Histogram not tightly enclosed by raw-bars as in manual

[Brian Boyer]

I get what you're saying but the only look parameter is 5000K, iOS 400, and redgamma2. The problem is that both redgammas compress the highlight range from sensor clip and represent them below 98 IRe. It's the reason there is a gap in the histogram as well. Hoping someone like Graeme can chime in and answer why the redgamma curves do not represent clipped highlights at 100+ IRE or far right on the histogram.

I can dig it. An official response from a RED team member is always the preferred source of info.

With all the spouting off I'm doing on this subject it would be nice to get an official nod that I'm on the right track and not unintentionally spreading misinformation. I hate when I do that.

Having said that, I realize that I'm not fully addressing portions of your question but I'm going to totally take a guess anyway based on something I remember Jim stating a little while back.

Are you talking about the same shot in RCX? Or monitoring path in the camera?

RC3/RG3 should induce you to expose the same in camera. RC3/RG3 can force a bit quicker clipping (if the shot was clipped) in RCX. As I stated a few times, critical DI will probably be done with RC3/REDlogFilm. That combo gives you the most amount of info.

RC3is intended to be your primary color option in all cases. RG3 is intended to be a monitoring path option and good/fast grading option.

Remember that you are shooting RAW (REDCODE RAW) and your monitor path choice has no bearing on anything in post.

Jim

The link is here: http://www.reduser.net/forum/showth...-3-0-(release)&p=961099&viewfull=1#post961099


The monitor path choice has to conform to video standards. So, if I had to guess i'd say perhaps they're mapping the values to fit in the video space and maybe give a little headroom before clipping the video level as a good starting point, which could come in handy in some instances.

Then again, I could be WAY off.


Eric, please forgive me if I'm wrong, but it seems that you may be using and thinking about the zebra tool and IRE scale in the wrong context.

As far as I know, there's no way to look at the IRE scale and tell whether or not the sensor was blown. It's measuring something entirely different. I think it assumes we used our in-camera exposure tools that are built for that specific purpose.

In fact, the IRE scale in RCX is also dependent upon the metadata settings and ultimately tells you nothing about the clipping status of the sensor. All it cares about is showing whether or not the signal passing through it is broadcast legal. It doesn't know any more about the signal coming into it than an external waveform monitor does.

You've asserted that the zebra tool should be reflecting what the exposure tools are indicating but, as I said before, the tools measure different things at different points in the signal chain.

The IRE scale is decades old...older than RED and raw sensor recording combined. It's based on broadcast video levels and represents far less dynamic range than what we're capable of capturing.

We have to squeeze that increased dynamic range into the broadcast standard in a way that preserves the integrity of the original image. Audio engineers face a similar challenge when compressing the wide dynamic range of a full orchestra to fit within the 16 bits of audio CDs. (I know, I'm on an audio analogy kick).

If you're using the zebra to reaffirm you've clipped the sensor then, to paraphrase "Raiders", you're digging in the wrong spot.

RED gave us no less than three at-a-glance tools to illustrate that info - the right goal post, the traffic lights and the exposure check - that you now seem to discount (because of the gap) in favor of the tools that are meant to analyze videocentric data that's at the mercy of metadata.

The irony is that the histogram (and therefore the gap) doesn't give you info about sensor clipping either. It's also dependent on metadata settings.

As for the gap on the right, I have no clue why it's there, even after all the explanation. What I do think I understand is that eliminating the gap would cause issues with how other data and info is represented.

I know you're sick of hearing from me but, as I said before, it was more about me wanting to make sure I understand what's going on more so than go back and forth with anyone. I think I'll keep mum on the subject for a while until someone from RED has more input.


Cheers,
Boyer

 

[Sidney L. Plaut]

now im getting confused... it was so simple on the red1 before the mx upgrade. - you had your falsecolors and they showed the colors up to and including RED.

ok I understand that FLUT (incl red gamma vxxx) never clips high lights therefore RED is gone from VIDEO (old falsecolors) - except you can get it to show RED when you go higher ISO than 800???

I seemd to remember the histogram being a RAW histogram on the RED 1 ?
(which would make a lot of sense - since we are always exposing for the RAW sensor in terms of CLIP and Shadows).

On the RED 1 we had the awesome vertical RAW exposure bar - that coupled with the RAW traffic lights and a skin tone test with "false color" gave you a fantastic accurate exposure! - now everything is more... uncertain.

because of that uncertainty mistakes happen, especially on the Rim lights and side kickers, I seem to overexpose, because in the "VIDEO" they look fine, but actually they are actually waaayy to close to "CLIP", BUT because I used a low ISO they never turned YELLOW.... uncertainty is not a friend of a DP.

 

[egon touste]

im just reaquaiting my self with a dragonized epic ..

..give me the option of a raw histogram, with the vertical clip bar (that starts when the first pixel clips) .. and im all good with exposure.. im hoping thats still an option .. otherwise i may have to view in linear /raw light.. if that is in there.. whatever makes the histogram useful again! ..must revise now.. clients coming soon..

 

[Phil Holland]

im just reaquaiting my self with a dragonized epic ..

..give me the option of a raw histogram, with the vertical clip bar (that starts when the first pixel clips) .. and im all good with exposure.. im hoping thats still an option .. otherwise i may have to view in linear /raw light.. if that is in there.. whatever makes the histogram useful again! ..must revise now.. clients coming soon..

Gio Scope. And it's selectable and useful false color "guides". Red is the color represented by the highlights and it dips to black towards pure overexposure. I'm using this too a lot these days.

The "stop lights" go off when your first pixel clips in any channel.

The historgram is the tonal distribution for your given ISO, Color Space, Gamma Space, and Look.