FEATURE REQUEST: Ed Lachman Zone (EL Zone) Exposure System

As professional cinematographers out here using Blackmagic cameras to shoot high-end content, it’d be a great addition to have the integration of EL Zone Exposure System (Ed Lachman Zone). A quick recap:

EL Zone is an exposure tool that is based on stop values that relate one-to-one with lenses and light meters. It is similar to false color, however, it allows for higher precision, greater consistency, and more intuitive on-set communication. Source: NewsShooter By Matthew Allard ACS

As Blackmagic has been quite ahead in terms of cinematography tools, I don’t see any reason why it shouldn’t be integrated in Blackmagic cameras.
Sam

+1

+1

Should be simple enough to do this with a LUT, and can customize too.

Being able to see stops, as this system allows, would be a great help. False color is good, but a bit too crude.....

+1

I like this so much better than False Color (especially since False Color on one system can be completely different than on another), wish every cam, monitor ect would just universally use this.

https://www.fdtimes.com/2021/04/25/el-z ... chman-asc/

I’m not so sure about recommending an exposure guide if it’s based on full stop variations. Please correct me if I’m wrong but in the current BMD False Colour the colours are not solid but represent a shading within the colour. The centre of the individual False Colour is solid but the edges blend somewhat with the colours that bound it.

And the steps between the solid colours are not exactly a full stop. For example, the green is based on the mid-tone and the pink we are told is one stop above the green. However there’s a medium grey between those colours representing perhaps a half-stop graduation.

[The proposal has ½ stop above and below middle grey.]

I don’t think we know how the dark grey and light grey progress in terms of stops but they’re both areas that give a good exposure. And we don’t know if blue and yellow are precisely a stop before potential ‘clipping’ with purple and red.

And red is only a warning that clipping may occur especially as you progress to solid red. If you are just barely into the red, you still may not be clipping. So the BMD False Colour has a lot to recommend it. Not if you’re only interested in full stop transitions to repeat a lighting setup, but if you are experienced with working with its gradual transitions, it’s very efficacious. The Pocket 4K/6K cameras have the added advantage of those three little boxes that warn of clipping in each RGB channel and that’s even more useful than False Colour red.

Edit
Revised to reflect that the proposal includes ½ stop variations above and below middle grey. Pardon my error.

+1
As a long time (45+ years) large format photographer, I would find this a great advantage over the brain damaged false color (IRE) system. IRE was based on a I Kid You Not six-stop exposure range (the limit at the time of the crappy vidicon tube.) RGB was right out. Zone system (as developed by Ansel Adams) gave photographers a way to place the values of concern in a photograph on a scale familiar to all.

So Ed has gifted the world with a realistic way to evaluate exposure ON the screen, and get it right at the start.
I'd happily pay to put EL Zone on my BMPCC 4K, even if I had to take out the ProRES.

(written on a Mac )

For reference, this is the EL zone scale

EL Zones.jpg (47.5 KiB) Viewed 9356 times

The color coding has a number of advantages over false color:

- it's built around middle grey
- middle gray is gray
- the color scale is clear yet agnostic, rather than the misguided design in false color that tries to tell the user where skin tones should be exposed

EDIT: And as has Hook and others have noted, as I should have done here, methods for determining sensor clipping are not part of EL zones. I didn't mean for this to come across as that I meant EL zones was the only tool, or even the best tool, with which to judge exposure.

+1


Sent from my iPhone using Tapatalk Pro

+1

https://creativityinnovationsuccess.wor ... chman-asc/

+1

+1

+1

Should also be added to the VA's

+1

EL Zone is an exposure tool that is based on stop values that relate one-to-one with lenses and light meters. It is similar to false color, however, it allows for higher precision, greater consistency, and more intuitive on-set communication. Source: NewsShooter By Matthew Allard ACS

Matthew Allard’s articles on Newsshooter are some of the most comprehensive and insightful ones I’ve come across. I often find myself pretty much concurring on his many schools of thought, including voicing perplexity as to BMD not including OLPFs in their cameras:

https://www.newsshooter.com/2020/07/17/ ... -thoughts/

Now that the URSA Mini Pro 12K OLPF exists, I think BMD should consider having Allard on board as a consultant for implementing better design and feature ideas in the future in order to minimize the time it takes to finally realise them.

+1

+1

+1

The EL system starts off great, use color to delineate stops in a scene referred environment. Where it goes awry is by using the rainbow, 450nm-750nm, as the basis for the color scheme. The same way IRE ignores stops in scene referred, the EL system ignores color luma values when display referred. An ideal system would pair colors with similar luma to the source. The image would transition naturally from dark to light rather than being tiger striped.

Good Luck

+1 here too.

Does anyone know if implementing this would be a licensing issue for Blackmagic?

I don't know if the patent was granted, but I believe it was applied for. It's not a public domain thing. And since third-party monitors would need a set of references for every camera sensor, it's not likely to become universal.

… An ideal system would pair colors with similar luma to the source. The image would transition naturally from dark to light rather than being tiger striped.

Good Luck

Worth a try at least to test if the stops would be easily distinguishable especially if used outdoors.

But even the proposed system may be difficult to actually work when a few of those adjoining shades of the same colour will be difficult to assess quickly outside a studio. BMD’s False Colour has already been criticized for the difficulty in determining the smooth grey graduations where the medium grey tones between green and pink are difficult to discern from the darker grey below the green. And the long light grey above the pink is also difficult to know where exactly you are on the scale.

I don't know if the patent was granted, but I believe it was applied for. It's not a public domain thing. And since third-party monitors would need a set of references for every camera sensor, it's not likely to become universal.

External monitors wouldn't need exposure values per sensor, just per log transfer function. Those values are published. You can find virtually all of them helpfully listed here:
https://antlerpost.com/colour-spaces/

+1

I already sent this feature request a year ago to a BM representative I met and I am happy I'm not alone with this. If of interest for anybody, here's the whole feature request list, I sent him:

1. More Anamorphic Desqueeze factors
– more preset ratios, e.g. 1.42 (Isco), 1.5, 1.75, 1.79 (Caldwell Chamelion), 1.8 (Vazen)
– custom ratios (like custom crop marks, where the user can dial in the desired desqueeze factor)

2. Crop & Scale function (like in SmallHD monitors)
– add Crop & Scale function (like in SmallHD monitors): https://guide.smallhd.com/a/814247
– if done in SmallHD monitors, the on-screen information from Blackmagic Camera is hidden behind the bars (e.g. Timecode, Cards, ISO, WB, etc.)
– therefore Crop & Scale in-camera is needed, to preview on external monitors without hiding on-screen information

3. Image overlay
– option to overlay a still image file (with support of alpha channel for transparency) from SD card
– for example to match shots
– or to have a logo as a preview overlay to be able to frame image correctly for post production, where logo will be added

4. On-screen record button appearance like in pre-7.9 firmware
– old appearance of standby record button (no red color) like in firmware versions before 7.9 to better distinguish between recording and standby.
– this is wished by many users in the forum: viewtopic.php?f=2&t=162668&start=100#p864452 or: viewtopic.php?f=2&t=164806#p869953

5. On-screen Detail Sharpening indicator
– like with LUTs and ND-Filters there should be a small icon on-screen that reminds the user that Detail Sharpening is active

6. Different False Color modes
– option to choose from different False Color modes like in Davinci Resolve
–e.g. ARRI ALEXA false color or “EL Zone”:

External monitors wouldn't need exposure values per sensor, just per log transfer function. Those values are published. You can find virtually all of them helpfully listed here:
https://antlerpost.com/colour-spaces/

Lachman himself isn't that hopeful. Here's what he said about it: "It has to be calibrated specifically to each specific camera sensor’s gamma and latitude, so it would only work if the monitors reflect the different camera’s sensors and settings which I’m not sure would be possible.

Possible or not, it still represents a level of integration which may not be easy to attain, and that assumes monitor manufacturers see money in it.

Screw monitors, just put it in the cameras already.

External monitors wouldn't need exposure values per sensor, just per log transfer function. Those values are published. You can find virtually all of them helpfully listed here:
https://antlerpost.com/colour-spaces/

Lachman himself isn't that hopeful. Here's what he said about it: "It has to be calibrated specifically to each specific camera sensor’s gamma and latitude, so it would only work if the monitors reflect the different camera’s sensors and settings which I’m not sure would be possible.

Possible or not, it still represents a level of integration which may not be easy to attain, and that assumes monitor manufacturers see money in it.

I wonder how companies like ColourlabAI implements it then. There you just select the camera mfg you are using and it selects the right one to use.

External monitors wouldn't need exposure values per sensor, just per log transfer function. Those values are published. You can find virtually all of them helpfully listed here:
https://antlerpost.com/colour-spaces/

Lachman himself isn't that hopeful. Here's what he said about it: "It has to be calibrated specifically to each specific camera sensor’s gamma and latitude, so it would only work if the monitors reflect the different camera’s sensors and settings which I’m not sure would be possible.

Possible or not, it still represents a level of integration which may not be easy to attain, and that assumes monitor manufacturers see money in it.

Yes, ideally it should be implemented in camera, for the best result. But if the user was able to enter the log gamma curve + ISO setting into the external monitor, that should be enough information to display a useful EL zone scale on screen from the camera signal.

In any case, this thread is asking BMD to implement EL zones in their cameras : )

[CaptainHook]

I have some thoughts/questions and please don't take this as any inference to intentions (or not) of adding this.

Say the camera/ISO has more than 6 stops over middle grey, do you just show white/black and ignore that you have might more range available? Or say the camera has less than 6 stops at a particular ISO, if the max shown was an orange how would the user know they have clipped the sensor versus have more range available? Remember that on just one of our cameras depending on your ISO you may have something like ~3 stops over middle grey or ~8. We also have our video modes to consider.

One option might be to show the scale with only the available colours for your current camera/settings. But this might also seem to go against the idea of a 'consistent scale' that is EL and is IMHO one area where false colour is still useful/intuitive, you not only have mid grey reference (and another stop over) but clipping points at each end that are always scaled for the camera/ISO.

Personally I feel using just EL Zone I would still need another exposure tool to indicate clipping, given the varying range our cameras/ISOs have.

Again, please dont read this as any indication about this being implemented by us or not. Just some things I dont see being discussed.

IMHO, a clipping indicator should exist independently of such a false color system.
I don't know if Red has patented even that, but the little 'traffic light' will show you which of the color channels is actually clipping. Something like that would be OK. Depending on the color of light, you may have a half or two/thirds of a stop to use by Highlights Recovery.

[CaptainHook]

We have highlight clipping indicators on histogram, still not as useful IMHO as the two colours used in false colour at each end (4 in total). But everyone has preferences and this is another reason why this stuff is never that straight forward.

clipping.png (252.83 KiB) Viewed 8643 times

I thought I've seen it somewhere!
But I have not operated a recent BM camera in a while. On my sturdy, old UMP 4.6K with firmware 6.9.4
the three dots on the right of the histogram don't do anything

Maybe you are always over exposing a little

Maybe (and I noticed the smiley). I'm still old-school enough to use a light meter.

But, seriously, I'm afraid that feature really doesn't exist in 6.9.4, maybe the three dots of two black pixels that don't do anything were in preparation for it. If somebody would join to explain where it can be activated, I'd be forever grateful. Or, at least as long as that camera lives.

Say the camera/ISO has more than 6 stops over middle grey, do you just show white/black and ignore that you have might more range available? Or say the camera has less than 6 stops at a particular ISO, if the max shown was an orange how would the user know they have clipped the sensor versus have more range available? Remember that on just one of our cameras depending on your ISO you may have something like ~3 stops over middle grey or ~8.

Oh wow. Good point! So what's the answer?

What I like so much better about the EL system is the intuitive use of color compared to False Color, where they are all over the place and you have to learn which color means what. On top, different manufacturers use different color codes.

You may want to have a look at this: timeinpixels.com/false-color-plugin/

Say the camera/ISO has more than 6 stops over middle grey, do you just show white/black and ignore that you have might more range available? Or say the camera has less than 6 stops at a particular ISO, if the max shown was an orange how would the user know they have clipped the sensor versus have more range available? Remember that on just one of our cameras depending on your ISO you may have something like ~3 stops over middle grey or ~8. We also have our video modes to consider.

If you're mapping the colors into the space it seems pretty straightforward. ISO is definitely where it gets tricky, I think you'd probably need a 1D LUT for each ISO. I guess you just have to know your camera and it's over/under distribution.

Here's something that I was working on while trying to figure out the native ISO for the Komodo. The plot is 11 over/11 under in Log3G10. I Just drew the EL colors on the scope. Above mid grey the distribution of stops is pretty even, below that is a very different story.

This is at the theoretical "native" ISO but you can see how when shooting a lower ISO the EL system wouldn't be super helpful as the zone system taps out (white) before the sensor is clipped. In fairness though, the EL system isn't about ETTR.

Good Luck

Log3G10.jpg (58.5 KiB) Viewed 8530 times

You may want to have a look at this: timeinpixels.com/false-color-plugin/

Oh wow, I followed that guy 10 years ago for his After Effects tutorials and scripts. Great to see him still around and thanks for the tip Uli, looks perfect.

His Omniscope is great.

Personally I feel using just EL Zone I would still need another exposure tool to indicate clipping, given the varying range our cameras/ISOs have.

Fair point.

However, if there’s one idea that must be borrowed from the EL Zone – if the EL Zone itself cannot be practically tailored to BMD cameras – is that 18% middle grey NEEDS to be grey.

The green of 18% middle grey is the one aspect of the IRE scale of which I’m personally not a huge fan. On the other hand, overexposure in the EL Zone can be flipped to accommodate the red and yellow of 95% white and 80% near-white clipping, instead of just plain white.

Maybe someone needs to reconfigure the IRE scale and hybridize it with the core concept of the EL Zone to create something new.

The tool I referenced above can do this:
"Export your fine-tuned False Color overlay as 3D LUT and use in your camera monitor on set!"

I have some thoughts/questions and please don't take this as any inference to intentions (or not) of adding this.

Say the camera/ISO has more than 6 stops over middle grey, do you just show white/black and ignore that you have might more range available? Or say the camera has less than 6 stops at a particular ISO, if the max shown was an orange how would the user know they have clipped the sensor versus have more range available? Remember that on just one of our cameras depending on your ISO you may have something like ~3 stops over middle grey or ~8. We also have our video modes to consider.

One option might be to show the scale with only the available colours for your current camera/settings. But this might also seem to go against the idea of a 'consistent scale' that is EL and is IMHO one area where false colour is still useful/intuitive, you not only have mid grey reference (and another stop over) but clipping points at each end that are always scaled for the camera/ISO.

Personally I feel using just EL Zone I would still need another exposure tool to indicate clipping, given the varying range our cameras/ISOs have.

Again, please dont read this as any indication about this being implemented by us or not. Just some things I dont see being discussed.

I found an article that deals with your question. Here is the answer Ed Lachman gave regarding it:

“Now, it’s true that manufacturers may well produce cameras with more than six stops latitude from 18-percent grey. But that isn’t really a problem with this system, because 18-percent grey remains the most important part of the exposure, and that is largely because that is what will affect how the face looks the most.
“When you move out to extremities of the body, therefore, you can just make those two-stop values – minus four can represent two stops, and so on. The reason I felt there was no point in going further is that, if you have too many colours in the viewfinder, it can get confusing. So, the cinematographer can just designate and say the important area is where I will lose or keep detail, and they can calculate from there.”

You can find the entire article here: https://www.cinematography.world/exposure-control/

Great find!

EL: "The reason I felt there was no point in going further is that, if you have too many colours in the viewfinder, it can get confusing."

I agree with him.

EL: “When you move out to extremities of the body, therefore, you can just make those two-stop values – minus four can represent two stops, and so on."

I don't quite understand, can someone please explain?

... one area where false colour is still useful/intuitive, you not only have mid grey reference (and another stop over) but clipping points at each end that are always scaled for the camera/ISO....

Yes, and this works perfectly fine. I have a 7" monitor for viewing and focusing and use the small viewer of the URSA 12K with false color (same with the Pocket 4K, when using an external monitor.) I find the most important information to be that of clipping highlights, second to be that of noise floor, and to a much lesser extent third where mid grey might be.

I spent decades spot metering while exposing negatives for platinum/palladium printmaking. And with large format films costing plenty for the amount of photographs I was shooting, I would be using one film per image. The spot meter would help me determine the development to get the contrast I wanted. As to exposure, at first it was useful but over time with experience I had my aperture and shutter speed already set in my mind before putting the lens onto the view camera.

Same now with my BMD cameras... I have used them enough and have worked with Resolve enough that exposure is the least of my concerns. And I certainly do not need any other distractions to take away from my getting the moment captured. Simpler the better... and the current False color is simple and does the job.

I mostly shoot in natural light... but if using lights (and I have done that some) I will use a meter if I want a certain ratio. And I certainly would not rely on a zone wide block of false color (or even half zone) to expect any accuracy out of a ratio calculation. (Fill could be on one side on the block and main on the other side in effect adding almost two blocks (at least a full stop) of error.

I guess I would rather the efforts of BMD go into other improvements of which there are many (and catch ups, like I would rather the 3 color histogram with clipping indicators on the 12K.) The current False color works great... don't fix what ain't broke.

I have some thoughts/questions and please don't take this as any inference to intentions (or not) of adding this.

Say the camera/ISO has more than 6 stops over middle grey, do you just show white/black and ignore that you have might more range available? Or say the camera has less than 6 stops at a particular ISO, if the max shown was an orange how would the user know they have clipped the sensor versus have more range available? Remember that on just one of our cameras depending on your ISO you may have something like ~3 stops over middle grey or ~8. We also have our video modes to consider.
*snip*
Again, please dont read this as any indication about this being implemented by us or not. Just some things I dont see being discussed.

Keep in mind, this is more of a system to nail down Skin Tones than absolute exposure. One could use zebras to show the absolute clipping.

Check this article out by Matthew Allard ACS on News Shooter it talks on some of about many of your assertions, he also tests it with an Arri Alexa 35.

https://www.newsshooter.com/2023/03/06/ ... munication.

EL: “When you move out to extremities of the body, therefore, you can just make those two-stop values – minus four can represent two stops, and so on."

I don't quite understand, can someone please explain?

It sounds like he expects shooters to recalculate the value of the stops at the "extremities", for cameras which don't conform to the existing scale. So in his example, -4 isn't two stops away from crushed blacks, it's more like four. But since -6 and +6 are no longer absolute measures, there's no way to see, graphically, when the sensor is clipping, without some additional indicator on the screen. And recalculating the value of stops at the edges becomes more complicated as ISO values change the distribution of stops above and below middle grey.

I think, anyway.....

The green of 18% middle grey is the one aspect of the IRE scale of which I’m personally not a huge fan. On the other hand, overexposure in the EL Zone can be flipped to accommodate the red and yellow of 95% white and 80% near-white clipping, instead of just plain white.

I think the big difference is that false color is at least a somewhat manageable shooing mode. Shades of grey represents the tonality of the image and the colors are basically a six point zebra. False color was never intended to be a spotmeter. Most of the image using false color will be green, pink, and grey.

The EL system by contrast is completely disconnected from the tonality of the source image except at middle grey. I'm looking for a system that supports the tonality with the color.

I use waveform monitors with stop scales. Instead of being delineated 10-20-30... they're...24%, 50%, 74%, 102% , in 709. IRE directly correlates to data, and voltage, 100IRE=700mv=940(10bit). I don't think there's an IRE scale per se, Resolve definitely doesn't think so, it changes every couple years.

Good Luck

Good Luck

With CaptainHook’s observations, Ed Lackman’s scale is a bit of a PITA. Be honest with yourself, if BMD had originally implemented this scheme in 2013, there would be ten years of complaints about it too.

The proposal suggests stops, but it’s not stops given even the default proposal has half-stops bounding middle grey. With Ed’s suggestion to consider the colours near the end points as two stops. having to remember whatever variations there are is something extra to remember and apply to a given sensor for a given ISO. The point is that sometimes the maximum stops above middle grey may only be 3 stops and could be 8 stops as an example. Stops below middle grey have similar variations.

I’m quite comfortable with the False Colour we have that should have the three traffic lights indicating true clipping in each channel as found on the Pocket cameras. I already know the False Colour display varies as you change ISO. I’ve seen it in my controlled tests where something may be slightly red at one ISO and only yellow at another ISO. Sure I don’t see how the actual stops change. If that’s important, as it well may be within a scene with several shots at different angles, in a controlled shoot, one can rely on an exposure meter.

Luckily all I care about is keeping a face in that green-grey-pink-light grey range.

The excellent detailed RGB histogram on the BMVA12G7 is very helpful and the traffic lights of the BMPCC4K etc ensures I know if channel clipping exists. The waveform shows me (by inference) what is clipping in the frame and the histogram indicates how prevalent the clip is. I often have to identify something is clipping in a specular reflection or because I’m looking at a LED source. I don’t worry about those things… we’ll okay, the LEDs are irritating if I can’t edit them out of the frame in post.