Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Menneisyys2

macrumors 603
Original poster
Jun 7, 2011
6,004
1,108
Yes, it's always worth shooting with 48MPixel PNGs. And if you can't handle the speed - use 12MPixel ones. Examples below.

I did some extensive experiments, with a tripod in front of one of my (under-construction, LOL) game / book / CD / camera / clock etc. shelf, with my new precious and found out that:

1, if you do shoot compressed, JPG and HEIC files result in equal(ly bad)-quality images. They're, as have always been, an oversharpened mess.

2, you should AT LEAST shoot 12 Mpixel DNGs. (You WILL be forced to shoot them if you for example use the flash, even in good light where you only use the flash to cast some light in shadows!) The JPG exports done in MacOS' Preview are FAAAR better than the compressed garbage coming out from the camera - and even at a lower filesize!

(And the same stands for Preview's HEIC output. They can safely be used for, say, archiving - no quality loss there compared to JPGs, and the file sizes are vastly lower.)

3, and, of course, whenever you can do (you can put up with the slowness of even M1 Pro MBPs scrolling thru dozens of 48 Mpixel DNGs etc.), shoot 48 Mpixel. There is a vast resolution advantage particularly with contrasty, black-and-white content. (With low-constrast, colorful and/or very low-light content, the difference isn't as huge.)

OK, some examples. Note: the full-res images are all at https://www.flickr.com/photos/33448355@N07/albums/72177720302306542 . The image/filenames do have the original file format as their prefix (jpg, heif, 12, 48.) Do use the Download icon particularly with the 48 Mpixel image ( ); then, select "Original (8064 x 6048)" to get the original files for pixel peeping - flickr, even with Pro accounts (like mine), only renders around 6 Mpixel worth of data when fully zoomed in, even on larger, hi-res displays if you "just" use their own interface instead of downloading the original file and zooming in on your own computer.

The four attachments show a heavily zoomed-in, center area of the screen with stacked CD's partly in light and partly in shadows. The CD inlays are variously colored. It's definitely worth notificing that black-and-white (or other kinds of very constrasty) content is VASTLY better in 48 Mpixels than in lower resolution. Not so with lower-contrast stuff, however. Check out for example "Pohjanmaan Sotilassoittokunta • Kytösavun aukella mailla" CD (dark blue writing on top of light blue background) just on top of the "STRAY CATS • CHOO CHOO HOT FISH" CD (the one with light blue letters on black background). It's equally unreadable on the 48 Mpixel image as is on the 12 Mpixel ones, while contrasty stuff like the just mentioned "STRAY CATS" does fare MUCH better on the 48 Mpixel images. This should be kept in mind - after all, it's not true raw image that you see here but a result of computational photography, which can't do REAL wonders.

Note: the "with flash, from close IMG_8381" file at shows a closeup shot of these CDs, lit with the flash (on the iPhone11) so that you can see what they look like in reality.

The attachments are very good-quality JPG exports of the original PNG MacOS screenshots.
SS 12raw.jpeg
SS 48raw.jpeg
SS HEIC.jpeg
SS JPG.jpeg
 
In addition, if you wondered how the situation is different when NOT shooting from a tripod but handheld: here's another album showing two images, one originated from a 48 Mpixe raw, the other a bog-standard HEIC produced by the phone: https://www.flickr.com/photos/33448355@N07/albums/72177720302307342 .

The 48 Mpixel images have the same resolution advantage handheld as on a tripod. That is, the camera doesn't need to move for better-resolution images to be taken.

Check out for example the VAST difference shown in the attached two crops!

(again, this is (also) because of the high constast. With low-constrast stuff, particularly in low light, the advantage wouldn't be as pronounced!)
 

Attachments

  • hqcoll-12Mpixel.jpeg
    hqcoll-12Mpixel.jpeg
    246.6 KB · Views: 163
  • hqcoll-48Mpixel.jpeg
    hqcoll-48Mpixel.jpeg
    171.1 KB · Views: 167
So just to confirm, as a 13 pro user, proraw is always the better option?
I haven't personally compared the 13's ProRAW to the HEIC/JPEG output of these models but knowing that the 13/P/M produce equally bad, oversharpened HEIC/JPEG files than any modern iPhone, I'd say you will ALWAYS get better output with ProRAW even if you "only" use the built-in Preview in MacOS. No oversharpening for a starter.
 
Here is a 12MP JPG vs 48MP ProRAW in medium light. Clear detail advantage for 48MP ProRAW. I took the photos from about 15 feet away. These are cropped in to see the difference.

12MP JPG 14 Pro
View attachment 2076413

48MP ProRAW 14 Pro
View attachment 2076415
Yup, excellent comparison. The resolution of the lens in the non-extreme areas (corners etc.) is excellent. While I haven't officially compared it to my (two!) Nokia 808's, based on my past experience (pixel-level sharpness) with the Nokia, I'd say the iPhone is better. (And it should be, given that there has been 10,5 years(!) between the two models.)

The corners, however, may NOT seem to be able to deliver 48 Mpixels (I'll verify this later). Which isn't completely unheard-of: a 24mm equiv ((under around 40mm equiv) the wider, the worse corner sharpness) mobile lens serving a BIIIG sensor - that's a bad omen with regards to corner sharpness ;)

What I'm also VERY interested in are 'pure' RAW files completely outside of the imaging chain (staking etc). While I'm certainly aware of the problems of such a small sensor (there'll be TONS of noise etc., as opposed to the 'clean' ProRAW files), in certain conditions, it MAY be better (with proper, manual post processing) than Apple's post processing "baked in" in these ProRAWs.
 
48.jpg
12.jpg
BTW, here are the two (now-inline so that you don't have to open and scroll the attachments yourself) crops you DEFINITELY should compare. All I can say is WOW!
 
  • Like
Reactions: Kataran
Great examples - have you

a) Tested alongside iPhone 11pro?
b) Tested 3rd party ProRaw / 48mp JPEG such as Halide?
 
Great examples - have you

a) Tested alongside iPhone 11pro?
Nope, it would be futile - the 11 Pro has exactly the same cameras as the 11, except for (obviously) the tele. (This is why I went for the 11 and not the Pro back in 2019, BTW.)
 
  • Like
Reactions: nemofish
Great examples - have you

b) Tested 3rd party ProRaw / 48mp JPEG such as Halide?
Will definitely do it when I have some time. I'm particularly interested in the "real" RAW (not the "baked" ProRAWs!) output of the sensor if it's at all accessible.
 
  • Like
Reactions: nemofish
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.