How the Alexa really works

[Tom Yanowitz]

One guy I enjoy a lot is John Seale.

For Mad Max, he actually flat out admits he's no digital expert and thus trust completely DITs, VFX guys and colorists.

 

Here, they discuss about the proposition by Andrew Jackson to overexpose day for night by 2,3,..,4 stops with the Alexa.

The VFX supervisor decides the F-stop now! I mean it's great.. but strange.

Edited March 7, 2016 by Tom Yanowitz

[Stuart Brereton]

An other huge reason for overly underexposed footage (I'm talking to the point where the DIT or DP are fired once people see the dailies) is trusting the monitor for exposure.

 

 

You guys trust your eye way too much.

 

I use my monitor to judge exposure all the time. It's properly calibrated, and i know what I'm looking at. My exposures seem to come out exactly where I want them.

 

As for trusting our eyes, well Cinematographers have been doing that for generations and it seems to have worked out ok. Some of the greats didn't even use light meters, trusting solely in their eyes.

 

You've created a chart based on information which is fairly basic to anyone with more than a passing interest in digital cinematography, and seem to be passing it off not only as new research, but also as some kind of secret("why doesn't Arri share this info?")

 

You then accuse other cinematographers, up to and including Roger Deakins and Emmanuel Lubezki, of either not knowing or not caring about the characteristic responses of their chosen camera system.

 

Finally, you call anyone that disagrees with you 'anti-technique' and say that they are 'looking down' on the technology.

 

You are coming across as very arrogant, which is hardly an approach that's likely to endear you to other members, particularly those with far more experience than you. I always thought the purpose of being a student was to learn from others, not to rashly tell them that they are wrong.

[Satsuki Murashige]

You seem to be very afraid of white clipping, as everyone should, but black clipping is actually the more common problem isn't it. With footage having no details whatsoever in the shadows.

 

What you are missing is that the general approach most cinematographers who came up shooting film take to arrive at a well-exposed negative is to expose to hold highlights and then add light to bring up midtones and shadows where we want them to be. Where we want black, we simply let the shadows go. In such a working method, extreme highlights are much more difficult to control than shadows, hence the need to protect them from clipping. When shooting and printing film, we never expected to see detail in shadows darker than 3 stops under middle grey. So we had to light for that limitation. However, we expected highlights to retain enough detail to roll off gradually to white 4 stops over.

 

I say 'most cinematographers' because some like Harris Savides were famous for playing in the toe of the characteristic curve, but that was very atypical for the era.

 

The digital paradigm is obviously different. But a lot of cinematographers still follow the same old-school approach. Many who follow the new-school rely instead on the dynamic range of the recording medium and push under-exposure to its limits, much like Savides did. What made Savides's work so bold is that he worked on the edge of an acceptable image while not being able to see the result until the next day. Nowadays when you can actually see the image you are recording live, others are emboldened to follow his lead, sometimes with disastrous results. Other times in the hands of a skilled DP, the results are stunning.

 

I would submit that it's actually the flexibility of digital post-production which has allowed a greater degree of exposure 'adventures' to become acceptable. Cinematographers who shot regularly for broadcast back in the day had to be even more precise with their exposures because there was often no post-production grading available - 'what-you-see-is-what-you-get,' or WYSIWYG.

 

The bottom line is that if there are exposure mistakes, it is the cinematographer's responsibility. You also have to account for individual taste.

[Tom Yanowitz]

 

I use my monitor to judge exposure all the time. It's properly calibrated, and i know what I'm looking at. My exposures seem to come out exactly where I want them.

 

As for trusting our eyes, well Cinematographers have been doing that for generations and it seems to have worked out ok. Some of the greats didn't even use light meters, trusting solely in their eyes.

 

 

We mostly agree here. But I'm pretty sure the "eye only" DPs had a pretty in depth knowledge about the film stock they were using.

 

 

 

You've created a chart based on information which is fairly basic to anyone with more than a passing interest in digital cinematography, and seem to be passing it off not only as new research, but also as some kind of secret("why doesn't Arri share this info?")

 

First time I've seen these curves online is here, with this topic. And I'v looked for it a lot before deciding to do it myself.

I'd say 5% of Alexa DPs may know about the LogC curves, maybe 0.01% know about the DGA and Arriraw ones, hence why I kindly shared them.

To get the Arriraw curve you actually have to get the algorithm in some 150$ smpte pdf and then program it to visualize it. So you gotta be pretty obsessive about the topic to do that stuff(which I am).

Or you can do it by exposing a dynamic range chart, but if you watch it on

- a waveform monitor

- an Arri software

- any software that accept Arriraw

You'll have the LogC curve instead, hiding the actual curve of what's been recorded.

 

 

 

You then accuse other cinematographers, up to and including Roger Deakins and Emmanuel Lubezki, of either not knowing or not caring about the characteristic responses of their chosen camera system.

 

Finally, you call anyone that disagrees with you 'anti-technique' and say that they are 'looking down' on the technology.

 

You are coming across as very arrogant, which is hardly an approach that's likely to endear you to other members, particularly those with far more experience than you. I always thought the purpose of being a student was to learn from others, not to rashly tell them that they are wrong.

 

Your progression ends once you cease to see yourself as a student.

We should all be students until we die, if not what's the point.

That's probably how Deakins and Lubezki think and what made them so great in the first place. I don't think they would reject someone's idea or information just because it's given by some beginner student.

I've done a few masterclass/workshop and talked to DPs (well known or not) and every-time I'm thinking "if only he would let me give back some information about the things I know, so that I'm not the only one getting something out of the conversation".

So go ahead, ignore me, and don't progress, while I keep absorbing what every DP and his mother does. I just learned a lot a minute ago thanks to Satsuki Murashige and his great post. :wub:

 

As for "people that disagree with me", I highly respect them, but no one (here at least) has yet disagreed on the actual facts.

 

If anyone want to discuss the actual content of the chart, I'm here :ph34r:

Edited March 7, 2016 by Tom Yanowitz

[Tom Yanowitz]

To follow on the debunking of the myth that every big time DP understands the Alexa, I just randomly found an article on "Race".

Peter Levy talking

 

 

 

“In the past I’ve been very reluctant to deviate from the camera’s native 800 ISO,

 

Ok, let's pass on that.

“but I found when I knocked it down to 400 and 200 it actually changed the contrast of the image, [such] that the 200 ISO is a lot less contrasty than the 800. I think the Alexa at 200 ISO is more like a pull process; it’s a softer image. That’s a revelation I discovered on Race and a technique I’ll use again. I used it on all my daylight scenes and whenever I had stop to spare — much to my focus-pullers’ chagrin. But I had a wonderful first AC from Toronto, Kerry Smart, and she managed to run a very tight ship under very unforgiving circumstances — and I don’t think we had a soft shot in the whole movie.”

 

A lot less contrasty at 200 ?

When people talk about contrasty images, do they mean crushed blacks ? In that case yeah, I guess it'll be less contrasty at 200.

But as for the actual definition of contrast for an image here (difference in levels between parts of the image), then 200 is more contrasty because you avoid the tow of the curve more.

Again, the only thing that changes by going from 800 to 200 is the f-stops that you probably open by two stops.

That's it that's the only thing that changes, no camera sorcery as far as what's being recorded goes.

So I don't know about inherent characteristics changes like 'softness' (yeah sure there's less depth of field I guess).

"Between the Truelight system and a waveform monitor, I’m very confident about my image and getting what I want. I’ve pretty much forsaken the use of my beloved light meters these days — I find that monitoring the image on a waveform monitor reading the raw file gives me a very accurate picture of what my ‘negative’ looks like, and allows me to light by eye."

 

The cake is a lie.

Hate to be nitpicking but waveform monitors don't show the raw data (I wish).

Edited March 7, 2016 by Tom Yanowitz

[Satsuki Murashige]

But how does the actual movie look? Isn't that what actually matters?

[Tom Yanowitz]

But how does the actual movie look? Isn't that what actually matters?

 

Sure it's what matters.

But thinking like this for doesn't help for learning and improvement I think.

 

When I first bought a DLSR, I thought pictures taken in Automatic mode looked great.

Then I learned about aperture and exposure time, the possibilities got wider, the pictures got better.

 

Every DP that makes a good looking film can't be shielded from fact checking. Because there's going to be students reading or listening to them and believing every word.

The "problem" with the Alexa is its greatness, it's so forgiving you can't really screw up in a definite way, so it's the worst camera to learn digital cinematography and its limitations.

It's a lot trickier on the lower end.

 

Most DPs say the same thing "the camera you choose doesn't matter", but always end up choosing the one that will cost the production the most, the Alexa.

Edited March 7, 2016 by Tom Yanowitz

[Robin R Probyn]

"So go ahead, ignore me, and don't progress"

 

hehe now come on Tom.. this is a wind up post isnt it.. :)

[Stuart Brereton]

I've done a few masterclass/workshop and talked to DPs (well known or not) and every-time I'm thinking "if only he would let me give back some information about the things I know, so that I'm not the only one getting something out of the conversation".

 

 

“When I was a boy of 14, my father was so ignorant I could hardly stand to have the old man around. But when I got to be 21, I was astonished at how much the old man had learned in seven years.”

 

Mark Twain.

[David Mullen ASC]

The problem with rating the Alexa at 200 ISO is that instead of 7-stops over and under, you have only 5-stops of overexposure detail but 9-stops of shadow information... And few cinematographers complain that the problem with the Alexa, or most digital cameras, is a lack of shadow detail. Film has more latitude for overexposure but less for underexposure, so if your goal is a film look, the last thing you'd want to do is trade overexposure detail in favor of more shadow detail when you already have more flexibility to bring up shadows with most digital cameras compared to film. Not to mention, if you ever have to make an HDR master from something shot on the Alexa, you are going to want that extra highlight information.

 

One of the most telltale signs that something is shot on a digital camera is how and when thing clip in the highlights. Often the amount of shadow detail is controlled by lighting, but it is harder to use lighting to control naturally hot areas of the scene.

 

Most people find that 7-stops of shadow information with the Alexa at 800 ISO are already enough. Certainly lower ISO's are cleaner but it's the same scenario in one way with film, you'd get less grain if you shot on low ISO stock but that may limit you creatively. And where it is not the same is that all the different speed stocks have the same gamma, you don't trade highlights for shadows by choosing a slow speed stock.

[Satsuki Murashige]

 

Sure it's what matters.

But thinking like this for doesn't help for learning and improvement I think.

 

I think the difference of opinion we are having has to do with theory versus practice. Sure, in theory overexposing the Alexa by several stops would lead to cleaner and thus 'better' images overall.

 

On the other hand, you have a bunch of cinematographers who have actually used the camera telling you that the trade-offs are not worth it. If we can agree to disagree, then fine. But if you're going to keep insisting that our practice is wrong, then the onus is on you to provide compelling visual evidence that backs up your claim.

 

So. Care to share some frames? My mind is open to change, but I need to see these superior 200ISO images for that to happen. :)

[Satsuki Murashige]

I suspect the reason Mr. Levy percieved the 200 ISO images to have less contrast is partly due to the fact that we generally view images on set in Log-C or Log-C with a Rec709-ish LUT applied. These gamma curves have a shoulder with highlight compression to mimic the way that film responds logarithmically to exposure. When the sensor was overexposed by two stops, more midtone information was pushed onto the shoulder resulting in less visible differentiation between these tones. Since we are all visually trained to respond to these all-important midtones (skin tones, blue skies, bright colors), the compression of these tones is immediately noticeable.

 

Now, I suppose you can question whether changing ISO in-camera to remap midtones would still lead to compression of those tones, or whether developing raw with linear exposure data and a custom gamma curve would change any of these things. But that's where practical testing would be relevant.

[Tom Yanowitz]

The problem with rating the Alexa at 200 ISO is that instead of 7-stops over and under, you have only 5-stops of overexposure detail but 9-stops of shadow information... And few cinematographers complain that the problem with the Alexa, or most digital cameras, is a lack of shadow detail. Film has more latitude for overexposure but less for underexposure, so if your goal is a film look, the last thing you'd want to do is trade overexposure detail in favor of more shadow detail when you already have more flexibility to bring up shadows with most digital cameras compared to film. Not to mention, if you ever have to make an HDR master from something shot on the Alexa, you are going to want that extra highlight information.

 

One of the most telltale signs that something is shot on a digital camera is how and when thing clip in the highlights. Often the amount of shadow detail is controlled by lighting, but it is harder to use lighting to control naturally hot areas of the scene.

 

Most people find that 7-stops of shadow information with the Alexa at 800 ISO are already enough. Certainly lower ISO's are cleaner but it's the same scenario in one way with film, you'd get less grain if you shot on low ISO stock but that may limit you creatively. And where it is not the same is that all the different speed stocks have the same gamma, you don't trade highlights for shadows by choosing a slow speed stock.

 

 

I agree with you on the highlight clipping David, I hope that at no point I was seemingly advocating for it !

This is of course more important to preserve highlights than shadows (I mean in most cases anyway).

 

If we realy want to make some analogies with film, I'd go with this :

The Alexa records with a unique film stock.

We underexpose it more or less depending on the ISO we choose.

And the raw-to-lin-to-LogC image we see on a monitor or on scope is varying degree of push-processing, so that in the end, a middle gray chart falls at 39% of the maximum value (so 1597 in LogC 12bit and 400 in LogC 10bit) and actually it stays there for the Rec709 image as well if I'm not mistaken.

At what ISO is not pushed ? One between 125 and 160, And you have exactly 5 stops of highlights from middle gray at this ISO.

Let's call it 160.

Everything from there is rating the camera above what she actually is and then doing some digital push processing magic.

From file value (arriraw) to display value (LogC) :

ISO 200 (and EI 200 on the camera) : your mid gray is pushed from around 1333 (33%) to 1600 (39%)

ISO 400 (and EI 400 on the camera) : your mid gray is pushed from around 900 (22%) to 1600.

ISO 800 (and EI 800 on the camera) : your mid gray is pushed from around 625 (15%) to 1600.

ISO1600(and EI1600 on the camera) : your mid gray is pushed from around 400 (10%) to 1600.

 

 

So in way, most digital cameras don't have a higher sensitivity than film stock, they are "just" a lot better at push processing.

 

 

Let's check the number of numerical values given to each stop depending on the ISO I choose (It's all half a stop from the truth sorry) :

 

ISO 200 : +5(508) +4 (508) +3(504) +2(496) +1(480) -1(448) -2(384) -3(256) -4(128) -5(64) -6(32) -7(16) -8(8) -9(4) -10(2) -11(1)

total : 2495 values above mid gray / 1343 below

 

ISO 400 : +6(508) +5 (508) +4(504) +3(496) +2(480) +1(448) -1(384) -2(256) -3(128) -4(64) -5(32) -6(16) -7(8) -8(4) -9(2) -10(1)

total : 2944+ / 895-

 

ISO 800 : +7(508) +6 (508) +5(504) +4(496) +3(480) +2(448) +1(384) -1(256) -2(128) -3(64) -4(32) -5(16) -6(8) -7(4) -8(2) -9(1)

total : 3328+ / 511-

 

ISO1600: +8(508) +7 (508) +6(504) +5(496) +4(480) +3(448) +2(384) +1(256) -1(128) -2(64) -3(32) -4(16) -5(8) -6(4) -7(2) -8(1)

total : 3384+/255-

 

So basically, it's not just about the number of stops above and below the middle gray, it's about how good these stops are.

A stop with 32 values is a lot crappier than a stop with 500.

Given that the number of values decreases as you get further away from sensor saturation, you have to decide : what is my threshold number of values per stop that I accept as part of dynamic range ?

For some people it could be 16 for example, for other that want highly grade and VFX-friendly shadows, it may be 100. Etc etc.

 

For all the +7/-7 etc on the Arri website, they count a total DR for the Alexa at 14,5 stops.

So it implies that they include every stop as proper dynamic range until you reach 2 values per stop, which is crazy.

No one in the world believes a two-value stop counts as dynamic range, there's hardly anything 'dynamic' about it.

 

A lot of people including me, experienced the feeling that shadows, even though not clipped "feel" less rich and "grade-able" than highs once you are manipulating them in a grading software. And this is why.

 

Now then, once you know this you can self-impose a threshold for minimum number of values per stop in the shadows :

 

If your threshold is around 16 (no grading for the shadows), the distribution for usable and detailed stops becomes this :

ISO 200 : -6,5/+5,5

ISO 400 : -5,5/+6,5

ISO 800 : -4,5/+7,5

ISO1600: -3,5/+8,5

(total : 12 usable detailed stops)

Here ISO 320 seems to be the one with proper +/- distribution.

 

If your threshold is around 64 (heavier post-production) :

ISO 200 : -4,5/+5,5

ISO 400 : -3,5/+6,5

ISO 800 : -2,5/+7,5

ISO1600: -1,5/+8,5 ('good luck VFX' as we call it)

(total : 10 usable detailed stops)

Here ISO 160 seems to be the one with proper +/- distribution.

 

So there it is, the trade off of working at higher ISOs to get more highlight headroom.

So is it worth it ?

Who can answer that ? Labs. They could check how often with the footage they receive the gigantic highlight range of ISO 800 is actually put to use.

Each time a .ari file has a max value between 3500 and 4094 and was shot at 800, that highlight would have been burned at 200. So you can stamp it "worth"

(I say below 4094 to exclude things that have been clipped at 4095 even at ISO 800, which are probably light bulbs and stuff like that that shouldn't be considered)

Every time it's isn't, not worth (unless you want some of the things I mentioned earlier in the topic like noise, deep focus, not enough light anyway etc...)

 

That leads me to this question :

With film stock : where would DPs usually stop in terms of highlights being a lot stronger than middle gray ? I'm actually curious to know.

Was it enormous like the Alexa at 800 (so +7,5 give or take) or were DPs just less concerned anyway with loss of detail and contrast in the highs than with digital

(fully understandable given there isn't a curve shoulder in digital so it's pretty brutal, even though LUTs now do a fair job at producing pleasing film-like roll offs).

 

Now. Arri could be explaining all this, nothing here makes them look bad or anything (ok, except the 14,5 stops debunking), it's just how digital works.

I think it's pretty cool that exposure is not as simple as "use that ISO, its the best mate", and more like trade off here and there for each and every one.

It's less boring this way anyway.

Edited March 8, 2016 by Tom Yanowitz

[Tom Yanowitz]

I think the difference of opinion we are having has to do with theory versus practice. Sure, in theory overexposing the Alexa by several stops would lead to cleaner and thus 'better' images overall.

 

On the other hand, you have a bunch of cinematographers who have actually used the camera telling you that the trade-offs are not worth it. If we can agree to disagree, then fine. But if you're going to keep insisting that our practice is wrong, then the onus is on you to provide compelling visual evidence that backs up your claim.

 

So. Care to share some frames? My mind is open to change, but I need to see these superior 200ISO images for that to happen. :)

 

So for this, I guess I could do some test sure, but :

- the noise factor would be mostly visible on moving images and during grading, so a screenshot wouldn't be very useful.

- as for the dynamic range, it's complex as well : If I work at 200, it's because no whites are clipped, if they are an I can't do anything about it (or don't have the time), I would be at a higher ISO in the first place. (that's my personal method)

And for the shadows : if you work at 800, you can just "light them more" I guess and you get the same thing visually.

For the fact that your image will be recorded on a lot more of gray values at 200 than at 800 : an 8bit screen won't show the difference probably (except of course the noise that comes with it). But data manipulation will, so vfx and grading guys will feel it in how fast they can work or how far they can go before things start to get ugly.

Edited March 8, 2016 by Tom Yanowitz

[Stuart Brereton]

Choosing an ISO is always a trade-off between speed and noise. Any competent DP understands that. One of the ironies of digital sensors is that on the occasions when you have enough light to shoot at 160ASA (Day/Ext) you also generally need the highlight protection of 800 ASA. Conversely, the occasions when that extra shadow detail would be really handy (Night/Ext) are also the times when you need the most speed.

Regardless of what the science says (which no-one here is disputing, as it's fairly well known), the reality is that for the vast majority of shoots, 400-800 ASA represents the best compromise.

[Carl Looper]

The main things about Tom's work is that it explores the Alexa's potential without the need of any pictures. This work is not everyone's cup of tea, but it's certainly mine.

 

I've done a lot of this sort of thing when designing a DI workflow for a bespoke system - to find the optimum settings required to ensure maximum signal integrity between original scene, exposed film, a subsequent DI, a render back to film, and subsequent projection. I ended up writing quite complex software to handle all the numerical details of such. While one could certainly do it by eye (trial and error) it's just so much more easier, satisfying, and accurate to do it by the numbers. The numbers allow one to convince oneself that a proposed practical implementation will be inevitable, rather than just a wait-and-see possibility.

 

Of course, errors are also inevitable, and perhaps that's where our eyes come into their own - they can pick up things our brains miss.

 

One of the interesting points in Tom's work is debunking the DR hype. I found that great.

 

C

Edited March 9, 2016 by Carl Looper

[Robin R Probyn]

"While one could certainly do it by eye (trial and error) it's just so much more easier, satisfying, and accurate to do it by the numbers. The numbers allow one to convince oneself that a proposed practical implementation will be inevitable, rather than just a wait-and-see possibility."

 

Well thats the difference between a DIT and DP then I guess.. of course every DP who is working with digital should have a good knowledge of what Tom has pointed out.. its not news .. its been documented before Toms discovery.. even I know about it.. in fact its pretty basic ..

 

But surely as a DP your eyes are more important than numbers.. thats like saying a pilot would fly into a mountain because the auto pilot computer says its not there..

[Carl Looper]

"While one could certainly do it by eye (trial and error) it's just so much more easier, satisfying, and accurate to do it by the numbers. The numbers allow one to convince oneself that a proposed practical implementation will be inevitable, rather than just a wait-and-see possibility."

 

Well thats the difference between a DIT and DP then I guess.. of course every DP who is working with digital should have a good knowledge of what Tom has pointed out.. its not news .. its been documented before Toms discovery.. even I know about it.. in fact its pretty basic ..

 

But surely as a DP your eyes are more important than numbers.. thats like saying a pilot would fly into a mountain because the auto pilot computer says its not there..

 

The fact that Tom's work is "basic" doesn't mean it can't be discussed. I mean basic stuff is discussed all the time on this forum. All the time. And the reason is simple enough: it provides an opportunity for those who don't know these things to learn such. Or for those who already do understand it to elaborate it in some way.

 

Or indeed for those who don't get into at all to make some silly quip about autopilots. :)

 

Numbers are not more important than one's eyes. They are just a different way of speaking about DP work. And that's all one can really do on a forum anyway. Speak. Numbers are one way of doing that. Just as words are. Or pictures. It's a way of theorising or thinking. What else does one do on a forum? Certainly not film making.

 

C

Edited March 9, 2016 by Carl Looper

[Robin R Probyn]

Agree totally.. the trouble is its being put forward as stuff that no one has known about.. even Oscar winning DP,s..

 

You thought the auto pilot quip was silly ? :)

Edited March 9, 2016 by Robin R Probyn

[Carl Looper]

You thought the auto pilot quip was silly ? :)

 

 

No not really. The autopilot quip was actually quite good. :)

 

C

[Robin R Probyn]

I knew you liked it .. :)

[Freya Black]

You seem to be very afraid of white clipping, as everyone should, but black clipping is actually the more common problem isn't it. With footage having no details whatsoever in the shadows.

It's a problem we see over and over.

 

How is this a problem?

 

If we are talking about shadows then that would seem to imply darkness to me and if an area is dark then it would seem to me that it would be fine if you can't see into the darkness.

I don't see it as a problem if you can't see into the shadows.

[Stuart Brereton]

I don't see it as a problem if you can't see into the shadows.

 

It depends on whether you see shadows as an indispensable part of lighting, and the thing that gives light shape, or you see them as a math problem.

[Carl Looper]

When we speak of details, be they within shadows, mid-tones, or highlights, we're really talking about contrast, without which there would be no details. It is in the variation of light that we appreciate anything we might call an image.

 

Without such variation (or contrast) we would have nothing but the same value across the entire visual field.

 

 

In film noir, details were elaborated in the midtones to highlight end of the range, and otherwise 'suppressed' in the shadows. This had a lot to do with the nature of photo-chemical film and the cost of lighting - and particularly so during the war. But the filmmakers, instead of treating such as some sort of compromise, would exploit such constraints to find what would become a virtue within such constraints. There was a particular kind of reality that could be elaborated, which could resonate with otherwise independant cultural fascinations, such as crime.

 

 

The shadows become devoid of information. The world beyond the walls disappears into darkness. It is as if there is no world out there at all. Or if there is, it is all just some sort of illusion, which, if we turned elsewhere, would return to the void from whence it came. There is a heavy emphasis on the psycho-logical as the ultimate if equally doomed frame of reference.

 

But details operate at all scales - not just at the level of the small. While the darkness will be devoid of details, the darkness itself will become a "detail", in the sense that it will contrast with the remainder of the field. The shape of the blacks will become just as important as the lack of detail within such blacks.

 

 

 

Now, while all of this is easy to understand using one's eyes, it's not impossible to represent these same insights (as well as many others) in numerical terms. Indeed it becomes completely necessary to do so in the development of digital post techniques. HDR toning will find it's basis in the mathematical representation of the various factors in play within such an otherwise sensory-perceptual field. Even in the analog domain a mathematical representation will facilitate analog precursors to such things. Things like digital sharpening, and indeed motion tracking (!) will find it's basis in the mathematics of analog techniques, developed during World War II aerial reconnaissance photography analysis.

 

Does any of this matter, or should it matter to the cinematographer?

 

No, not at all. But neither does it hurt.

 

It's not in opposition to cinematography. It won't in anyway make you any worse at cinematography.

 

C

Edited March 10, 2016 by Carl Looper

[Stuart Brereton]

Carl, I am never one to advocate 'dumbing down', but is there any way you could write a post that doesn't need a thesaurus and a dictionary to understand?