Remix.run Logo
markhalonen 7 days ago

woah I am the author. I don't even have analytics set up on this site, but hope everyone enjoys it!

dataflow 7 days ago | parent | next [-]

Feedback: I absolutely love the idea of doing analysis like this, but it's incredibly frustrating to be shown photos that were clearly taken at different times when the subjects naturally don't look exactly the same. Like for example who's to say that player isn't actually leaning? The second photo sure doesn't prove anything. And comparing them side by side feels like an exercise in frustration.

I would probably (if possible) repeat this idea but with photos taken at the same time, with cameras as close to each other as possible. If at all possible I would also try to use as similar of a lens as possible, if only as a 3rd comparison point to compare the other two to.

multiplegeorges 7 days ago | parent | next [-]

The building shot perfectly illustrates all his points, very little difference between them.

The child in the surf is almost identical. Maybe a few ms of difference, look at the foot position.

The facial structure differences in the players were striking despite not being identical shots.

markhalonen 7 days ago | parent | prev | next [-]

you'll have to believe me when I say they are not leaning. They were just standing there posing for the photo.

Would love for someone else to get more scientific about it, but I think the results would be the same.

dataflow 7 days ago | parent [-]

> you'll have to believe me when I say they are not leaning. They were just standing there posing for the photo.

I mean, if believing your words were enough to convey the message, then there'd be no point in taking the second photo and comparing them.

The point here isn't whether you're telling the truth (of course you are), it's about being able to see what's going on and get an intuitive feel for what changes and what stays the same. When I said "who's to say they're not leaning" my point wasn't to call you a liar; it was to say that that question is what immediately arises in your audience's brain, and it's completely distracting. Trust can't correct for the visual discrepancy, even if I had taken for the photo myself.

fallinghawks 7 days ago | parent | prev [-]

I was pretty irked by that as well. The change from smiling to not smiling affects face shape. But at least the building and car photos were stationary enough to illustrate the fisheye quality.

turnsout 7 days ago | parent | prev | next [-]

I like the comparisons! I think it's 100% fair to compare the "out of the box" images from the iPhone to other cameras. With that said, some notes:

I think a lot of the differences you're seeing are the result of FOV differences; the iPhone camera is a ~24mm equivalent, which is much wider than most people would shoot on a dedicated camera. That wide-angle distortion is just a natural part of the 24mm focal length, but not really the iPhone's fault.

The other effects you're seeing are related to Apple's default image processing, which, at this point, most people would agree is too aggressive. This difference goes away if you shoot in ProRAW and process your photos in an app that allows you to dial down (or ideally turn off) local tone mapping.

If you have an iPhone that shoots 48MP ProRAW, don't be afraid to crop the image significantly, which increases the effective focal length and makes the image look more like a dedicated camera. It also increases the apparent bokeh, which is actually quite noticeable on close-ups. With the RAW you can then quickly edit the image to end up colors which are much more faithful and natural.

If anyone out there doesn't have a Pro model, they can shoot RAW photos in 3rd party camera apps, including Lightroom, which is free.

sturza 7 days ago | parent | prev | next [-]

One observation i'd expected to see is sensor size versus apparent focal length - this might be at least one of the reasons for distorsion. iPhone camera is ±7mm, which is ±4x crop factor in 35mm terms - but it's marketed as ±26mm.

shinycode 7 days ago | parent | prev | next [-]

There is apps like Halide or Photon that have a Process Zero or TrueRaw mode that is more natural. Of course a phone is just an other tool with different constraints. I gave up paying 2 or 3 times the price of my phone for a dedicated camera. I like the lightness and integrated software to edit photos and share them on the spot. I made that sacrifice knowing I’ll never have the same quality but I don’t have to carry a big camera now. But for passionate people who want the best you can’t replace a dedicated camera with a phone

datadrivenangel 7 days ago | parent | prev | next [-]

What kind of camera was used for the non-iphone shots?

markhalonen 7 days ago | parent [-]

sony a6400 with sigma 30mm f/1.4, but then the child one is a 2004 Digicam I think a Sony Cyber-shot DSC-W5

ticklemyelmo 6 days ago | parent | prev [-]

Much of the criticism of the Iphone photos is the fisheye effect. This is exaggerated, because you took the photos from different distances. If the Iphone photos were taken at the same distance, a cropped version of the Iphone photo would have identical perspective.