[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/p/ - Photography

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • There are 39 posters in this thread.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


I just had the idea of getting a color checker, taking a shot of it with your digital camera, and then with a film camera with let’s say Ektachrome, and calibrating Lightroom color profile to exactly fit the renders of the film color on the digital photo. Has anyone done this?

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width1280
Image Height862
>>
>>4085286
yes, they are called fuji users and the fujifilm corporation
>>
>>4085286
What you think of is called a color lookup table and yes there are some that aim to emulate film colors
https://marcrphoto.wordpress.com/film-simulation/the-largest-collection-of-film-simulation-haldclut-luts-brought-together/
>>
>>4085289
I mean like actual color science, I think we all know fuji colors aren’t trying to “exactly” replicate film to exact hues digitally. (though they probably could do it if they wanted to)
>>
>>4085301
What are you talking about. It is not possible to replicate film exactly. Even the best look up table will not capture the difference of contrasts and how the color changes according to exposition. You can either use Fuji or buy presets that will fit most of your need, but if you want exactly film, you have to shoot film.
>>
>>4085286
Why not just shoot film instead?
>>
>>4085286
Ages ago, made a reverse profile with some open source software. There’s plenty of photos of colour checkers on film. Just white balance it off grey, and profile that. And profile your camera and lens combo then you have to subtract them into one profile, I forget exactly.
>>
>>4085344
Shooting film is about the spectacle of shooting film and is really only shared and commented on to gearfag about the specific film it was shot on, the content of the actual photography is typically ignored with film.
>>
>>4085360
>Shooting film is about the spectacle of shooting film and is really only shared and commented on to gearfag about the specific film it was shot on
Then why not just shoot film and not care about that shit?
>>
>>4085364
Why don’t you just shoot film instead and not care about asking others to shoot film?
>>
>>4085368
>Why don’t you just shoot film instead
I do.
>and not care about asking others to shoot film?
Where did I ask you to shoot film? learn2read
I'm asking why you're not shooting film if you want the color of film?
>>
>>4085344
Film prices are too extreme right now. What I like most about film is the colors rather the experience of analog. Technology is definitely there to replicate film colors but I just don’t think anyone has put real work into that.
>>
>>4085339
I’m not trying to copy exactly film in it entirety, i really mean film colors not so much it’s grain or texture. I believe film isn’t as mysterious or random as it seems. Film has been around for like 100+ years. Fuji, kodak definitely knew why and how there film was rendering colors. Digital is definitely in a state it can copy at least the color of it
>>
>>4085372
>where did I ask you to shoot film?
>why don’t you shoot film?

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Horizontal Resolution72 dpi
Vertical Resolution72 dpi
Color Space InformationsRGB
Image Width720
Image Height960
Scene Capture TypeStandard
>>
>>4085394
Yes, I didn't ask you to shoot film. I asked why don't you do it.
learn2read, ESL friend.
>>
Can anyone point me to some free Capture One film presets?
>>
>>4085446
Yeah learn how to use it, fool
>>
>>4085286
Yeah that will get you about 50% of the look for that one shot.
>>
>>4085382
>Digital is definitely in a state it can copy at least the color of it
That's where you are wrong.
If you look at the spectral responses of digital sensors and a given color film it is easy to show that there exists a set of two lighting conditions that will yield the same color in one, but two different colors in the other, which makes it impossible to create a color profile that is 100% accurate across the whole spectrum.

The most you can achieve is a profile that is reasonably accurate within a given lighting situation and for a specific set of subjects.
>>
>>4085377
maybe theres an enhanced AI film emulator now that does the job?
>>
File: PSX_20220406_100626.jpg (1.02 MB, 1500x1000)
1.02 MB
1.02 MB JPG
>>4085301
>Buy Fuji camera
>Dial in a few presets using daylight WB with tweaks to the colour balance to achieve the look you want
>Dial in shadows/highlights to get desired contrast

You can already do exactly what you want to do.

t. some looser who has been shooting film for nearly 20 years and now shoots Fuji digital happily this way.
>>
>>4085496
Veryyyy interesting. Is there anywhere I can learn about about this? Also have you tried it yourself?
>>
>>4085492
What do you think is the other 50%
>>
>>4085286
Ive thought about the problem. Pretty easy for reversal stocks, more difficult of color negative. Basically, you photograph cc24 with 4x5 e100 or whatever film slide film you want to emulate, then get it developed and photograph it on a high cri light table. Take a spectrophotometer and measure the target patches on a light table to generate a .ti1 file using the film's CIEXYZ values as objective instead of the target's.
If you really want it to look exactly like the film does on a light table, build a 3d LUT, but this will mean if you over expose your shot, it will be blown out like film. 2.5d lut or matrix based profile would be more general, but look less like film. You'd also just want to make a daylight profile since slide film looks like piss under anything but high cri daylight LEDs or daylight.
Emulating the grain structure could be done by matching it visually in PS or LR to a scan.
>>4085339
This isn't true. A given film's response to light/color is very well understood problem. Kodak and fujifilm engineers and chemists publish these findings in the datasheet. Its probably possible to build a one-to-one sim(3d lut) of film this way, There are some complications bc films use a subtractive process with digital uses additive. However, additive processes are better in practice for a variety of reasons. So I think its possible.
>>
>>4085588

Im going to try this, do you have any socials or email I can contact you on, I want to stay in touch with this process
>>
Has there ben any news on Fugufilm after the initial announcement?
>>
>>4085286
looks like shit
>>
>>4085588
I might be wrong but I always thought it was strictly impossible with a LUT because, to express it mathematically, there is no linear relationship between film color and digital color. It always seemed to me that film color was only one part of the equation. When you shoot film, the output is way more sensitive to exposition and technique than in digital. I feel like you could get very different colors (but also contrasts ans textures) using the same film stock in different lighting conditions. The difference between film render and digital render is way more complex than a look up table. I think a properly trained neural network would be more adapted to solve this problem because it could model the non linearities.
>>
>>4085544
On /p the solution to your problem is always to buy a new camera
>>
>>4085878
That or maybe Panasonic's organic sensor.
>>
>>4085889
Yeah I don't know yet what it is but the images kind of illustrate my points, some things are just lost between film and digital, like when you irradiate another color light.
>>
>>4085878
>there is no linear relationship between film color and digital color
Not true, but even if it were it doesn't matter, nonlinear mappings have been a thing for well over a century. Mapping tonal shifts is trivial: use 3 mappings (1 per channel) using all 3 channels as weighted inputs, rather than 1 mapping and a fixed chroma offset. If you want to do it properly, you also take into account the temporal component (reciprocity failure), but as digital sensors are mostly immune to that issue you can choose you 'exposure' in post during the conversion so you can decide to just... not have any reciprocity failure.

As long as the dynamic range of your sensor is greater than that of the film you want to replicate, then a perfect 1:1 mapping is possible. If it is not, then you can still achieve an imperfect 1:1 mapping with either clipping or stretching of extreme values.

The sole issue you may encounter is with films that have a very LOW dynamic range, like Velvia 50: If you take a 12-bit digital image with a 15 stop dynamic range (for example) and only want to use 5 stops of that, your effective bit depth for the output will be dramatically reduced because you are throwing away so much of the captured data to achieve that desired mapping.
>>
>>4085878
Yes you are right, film has non-linear response to light where as digital is linear. There is no linear function between the two, however, a 3d lut allows for digital to behave exactly like film. 3d luts are non-linear (or linear) by design. They allow any RGB triple to be mapped to any other RGB triple. You can photograph a IT8 style target and fix the digital values to those slide film values measured with a spectro on a high CRI light table. If you have a good 3d lut that describes the way film reacts to light and color, it'll be just as sensitive to exposure as film is since lots of the highlights and shadows will be mapped the blown or crushed RGB values mathematically. 3d luts are very unforgiving, and often only used in studio settings when color and tonal accuracy are key.
>>4085940
That's just more work than making a high res 3d lut.
>The sole issue you may encounter is with films that have a very LOW dynamic range, like Velvia 50: If you take a 12-bit digital image with a 15 stop dynamic range (for example) and only want to use 5 stops of that, your effective bit depth for the output will be dramatically reduced because you are throwing away so much of the captured data to achieve that desired mapping.
You're retarded, this isn't how digital photography works. Thinking about it in terms of stops isn't useful or correct. Think about it in terms of noise. You don't 'throw away' the data, you map it mathematically. The noise stays the same. Perceptually, it will be reduced since its harder to see noise in dark areas and whites.
>>4085634
This thread has also made me want to try it. Youll need a spectro and know how to use argyllcms. Are you up to the task? Make a post here and ill reply
>>
>>4086011
I don’t have spectro nor understand know how to use argyllcms. I would love to help to anyway though. A lot of this stuff is flying over my head if I’m being honest
>>
>>4086011

>That's just more work than making a high res 3d lut.

It IS a 3D LUT. What do you think those extra Ds are for, fun?

>Thinking about it in terms of stops isn't useful or correct

Nope. Think about it in terms of photon counts (which is what stops are a shorthand for):
A given unit area of a sensor can accurately measure between 10 photons in a unit time or 40,950 photons in that same unit time and output a voltage in response, but fewer photons will not register and more photons will not register. That gives you your sensor's dynamic range. With 12 bit sampling, that gives you photon sampling in 10-photon steps (because sensor response is close enough to linear for practical purposes).
A given unit area of a slide film will transform some quantity of halide into metallic silver on exposure to light. Say, anything less than 1,000 photons won't even get any grains to switch, and beyond 20,000 photons all your halide is exhausted. To oversimplify, the number of grains of halide in that unit area is the effective bit depth (in a sampling-theory sense) - granted that over an image there is addition data in the spatial distribution of the grains that is not present with quantised pixels to actual bit-depth will be higher ion practice) - and the range of photons that produce a useful response is the dynamic range.

The problem comes in that your digital pixel is outputting a number between 0 and 4095, but you only care abut numbers between 10 and 2000. You MUST discard 0-10 and 2000-4095 if you are aiming for a 1:1 mapping, because those values should be clipped if the same unit area of photons fell on a piece of film rather than on the digital sensor. That's exactly the effect you want to replicate. The nonlinearity of the film grain response within that range is just a mapping problem from the linear digital space, but outside that range that captured data is worthless.
>>
>>4086011
>3d luts are very unforgiving, and often only used in studio settings when color and tonal accuracy are key.
Uh, no. Even movies shot on film use luts for live preview of dailies. Shit gets digitised and graded on the fly.
>A given film's response to light/color is very well understood problem. Kodak and fujifilm engineers and chemists publish these findings in the datasheet. Its probably possible to build a one-to-one sim(3d lut) of film this way,
It was tried before, and It's simply not possible. You're only emulating film, digital sensors have inherently different process of capturing light. What you'll get is something that will look the same on one scene, but as soon as you'll take that same grade to a slightly different light, it'll start falling apart. It'll need to be adjusted again and again to equal film. But that's not the real problem with grading, it's destructive process. It harms image quality by introducing noise and reducing dynamic range. This happens even when it's done on raw data. Bottom line is that digital is inherently different process, it allows for many things that aren't possible on film, at the same time it's still not possible to emulate it.
>>
>>4086079
what are you refencing when you say "it was tried before", and could you not make multiple 3D LUT for different types of scenes? Like the fin incandescent, fluorescent, daylight, flash, cloudy, and shade. And would it change enough where anyone would even notice?
>>
>>4086101
Are you really not aware of rni films, fuji's film simulations, dxo filmpack, adobes luts, mastin labs, vsco, and about a hundred youtoubers and blogers each with his own and only true film simulation? They all have this same problem. They're capable of reproducing film in one scene, but never across different ones. You'll notice that those that put forward samples, will only show one comparison. Which makes sense, it's not that hard matching one specific scene and making a cool promo shot out of it, on other hand, it would expose them when going for a set done in various different situations. As it is, all colour film simulations are colour grades. They take inspiration from film look, and aren't a bad thing in itself, but inspiration is all they can be.

I also don't see a reason why some of you are putting such a strain on 3d luts, they're not some advanced magical knowledge. They're just a preset storage format. A set of instructions from HSL tool in photoshop, RGB curves, or auto generated via a colour checker, that's it. In many ways they are more restricting than working with colours in raw editor.
>>
File: spectral.png (277 KB, 1246x839)
277 KB
277 KB PNG
All y'all need to take some information theory classes.

The reason you cannot create a perfect LUT is that when you map the spectrum of the incoming light onto RGB channels, the information which specific wavelengths the incoming light was composed of is lost. This makes it impossible to predict which color a film photo would have produced (unless the sensitivity curves had the same shape)
If you look at picrel for example, a 40% (of some arbitrary unit) exposure at 600nm would yield the very same "R" value as a 100% exposure at 650nm. The film on the other hand is much more sensitive at 650nm and would yield ~10 times more cyan (which becomes red after inversion) at the latter exposure. You have no way of telling that from the digital image.

This has nothing to do with linearity or dynamic range whatsoever.
>>
>>4086180
Since any given scene will generally feature only one type of illumination (e.g. sun = effective blackbody with known spectrum, colour-balanced studio LEDs will have a known spectrum with specific peaks, etc) then all you need in addition is either a colour meter and knowledge of the spectral response of your sensor to get an arbitrary mapping, or a-priori estimation of the illumination spectra (e.g. "direct sun" or "cloudy shade" or "indoor incandescent") to account for 99% of lighting situations. The only time that won't work is when illumination type varies dramatically across a scene (e.g. if you have half of a scene lit by tungsten and half by a fluorescent tube) but with comparable illumination levels, which is pretty rare.
>>
>>4086294
lmao this guy thinks “direct sun” is a metric.
>>
File: 41334486611_99f086a45d_h.jpg (225 KB, 1600x1184)
225 KB
225 KB JPG
>>4086294
>to account for 99% of lighting situations
Maybe when incandescent lights (blackbody radiation, best radiation) were widespread, with these new LEDs, its a crapshoot.
>>
>>4086299
It's a near blackbody radiator (with some spectral lines, inconsequential as not even speciality films have a sharp enough cutoff to discriminate them), filtered by the atmosphere. There's a good reason the D65 illuminant is a standard.

>>4086304
And how often is a spectrally critical film replication going to be needed outside of landscapes (Fuji is going to kill Velva eventually) and studio work? If you're just snapshitting in random people's homes then the existing filters are more than adequate.
>>
>>4086394
Lol he mad
>>
>>4085286

Get a sigma sd 9/10/14/15. It's like shooting slide film.
>>
>>4086073
This is so incorrect I'm not even going to try to correct it. Id checking out 'The reproduction of color' for the film stuff and https://www.photonstophotos.net/GeneralTopics/Sensors_&_Raw/Sensor_Analysis_Primer/Engineering_and_Photographic_Dynamic_Range.htm for the digital stuff.
There isn't a mapping problem going from high dynamic range digital sensor to a low dynamic range slide film. The Shadows and highlights get crushed mathematically to 0 and 255, just like slide film.
>>4086079
You'd need different luts for different lighting, this is true for digital too. Color grading and 3d icc luts are completely different things. Unless you're really boosting the shadows, which you won't if you're trying to sim slide film, 3d luts don't introduce more noise. 3d luts aren't distructive at all. RAW files can be re-edited with a different .icc enabled with a few clicks.
Film is entirely able to be simulated digitally. Sure, digital is different, but digital is better than film in a few key ways which allow digital to simulate film.
1) Subtractive color is worse than additive color in practice. Phosphers in CRTs have a larger gamut than even the best slide films (Hunt, reproduction of color, pg 130)
2) Subtractive color is worse than additive color in theory too. "As far is color rendering is concerned, all subtractive color processes suffer not only from the unwanted color shifts present in additive systems (that were described in chapter 2), but also further defects of their own.
3) digital sensors have linear response to light. This is crucial, since to simulate film, the characteristic curves of the CMY curves must be perfectly matched. This can done via a LUT.
>>
>>4086770
>There isn't a mapping problem going from high dynamic range digital sensor to a low dynamic range slide film. The Shadows and highlights get crushed mathematically to 0 and 255, just like slide film.

That's literally what I just said. Glad we're in agreement.

There is no magic going on: if you sample a range, and then clip that range, then you're throwing away data. You don't get to retain data and clip it at the same time.
>>
>>4086180
This is why you have to measure the slide film on a high cri light table with a spectrophotometer then generate a 3d lut from that. If the gamut boundary and neutral axis are measured correctly, the digital sensor response after the lut is applied will be the same as film for that light source and the target colors.
>>
>>4086299
It unironically is you idoit.
https://en.wikipedia.org/wiki/Standard_illuminant
Tungsten light (S_A) is also a metric.
>>
>>4086771
Ok, misunderstanding. My point is that the data isn't 'thrown away' since the icc profile can be changed to something else down the line.
>>
>>4086770
>Color grading and 3d icc luts are completely different things. Unless you're really boosting the shadows, which you won't if you're trying to sim slide film, RAW files can be re-edited with a different .icc enabled with a few clicks.

They're not the same, but they are very close. Luts are destructive and they do introduce noise. They're just very light on it. icc profile is a separate thing that does the same job as well. It's not advised for it to be used in an editing workflow, but it can be. It'll just do the same thing that luts or grading do. It's the first step that overrides everything else, that's the problem with it.

>>4086777
>My point is that the data isn't 'thrown away' since the icc profile can be changed to something else down the line.
Oxymoron. It's like saying that engine is still naturally aspirated, after turbocharger is attached to it, since it can be removed at a later point. Undestructive editing means that you can redo steps and change them as you work, applying icc profile is destrictuve, it's either on or off. Luts are samey, but they're can at least be applied at a later point.
>>
somebody just do it and prove it to be right/wrong

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Horizontal Resolution144 dpi
Vertical Resolution144 dpi
Color Space InformationsRGB
Image Width599
Image Height763
Scene Capture TypeStandard
>>
>https://www.dehancer.com/store/photo/pslr
>>
File: NZ road safety.jpg (188 KB, 1000x664)
188 KB
188 KB JPG
>>4085286
If you take pictures with road signs in them, they should be more fun.

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Equipment MakeOLYMPUS IMAGING CORP.
Camera ModelTG-3
Camera SoftwareVersion 1.0
Maximum Lens Aperturef/2.8
Focal Length (35mm Equiv)70 mm
Image-Specific Properties:
Image OrientationTop, Left-Hand
Horizontal Resolution314 dpi
Vertical Resolution314 dpi
Image Created2015:02:09 18:07:33
Exposure Time1/400 sec
F-Numberf/5.6
Exposure ProgramNormal Program
ISO Speed Rating100
Exposure Bias0 EV
Metering ModePattern
Light SourceUnknown
FlashNo Flash, Compulsory
Focal Length12.60 mm
Color Space InformationsRGB
Image Width1024
Image Height680
RenderingCustom
Exposure ModeAuto
White BalanceAuto
Scene Capture TypeStandard
Gain ControlNone
ContrastHard
SaturationHigh
SharpnessNormal
GPS StatusMeasurement Interoperability
>>
>>4086825
>icc profile is a separate thing that does the same job as well.
You're probably a retarded cine fag that doesn't understand how luts work in still imagery. Luts (and matrices) are implemented through camera input profiles, .icc and .dcp file extension. RGB curves adjustments are 3d luts, but require lots of effort to implement (I would know I use RGB curves layer to align the neutral axis for art repro work). Camera input profiles are non-destructive in the sense that you can change them with a few clicks and they don't change raw file data, only the rendered tiff. The raw file is the raw file, and you can apply a different input profile anytime you want to get a different rendered tiff.
>>
>>4086777
>My point is that the data isn't 'thrown away' since the icc profile can be changed to something else down the line.
Well, it is thrown away in the final image. If you remove the mapping, you no longer have the film-imitating image you wanted, and when you apply the mapping, you throw away the light values the sensor measured that are outside what the imitated film would have been able to capture.
You can either perform a correct mapping and throw away some data (data that is not needed in the final image), or not perform a correct mapping. You don't get to have it both ways.

Remember, you can't tell your sensor to capture the same bit depth across a narrower dynamic range. That's not how sensors in imaging digital cameras work. A handful of esoteric scientific cameras ay be capable of that sort of dynamically variable amplification range, but if you have access to those you have call-up-Fuji/Kodak-and-tell-them-exactly-what-film-you-want-them-to-make money (and probably work for the NRO).
>>
>>4087104
>You're probably a retarded cine fag that doesn't understand how luts work in still imagery.
Completely clueless, and then he starts with name calling... Luts don't work well for photography. Nobody uses them because they're just presets. If you knew how to make luts, you'd use the process involved as a preset because it's not destructive. But you're clearly just bullshitting. Smart photographer will simply use lightroom or photoshop presets instead. They do the same thing, only they allow for larger degree of customization and adaptation. Dumber will buy presets and buy luts, and fool around with them, thinking that he's doing something.

>Luts (and matrices) are implemented through camera input profiles, .icc and .dcp file extension.
You're talking about raw engine camera profiles as if they're luts. They are not. They are destructive even in raw editors. You can only select one, and then work from it. You can not modify it.

>RGB curves adjustments are 3d luts, but require lots of effort to implement
They're not restricted to that. Luts can be simple rgb curve adjustment, or aggressive individual hue adjustment.

>(I would know I use RGB curves layer to align the neutral axis for art repro work).
I really don't believe you at this point. Your knowledge is severely lacking. You're applying someone else's presets on your images thinking that they're giving you film look, and you've no idea how these presets work.
>>
>>4087104
what the fuck-shit are you on about? how ancient is your workflow that you still need to use curves like this?
>>
>>4086772
well yeah
my point was that a LUT that is accurate across all situations is not possible (which is what OP wants as I understand it?)
If you do it like you said you can indeed get nearly perfect results limited to that scene (except for corner cases like I described)

>the digital sensor response after the lut is applied will be the same as film for that light source and the target colors
I guess I'm being pendantic but the correct word is 'system response'.
'sensor response' strictly refers to how incoming light is translated into a voltage (and then digital number)
>>
>>4087226
>Completely clueless, and then he starts with name calling... Luts don't work well for photography. Nobody uses them because they're just presets. If you knew how to make luts, you'd use the process involved as a preset because it's not destructive.
This is what I have been saying the whole thread. Luts work very well for color critical work. I use them for art repro every time I have a job. I've made several custom luts, including 3d .icc luts using argyllcms, 2.5d luts(corrects chromaticity, but not brightness) using dcamprof. Custom luts to dial in color accuracy is standard in my industry.
>You're talking about raw engine camera profiles as if they're luts. They are not. They are destructive even in raw editors. You can only select one, and then work from it. You can not modify it. They can matrix based or LUT based. You're an idiot. https://argyllcms.com/doc/Scenarios.html#PS1
>To create a matrix/shaper profile, the following suffices:

colprof -v -D"Scanner A" -qm -as scanner

For an XYZ PCS LUT based profile then the following would be used:

colprof -v -D"Scanner A" -qm -ax scanner

For the purposes of a poor mans colorimeter, the following would generally be used:

colprof -v -D"Scanner A" -qm -ax -ua scanner

Here are the commands to CIEXYZ lut based, shaper matrix, and linear matrix icc profiles using argyllcms. https://rawtherapee.com/mirror/dcamprof/dcamprof.html#workflow_dcp
Here are the commands to make a 2.5d .dcp or .icc lut using dcamprof. Lets see how much seethe you have now that I've shown you that you're wrong you fuckwit.
>>4087293
I've done tests. Using a 2.5d lut (with custom illuminant) to correct chromaticity and then using RGB curves to correct the neutral axis (20 step tiffen grey scale) using spectro L*a*b measurements of the grey scale converted to AdobeRGB using matlab. It works better than making a 3d lut in terms of accuracy.
>>
File: DSCF5588.jpg (1.65 MB, 2160x1440)
1.65 MB
1.65 MB JPG
>>4085544
not op but i like your pic, do you have any recipes you'd like to share?

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationLeft-Hand, Bottom
>>
>>4087388
all that sounds really complex and ridiculous. i just toss the work and the color target under the phase one, snap it, click around the target a lil bit and it's perfect. i feel like you're doing too much man.
>>
>>4087293
I've come to conclusion that this dude is 100% larp. The shit that he spews is pure fantasy.
>>
bump
>>
ok so I found a Ektachrome 3DLut and here some examples I did. I did a pretty complicated setup to get this right. I did run into the problem of everything being blown out but I figured it out kinda but let me know what you all think. This is just color not texture or grain


before

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4089114
After

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width857
Image Height1280
>>
Before

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4089119

After

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width857
Image Height1280
>>
>>4089121
>>4089116
did it add the black borders too lmao
which one is this?
>>
>>4089121
Honestly this second one isn’t bad. If you want a really faithful slide reproduction just shoot any olympus point and shoot from the early 2000s.
>>
>>4089490
i added those for aesthetic purposes
>>
(OP and the one who posted the before and afters)

as of right now i think i have some perfect replications-- i think. Currently waiting for some ektachrome to be developed to compare. I did a whole session of shooting with my digital and my Olympus OM4ti, though did not use the same lens, but i did use the same focal length, and every setting was the same. will check back in.
>>
>>4089752
So, what happened?
>>
mastinlabs.com
>>
>>4089752
bump! it takes a man to fess up his screups anon.
>>
bump!
>>
Does anyone know what happened to op?
>>
>>4094746
mastinlabs.com
>>
>>4089121
thats quite impressive
has a weird quality to it that I can't quite figure out though
>>
>>4089116
Sorry anons, it was a failure and I can’t figure out why honestly.

There is no objective film look …

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Horizontal Resolution96 dpi
Vertical Resolution96 dpi
Color Space InformationsRGB
Image Width1140
Image Height756
Scene Capture TypeStandard
>>
>>4089121
The film ones look more like digital than the digital film emulations hahaha how? Idk

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Horizontal Resolution96 dpi
Vertical Resolution96 dpi
Color Space InformationsRGB
Image Width1140
Image Height756
Scene Capture TypeStandard
>>
I have learned the Ektachrome references I was looking at are usually bad scans. The purple / blue tint is just bad scanned film. Ektachrome is extremely true to life. When you shoot Ektachrome, expose perfect, scan perfect, it looks borderline digital. Especially if it’s nature. There isn’t much contrast either … Here’s more examples.


Digital photo edited for ekatchrome.

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Camera SoftwareInstagram
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width1440
Image Height1700
>>
>>4099424

Film shot…

There is a slight green hue though on this film shit

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Horizontal Resolution96 dpi
Vertical Resolution96 dpi
Color Space InformationsRGB
Image Width1140
Image Height756
Scene Capture TypeStandard
>>
Digital edited for ekatchrome.

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Camera SoftwareInstagram
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width1440
Image Height1659
>>
>>4099426
Film

I just don’t get it. I think it literally might be impossible with the tools and knowledge I have. Film is not at ALL consistent. I’m afraid the anons telling me it was impossible were right.
>>4087338

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Horizontal Resolution96 dpi
Vertical Resolution96 dpi
Color Space InformationsRGB
Image Width1140
Image Height756
Scene Capture TypeStandard
>>
I took a film and digital shot of this subject. This is the film version. I put both in Lightroom and tried to get the digital to look exactly like the film shot. But it just wouldn’t work no matter what I did. Maybe I’m just not that good at editing. I can’t tell what I was missing. Maybe it’s just —soul

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Equipment MakeNORITSU KOKI
Camera ModelEZ Controller
Camera SoftwareEZ Controller 6.60.031 (160615)
Image-Specific Properties:
Image OrientationTop, Left-Hand
Horizontal Resolution72 dpi
Vertical Resolution72 dpi
Color Space InformationsRGB
Image Width3089
Image Height2048
Scene Capture TypeStandard
>>
>>4099433

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Equipment MakeFUJIFILM
Camera ModelX-Pro3
Camera SoftwareAdobe Photoshop Lightroom Classic 10.1.1 (Windows)
Maximum Lens Aperturef/1.4
Sensing MethodOne-Chip Color Area
Focal Length (35mm Equiv)53 mm
Image-Specific Properties:
Image OrientationTop, Left-Hand
Horizontal Resolution300 dpi
Vertical Resolution300 dpi
Image Created2022:10:18 15:21:37
Exposure Time1/60 sec
F-Numberf/1.4
Exposure ProgramManual
ISO Speed Rating100
Lens Aperturef/1.4
Brightness1.6 EV
Exposure Bias0 EV
Metering ModePattern
Light SourceUnknown
FlashNo Flash
Focal Length35.00 mm
Color Space InformationsRGB
Image Width4160
Image Height6240
RenderingNormal
Exposure ModeManual
White BalanceManual
Scene Capture TypeStandard
SharpnessNormal
Subject Distance RangeUnknown
>>
>>4099433
"soul" is just the name you give something you don't know enough about to explain
>>
>>4099435
i can get the colours looking much more similar by spending a minute tweaking the rgb curves

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Equipment MakeFUJIFILM
Camera ModelX-Pro3
Camera Softwaredarktable 4.0.0
Maximum Lens Aperturef/1.4
Sensing MethodOne-Chip Color Area
Focal Length (35mm Equiv)53 mm
Image-Specific Properties:
Image OrientationTop, Left-Hand
Image Created2022:11:14 12:33:22
Exposure Time1/60 sec
F-Numberf/1.4
Exposure ProgramManual
ISO Speed Rating100
Lens Aperturef/1.4
Brightness1.6 EV
Exposure Bias0 EV
Metering ModePattern
Light SourceUnknown
FlashNo Flash
Focal Length35.00 mm
Color Space InformationsRGB
Image Width4160
Image Height6240
RenderingNormal
Exposure ModeManual
White BalanceManual
Scene Capture TypeStandard
SharpnessNormal
Subject Distance RangeUnknown
>>
>>4099608
the green hue still isnt "rich" enough in my eyes. but you did better than me
>>
File: cover3.jpg (126 KB, 1920x1080)
126 KB
126 KB JPG
>>4099606
figure it out then anon
>>
some more Ektachrome emulation

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4102677

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4102677
If anyone with lots of experience with Ektachrome plz lmk some consistent characteristics Ektachrome has that I may be missing

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4102677

Right now I feel like I’m lacking proper grain for sure. Just been using capture one’s grain which is mono (not sure why they don’t do colored grain) if anyone has a good idea how to get good positive film grain plz lmk

Also I think I mostly understand how ektachrome works outdoors under good conditions. But when it comes to indoors i really don’t have a clue Or just any complex lighting.

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width1266
Image Height1280
>>
>>4085286
Yeah there is the Film Pack 6 by DXO. There are probably more but thats one that I know of.
>>
Ektachrome 100 on 120 is by far the best engineered film stock. Perfect balance, realistic colors, 0 grain, a bit contrasty at times but other than that this is a borderline digital image.

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationUncalibrated
Image Width879
Image Height1280
>>
File: e100.png (1018 KB, 923x731)
1018 KB
1018 KB PNG
>>4102677
the grey axis on e100 is pretty damn neutral. The blues look good tho
>>
>>4085286
First you'd want a camera with a strong CFA from the 2000s. Modern ones sacrifice color separation for low light performance.

Use an old lens lens flares, soft focus, chromatic aberration and vignetting.

Then you'd apply a film LUT and add grain.
>>
>>4086304
>flashlights optimized for brightness have bad color reproduction
shocker
>>
>>4087226
>Luts don't work well for photography. Nobody uses them
(X)
>>
>>4085889
wtf I'm becoming a filmfag now
>>
OP here

well I did it.

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4106422

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4106422
elaborate?
>>


[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4106424
I can emulate Ektachrome pretty well. Probably not 100% accuracy, there is no “true” Ektachrome look since the look and colors are changed by scanner, emulsion, development, camera, lens etc etc but there is ofc a theme between them. I definitely got a good idea how it reacts in bright daylight. Just need more references for under different conditions. I’ve been getting scans from the darkroom lab with 0 edits, the rawest the film scan can be to use as references.

Film is almost “random”, I think there are some artistic liberties you can get away with. For example shadows being slightly purple or blue or even green is *random* on ektachrome for every scene, so just choosing whatever you think looks good isn’t “inaccurate” cuz the film is 100% capable of doing that. But the shadows being red doesn’t really happen.

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Image-Specific Properties:
Image OrientationTop, Left-Hand
Color Space InformationsRGB
Image Width853
Image Height1280
>>
>>4085286
>trying
yes
>succeeding
never
>>
ok hear me out

I shot a roll of Ektachrome 100 on a Fujifilm GW690 camera, simultaneously with my XPro 3. I got back 8 shots and I exactly match them with the digital ones (classic chrome is a huge help with matching). Most of these photos were all in the same conditions, bright sunlight. So I edited a digital photo exact as the film and copied the adjustments from the digital photo and pasted it on another unedited digital photo, and they almost exactly matched its own film photo. The only thing that need to be changed slightly sometimes is usually the color temperature. Capture One has a great smart adjustment tool for matching images.

It's very hard to explain but all the film photos were taken on a sunny day around 1pm-4pm, I specifically have a preset on Capture One now that can emulate almost exactly how ektachrome would react during that time for my digital photos. And i think it works pretty well. When i use this preset it match very very well with a ektachrome counterpart. I hope all this makes sense.

i feel as film presets fail to realize exposure doesn't shift THAT much with film at all if at all, but its colors majorly shifts and reacts so differently under different light conditions, i have never had a preset pack categorized into "Bright outdoors" "Cloudy Outdoors" "Indoors" "Night" which i think they should be.

i have yet to test out cloudy, indoors or any of that
>>
>>4106484

i will do the ultimate test.

4:00 on a sunny bright day, I will take a picture with my digital camera, post it here and then take a photo with Ektachrome 100 of the same subject 4:00 on a sunny day, apply this method without even seeing the film counterpart to the digital image and we shall see if this is all farts in the wind when i have both images.

will report back ...
>>
>>4106484
You realise ektachrome was created in an attempt to emulate digital look on film right? lol
>>
>>4106487
Ektachrome was invented in the 40's you fucking mongoloid.

God DAMN i hate this fucking board.
>>
File: anime-cry.gif (44 KB, 220x124)
44 KB
44 KB GIF
>>4106487
gotta be bait ...right anon? Your not that dumb are you?
>>
>>4106488
>he thinks he's shooting le epic antiq 40's film
the brand name yes... the film in your camera not so much, try re-formulated in 2018 numbnuts

But stop deferring from the fact you haven't posted these so called results you've already taken, LARPing again?
>>
>>4106486
>>4106484
>>4106431

ps:

just realized this sounds all over the place like different ppl but this all me (OP)
>>
>>4106491
nigga you said "Ektachrome was created ..." like a retard and didn't at all specify the 2018 E6 emulsion

sounds like you just dont know what your talking about
>>
File: 000452580004 1.jpg (4.49 MB, 2900x3924)
4.49 MB
4.49 MB JPG
>>4106491

I cant just post an image and say "this is exactly how ektachrome would look" youd need an actual film image too to compare and for transparency i'd need to take the digital image first , apply the preset, then take the image on film, wait for the film to be developed comeback and see if it matches. Thatd be the truest way to show results, which i already said here>>4106486


as of right now i can show you the preset on an already existing image.

The top is film and the bottom is digital


also where did you see it was made to emulate digital?
>>
File: DSCF7783 2.jpg (4.53 MB, 6240x4160)
4.53 MB
4.53 MB JPG
>>4106500
and here is the digital image with 0 adjustments

[EXIF data available. Click here to show/hide.]
Camera-Specific Properties:
Equipment MakeFUJIFILM
Camera ModelX-Pro3
Camera SoftwareCapture One 16 Macintosh
Maximum Lens Aperturef/1.4
Sensing MethodOne-Chip Color Area
Focal Length (35mm Equiv)53 mm
Image-Specific Properties:
Horizontal Resolution300 dpi
Vertical Resolution300 dpi
Exposure Time1/60 sec
F-Numberf/10.0
Exposure ProgramManual
ISO Speed Rating100
Lens Aperturef/10.0
Brightness8.1 EV
Exposure Bias0 EV
Metering ModePattern
Light SourceUnknown
FlashNo Flash
Focal Length35.00 mm
Image Width6240
Image Height4160
RenderingNormal
Exposure ModeManual
White BalanceManual
Scene Capture TypeStandard
SharpnessNormal
Subject Distance RangeUnknown
>>
>>4106496
>nigga you said...
>what is context
i'm sorry you can't comprehend at a 3rd grade level anon
>>
>>4106507

hey buddy you never added context, next time add a little more information if you want people to understand you and have meaningful conversation, like an adult.
>>
>>4106513
context was provided by the poster
>OP: I'm currently shooting a type of film
>rest of /p/ OMFG you're shooting 1940's version of film OMG WOW! Can I suck your cock
seriously... its pretty obvious he's buying new E100 off shelf which is 2018 emuls version, so shut up and get a grip on reality
>>
File: db7.jpg (83 KB, 680x847)
83 KB
83 KB JPG
>>4106487
>>
>>4106520
do i really have to explain basic communication skills? The context is Ektachrome, OP never said E100 2018 emulsion, this whole time they could be talking about the classic 70s ektachrome look, or they could be using expired film which is a popular thing. That why we say "E100" which directly refers to the 2018 emulsion, just like other people in this thread are doing. When you say "ektachrome was created..." you are including the entire brand because you simplified it by just saying --the brand. Its pretty simple really to understand. basic common sense
>>
>>4106553
>t. can't read
He literally said he took photos recently in ektachrome and was trying to emulate it in digital, god /p/ smells so much like reddit today ffs i'm out
>>
>>4106491
Got a source for "IT WAS FORMULATED TO LOOK LIKE DIGITAL" or are you just pulling that out of your mongoloid ass?
>>
>>4106560
Sorry for the delay princess, I had to dig up my old thesis where I referenced this little tidbit back in 2019, unfortunately it appears to be down...

https://imaging.kodakalaris.com/professional-photographers/professional-films

and even the archive seems to be scrubbed (picrel)...holy fuck, did they rescind cause you reddit faggots got all uppity about your film emulations?
>>
>>4106564
ok so no source, now stfu
>>
>>4106570
ok, now show me your source that OP is shooting 1940's ektachrome?
>>
>>4106564
oh okay so right out of your ass it is then
>>
>>4106555

lol literally said ektachrome was created to look like digital and now your crying cuz you misspoke, can you even admit you should have worded it different anon? Can you admit that?

your just like your father
>>
>>4085286
>groups or communities
could you fuck off and be your own person?
die if not
>>
>>4106577
>using a hobby related imageboard for the purpose of communicating and sharing ideas and opinions

kys
>>
>>4106574
can you at least admit that you're overreacting over such minor grammatical error? sorry i wasnt born in a priveleged white country
>>
>>4106582

what the hell kinda excuse is that? do you want me to feel bad cuz your a retard? English isnt even my first language, nor did you even make a "grammer" mistake really, you just dont know what the fcuk your talking about, grow up
>>
>>4106589
>>4106592
wtf why u delete my posts? you a fucking mod schizo?
>>
>>4106581
I know what you are.
>>
File: maxresdefault.jpg (158 KB, 1280x720)
158 KB
158 KB JPG
>>4106589
your starting to make me feel bad cuz you might actually be retarded, heres something maybe even you can understand.
>>
File: keyboard-kitten.jpg (73 KB, 785x355)
73 KB
73 KB JPG
>>4106596
you caught me
>>
>>4106602
Your inability to elicit even the most minor chuckle is evidence of exactly what I suspect you to be.
You do not belong here. Leave. Now. Forever.
>>
>>4106612
I don’t know about you but I chuckled a bit.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.