Truth or .... False Color

[David Battistella]

This would be very useful!

But above that we need to know what that is actually referencing in terms of light because right now with epic those tools are based on 320ISO. Not 800iso (which is supposed to the base iso of epic)

Battistella

 

[Mark L. Pederson]

But above that we need to know what that is actually referencing in terms of light because right now with epic those tools are based on 320ISO. Not 800iso (which is supposed to the base iso of epic)

Battistella

No you do not. It should not be referencing any ISO - but rather - when the photons get too much light or not enough - just like the stoplights David.

 

[Mark L. Pederson]

But above that we need to know what that is actually referencing in terms of light because right now with epic those tools are based on 320ISO. Not 800iso (which is supposed to the base iso of epic)

Battistella

This is not correct David. The RAW mode RGB representation of the MX is "based on 320ISO" - but the tools are NOT. Which further proves IMO that RAW view causes more harm (confusion) than good.

 

[Zeb B]

The RED engineering team gets what you are suggesting Mark - even if some of us don't. Solid idea

 

[David Battistella]

With respect to dropping RAW view - we already have a LOG view - so ... I would love a real argument why we need RAW view - and who actually uses it - and why? (other than using it to get the exposure meters not effected by ISO)

It's not that we need a RAW view, that could easily be replace with a log view or "wizard of OZ view". The point is that all expose tools on red since the RED one have been based on a 320ISO sensitivity. So why do I rate at 320, because that is the stated base sensitivity as confirmed by the exposure tools.

When epic came out and they moved to a "recommended base" of 800 to move the middle grey and widen the exposure latitude, the tools and raw view were left at 320. Why? Well mainly because MX was less noisy at 800 than M but not really more sensitive. This looks to have changed with dragon, so it's probably a good idea to move all that to 800 with dragon.

My my ultimate point.

All RED really need to do is state: all in camera exposure tools are based on a rating of ISOXXX and let people work out what is best from there (which people have done) but when you use the tools, no matter how the cameras metadata is set, you are measuring light hitting the sensor, not post processed light coming through the monitoring path.

I want to know what is hitting the sensor, that should be the priority, even if it means getting rid of false color, or histograms or stoplights or whatever is not showing you a true and accurate sensor reading.

But this is why I have raw view mapped to a button with exposure right beside it, so that with two quick clicks I know what's going on.

Battistella

 

[David Battistella]

This is not correct David. The RAW mode RGB representation of the MX is "based on 320ISO" - but the tools are NOT. Which further proves IMO that RAW view causes more harm (confusion) than good.

Sorry, should have added, while in RAW view....

 

[Gunleik Groven]

but when you use the tools, no matter how the cameras metadata is set, you are measuring light hitting the sensor, not post processed light coming through the monitoring path.

Battistella

Problem is that this is not currently the case. And thus people think they have a good exposure (due to some ISO/WB messup + exposure tools), while they do not really.

 

[Mark L. Pederson]

Sorry, should have added, while in RAW view....

Still David - the RGB representation is - but the tools are NOT (based on 320ISO)

 

[Mark L. Pederson]

Problem is that this is not currently the case. And thus people think they have a good exposure (due to some ISO/WB messup + exposure tools), while they do not really.

Amen.

And further - advanced users - who understand exactly what is and what is not happening regardless of documentation - can't use this tool the way it does work correctly without screwing up the monitor path.

 

[David Battistella]

No you do not. It should not be referencing any ISO - but rather - when the photons get too much light or not enough - just like the stoplights David.

Nope. I disagree. All that is telling me is that some pixels are blowing out. That happens on film all the time. Things blow out and you let them go. That's no problem. All those stoplights tell me is that there are unrecoverable pixels in one channel being recorded. That does not mean I am not choosing to let those pixels go.

I could post a lot of example of where it's perfectly fine to let pixels blow out in a channel. I'm sure even Gunliek can attest to some images he has posted where the stoplights may have all been lit.

Like it or not that digital sensor is like a fixed electronic piece of 35mm film. Would you load any kind of film in a magazine and tell a cinematographer to just go shoot, don't worry if it's 50D, 320, 500, 800, etc? , "don't worry this little light will go off if ten grains of silver blow out on the neg."

It's nuts to take years of photographic scientific ways of measuring light and simply state that three dinky stoplights replaces that.

You should rethink your position on this one. A whole creative industry depends on ISO as a way of measuring light values.

 

[David Battistella]

Still David - the RGB representation is - but the tools are NOT (based on 320ISO)

Then I should have added raw view and the current tools red gives us for measuring light hitting the sensor, which we agree are flawed. Histograms from camera to REDcine don't always match up that well either...

battistella

 

[David Battistella]

Amen.

And further - advanced users - who understand exactly what is and what is not happening regardless of documentation - can't use this tool the way it does work correctly without screwing up the monitor path.

And monitoring path is a different thing altogether from the accuracy of what the tools are measuring.

 

[Gunleik Groven]

And monitoring path is a different thing altogether from the accuracy of what the tools are measuring.

It should be, it currently is mostly not.

 

[Mark L. Pederson]

Nope. I disagree. All that is telling me is that some pixels are blowing out. That happens on film all the time. Things blow out and you let them go. That's no problem. All those stoplights tell me is that there are unrecoverable pixels in one channel being recorded. That does not mean I am not choosing to let those pixels go.

I could post a lot of example of where it's perfectly fine to let pixels blow out in a channel. I'm sure even Gunliek can attest to some images he has posted where the stoplights may have all been lit.

Like it or not that digital sensor is like a fixed electronic piece of 35mm film. Would you load any kind of film in a magazine and tell a cinematographer to just go shoot, don't worry if it's 50D, 320, 500, 800, etc? , "don't worry this little light will go off if ten grains of silver blow out on the neg."

It's nuts to take years of photographic scientific ways of measuring light and simply state that three dinky stoplights replaces that.

You should rethink your position on this one. A whole creative industry depends on ISO as a way of measuring light values.

David - you are not understanding me. I'm the guy who preaches the importance of ISO for measuring light.

We are talking about EXPOSURE TOOLS - and defining and understanding what they do.

In RAW mode on MX - the TOOLS are NOT based on 320 ISO - that is what you said - that is what I tried to correct.

You are also proving my point about how valuable a tool FALSE COLOR can be when NOT TIED TO ISO. Your stoplight indicate clipping - you don't know what and how much - you check the false color - and adjust so only the exact part of the image clips that you want, etc.

VIDEO mode gives you complete ability to have all tools tied to ISO - you wanna work like video - you can.

 

[David Battistella]

I am totally getting the discussion about exposure tools.

Let et me explain.

At iso320 raw view all the exposure tools give you an idea of light hitting the sensor. (Known value)

when you switch the camera to iso800 rg3/rc3 (meta data view) these tools are now reading the meta data values, so when you are using them they are inaccurate according to what is hitting the sensor.

What I hear you saying, and what I agree with, is that when you use any of the exposure tools, they should be giving you back a known value (what you would see in raw view) and not what you are monitoring in the metadata. Without having to switch to RAW view to do so.

Also, these tools should not affect the monitoring path, only the operators view should be affected by engaging these tools.

Is that correct?

 

[Hans von Sonntag]

if someone can't drive in F1 they move to F2 or F3. If you can't drive you hit the wall. If you don't know the car, that is not reds problem. It's a professional tool, not a handy cam. Some want it basic, some want all the tools.

Next problem becomes, "well I had it in basic mode and my picture still look like shit" well, the pictures look like shit because you have been using cameras that compensate for a lazy lack of talent. Red is not a complicated camera, it's actually an honest camera. Many students blew a few hundred feet of film because of lack of knowledge, not because the film was no good. That's good, that's a learning process.

Red should not not be holding peoples hands so that everyone loves them. I think it's laid out pretty well now, based on a lot of user input from the start.

battistella

David, I heartily and humbly disagree. Red's cameras are supposed to be made for filmmaking and not for nerds. I don't know how your fellow filmmaker are "set up". My, at least, are good in their craft and not so good in menus and the understanding of metadata conceptions. Actually, I know at least one very successful director who is getting very impatiently when things are complicated and not paying directly into his work (and it's not me, although I have the same approach). Let alone the thousands of post houses that advise producers to go the Alexa route for reasons that are beyond DR, colour, resolution and such but funky metadata, raw development, render times etc... Only a few postproduction supervisor I know understand the Red approach fully. The majority live from hearsay and tips that are often way out of meaningfulness.

INHO, the biggest obstacle Dragon has to face is an easy access for all those folks who are not F1 driver but very good chauffeurs of luxury sedans that carry demanding clients. Please read some lines in one of the endless "exposure" threads popping up every here and then. They all show that even seasoned user oftentimes don't really understand what is going on when shooting with a Red Epic.

Regarding learning process: In 2010 I shot my last roll if 35mm. I started shooting digital in 2008 on my RedOne; it took me some time to understand the route Red wants be to go. I know Alexa and some other cameras. They are all different and need all a deeper reading of the guide-book. If there was a standardised method life would be so much easier. In this regard film was as easy as it gets. Red invented the area of digitale film acquisition but Arri made it accessible to the seasoned DPs.

Red should take their leadership back. DR and resolution is one thing, easy access the other vital part. Frankly, Red's menu works for the adept well, not so much for the non-owner/DP who has to rely on the knowledge of DITs and third parties. This was very different in the age of film.

The easier the access is the more people rent the Dragon or buy it for them selfs. Marks's comments are good comments and a helpful step in the right direction but I guess for KISS' sake Red should investigate much deeper how things could be simplified.

Hans

 

[Nick Shaw]

I think this might just be a BUG/ERROR/OVERSIGHT on RED's part - since the description that Trent from RED posted in this thread from the HELP menu is EXACTLY what I want.

Pretty sure I'm not insane and it only works correctly in RAW mode. Anyone have a camera handy?

You're not mas Mark...

False color is affected by ISO/WB

Possibly this is a bug that came in when the green colour was added to Exposure Overlay. My memory (which could of course be completely wrong) is that originally the purple and red overlays were RAW based as advertised.

 

[John Marchant]

This is something that I spotted a while back on Dragon, but its almost certainly tied to calibration so I've just been making do for now.

Easiest solution to getting misled is just to map your preferred key with a couple of presets so that when you toggle exposure check you automatically toggle raw view too. Very easy to setup, and its how our Dragons are currently

 

[Mark L. Pederson]

This is something that I spotted a while back on Dragon, but its almost certainly tied to calibration so I've just been making do for now.

Easiest solution to getting misled is just to map your preferred key with a couple of presets so that when you toggle exposure check you automatically toggle raw view too. Very easy to setup, and its how our Dragons are currently

It's also the same on MX - has been for a while. Serious doubt it's tied to calibration.

Yes - I do the same hack you do - mapping a key to toggle to RAW and turn on Exposure - but clients sometimes totally freak out when they see the ugly-ass RAW view hit the monitor path. AND AC's forget to change back - and it causes drama and confusion on the set.

 

[Alexander Christ]

Hey Trent - please try this on a camera. Swing your ISO and watch the Purple and Red change. Then - switch to RAW view - and try the same thing. You will see that it only works correctly in RAW view mode.

Unless I have completely lost my mind ... which is always possible.

one more vote for: RED and PURPLE in EXPOSURE should stick to the raw metering and not be affected by ISO. I like the GREEN (middle grey) to stay in EXPOSURE and be tied to ISO. In this way I can judge exposure and clipping with pushing one button.