Thread 4449731

So Canon have just announced a 410mp camera sensor.

Wonder what the resolution would be like. Would love to see some landscapes shot through it.

Is the solution to Bayer color fuckery just pumping resolution so hard that it becomes negligible?

Canon will make a special 24mm f/1.4 lens that they will call "universal prime" with a ring that controls digital zoom and sell it to use on this sensor

>>4449733
I mean, the entire point of bayer is that it reduces color information by roughly 1/2, so if you want to get equivalent MP you need to have double the resolution

>>4449731 (OP)
No-photos will buy it in a second.
The more megapixels the better, right?

>>4449736
No photo here. Yes I will buy it, I need it to take more photos of pregnant black women for amazon ads that will be postered on building outside of your metropolitan bughut

>>4449734
Isn't that just the VCM?

Per pixel snr will be ass and fuck tonality without lots of light. Sensors are not 100% efficient. As resolution increases total photosensitive area decreases.

>>4449731 (OP)

you could pixel bin one of those photos and still too big for /hr/

>>4449731 (OP)
>resizes to 8.3MP for le 4k experience
>it's-all-I-sneed.jxl
>Bayer artifacts still appear

>>4449733
>*clap*
YOU
>*clap*
DO
>*clap*
NOT
>*clap*
>NEED
*clap*
ANY
*clap*
MORE
*clap*
PIXELS
*clap*

The human eye can't see berries, let me guess, you need more?
You're literally cherry picking at this point but they're not even cherries here.
Bayer is statistically nyquist diagonally technically measurably scientifamagically incapa-I mean it actually delivers its advertised resolution you're just hard to please!

For real though we're doomed to suffer until cameras start giving us gigapixels in full frame. Anything less is pathetically low resolution. It's bad enough we rarely get anything below ISO100 but the fact that it has such low dynamic range, fake colors, aliasing, and can't even capture grass right is just sad. We should have ISO1 digital sensors but marketers are too obsessed with delivering low res bayer slop (and even doing QUAD bayer, looking at sony...) and claiming it's night vision all while refusing to give us actual infrared that would actually give us night vision like CCTV cameras have...


>>4449731 (OP)
A bit more than 24000x16000 pixels. Basically just two linear double steps above a 24MP sensor. 24MP sensors currently alias like fucking crazy so I doubt it would fix much. It would definitely be better but not enough to eliminate issues unless you shoot at something like f/16.

>>4449981
A 45mp bayer camera outresolves 645 film. You literally do not need more. You need to learn to accept that cameras are not tuned for SOOC output and shooting jpeg is neither a flex nor a goal. Shoot raw and competently downscale to 75%. That is the actual resolution of your camera. Aliasing happens at the nyquist limit so do a once over (it's like 5 minutes of brush+moire tool in C1), add a grain sim to disrupt any flat spots, and downscale.

And it still outresolves 645 film.

You are legitimately hard to please when you can buy a camera that outdoes 645, for the price of a 645 shitbox and 10 rolls of film shot and developed, and yet still complain and demand gigapixels. Like what, you want FF to outresolve 4x5? With what fucking lens?

>>4449733
What the hell is going on with that picture? Doesn't look like a normal 5D2 pic to me. 100% crop with maxed out sharpening?

>>4450019
It's just a crop of a bush with sharpening increased to highlight the number of berries

A 35mm film version would be more informative because despite the low resolving power it would still have some more red dots in there. This is just a hyper specific bayer issue with the red/green pair.

>>4449731 (OP)
all to be mostly displayed on a less than 1440p smarthphone screen :)
0,5GB per RAW btw

>>4449731 (OP)
>410mp
This should have been the R1

And what they currently as an R1 should have been the next mark in the R3 line

>>4449981
>>4450019
There are 2 people on this board that need to shut the fuck up
1. The guy with the Canon 5Dmk2,
2. The guy obsessed with bayer layouts
You're both subhuman degenerates and need to fuck off

Now that I think about it the guy who keeps reposting "lumix vs snoy" colors with a wrong whitebalance pic of his sister can fuckoff too

>>4450720
based

>>4449981
The only thing shooting high res cameras made me realize how how much I actually don't need that resolution

>>4449733
the 5d shot looks really compressed. Also 4x5 is like 150mp. Not a great comparison

>>4450687
The R1 confused me, but then so did the R5II and then the R50V (at least they're trying *something* different with that one). All Canon needs to do is release an R8II with a BSI sensor, a full shutter, and fix the overheating and it'll sell like crack. The most sold canon units currently on offer are the small and lightweights, but they cripple the fuck out of them.

The R3 and R1 are of no use to basically anyone on the board, but at least they're not suffering from the R5II's special brand of bullshit.

>>4450687
The 1D line was always a lower resolution than the contemporary 5D as far as I'm aware.
>The 1D is for taking hundreds of photos of a heron diving for a fish
is actually true
Also >>4450591

>>4450932
The r5ii makes a lot more sense if you forget about it being a photo camera and think of it as canon tossing a video bone to fags who would otherwise not buy a C series camera, but think themselves above the “camcorder” lines. Hybridization was the biggest mistake and the downfall of modern photographic devices

>>4450939
They at least did the R5C for videofags which was physically different enough to seperate itself from the mainline R5. It's like for the R5II they just went,
>we don't have anything meaningful to add to make an R5II (despite 4 years to innovate) but we need a new model
>let's just forget about photography and just rebadge the R5CII, who's going to notice the comprimises? Autistis on the internet? Lmao nice one cANON.

>>4450746
>Also 4x5 is like 150mp

4x5 is 10 times that at minimum.

>>4449735

It's actually 3x since it has a 3 color matrix.

>>4450017
>A 45mp bayer camera outresolves 645 film.

Your findings are based on current scanner limitations.

>>4449731 (OP)
This is only useful in big telescopes for capturing distant galaxies and nebulae

>>4451022
and you are mother.

>>4450687
>>4450938
What's the difference between the r3/3D and the r1/1D line then?

They're both lower res but faster than the r5/5D line. They're identical. This sucks. They should differentiate one of the two lines.

My suggestion:
r5: High res
r3: Fast
r1: Flagship that is both high res and fast no matter the cost

>>4451118
anon canon does not make cameras for people who buy cameras. canon makes cameras for businesses that buy crates of cameras with service contracts, and hands them to retards. everything makes sense when you realize this.

>>4451015
no, that's wrong.

people have been scanning film with phase one IQ backs and top tier optics for years. 700mp scans of 6x9 simply enlarge grain and the lowest speed slide film on the greatest camera, shooting a high contrast subject (film requires high contrast for high resolution) produces an image that is pointlessly large if scanned beyond 100mp. unless you enjoy looking at the shape of each individual dye cloud?

645 slide film might be a touch better than 45mp bayer if ran through an ideal camera with an ideal, brightly lit, contrasty subject but that doesnt mean shit to you. you aren't going to buy a gfx100s or p1 iq4 to scan film, ever, and with the film gear and film stocks people actually use 6x9 struggles to best a d850. your mamiya 645/bronica isn't going to milk peak resolution out of shit and kodak gold 200 isn't anywhere near the actual real world performance of velvia (discontinued!).

if you want resolution autism film is not it. use it for the unique color palette or to do real life darkroom work. shooting film to make high resolution scans of film is just worse digital with extra steps, or a scientific exercise best left to kodak's engineers.

>>4451129
Watch out, the filmboys will now start talking about CMS20 despite 1: never shooting CMS20 2: being unable to use CMS20 with its specific developer 3: not owning a camera that can keep up with CMS20 4: not having the skill to even hit focus or shoot without camera shake/subject motion blur

>>4451129
>canon makes cameras for businesses that buy crates of cameras with service contracts, and hands them to retards. everything makes sense when you realize this.
Ik, I actually worked for a business like that. After Abitur I jobbed for a wedding photographer and the business just had a dozen 5Dmk4s. That's how I got into this hobby.

>>4451129
In real world photography, settings are rarely ideal enough for even "200mp" 4x5 to outdo an 80mp digital

4x5 at 200mp is 4x5 shot at f8 with a short exposure, and the camera on a concrete pad, and then scanned on a digital camera making the entire exercise pointless. Anyone here can take 200mp photos by doing handheld panoramas with a $300 D800E.

>>4451130
I've shot cms20 4x5 and I'm never doing it again. Way too annoying and the developer is expensive. The result is quite nice in terms of pure resolution, but larger grained film offers more pleasing tonality.

>>4451015
>maybe some day scanners will be good enough for film!
scanners have been outresolving film for ages given competent drum operators and quality facilities
the real world difference between 61mp and 4x5 is negligible.
https://www.mountainphotography.com/gallery/4x5-film-vs-digital-resolution-comparison/
do you have high hopes for 645 then?

>>4451135
>scanners have been outresolving film for ages given competent drum operators and quality facilities

No they have not. Anyone who has viewed film under a microscope will tell you that no digital scanning device even comes close to the true resolution of film.

>>4451169
take your meds

digicuks btfo once again by a piece of funny plastic

>>4451171
>a piece of funny plastic
>Film?

>>4451169
lmao color film has been scanned under microscopes and there was no additional discernible detail. its just an expensive way to upscale images and look at grain structure. black and white film is impressive but only because real monochrome digital cameras are rare. a gfx100s mono would be cms20-destroying.

>>4451169
>filmcucks coping because using film does not make a photo better
lol

>>4449733
>Bayer color fuckery
This is a solved problem.

>>4451174
Silver particles are literally smaller than the planck length, film has infinite resolution. If you take a picture of a beach you can zoom into the film negative and see the electrons in the grains of sand. Keep coping digicuck

>>4451177
its also a non issue. bro is comparing a 20mp camera to an ~80mp (optimistically) camera and saying bayer is the problem.

maybe darktable and lightroom users have bayer issues lol

>>4451174
It wouldn't even beat 35mm CMS let alone medium format CMS which is like 1500MP shot with 645 gates.

>>4451181
>nophoto copes
talk all you want but you will never post evidence, just your busted logic based on made up facts and zero photographic experience

>>4451178
>Silver particles are literally smaller than the planck length

They are not.

But for the sake of this argument they might as well be.

>>4451178
take your meds anon

>>4451175
Quality wise no, character wise, yes.
Plus, rilm actually takes skill and is a symbol of status.

>>4451208
you cant tell film from edited digital. remember /fag/? post exif /p/ merged it with /fgt/ lol

>>4451234
the grain adds sovl to pictures, it's not a REAL picture without grain and inaccurate colors

>>4451208
>5 stop exposure latitude takes skill

>>4451182

He is right. You are WRONG! I have conducted such tests. CMS20II is capable of resolving 250-lp/mm. The color GFX100 maxed out @ 74-lp/mm which is less than the lowest resolution color film @ 78-lp/mm.

I estimate that the monochrome GFX100 isn't likely to exceed 120-lp/mm.

>>4451238
>the grain adds sovl to pictures, it's not a REAL picture without grain and inaccurate colors

Tell that to the worlds most accurate and realistic color medium, Ektachrome.

>>4451132
The truth everyone is missing everytime is that What's good about digital camera is not necessarily superior IQ. It's conveniency. Before digishit you have to wait to even preview it until getting wet and dried the film you shot.

>>4451353
Real photographers can preview images entirely in their mind. No screens necessary.

>>4451356
This. The real photographers are the painters.

>>4451357
A real painter only needs light to paint.

>>4451348
GFX 100 with a ~33mm tall sensor at 8736 quite literally means the highest theoretical possible atom-aligned ultra perfect ideal scenario is a max of 132l/mm
In practice you get maybe half of the theoretical maximum so it's cloer to 66 than people want to admit. Remember these are line pairs, meaming white/black not just detecting white or whatever, line pairs require dividing sensor pixels per mm by at least 2 in theory but usually 4 in practice because pixel grids don't actually measure lines to begin with and the more pixels you have, the better the line pairs get resolved (more contrast)

Add a little distortion and a rotation and imagine a 1920x1080 display showing you 540 white and 540 alternating lines. It's impossible. You need 4K to show 540 diagonal line pairs well and that's how digital sensors work. Film doesn't care about lines being diagonal or distorted they just deliver that line pair resolution for whatever you feed it becuse the film structure isn't a finite grid of individual pixels.

>>4450720
t. rents his pro sony body

>>4451348
People keep saying this but it has never been proven with a PHOTO

And they won't because film requires an absurdly high contrast ratio to reach its peak resolution, and the cameras would miss focus slightly anyways because focusing to 250lp/mm precision, and finding a film camera lens that can keep up, is unlikely

People also claimed 6x9 velvia (which isnt even made) was 200some mp but a guy shot a flashed chart with a god tier mamiya 7 and got 80mp, and P1 Iq4 based scans agreed, more than that is just bigger and blurrier

>>4451731
that's the way to do it. anyone who buys a pro body as a hobbyist/gig worker is peak dumb.
>yeah i wear a $4000 necklace but it's really ugly. i need to shoot at 40fps.
is the magic autofocus of lesser cameras not enough for you? a single shutter press will basically always be technically correct, but for real this time because we all have 90%+ AF coverage.

>>4451734
Nah I shoot M-mount lenses adapted to a body with thin sensor glass.
Anyway I'm glad I can identify your posts even without a tripcode, but you should consider using one for the convenience of others.

>>4451736
who's "you"?. im not even >>4450720

you are schizophrenic. not everyone who's against buying a high end snoy is the same person. sony and canon make great professional cameras but they suck for anything else.

>>4451733
>People keep saying this but it has never been proven with a PHOTO

Because this is currently impossible outside of lab grade tests.

>>4451750
>put film in camera
>take 36 pictures
>develop film
>scan normally (ie: hit it with the gfx100s and pray to god it doesn't fall apart at "just" 100mp - you had 36 frames to nail it bro)
>pick a landmark in the best image
>scan that with a microscope and include a little micron scale
>do your maffs
>???
>Proof outside the 9,000,000:1 contrast data sheet claim
or do you mean it can only resolve that detail with lab grad tests (ie: bare film against a finely machined metal comb exposed with flash)

>>4451751

I am going to say this again. Scans are WHOLLY INSUFFICIENT for testing the resolution of film!

>>4451761
>film's resolution is magic, put a digital camera in front of a microscope and it magically turns back into 12mp because uhhhh bayer
films resolution is like doghairs talent

^rent free

>>4451761
>>4451750
You are absolutely correct.

>>4451750
>film's resolution can somehow not be proven when we have electron micoscopes used for looking at ant faces and laughing at how funny they look
Even a 100mp scan of cms20 35mm would be acceptable, alas, no one can pull it off. it might as well not exist.
>here's my 1 gigapixel camera, but if you don't say the magic word, it's only one megapixel
>whats the magic word
>iunno
>iunno?
>*1mp*
>no, i dunno
>i dunno?
>*1mp*
>dunno?
>*1mp*

>>4451806
You people asking for examples are legitimately retarded.

CMS 20 is rated for 800lp/mm.
One pixel can't resolve a line pair. You need at least two, and probably want more. More is always better. For the absolute minimum you must have two. This means you need 1600 pixels per millimeter to resolve 800 line pairs, under ideal tests.

Simple math tells you that means you need 2.2GP and will probably want 8.8GP to properly digitize it. That's for 35mm.

36mm width gets you 57600 pixel width at 1600 pixels/mm
24mm height gets you 38400 pixel height at 1600 pixels/mm
Sampling at 3200 pixels/mm will give better contrast resulting in 115200x76800 pixels of 8.8GP.

CMS 20 is fine and other films resolve less but some films resolve even more.
Arguments about limits are only relevant in respect to 35mm format. What film offers for larger formats like 4x5 are effectively infinite even for high ISO films. Your optics and shooting outside of a lab will never saturate film's resolving capabilities and film will never be the limiting factor. Digitizing it is.

>>4451813
But you can't do this >>4451751
Because the lp/mm rating is at an absurd contrast ratio, and film does not perform in real life.

Most medium format is outresolved by a canon 5ds. Not even the r, the one with the AA filter. The differences between 4x5 and an a7rv are negligible. And that is with discontinued ultrafine slide film.
>film has infinite resolution but nothing can use it
So it's a theoretical spec at a theoretical contrast ratio
>digitizing it is the limit
You fags can't get one person take a photo of a woman and microscope scan her eye to oo and aa at iris fibrils. Every time someone tries it's like, 6x9 = 80mp optimistically, <50mp usually, 4x5 200mp extremely optimistically, <80mp usually. There are 400mp scans of 4x5 tmax out there where the scanner insisted on "extra detail" but it isn't discernible as detail, it's very faint dotted lines.

Films resolution goes to another school

>>4451813
Correct.

>>4451813
Wow. It all makes sense now. Thanks for the explanation.

>>4451825
>>4451819
most pathetic samefag on earth lol

>>4451806
What an embarrassing post. Completely braindead and woefully ignorant! You are like a preschooler talking to a phd graduate.

Why don't people just post good scans of film they're shooting
Seems a lot easier to prove a point about how a photo looks with pictures instead of hypothetical text

>>4451751
Just use multi-shot mode. With a GFX100 it'll be 400MP without any Bayer memes. Compare it to a regular shot taken at 100MP and for good measure a regular 24MP fullframe camera.

>>4451879
Dishonest retard or ignorant retard.

>>4451950
Care to offer an amendment to my suggestion, O wisefag?

>>4451877
Because hypothetical text is all they have

I have seen phase one IQ 14 scans of kodachrome, made by the world's first and foremost scanning experts, and "glorious slide film" (ISO 50 or something) is at about the level of a canon 5diii, if the 5diii were shot at ISO 400

I have seen and pixel peeped an 800mp IQ 4 scan of 6x7 (or was it 6x9?) and despite how fucking huge it was the smallest details on it were enlarged so much it actually looked better <100mp

Pixel peeping experts claim exactly 80mp

I'm sure there's some cope coming now