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 :)

REC709 Output and the need for 4K Monitors

Stuart English said:
Verify that your monitor can reproduce all 11 steps - if it can't do that then the monitor is either clipping the highlights or shadows or both.
Click to expand...
The only problem with that is: it still doesn't resolve the issue of calibrating monitors on location, in the edit bay, and in the final color-correction room. It also won't confirm that accurate color is being displayed on location -- only that the whites and blacks aren't clipping.

We have non-trivial problems in post just getting well-adjusted monitors to look similar -- say, a broadcast CRT, a high-end plasma, and a D-cinema projector. Trying to do this with a $2000 LCD display on location is a tough challenge. It's a concern when DPs are making critical changes to LUTs on location, and expecting this look to translate throughout the post process.
 
Marc Wielage said:
The only problem with that is: it still doesn't resolve the issue of calibrating monitors on location, in the edit bay, and in the final color-correction room. It also won't confirm that accurate color is being displayed on location -- only that the whites and blacks aren't clipping.

We have non-trivial problems in post just getting well-adjusted monitors to look similar -- say, a broadcast CRT, a high-end plasma, and a D-cinema projector. Trying to do this with a $2000 LCD display on location is a tough challenge. It's a concern when DPs are making critical changes to LUTs on location, and expecting this look to translate throughout the post process.
Click to expand...

You are right about this. Different display technologies don't track or display color exactly alike. Calibration for white balance and gray scale with a good colorimeter and proper software can get you close though. Not something that can be casually done by eye with test charts.
A new 1080p DLP projector I recently installed offered the following color gamut choices.

CRT, REC709, SMPTE-C, EBU, and native.

Most displays don't give you a choice and you may not even know what its mapped to. It is not common to see it in the specs.
 
If we're talking digital video we can't talk IRE. IRE is old analogue video and irrelevant to digital video. I know such terminology is still used, but it's wrong. Talk code values for digital video.

Digital video never had, never has and never will have "setup" or 7.5IRE, not least because IRE is analogue and digital is digital, and because digital has a spec based on code values that is utterly the same the world over. In 8bit terms black is a 16, white at 235, some of the code values beyond those limits are "reserved" and the rest are for super-blacks and super-whites.

Finally, REDColor and REDGamma are utterly REC709 compatible.

Graeme
 
Graeme Nattress said:
If we're talking digital video we can't talk IRE. IRE is old analogue video and irrelevant to digital video. I know such terminology is still used, but it's wrong. Talk code values for digital video.

Digital video never had, never has and never will have "setup" or 7.5IRE, not least because IRE is analogue and digital is digital, and because digital has a spec based on code values that is utterly the same the world over. In 8bit terms black is a 16, white at 235, some of the code values beyond those limits are "reserved" and the rest are for super-blacks and super-whites.

Finally, REDColor and REDGamma are utterly REC709 compatible.

Graeme
Click to expand...

Agreed, but the hardware and software scopes used for displaying the range of signal values for calibration are still IRE referenced to those digital values, and not always correctly. IRE still represents the analog logarithmic value scale for the image that we record and ultimately see.
 
http://en.wikipedia.org/wiki/IRE_(unit)

Digital scopes should be in code values. Composite analogue voltages expressed as a percentage make no sense. It might be what is used, but it doesn't make it right unless you're measuring a composite output to check that it's in spec.

Graeme
 
Hi Graeme, thanks for joining this discussion.

What does "REC709 compatible" mean? Does it mean that REDColor is mapping the camera's output to a standard REC709 color space in the same way that ColorSync maps the output of Photoshop to the profile my printer uses? Or is something different at play here?
 
REDColor maps the camera output so that the output colours are compatible with viewing under a REC709 monitor.

Graeme
 
Graeme Nattress said:
http://en.wikipedia.org/wiki/IRE_(unit)

Digital scopes should be in code values. Composite analogue voltages expressed as a percentage make no sense. It might be what is used, but it doesn't make it right unless you're measuring a composite output to check that it's in spec.

Graeme
Click to expand...

But its not about what should be used its about what is being used at the moment, we can discuss what the correct terminology is but scopes still show IRE and broadcasters still define deliverables in IRE.

I understand where your coming from but RED seem to have an attitude of telling people what they should be using as opposed to what is being used which I don't understand.

Just try and work with people as opposed to against them and there wont be an issue here with it..
 
Color and REC709

Color and REC709

Well O.K, forget IRE for the moment, as that's not the main issue.

REC709 calls out a specific gamut (color space) and gamma (tonal response curve) - but note these were designed for broadcast systems.

So that's what we followed with the REC709 choice in Build 21. Unfortunately, that doesn't look too good - skin tones and highlights both suffer.

So we adjusted the code to better fit the reality of mapping our sensor to a REC709 monitor's boundaries, and called that REDcolor.

So you can think of REDcolor as a REC 709 LUT if you like, it is 100% compatible.


Now, that said what is the perceived issues with IRE?

Because, if you select SMPTE Bars as your camera test signal output you can see that it aligns with monitor, waveform monitor or vectorscope, and if you then switch to output video and adjust either aperture or ISO you can see the the peak whites do not violate those maximum IRE levels ..
 
Marc Wielage said:
It's a concern when DPs are making critical changes to LUTs on location, and expecting this look to translate throughout the post process.
Click to expand...

Anyone making "critical changes to LUTs on location" is not being realistic. Viewing conditions on location are never appropriate for final color decisions, regardless of how much people want to believe they can be. The terms "critical color" and "location" don't belong together.
 
OK - anecdote time....

When I moved to Canada I had to learn about NTSC, 7.5IRE setup and all that stuff, and also about how digital video worked. I came from the UK which is PAL and doesn't have any of that setup nonsense.

What I learned was up-to-date knowledge of the specs of how digital video worked, and how digital video doesn't have the "setup", and that when dubbing from NTSC analogue to digital, you had to remove the setup so that blacks lined up correct for the digital spec, and when outputting your digital video to analogue NTSC, you had to ensure setup was added in again. I also learned that Japan had dropped the use of setup in their NTSC way back when as not only was it no longer useful, it was detrimental to the image quality.

So when I asked a post services house here in Canada to dub some 1" to DV for me, I was shocked when they didn't remove the setup and hence the black levels on the DV were wrong.

I also learned that some broadcasters had set their digital cameras to add a "digital setup" in the digital domain, effectively "fixing" an issue elsewhere in their chain, but making each and every digital tape they recorded incompatible with established standards.

OK - what's that got to do with above.... Someone "knows" blacks go at 7.5IRE, and hence, even though it's digital and digital doesn't have setup, sticks their digital camera output on a scope and twiddles things until it's "right", but we all know that now blacks are not 16 in the digital codes and the tapes are wrong. Using poor terminology is part of what lead to that.

Graeme
 
Stuart English said:
Now, that said what is the perceived issues with IRE?
Click to expand...

Graeme seems to be overlooking the fact that nearly all monitoring today is done in HD formats, which don't and never did have the issues he's bringing up.
 
Stuart English said:
REC709 calls out a specific gamut (color space) and gamma (tonal response curve) - but note these were designed for broadcast systems.
...
So we adjusted the code to better fit the reality of mapping our sensor to a REC709 monitor's boundaries, and called that REDcolor.

So you can think of REDcolor as a REC 709 LUT if you like, it is 100% compatible.
Click to expand...

Thank you Stuart. That's exactly what I wanted to know. (and what I was hoping was the case.)
 
I think there's just an internet barrier here with getting points across, Stewart thats good info about the RedColor info thanks for that.

I still think that although its a better system people would like an option/LUT that brings back the old responses as they were better suited to some peoples applications and perceptions and experience...

I think this is a case of Red going too far in their improvements for some people working in confined productions and workflows. I am still undecided if thats a good or bad thing.

Its important to remember while being a digital cinema camera people are using it for broadcast productions.
 
Jay Gannon said:
I think there's just an internet barrier here with getting points across, Stewart thats good info about the RedColor info thanks for that.

I still think that although its a better system people would like an option/LUT that brings back the old responses as they were better suited to some peoples applications and perceptions and experience...

I think this is a case of Red going too far in their improvements for some people working in confined productions and workflows. I am still undecided if thats a good or bad thing.

Its important to remember while being a digital cinema camera people are using it for broadcast productions.
Click to expand...

OK - what (specifically) is unsuitable as things are now for broadcast production use?

Graeme
 
Broadcast use

Broadcast use

I still think that although its a better system people would like an option/LUT that brings back the old responses as they were better suited to some peoples applications and perceptions and experience...

Its important to remember while being a digital cinema camera people are using it for broadcast productions.
Click to expand...



Let's keep going down this path Jay; Graeme you beat me to this question :-)

Why would it be better to use the original "REC709" selection for broadcast uses ?


scopes still show IRE and broadcasters still define deliverables in IRE.
Click to expand...

Are IRE levels for reference exposures / levels the concern, and if so can you amplify on that. i.e. the specific problem is perceived to be ?
 
Last edited:
No the IRE was a tangent. Not really important.

The REC709 (Old) was the closest monitoring that we had to a finished REC709 deliverable, this was a good thing as when you had a client on set it gave them the look of roughly what the finished product for broadcast was going to be.

Yes it was not the best for actually judging exposure but it was very pleasing to the client to be able to look at something that they found pleasing and a good approximation of what the final product was going to be.

I am not saying that the current system is not usable I am saying that the old system was easier to work with in my own situations, I don't profess to speak for others.
 
You must log in or register to reply here.
Back
Top