Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It only takes a minute to sign up.
Sign up to join this community
I've been wondering this for a while now, and I can't seem to find a straight answer about this.
If we are talking about the exact same sensor size, (lets say FX) and we are using a 50mm lens.
Camera A has 24 Mpx, and camera B has 12mpx.
If we aim to have a print of at least 300 DPI 8x10 picture, would the 24 PX not give us more "crop room" which could make it look like it was taken by a longer lens?
Thanks
I will just do some math (I will avoid all the sharpness of the lens and noise things aside).
12 Mpx = 2829 x 4243 px
24 Mpx = 4000 x 6000 px
If I divide 4000/2829 I get a 1.414 ratio
This means If I crop a photo, taken with a 50mm lens on a 24Mpx camera to a size equivalent of a 12Mpx photo (in pixels) will look as If I had photographed that with a 70mm lens.
(1.1414x50=70.6)
Oh. I'm not comenting anithing about printing resolution either.
Yes, and that is the maybe the major advantage of high-resoluton sensors, for typical print sizes. You can crop the image and still get an image with reasonable detail.
That said, the actual resolution of the image depends on the quality of the lens, too. Only quite high-end lenses will actually make good use of a 24MP sensor. if you look at a lens comparison like dxomark, you see that the top lens in this selection has a sharpness rating of only 21 "perceptual MPix", as they call it.
Following up on the comments, i'd like to forward to this extended discussion on the relationship between sensor and lens resolution.
It will give you some cropping room - if you use the same lens and a couple of steps back. If you're already stood against the back wall of your studio then more pixels won't help you.
12-24Mp also isn't as big a jump as you might first think. If you're using the extra pixels as a lever to get someone to break out their wallet then you might have to try another angle as you're only adding around 1/3rd more pixels in any one direction. You definitely shouldn't expect CSI style 'infinite zoom' capabilities from the higher-res body.
Yes, if you crop you'll have more pixels to play with but there is a trade off that zooming in may well start to reveal limitations in your glass. Supersampling and clever processing can help address some lens aberration problems and you can only work with the data you've got. Eventually you start to get into the realms of wanting better glass - and better glass is always expensive.
Only certain subjects really benefit from having the extra pixels and that goes some way to explaining with why the workhorse D4 isn't a near-40Mp monster like the D810.
tl;dr - yes you'd get some crop room but your clients probably wouldn't notice much difference in quality if the rest of your workflow is good.
In theory, yes. In practice, one tends to view the image quality including the available resolution and per-pixel quality as an underlying property of the camera. This also expresses itself in what kind of crops may still work in print. Some of the pixel resolution may not really be fully backed by optical resolution. That makes the pixels still useful as backup for the kind of interpolation required for geometric corrections: lens errors are ubiquitously corrected with today's lenses and cameras, but also correction of rotation and perspective distortion benefit from higher resolutions even if not fully backed optically.
The ability to choose judicious crops in post-production is viewed different from zooming "proper" since it is rare that an object is getting shot from the start with an intent to crop. It's more typical to take in a scene and then pick details afterwards. While this is sort-of "post-zoom/crop" in a manner similar to how picking an image off a focus stack with "post-focus" capable cameras works, it is not perceived in the same manner.
Early mirrorless cameras with large CMOS sensors suffered from comparatively low sensitivity (since the light sensitive areas of the CMOS sensor were interspersed with electronics that nowadays is placed on the opposite side of the sensor) and consequently had comparatively low digital resolution. When combined with high quality optics, you had a staggering amount of detail at the pixel level, but doing geometric transformations or cropping was taking quite a toll. Modern large sensor cameras don't have the same quality of per-pixel rendition (though quite better sensitivity) but those kinds of transformations and crops are much more affordable regarding the resulting hit in quality.
One thing you have to keep in mind is that anything but a central crop is not equivalent to zooming since focal plane and geometric distortions are no longer based on the center of the picture. You will not, in general, crop a portrait from the side of a larger photograph (also image quality tends to be worse in some other respects). So zooming and "free-range" cropping ability are usually seen as separate features.
More "reach" with the same lens is "achieved" with a smaller sensor (DX), which is simply a crop, but it has to be enlarged more, which is a telephoto effect ("reach"). You can easily see this in your editor, just zoom any image larger, which shows as a crop, which is enlarged more. Same visual effect as zooming with a telephoto lens (except lens zoom could show more detail, and we have fewer pixels if cropped after exposure).
More lens resolution shows more image detail. Digital sampling resolution (pixels) simply tries to reproduce that detail, more is better. More pixels is not "size" or "reach", and sampling cannot add detail to the original lens image, but more pixels will reproduce it better, with benefit to greater enlargement or cropping. Enlargement is what gives "reach" (done with longer lens, or via crop and enlargement).
Megapixels is sampling resolution of the lens image. Nyquist rule says 2x sampling is MINIMUM requirement. More is better. Those suggesting otherwise probably never saw it, and DxO sort of has their own numbering ideas.
Few of us have the necessary cameras to compare, but probably we have a scanner (which does the same digital sampling as the camera does, same concept, and can easily show us).
Scan a photo in a glossy magazine, printed halftone which is likely 150 dots per inch.
If scanned at 150 dpi or less (1x sampling or less), we get moire, which is aliasing, false detail due insufficient sampling, less than Nyquist 2x minimum.
Scan at 300 dpi (Nyquist 2x Minimum), and it looks better, with no moire. Zoom it to 100% size and examine detail of the halftone dots... dot reproduction is vastly better than at 150 dpi.
Scan at 600 dpi, and zoom it to 100% size and examine detail of the halftone dots... better yet than at 300 dpi.
Scan at 1200 dpi, and zoom it to 100% size and examine detail of the halftone dots... better dots than at 600 dpi. More is better. Starts looking fair at 8x sampling.
Previous scans look OK at "normal" size, but 100% of the detail shows that 8x sampling is better, and more might be better yet.
In this case, the goal of sampling is to reproduce those dots, which is the only detail that exists in this image.... 150 dots per inch.
If our photo image has lens detail at 100 line pair per mm, then just 8x sampling is 800 pixels per mm. Our current FX/DX cameras only do more like 200 pixels per mm (barely 2x, except little of our data is at 100 lp/mm). We are not yet near any limits. Our goals for using the image may be a limit for this instance (we can only print so big), but is not a technical limit.
Newbies imagine 1x sampling suffices, but far from it, it is not even Minimum. Digital sampling is a new concept for many.
Digital resolution - more is better. Up to a point (well past 2x sampling), because the best result possible is merely a reproduction of the lens image. Digital sampling cannot add new detail that the lens did not see.
This is 36 megapixels, Nikon D800 camera, which is available, and pretty nice.
(click to enlarge)
Not necessarily.
There are two factors limiting image quality.
One of the factors is the sensor pixel density.
Another of the factors is lens sharpness.
For example, the first version of the Canon 100-400L lens doesn't necessarily give a better picture on a higher-density crop sensor (as opposed to a lower-density full frame sensor), because the lens is the factor limiting image quality. On a typical full frame sensors, the 1st 100-400L is quite balanced. But, if you start adding more pixels, you won't find the true resolution increase.
Obligatory DxOMark which will surely cause a bunch of downvotes for me:
Also, you have to keep in mind diffraction when cropping. Cropping magnifies the effects of diffraction. So, use a large enough aperture to overcome the effects of diffraction, but then again cropping magnifies the effects of limited depth of field too, which is exactly what a large aperture gives you.
