Blackmagic Forum

I think your earlier frame would look great with one side color going to completely monochrome the other side with the light level. I would even do a whole movie like that. I want to do worse. Imagine bald executives trying to pull their hair out over that.

John Paines wrote:

Roberto de la Torre wrote:Just keep in mind the subtractive watercolor concept.
As more dark= more color. (not grayish shadows)
The brighter= less color. Like transparent film print areas (not weird roll-off colors when clipping)

That's an interesting theory from Art Adams, but performing it after-the-fact, by desaturating highlights and saturating shadows in post, doesn't achieve an Alexa miracle.

As I recall, some youtubber a while back claimed that desaturating shadows (not highlights) was an instant and wonderful film-look technique. And the commenters enthusiastically agreed and were full of thanks at the miracle performed in front of their very eyes.

The Emperor was wearing a baseball cap, but no clothes.

yeah, absolutely!
just clarify, I'm talking about shadows a few stops below te key, not pure blacks.

It's kinda subjetive because there's a lot of movies who tint the highlights warm or whatever, but the subtractive roll-off still there smoothing the look

Wayne Steven wrote:
BTW, the girl in that bus, I think is on the latter android keyboard gifs, doing an impression of a friend's daughter. Hilarious.

just was I search a frame for another blonde kid with angry gaze as we're seeing in this thread

John Paines wrote:That's an interesting theory from Art Adams, but performing it after-the-fact, by desaturating highlights and saturating shadows in post, doesn't achieve an Alexa miracle.

As I recall, some youtubber a while back claimed that desaturating shadows (not highlights) was an instant and wonderful film-look technique. And the commenters enthusiastically agreed and were full of thanks at the miracle performed in front of their very eyes.

The Emperor was wearing a baseball cap, but no clothes.

Depends on which film process is being emulated. If what's being mimicked is print film stocks, then the highlights are desaturated. If what's being emulated is a negative film scan into color grading, then the shadows are desaturated. Art explained the difference quite well in a later article.
https://www.provideocoalition.com/film-look-two/

Jamie LeJeune wrote:Depends on which film process is being emulated. If what's being mimicked is print film stocks, then the highlights are desaturated. If what's being emulated is a negative film scan into color grading, then the shadows are desaturated. Art explained the difference quite well in a later article.
https://www.provideocoalition.com/film-look-two/

Whatever the theory, the technique as implemented didn't make the material look remotely like film, or more like film. That's my only point. I can't judge the theory, or comment on what the Alexa sensor is doing.

John Paines wrote:Whatever the theory, the technique as implemented didn't make the material look remotely like film, or more like film. That's my only point. I can't judge the theory, or comment on what the Alexa sensor is doing.

Then that's on the colorist. Not sure what it would have to do with Alexa because the Alexa sensor itself is doing neither of those things. ARRI's default LogC to REC709 LUT does rolloff and desaturate the highlights more than most other REC709 LUTs, though BMD's new Extended Video LUT also treats the highlights somewhat similarly.

John Paines wrote:

Jamie LeJeune wrote:Depends on which film process is being emulated. If what's being mimicked is print film stocks, then the highlights are desaturated. If what's being emulated is a negative film scan into color grading, then the shadows are desaturated. Art explained the difference quite well in a later article.
https://www.provideocoalition.com/film-look-two/

Whatever the theory, the technique as implemented didn't make the material look remotely like film, or more like film. That's my only point. I can't judge the theory, or comment on what the Alexa sensor is doing.

But it's one step.
It's a fact. Film have a transparent film layers, negative or reversal there's one of those transparent = not color on it.

Cheap film stock without good anti-halation layer makes the best roll-off. Well, some panty in front of the lens can help as well if wants a foggy image..LOL
Also red halo in highlights sometimes gives a cold scene that warm touch that digital can't itself

Roberto de la Torre wrote:

Wayne Steven wrote:
BTW, the girl in that bus, I think is on the latter android keyboard gifs, doing an impression of a friend's daughter. Hilarious.

just was I search a frame for another blonde kid with angry gaze as we're seeing in this thread

Whoever the kid is, if you think that she looked angry there, you should hunt down a recent android phone and check the keyboard gifs.

Guys, a scientific discussion on "more pixels" and their experience on audience, made by Steve Yedlin, ASC, DP for Star Wars: The Last Jedi, Looper, and others.

The first video is a 10 minutes only. The second is 1 hour long but you can watch it in parts (its Youtube!) like I did.

He gives a pretty good demonstration of what we are talking about using cameras like Alexa 65, Red 8K and IMAX.

Part one:



Part Two:

Ulysses Paiva wrote:Guys, a scientific discussion on "more pixels" and their experience on audience, made by Steve Yedlin, ASC, DP for Star Wars: The Last Jedi, Looper, and others.

The first video is a 10 minutes only. The second is 1 hour long but you can watch it in parts (its Youtube!) like I did.

He gives a pretty good demonstration of what we are talking about using cameras like Alexa 65, Red 8K and IMAX.

Part one:

Part Two:

For you out there saying you can't watch it on youtube I herby give you the link to his personal website where you can watch it in high res:

http://www.yedlin.net/ResDemo/index.html

I think that the conclusion I get is that at a certain point you no longer need a better camera, you need a better colorist or post-production workflow.

16k+ like Pansvision (who has used a Red sensor) says?

There are many opinions, to be examined for facts.

Reports are James Cameron is shooting the Avatar sequels on the Sony VENICE - 6k.

Ulysses Paiva wrote:Guys, a scientific discussion on "more pixels" and their experience on audience, made by Steve Yedlin, ASC, DP for Star Wars: The Last Jedi, Looper, and others.

The first video is a 10 minutes only. The second is 1 hour long but you can watch it in parts (its Youtube!) like I did.

He gives a pretty good demonstration of what we are talking about using cameras like Alexa 65, Red 8K and IMAX.

Part one:

Part Two:

I remember those demo by Steve. And all that stuff showing digital cameras looks like film.
Much respect to Steve Yedlin, he's a master...but, he still choosing film to shot...why?

michaeldhead wrote:Reports are James Cameron is shooting the Avatar sequels on the Sony VENICE - 6k.

Indeed he is. Sony even developed a version of Venice that has a remote head so that the bodies can ride in a backpack or on a cart to make getting the inter-ocular distance between the two lenses easier to adjust. And so that it could be smaller.

michaeldhead wrote:Reports are James Cameron is shooting the Avatar sequels on the Sony VENICE - 6k.

No Red. I remember Duke Nukem Forever. So many delays, and technology switches. What's happening with Avatar now?

Wayne Steven wrote:

michaeldhead wrote:Reports are James Cameron is shooting the Avatar sequels on the Sony VENICE - 6k.

No Red.

It’s Venice.

JB

Watched the first video. Inadequate demonstration. If you want objective comparison you have to flip between image frames, rather than jar between a stupid repeating imitation sequence that wonders in movements. The perception is jarred on each reset of image sequence with complex movement, focusing etc, masking perception of image differences like resolution.

It speaks of upscaling to 4k, further hiding image differences. It speaks of adjusting the light getting through the lens, which amplifies noise on more sensitive cameras, reducing the other picture quality talked about.

I noticed the big black pixel looking thing in the book case, which reminds me. Viewing and judging resolution based on a small field of view, or poorer vision, may not be seeing the correct stuff.

It is as if more indirect talk makes it true. But NOBODY here seems to have picked up on any of this stuff, its fan stuff.

John, you know I read it's Venice.

Wayne Steven wrote:Watched the first video. Inadequate demonstration. If you want objective comparison you have to flip between image frames, rather than jar between a stupid repeating imitation sequence that wonders in movements. The perception is jarred on each reset of image sequence with complex movement, focusing etc, masking perception of image differences like resolution.

It speaks of upscaling to 4k, further hiding image differences. It speaks of adjusting the light getting through the lens, which amplifies noise on more sensitive cameras, reducing the other picture quality talked about.

I noticed the big black pixel looking thing in the book case, which reminds me. Viewing and judging resolution based on a small field of view, or poorer vision, may not be seeing the correct stuff.

It is as if more indirect talk makes it true. But NOBODY here seems to have picked up on any of this stuff, its fan stuff.

You should watch the second video.

Got a headache from the first one. . But, yes, people saying his good the first one was in isolation. But many things still stand, not an easy comparison, and upresing them all to 4k?

Wayne Steven wrote:Watched the first video. Inadequate demonstration. If you want objective comparison you have to flip between image frames, rather than jar between a stupid repeating imitation sequence that wonders in movements. The perception is jarred on each reset of image sequence with complex movement, focusing etc, masking perception of image differences like resolution.

It speaks of upscaling to 4k, further hiding image differences. It speaks of adjusting the light getting through the lens, which amplifies noise on more sensitive cameras, reducing the other picture quality talked about.

I noticed the big black pixel looking thing in the book case, which reminds me. Viewing and judging resolution based on a small field of view, or poorer vision, may not be seeing the correct stuff.

It is as if more indirect talk makes it true. But NOBODY here seems to have picked up on any of this stuff, its fan stuff.

I mean, you're critiquing the literal prologue of a video- the second part is like an hour longer and goes into much of what you're complaining about.

One thing to note is that the second part is actually mastered in HD so that he can compare a crop of the videos at one-to-one pixels without resampling. Furthermore, he actually goes into your point about upscaling, correctly noting that nowadays video will be resampled a dozen times through different phases of professional remastering and delivery (your blu-ray is in HD, DCI delivery 2k, acquisition 4k/3k/6k/4.6k/5k/etc.), and finally when watched on a screen that might not be the right resolution for the content. Every time this happens every pixel is completely remapped, meaning that the idea of preservering the 'original' pixels for image quality comparison is already moot, except for on a theoretical level (and this guy is a working professional).

Speaking to that last part, this is also why the images are matched visually (what you're criticizing when you say 'adjusting the light through the lens'): in an actual real-life professional situation you're shooting for that t-stop that gives you a particular Circle of Confusion. It's an aesthetic choice which informs the t-stop you're shooting at, which means that you'll have to adjust the t-stop you're shooting at for the particular format you're using (I think you're probably familiar with this concept but nonetheless, here's an article explaining this concept from, you guessed it, Yedlin: http://yedlin.net/lens_blur.html). This means that, what's interesting to a professional cinematographer (at the level where it can be accepted that he has the means to light for any t-stop he wants) it's not interesting to see how the cameras compare at the exact same f-stop, but to see how they compare when they're aesthetically matched. I understand that this is an affront to an engineer like you (at least I think you have an engineering backgroun, correct me if wrong), but this is because you are coming from different backgrounds.

Yedlin's moto is that cameras are tools, and there are ways to get them to look like film, if you put your effort on it. The truth is that 99% of digital cameras are not calibrated for the film look that we all so desperately want, except the Alexa, which was created specifically with film emulation in mind. However, there is a way to get the Alexa look, which will give you the film look "for free".

I'd highly suggest you look the EmotiveColor's GHa solution. It's a full solution: it asks you to shoot a certain way, to add their lut(s), and then to further change the image via additional nodes in Resolve, as per their instructions. Right now, their solution WORKS (I've tested it, I bought it a few days ago), but it only works in a per-camera/format basis. So far, it only works with Panasonic's vlog from GH5 and GH5s.

I emailed them, and they said that they can't wait to get a BMPCC 4k, so they can port their solution. It's their next go-to camera that they want to work with. Check their website for examples.

Film look is an old thing, but based on trying to get the image pleasing. I'm interested in natural look and scientific looks, as starting points, then grade as you will. Scientific look will give you the most flexibility to change look, but you need enough bit depth in each pixel to accurately record levels in full latitude. You effectively need ACES colour space relevant pixel level mapping. But you don't those things, you can make ro with what you have.

Sean van Berlo wrote:I mean, you're critiquing the literal prologue of a video- the second part is like an hour longer and goes into much of what you're complaining about.

One thing to note is that the second part is actually mastered in HD so that he can compare a crop of the videos at one-to-one pixels without resampling. Furthermore, he actually goes into your point about upscaling, correctly noting that nowadays video will be resampled a dozen times through different phases of professional remastering and delivery (your blu-ray is in HD, DCI delivery 2k, acquisition 4k/3k/6k/4.6k/5k/etc.), and finally when watched on a screen that might not be the right resolution for the content. Every time this happens every pixel is completely remapped, meaning that the idea of preservering the 'original' pixels for image quality comparison is already moot, except for on a theoretical level (and this guy is a working professional).

Speaking to that last part, this is also why the images are matched visually (what you're criticizing when you say 'adjusting the light through the lens'): in an actual real-life professional situation you're shooting for that t-stop that gives you a particular Circle of Confusion. It's an aesthetic choice which informs the t-stop you're shooting at, which means that you'll have to adjust the t-stop you're shooting at for the particular format you're using (I think you're probably familiar with this concept but nonetheless, here's an article explaining this concept from, you guessed it, Yedlin: http://yedlin.net/lens_blur.html). This means that, what's interesting to a professional cinematographer (at the level where it can be accepted that he has the means to light for any t-stop he wants) it's not interesting to see how the cameras compare at the exact same f-stop, but to see how they compare when they're aesthetically matched. I understand that this is an affront to an engineer like you (at least I think you have an engineering backgroun, correct me if wrong), but this is because you are coming from different backgrounds.

Yes, yes Sean. But it was put forward on it's own, and if people think I'm long winded, this 10 minute video is a classic example. Everything night have been done in 10 minutes instead of waffle on. When I go on, I might be explaining everything st once simply, rather explaining nothing unsimply. I can use the link as an test anytime somebody complains. Is what he has done bad, no not really, testing yes, in a metaphysical (is that the term) way, a chosen style. Is it the greatest style, I would not have thought so. If the technical schools, studios, and periodicals that define the visual look for the audiences people try to emulate (film often looks poor before grading in the old days, but had more performance at the time) do not write like this, that tells you something.

Now, true comparison is native resolution, native throughout (in terms of sensor noise). That would be grade A (what's really happening here) comparison. Grade B would be to adjust scene lighting or in post, maybe in camera through gain control (but again, that might sdjust other things). Grade C would be what was done in the 10 minute video, except of exact same scene and movement filmed simultaneously. See, that is a minimalistic succinct way of putting it. You would show all these grades together. That's is a practically minimalistic rather than simplistic way of metaphysically putting it. Grade A is more for charts and test shits to see what is actually happening. Grade B is an real world extension stress test in matching an image (if you were talking about resolution without regard to noise etc, ND filters wouldn't matter) as long as things like bit range don't choke the image etc. Grade C is showing how you can work the image to seem the same. Now. I'm starting to loose the strands here. But you get the direction.

Now, resizing everything at 4k or 2k, diminishes your resolution comparison even further and means fuzzy pixels. Why are you not either keeping the full resolution until you get to delivery or swapping to delivery resolution once somewhere beforehand (the lesser option)? But this is the most crucial thing about a resolution comparison, to truly show what resolution is doing by itself, you should shoot it on the same camera. This will mean one or two things. As practical sub resolutions are an issue get an 8k camera and 1) Shoot 8k, and downres evenly by fractions to 1/2, 1/3, 1/4 (2k) etc with best routine. But this still will show improvements in the fractional resolutions from starting with 8k versus a native resolution sensor of those fractions. 2) Resize with the camera to reproduce those fractions, but to keep the same lens and ND status you might never reach Grade A because the reframing changes the setup and available light. But you should be shooting a static scene/flat picture anyway, so you can easily adjust its lightning and remove lens setup biases as much as possible. And that is all the stuff I can come up with off the top of my head on the spot. To say something about resolution, you compare resolution keeping other things the same, otherwise you are comparing different camera setups at the same time, diminishing the measured effect of resolution.

Part B. But, it is important to measure how shooting a native camera compares, and doing uneven fractional down and up resting compares. These things are also done in isolation. So, you can say at what range percentage increases and decreases of re-resolution adjustment make the image more pleasing (aliasing, pleasingly software, more sharp etc). You can say here is the difference between resolution x camera and fractional resolution y camera, and after adjustments, and after matching to the image delivery, and blending through several adjustments in a post chain as mentioned here.

Such comparison tests should be difficult hair tearing experiences. Like writing an periodical article which should take a week or so, on the spot live.

Eugenia Loli wrote:.... I'd highly suggest you look the EmotiveColor's GHa solution.... they said that they can't wait to get a BMPCC 4k, so they can port their solution. It's their next go-to camera that they want to work with. Check their website for examples.

Eugenia, would you kindly follow up with EmotiveColor, since it seems it’s the codec that is also key? Let them know that BRAW is coming to the BMPC4K likely this year and shooters would likely be interested in seeing EmotiveColor include that in addition to CinemaDNG and ProRes.


Sent from my iPhone using Tapatalk

Eugenia Loli wrote:Yedlin's moto is that cameras are tools, and there are ways to get them to look like film, if you put your effort on it. The truth is that 99% of digital cameras are not calibrated for the film look that we all so desperately want, except the Alexa, which was created specifically with film emulation in mind. However, there is a way to get the Alexa look, which will give you the film look "for free".

I'd highly suggest you look the EmotiveColor's GHa solution. It's a full solution: it asks you to shoot a certain way, to add their lut(s), and then to further change the image via additional nodes in Resolve, as per their instructions. Right now, their solution WORKS (I've tested it, I bought it a few days ago), but it only works in a per-camera/format basis. So far, it only works with Panasonic's vlog from GH5 and GH5s.

I emailed them, and they said that they can't wait to get a BMPCC 4k, so they can port their solution. It's their next go-to camera that they want to work with. Check their website for examples.

That's interesting.

But why people try emulate Alexa? when Alexa try emulate film.
If people spend his time trying to emulate film instead Alexa, maybe Alexa stopping be the god of film cameras.

I took a couple screengrabs of the Log files from EmotiveColor, just make film look, not Alexa match. It took me few minutes to stop it, obviously jpg isn't the best to make this. Don't do that at home!

I think I virtually always like Alexa features look over film. Film is an ideal. But out of camera, what gives you an instantly usable feed? And Alexa is not the best which can be done.

But why people try emulate Alexa? when Alexa try emulate film.

The golden standard in Hollywood today is Alexa. Most films are shot with it. Which means that most luts and tools are developed with LogC in mind. Therefore, there's value to try to emulate the Alexa, and not a particular film stock. You color grade to get a specific film stock, AFTER you have transformed your footage to match Alexa's (and no, the Color Space Transform found in Resolve does not do the full work as GHa does). Then, color grading is just easier to get the film look you like.

So if EmotiveColor ports their solution to the BMPCC4k, I think that's tremendous news for us.

rick.lang wrote:Eugenia, would you kindly follow up with EmotiveColor, since it seems it’s the codec that is also key? Let them know that BRAW is coming to the BMPC4K likely this year and shooters would likely be interested in seeing EmotiveColor include that in addition to
CinemaDNG and ProRes.

Alex (the guy behind Emotive Color) wouldn't need to make changes for the different codecs. He does need to make different LUTs for each camera and each profile (gamma + gamut combination) under daylight and tungsten. I think everyone would be fine with just the BMD film profile.

btw, I've got the LUT, and it is a pretty damn good match. I'd dare say that within the limitations of the GH5's signal, it's about as good as you're gonna get.

But why people try emulate Alexa? when Alexa try emulate film.

Because we like the Alexa "look", just like we like the "look" of certain film stocks. I also hate the look of certain other film stocks. "Film" isn't one look.

From a 1992 patented technology that does over 28 stops while keeping color in balance throughout I think most of the range or more, from 1000th a lux to 500k lux. It has global shutter.



This is a picture from an old 25 or 27 stop version:



Imagine what would happen if these guys had some money to develop this. But this stuff is out of patent in the wild, and I saw an inspection camera claiming to do precise colour by not using a color filter (like Bayer). Haven't read up on that.

Better than Alexa is possible. What is BM spending their money on.

Eugenia Loli wrote:

But why people try emulate Alexa? when Alexa try emulate film.

The golden standard in Hollywood today is Alexa. Most films are shot with it. Which means that most luts and tools are developed with LogC in mind. Therefore, there's value to try to emulate the Alexa, and not a particular film stock. You color grade to get a specific film stock, AFTER you have transformed your footage to match Alexa's (and no, the Color Space Transform found in Resolve does not do the full work as GHa does). Then, color grading is just easier to get the film look you like.

So if EmotiveColor ports their solution to the BMPCC4k, I think that's tremendous news for us.

Make sense of point of view of the industry.

Just trying to do the role of developer not just a buyer. All this seems a consumer madness each year.
But is ok.

joe12south wrote:

"Film" isn't one look.

Really?

Have you shot many film stock or Alexa stuff? I'd love take a look at your works.

Roberto de la Torre wrote:

joe12south wrote:

"Film" isn't one look.

Really?

Have you shot many film stock or Alexa stuff? I'd love take a look at your works.

Do you think Kodak Vision3 200T 5213 looks the same as Fujifilm Superia Xtra 400? If so, remind me not to hire you as a colorist.

Wayne Steven wrote:I think I virtually always like Alexa features look over film. Film is an ideal. But out of camera, what gives you an instantly usable feed? And Alexa is not the best which can be done.

I remember when the first few features I saw shot on Alexa, I was blown away. I also preferred the look over film.

Okay, I think I'm getting somewhere...

v0.3


PS. I know my daughter is underexposed. I'm looking at how the LUT handles extreme highlight recovery in the doorway. Notice how the color shift is gone and more naturally desaturates.

joe12south wrote:Okay, I think I'm getting somewhere...

Out of curiosity, what are you using as a reference image? Were you able to shoot the same scene with both cameras on the same lens?

Jamie LeJeune wrote:

joe12south wrote:Okay, I think I'm getting somewhere...

Out of curiosity, what are you using as a reference image? Were you able to shoot the same scene with both cameras on the same lens?

No, unfortunately nothing so rigorous at this point. I'm eyeballing from an old Alexa shot I have of a test chart and a new test chart shot with the P4K. They were shot years apart, but with the same lens and the same key light.

If I were trying to make a super accurate LUT for intercutting the two cameras, this workflow wouldn't cut it, but right now my goal is to just make a pleasing "look" that I can sprinkle on the very last node of a project for a bit of Alexa-ness.

joe12south wrote:Do you think Kodak Vision3 200T 5213 looks the same as Fujifilm Superia Xtra 400? If so, remind me not to hire you as a colorist.

I see.
Man, you're a bit confused about motion film stocks, even what film can achieve from same stock.
Anyway, I'm in the same boat, we don't need be rude each other.

Roberto de la Torre wrote:

joe12south wrote:Do you think Kodak Vision3 200T 5213 looks the same as Fujifilm Superia Xtra 400? If so, remind me not to hire you as a colorist.

I see.
Man, you're a bit confused about motion film stocks, even what film can achieve from same stock.
Anyway, I'm in the same boat, we don't need be rude each other.

I said there is no one film look. You questioned that. I pointed out two film stocks with noticeably different looks. You then go on to seemingly disagree with yourself by pointing out that how you process a given stock can yield even more looks, reinforcing my point.

What am I confused about?

(PS. Apologies for the jab - this just seems like a nonsensical argument.)

joe12south wrote:

Roberto de la Torre wrote:

joe12south wrote:Do you think Kodak Vision3 200T 5213 looks the same as Fujifilm Superia Xtra 400? If so, remind me not to hire you as a colorist.

I see.
Man, you're a bit confused about motion film stocks, even what film can achieve from same stock.
Anyway, I'm in the same boat, we don't need be rude each other.

I said there is no one film look. You questioned that. I pointed out two film stocks with noticeably different looks. You then go on to seemingly disagree with yourself by pointing out that how you process a given stock can yield even more looks, reinforcing my point.

What am I confused about?

(PS. Apologies for the jab - this just seems like a nonsensical argument.)

That "really?" was ironic, was a joke.
And then I asked about to take a look at your work in friendly way (not ironic or kidding way) Also I tried to add my two cents for your LUT developing.

Apologies if I lighted your fire in some moment.

Empirically speaking, I can tell you that is much more easy match Eterna with Vision than GH5 with Alexa... but not impossible I guess

Roberto de la Torre wrote:That "really?" was ironic, was a joke.
And then I asked about to take a look at your work in friendly way (not ironic or kidding way) Also I tried to add my two cents for your LUT developing.

I missed you intent...sorry, I should pay more attention to emoji.

Red Gemini comparison: https://vimeo.com/295709062/186f03e88c?cjevent=6fe2a3e0d55b11e8807d038c0a240614

Pretty impressive considering the Red costs 15X more.

Carlos Garcia-Diaz wrote:Red Gemini comparison: https://vimeo.com/295709062/186f03e88c?cjevent=6fe2a3e0d55b11e8807d038c0a240614

Pretty impressive considering the Red costs 15X more.

Excelent! They are on par IQ wise, except the underexposed scene which the Pocket performed much better. Its just an impression or the Pocket 4k held the highlight details on the window better?

Carlos Garcia-Diaz wrote:Red Gemini comparison: https://vimeo.com/295709062/186f03e88c?cjevent=6fe2a3e0d55b11e8807d038c0a240614

Pretty impressive considering the Red costs 15X more.

Uploader does not mention whether or not the low-light mode was used on the Gemini for shooting at 3200. I would imagine with 16 stops dynamic range in that mode, it would do better.

This is something I do when take color palettes for my art work.
I'd take the palette of film by his 9 colors vs 7 of Alexa.

I'd like have the chance also to had some BM camera right there to see how many colors brings.
The funny thing is that in the full shot, Alexa looks warmer than the film shot which looks a bit cooler but at the same time the skin tones in film are more warm than Alexa.

As Wayne say. I think Alexa is pretty good camera but not so much that everyone wants to emulate it as hell.

Wayne, I smell a rat in the matter of your post of the photo of a girl behind a bright light. Wasn’t Isaac Newton driven mad trying to change lead into gold? You think if he succeeded he’d have a problem raising funds for his gold minting operation? Even more incredulous if there’s no active patent and yet we aren’t there yet. Who wouldn’t want 25 stops if it was available shooting middle grey at ISO 100 at 60 fps 180 degree shutter angle? I don’t even know if that’s a possible range with audio, let alone video, because I can’t recall seeing prosumer audio gear rated at 170 dB.

What do you think is the truth behind this claim and why it has withered on the vine?


Sent from my iPhone using Tapatalk

Re-edit: Sorry, wrong quote.

Kel Philm wrote:

Wayne Steven wrote:I think I virtually always like Alexa features look over film. Film is an ideal. But out of camera, what gives you an instantly usable feed? And Alexa is not the best which can be done.

I remember when the first few features I saw shot on Alexa, I was blown away. I also preferred the look over film.

Roberto de la Torre wrote:This is something I do when take color palettes for my art work.
I'd take the palette of film by his 9 colors vs 7 of Alexa.

I'd like have the chance also to had some BM camera right there to see how many colors brings.
The funny thing is that in the full shot, Alexa looks warmer than the film shot which looks a bit cooler but at the same time the skin tones in film are more warm than Alexa.

As Wayne say. I think Alexa is pretty good camera but not so much that everyone wants to emulate it as hell.

I might have started this off years ago, advocating if BM (and probably Red) could emulate the characteristics of Alexa, they would have a great image and camera worth buying. Having a poor image makes a camera purchase a sympathy buy. THEY DELIVERED, the pocket and mini was great, but $10million for the 4.6k. But I statedly meant this as a minimum specs going into the future. I've been drumming up support for better cameras since 2004, writing to various companies, including the one that made the Alexa sensor (I wanted to male my own commercial camera and they had bought two sensor companies I was interested in). Don't miss good opportunities to network the right people together, and get ideas in the right minds, it pays. If it wasn't for interference we could have had a $2k 720p camera nearly 14 years ago, or a $5k fullhd setup. But somebody had to try to divert thunder (attention) to himself and that probably put.the manufacturer off. Then Red came, and HDMI recording out on camcorders. We could be looking at cheaper than the pocket today if it wasn't for that. But, the old industry is reluctant to make such easy (which means it might not have happened simply because of that).

Anyway, you could argue that the Netflix look like that in the TV series, The Expanse, is a suitable look for a camera to emulate. But really, you just want three or do modes, a little bright, medium and lower contrast. That would cover most shooting before a colourist gets hold of it. Love watching some of that Netflix stuff.

We talk about an Alexa lut, but what about a BBC and Netflix look luts. These things look good because some very talented people set up those looks, like the old Panavision look. You guys talk about film look, but a lot is graded look. Even if a 13 stop camera can't match the Alexa perfectly, it doesn't have to be. Colour constancy (in balance)), dynamic range, noise and rolling shutter issue are issues where you get rewards. Nobody should have to grade, the camera itself should produce native footage good enough, which you may grade as you wish.

Why are we speaking about luts rather than newer transforms? Go and try that with BBC and Netflix look as well, if they will let you.

Why are we speaking about luts in camera, rather than the newer aces transforms? Now, that would be a selling point. A camera with official transforms for BBC and Netflix looks, and that at least let's you down load whatever transforms to get a seamless conformity in camera for post over a lut. In camera you can play with the sensor circuitry a bit too, to get the match even closer.

rick.lang wrote:Wayne, I smell a rat in the matter of your post of the photo of a girl behind a bright light. Wasn’t Isaac Newton driven mad trying to change lead into gold? You think if he succeeded he’d have a problem raising funds for his gold minting operation? Even more incredulous if there’s no active patent and yet we aren’t there yet. Who wouldn’t want 25 stops if it was available shooting middle grey at ISO 100 at 60 fps 180 degree shutter angle? I don’t even know if that’s a possible range with audio, let alone video, because I can’t recall seeing prosumer audio gear rated at 170 dB.

What do you think is the truth behind this claim and why it has withered on the vine?


Sent from my iPhone using Tapatalk

I haven't forgotten you Rick. I don't know the exact reason. I think the base technology goes back 1976 and was HDR back then. There was a commercial VGA car camera with this many years back. But it may have to do with how things are done, sensor size, resolution development. and I don't remember its noise profile, heat, energy drain, or politics. They have a, wait for it, latest sensor that can do 720p50. The sensors from memory are small. As I said, spend the money to develop it. I'm not an sensor engineer there, so am unsure what the exact limitations are or what can be further done with latest technologies from other sources. From memory, the technology went into the Europe sensor technology patent portfolio along with some other nice things, for access by EU sensor companies.

Now, about way things are done being a reason. A certain person advocated a few different sensor technologies and Jim over Red, said publicly that they would never use one of them, which might be related to this technology, so there must be some factor they didn't like. But for me, I'm willing to accept a few downsides to get great range and colour constancy.


So, I can't say particularly what the reasons are without having to spend time researching into it. But I did see a drone with a fullhd hdrc camera. I don't know if this version of the latitude extending technology, or if size restricted. But if on a drone it is low enough powered, it might suit people who simply want fullhd and are willing to accept a small format. I wonder if it is related to recent Sony sensor type (not the one in the pocket 4k) but O don't know what low light technology that uses. I should point out there are various Quantum Efficiency multiplying technologies out there. I think I saw a technology which multiplies the signal rather than the noise, used in astronomy, leading to extreme gain with very low noise.

Here is the wiki page I forgot to post l before.

https://de.m.wikipedia.org/wiki/HDRC