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 RAW Monitor On Camera Really RAW?

Yes, you're using lin to mean linear light, which is correct. But "lin" is a word that's been hijacked, so if we use it, people may take it to mean the hijacked meaning. We're right and they're wrong, but....

Sure, you'd be able to see what's going on in the highlights ever-so-slightly better, but at the expense of much, much worse shadow imaging. Hence the call for 709 curve.

Graeme
 
I would rename "RAW" mode to "709 RAW" or something to that effect and add the true "linear RAW" green darkness in there. There are a lot of times when seeing that brightest stop clearly is the most important thing that we need on set. It would be perfect for that final "what do we have and what don't we have" check before a take... without any pretty curves or matrix in the way. That last step is the most important part of getting that nice, fat exposure.

Many of us know the curve well enough to judge what the actual clip point is based on the other modes, but it would be nice to have a true linear mode to strip back the translation.

If somebody puts the camera into linear mode and exposes for the shadows like that... well... perhaps they're using the wrong camera. :) There could be a nice CYA blurb in the manual that says "do not shoot with this tool".

Adding access to the linear data would be nice in post as well. SCRATCH can start with the linear file, but the "free" apps can't. That's a separate thought, but it'd be nice as well.

I'd love to have a linear RAW mode in the camera that I can access with a single button push.
 
Sure the free apps can Brook. RedAlert Redcine, and the command line tools have always had the linear light option. The difference in how the highlights in the brightest stop appear is negligible, but the difference in the shadows is vast. The goal of the RAW mode is so that you can see what is happening across the whole dynamic range, and the way we're doing it meets that goal the best.

Graeme
 
There's quite a few situations where a "RAW/lin" preview would be a good realitycheck.

It is an issue that people will have to understand the tools - whatever.

before you added RedSpace, a lot of people were overexposing due to the thought that "it's really there, we just don't see it"

When you added RedSpace, a lot more underexposed images came in, because it pushes the highlights (to preserve them of course...)

For a lot of users, RedSpace is a very good starting point. For some, it would be nice to check the lin image before hitting record... :)

On another note (and I should take this to another thread, but just because you're looking over here Graeme...)

The LUT approach of the RED-Room is giving me some headaches, when I need to use a ... LUT :)

Any thoughts?
A Davio will solve it, of course, but what happens when you choose "remove display lut" before final render is... disturbing :)

Is there a way to use LUTs with RED/Color workflow without having the LUT applied to the projector/monitor?
 
Just to throw my opinion back into the ring...

The most important thing is that people understand that RAW output on the camera is not really what the sensor is seeing. That seems to be a common assumption that people are making and in some cases, they're underexposing because they thought it was exposed brighter than it actually was.

Just understanding that will solve most of the issues.

Some things might help solve this misconception. Calling RAW something like RAW 709 (like was previously mentioned). Or adding a linear light RAW as an output, but I can understand the concern that that might cause more problems than it solve.

For me, I can now explain to DPs what is actually happening and we can let them know some warning signs of when they'll underexposing their images (even though it might look dark, but acceptable on set).

For anyone who wants to see the difference between the two outputs, just open an R3D file in REDAlert and change the Output LUT/Gamma Space to linear light.

Thanks and I hope to see some clarification on this in the future.

-Russ
 
I'm with Gunleik; It would be useful to have a genuine Raw view. I use the monitor for composition and as a highlight exposure meter and therefore the option of wysiwyg would be a bonus. But I'm oldfashion in the sense that I always find it inconvenient when a system (such as Windows) tries to guess what I want to do.
 
David Mullen ASC said:
RAW, by definition, is unviewable as a color image.
Click to expand...

David, I think we use cameras in different ways and that's the explanation for the puzzle. For example, I can easily understand that in your case while shooting you "adjust" colors with lights, reflectors etc. and this is one of your main priorities and need for the monitor. But in my case as I shoot wildlife documents outdoors and it's not possible to control the light, the main priority is usually to make sure highlights do not get clipped.

Graeme, I wonder now, if RAW view and the zebras are on, do the zebras show the highlights after or before the gamma REC709 map?
 
David Mullen ASC said:
Sure, I understand... I'm just saying that a RAW image is not really viewable -- what you're seeing is a conversion of the RAW image into a color one.
Click to expand...

... yes indeed, and in this sense there are situations in which I would like also to see the "conversion into the R3D-file" --to know I'm storing the maximal amount of data-- and then worry about the colors and tones only while processing the files. For, if I have maximal amount of data, that should yield the best starting point in reproducing the colors in edit. Guess this is one of the main points behind storing RAW files. With no doubt, still the conversion to colors is vital and could not do without it.
 
The best tool, I find, for raw clipping, is that little bar graph you see next to the histogram. I find that works very well for me. When you're in RAW view mode, the histogram itself will show you clipping.

I've not played with the zebras on the RED, again, just the different ways we all use the same camera. However, it doesn't matter if the RAW is with a gamma curve or not for that as the gamma curve is not throwing away any levels. It's a full range curve.

Graeme
 
f.ck, I thought with rgb parade in raw view, false colour and a light meter I was overdoing it being on the safe side, now it seems I've still been misunderstanding what I was looking at. On the other hand, I wonder how academic the discussion is becoming for me, and for many others, the footage always turns out wonderful on the videomonitor
 
I've had zero problems setting an exposure, preserving shadows, and protecting highlights with available range of RED meters, my Sekonic, and my eye. Not sure what all the fuss is about.

I'll watch my contrast ratios with a meter/eye and pop over to the RAW view to double-check the chip's clipping point.

Recently, I've been setting a mild s-curve in the camera with the tone controls so clips start that way in REDCine, etc. Having the s-curve in my camera preview informs me even more about how much highlights will bloom and shadows will crush with that s-curve.

Michael
 
Michael Morlan said:
I've had zero problems setting an exposure, preserving shadows, and protecting highlights with available range of RED meters, my Sekonic, and my eye. Not sure what all the fuss is about.
Michael
Click to expand...

My hunch is that you've had some footage that wasn't optimally exposed and you didn't know it because you might not have checked to see how the RAW looked with a linear gamma. This isn't a personal attack. I just think that the RED camera offers enough flexibility that most people don't realize that their footage has been underexposed, especially if they're using a REDCine/REDAlert workflow that avoids showing a linear gamma in normal use.

Even the underexposed footage that I've worked with has been usable. It has more noise that it should, but it's never been unrecoverable.

Just as a reference though, if you're monitoring in REDSpace and your highlights are peaking at around 50 IRE, then the highlights in RAW are actually closer to 15 IRE. If you're monitoring in REC709 or the camera RAW and the highlights are peaking at around 35 IRE, then they're actually closer to 15 IRE.

Just understanding that makes problems avoidable. A DP might say that 50 or 35 IRE for a shot is dark, but acceptable for the circumstances because you can pull it up well in post. But when they think it's 35 IRE and it's really 15 IRE, they end up being surprised.

I completely agree with Graeme that trying to light a shot using linear light would be very challenging because it's so difficult to see what's happening in the shadows. But I do think there is value on set and in training DPs to being able to see an interpretation of the RAW data as linear light, or to at least understand that monitoring in RAW is actually monitoring in RAW with a RED709 gamma.

-Russ
 
False color zone system

False color zone system

Russ L said:
(snip)
Just as a reference though, if you're monitoring in REDSpace and your highlights are peaking at around 50 IRE, then the highlights in RAW are actually closer to 15 IRE. (snip)
-Russ
Click to expand...

Please explain this statement. Why would I consider 50 IRE a highlight peak?

FWIW, due to my extensive experience with the RedOne I am able to check several metering and monitoring displays to determine exposure quite reliably, but that is hardly scientific. In my perfect world there would be a "true" view for false color that re-mapped linear light to an approximation of Ansel Adams' Zone system with each zone keyed to a unique color.

I support the philosophy that led us to RedSpace as a functional real world working color/gamma space (particularly great for the client monitor). However, once I select an exposure tool like false color, human visual system compatibility no longer matters and I just want color bands to delineate 10 IRE segments.

More feature requests - now that's the holiday spirit :shifty:
 
Blair S. Paulsen said:
Please explain this statement. Why would I consider 50 IRE a highlight peak?
Click to expand...

It's less than ideal to have the highlights at 50 IRE, but in certain low light situations, 50 IRE would be considered acceptable. But if you're monitoring in REDSpace, then the actual IRE of the highlights would be closer to 15 IRE instead.

If a DP things the image is at 50 IRE, then they might go ahead with the shot. But if the DP knows it's at 15 IRE instead, many DPs might pull out more lights or use a faster lens.

Is that what you wanted me to explain?

-Russ
 
Russ L said:
It's less than ideal to have the highlights at 50 IRE, but in certain low light situations, 50 IRE would be considered acceptable. But if you're monitoring in REDSpace, then the actual IRE of the highlights would be closer to 15 IRE instead.

If a DP things the image is at 50 IRE, then they might go ahead with the shot. But if the DP knows it's at 15 IRE instead, many DPs might pull out more lights or use a faster lens.

Is that what you wanted me to explain?

-Russ
Click to expand...

Russ,

50 IRE seems awfully low for highlights. I set my zebra at 100, usually. The headroom usually doesn't need to be much. And some things (like lightbulbs, bright reflections) blow out to the eye anyway, so why worry about them blowing out on the camera?

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