooting with Build 17 about 14-15 months ago, I did get some blue noise when shooting under tungsten. That said, I was able to dial it down in post to an acceptable level.

The new colour science that came with build 20 reduced the blue channel noise substantially.

The next colour science, along with the improved sensor, will be even better. The sample photos and footage we've seen so far support this.

Do you have a shoot coming up, or is this just curiosity? A lot of people here will be able to suggest ways of minimizing noise, if you need help.

Edit: What Tim said!

 

[Stokestack]

Thanks for the info. It's not that I have a shoot coming up. The color rendition is a factor in my decision whether to buy a next-gen Red camera. I'd like to see more blue.

I saw the discussions in the past about strategies to reduce blue noise (or increase blue levels while lighting), and they were pretty informative.

G. Stokes

 

[David Mullen ASC]

Here are some frames from a Red feature I shot called "Stay Cool", Build 16. You can see that it's possible to get blues:

Of course, you'll get cleaner blues at a daylight color temp base setting.

These are jpegs heavily compressed for the web so don't get too concerned about the blockiness in the blue patches.

 

[Thomas Wright]

looks like a dream

 

[Cail Young]

I've seen what seems like tons of clips with an almost sepia-toned look; and while in some cases it's clearly exaggerated on purpose (or as an adaptive strategy), in others it just seems that blue is missing. Basically everything looks, to varying degrees, like Book of Eli.

You cannot judge a camera system's colour rendition on post-colour-corrected footage.

 

[David Mullen ASC]

Well, that's confusing because (1) all footage gets color-corrected eventually and (2) everyone here also says to not judge the RED look by how it looks in RAW, because it has to be color-corrected!

The truth is that a lot of RED movies have been timed on the brownish pastel side, for various reasons, and blue is not the most saturated element when working and color-correcting from RAW files, especially in tungsten lighting. But it is certainly possible to boost the blue saturation, especially if you shoot in daylight lighting where the noise problems in the blue channel will be minimized.

And now with the improved color science, etc. you have even more flexibility.

 

[Damien Molineaux]

Here are some frames from a Red feature I shot called "Stay Cool", Build 16. You can see that it's possible to get blues:

(see above post for the images)

Of course, you'll get cleaner blues at a daylight color temp base setting.

These are jpegs heavily compressed for the web so don't get too concerned about the blockiness in the blue patches.

Thanks for sharing Dave, those look nice.

Is the film out yet ?

Cheers,
Damien

 

[David Mullen ASC]

No, not yet.

 

[Stokestack]

Thanks for those shots, David.

I do notice that the overall color level looks pretty cranked on them; the concrete has a distinct color cast in the airport and locker shots. That is the first blue sky I've seen in a Red shot, however.

One thing I noticed in the "Dublin's People" short (from a 7D) was the ample blue, even under crappy streetlight conditions.

 

[Omar Saad]

That is the first blue sky I've seen in a Red shot, however.

Here are a couple of heavily compressed jpegs from some shooting I did last week. Nothing serious just some outdoor tests. Lots of blue skys. These were brought up in color and have some hard curves on them, but it's not too difficult to get some decent blue sky, especially since build 20, and when the conditions are right.

Best,

Omar

 

[Cail Young]

Well, that's confusing because (1) all footage gets color-corrected eventually and (2) everyone here also says to not judge the RED look by how it looks in RAW, because it has to be color-corrected!

Perhaps in my rush to be brusque I neglected to be clear.

I disagree with 2, specifically in the case of someone asking a question framed as "all footage I see lacks blue, is there a problem with blue on this camera?".

The solution is to get your hands on a camera and not make assumptions based on others' work.

 

[Stuart English]

Real Names

Real Names

........

A reminder Stokestack this is a real names forum. If you haven't contacted the administrator already please do so...

 

[Stokestack]

A reminder Stokestack this is a real names forum. If you haven't contacted the administrator already please do so...

I did, and got no response, so I set up a whole new account when the rule was first implemented, and that has disappeared. I put my name in an earlier post, so I don't know what further action to take.

G. Stokes

 

[Stokestack]

I disagree with 2, specifically in the case of someone asking a question framed as "all footage I see lacks blue, is there a problem with blue on this camera?".

The solution is to get your hands on a camera and not make assumptions based on others' work.

Asking questions does not equal making assumptions. In fact, it's just the opposite, and it's what forums like this are for. And you disagree with the framing of a question? What are your guidelines for question-framing?

 

[Rob Ruffo]

Here are some frames from a Red feature I shot called "Stay Cool", Build 16. You can see that it's possible to get blues:

Of course, you'll get cleaner blues at a daylight color temp base setting.

These are jpegs heavily compressed for the web so don't get too concerned about the blockiness in the blue patches.

Yummy clean Hi Sat - truly yummy. I love the look of this film.... Nice work.

 

[Mohammed El Sharqawy]

to me, all David's images above has some Orange/green and some warmy highlights and overall tonality,, except for the areas that is being lit by daylight..
I think I've seen some red footage the has more blue balance.. i mean seeing the image all white light sourced.. without those Orange/Green tints..
Btw, these are really great images.. its matters of deciding how do want the image to look like so you play on it while shooting and in post, working image controls, lights, and post stuff..

 

[Cail Young]

Asking questions does not equal making assumptions.

The only question in your first post is "will there be more blue in the new sensors?" - although you have included room for your hypothesis being wrong at the very end.

The rest is assumptions based on observation only - which is good if you can make those observations given the evidence but as I said (and which David Mullen has illustrated) you cannot assume the camera is blue-deficient by looking at processed footage. Blue is definitely a weakness, but keeps getting better with firmware and M-X looks to improve the situation even more.

In short, I don't mean to discredit your question; it's an important one. I have just seen too many people make judgements of an imager based on (undocumented) processed images without testing for themselves.

 

[David Mullen ASC]

to me, all David's images above has some Orange/green and some warmy highlights and overall tonality,, except for the areas that is being lit by daylight..
I think I've seen some red footage the has more blue balance.. i mean seeing the image all white light sourced.. without those Orange/Green tints..
Btw, these are really great images.. its matters of deciding how do want the image to look like so you play on it while shooting and in post, working image controls, lights, and post stuff..

These were only rough examples because the color-correction was done very crudely -- the chroma level was boosted in the camera's metadata, so the TIFF frames I grabbed in Red Alert had the boost, and then I played around with JPEG's in Photoshop, and then compressed them further for the web. So this in no way is a professional grade.