Hacker News Reader

Comment by muhfuhkuh | Hacker News Reader

There's got to be some shenanigans going on when the iPhone 4s shot of the city looks better than the 5DMKII, an essentially $3000 camera.

View on HN

Not so shocking, actually. The 5D is a pro camera, and you're expected to pull the RAW data into Lightroom and manually tweak and tune the final product. Its output is optimized for this workflow: minimum processing in-camera, maximum options later.

The iPhone's software, on the other hand, is doing some auto-adjustment before the file hits the SD card. It's optimized for a decent photo without any manual steps.

Also, notice the distortion in the iPhone photo: the buildings "lean" towards each other. The Canon 24-70mm f/2.8 is a _nice_ piece of glass, and is worth every penny of its $1500 asking price.

tl;dr: Take the 5D shot (in RAW, hopefully) and let a pro have a shot at it in Lightroom. That's a fair comparison.

The 5DII would have also been more accurately compared with the Canon EF 24mm f/1.4L II lens (according to the Exif data the sample photo was shot at 24mm).

The iPhone 4S's camera is a fixed focal length while the Canon lens they used is a zoom lens. Zoom lenses trade certain attributes for the flexibility of being able to zoom, one of which is corner softness at wide open apertures. The 24-70 is a decent all-around lens, and I have used it extensively, but I have never encountered a zoom lens that reaches the same image quality as a fixed lens for the same focal length.

With those little bitty scaled-down snaps, there ought to be no discernible corner softness with a big 2.8 pro zoom (and IMHO there isn't); we're mostly eyeballing center sharpness, white balance, and range of contrast. Of course there'd be less barrel distortion with a prime... but it doesn't stand out either way with these scenes.

You can see there's a little less blowout on the shiny roof with the 5D, and quite a bit more detail in the shadowed city streets. Certainly a higher color temperature, probably more accurate. But the 4S is quite impressive nevertheless. No apparent vignette issues.

These photos are not useful for assessing high ISO noise/grain...

Don't know about Canon, but for Nikon there are 24-70/2.8 and 70-200/2.8 which are on par with fixes.

Or blow up a small area (within the focus plane), and see which is sharper. Or better still, take a photo of a moving target at night.

One advantage of a DSLR is that you can get a sharper photo with less light and a shorter exposure time. On web-sized photo, taken on a sunny day, you just can't see the difference. If you take a photo at night even a complete amateur like me can tell the difference.

A DSLR isn't like a Ferrari, which blows the competition away in favorable conditions. It's like a Four wheel drive that can be used anywhere.

There is. As there is with most of the "4S vs 5DMkII" comparisons.

For example, the 4S photos are highly saturated (known to be a trait of the iPhone) and sharpened (whereas the S95 and 5D shots do not appear to have had any sharpening). In the first set of photos, the iPhone shots have a completely different DOF to that of the two cameras (and not even shot from the same position).

The "4S versus 5DMkII" video comparison going around conveniently ignores the little footnote in the accompanying description where the tester states that he "adjusted the 5D settings down to match the 4S options available", i.e. "4S video matches highly handicapped 5D video"...

No doubt, the 4S has impressive still and video functionality, but you're looking at a couple of lenspieces smaller than a penny versus 12+ pieces of glass in an L lens...

Aside from the points mentioned by others, the 5DMKII captures far greater dynamic range; look at the detail in the shadows. The 4S produces dark buildings with no definition.

The iPhone-processed result is warmer and more saturated, but it is technically less accurate, and contains quite a bit less dynamic range; you could get better with a JPEG out of the 5D and Picasa in a few seconds.

It feels so weird to defend Canon... I'm a Nikon man all the way.

It feels so weird to defend Canon... I'm a Nikon man all the way.

Does it matter these days? My impression for the last couple years has been that more or less any dSLR will work very well for the vast majority of what most people do. When people ask me about photography, I by and large say, "Get whatever your friends are using and/or whatever is cheap."

It does matter in a way that by purchasing the camera and a few lenses (which are most of the time more expensive then the camera body) you are locking yourself with that vendor for the future - lenses are not interchangeable between Nikon and Canon to the best of my knowledge. This lock-in is why you must be careful what you pick.

I own a Pentax, but would never choose this company (not hardware) if I was any serious about photography.

I own a Pentax

Oh, you're the other one. Glad to meet you :)

(FWIW, I chose Pentax because of on-board image stabilization. That should allow me to save money by not paying for it in every lens (as Canon and Nikon do). However, it seems to turn out that the big names get enough economy of scale that the lens price isn't better for Pentax anyway.)

Yeah, got sold on that one too. It was also a very good deal at that time in that price range (k200d), 11 focus points and water proof body with upper lcd.

A shame the selection of lenses seem to be that limited, many times with one type of make/model for given parameters only.

That is fantastic advice for any beginner up to someone relatively advanced. At the professional end it does make a difference, recent high-end Nikons have been truly excellent cameras and have had much better noise reduction and autofocus than the Canons. It still makes much less difference than the skill of the photographer, though. It's like Windows/Mac, Emacs/Vim etc etc - they each have their advantages that will probably only matter to a very small percentage of users.

What do you mean, "better"? To my eyes the 4S comes nowhere near beating the sharpness and clarity of the 5D. The colors are stronger in the 4S's photo, but do we know that to match reality?

The 5D uses a very large aperture f 2.8 which you'd normally never use for a landscape shot in daylight. This leads to a very shallow depth of field and much less detail.

That's not a particularly demanding setting, so the differences aren't so clear. However, clearly iPhone colors are overly saturated and detail is lost on darker areas.

Of course, this comparison is bullshit. They downsampled the 5D shot, no sharpening, they used f2.8 to shoot a landscape (not that it would matter much in this comparison).

iPhone 4S has no chance against 5D mk II.

Reply on news.ycombinator.com