0

I take photos of birds and frequently use most of the 50X zoom available on my Canon SX50 camera, which has a 1/2.5” sensor (about 25 square mm).

I want to compare the superzoom image I get to what I would get with a DSLR with a 300mm lens.

Let’s say I am 100 feet away from my subject and I zoom all the way to 50X and I get the exact image I am looking for (no cropping needed) with the superzoom. If instead I was using a DSLR with an APS-C sensor (about 370 square mm) and I am zoomed out to 300 mm I can get an image of the bird, but I will have to crop it to duplicate the one I got with the superzoom.

After the cropping gives me an image the same size of the superzoom, which image will be better (I.E. have more detail/information)?

How do you compute this?

Michael C
  • 175,039
  • 10
  • 209
  • 561
John Long
  • 1
  • 3
    You could just try it.... But honestly, there are so many different "superzoom" cameras, as you call them, and so many different DSLRs that you're not going to get any meaningful comparison without providing more info. – twalberg Jul 08 '20 at 17:33
  • 3
    Regarding "_N_ ×" zoom, see this question: How do I convert lens focal length (mm) to x-times optical zoom? – scottbb Jul 08 '20 at 17:34
  • 1
    See https://photo.stackexchange.com/questions/99906/nikon-coolpix-1000-how-to-decide-if-its-a-smarter-option-than-a-good-long-l & the shots of the moon in the answer. The super-zoom is nowhere near as good. – Tetsujin Jul 08 '20 at 17:35
  • Here is an example reference that happens to use an SX50 for comparison with a DSLR (Nikon 5200). https://www.discoverdigitalphotography.com/2013/how-does-the-zoom-of-a-superzoom-camera-compare-to-a-dslr/ – user10216038 Jul 12 '20 at 15:50
  • @Tetsujin - I looked at those referenced Moon pictures. Despite comments to contrary, the super-zoom clearly has more detail and is the better picture. I call, "The Emperor has no clothes." – user10216038 Jul 16 '20 at 04:50

4 Answers4

3

For a comparison you should first compare the field of view the different setups provide. You can do this by looking at the full-frame equivalent focal length of the lenses:

Canon SX50: According to the specs provided by Canon (https://www.usa.canon.com/internet/portal/us/home/support/details/cameras/point-and-shoot-digital-cameras/long-zoom-cameras/powershot-sx50-hs/powershot-sx50-hs?tab=technicalspecifications) it provides a maximum full-frame equivalent focal length of 1200mm (that is 215mm physical with a crop factor of 5.6).

Nikon D7500 with 300mm: Nikon DX cameras have a crop factor of 1.5, resulting in a full-frame equivalent focal length of 450mm.

Now you see a factor of 2.7 between these focal length. Cropping the image of the DSLR to the field of view of the Canon SX50 would result in a resolution that is only (1/2.7)2 of the original image, for the Nikon D7500 this would result in only 20.9*(1/2.7)2 = 2.5MP.

Even considering improved SNR because of the newer sensor with bigger photosites the quality is probably worse compared to the full 12.1MP image of the SX50.

If you would instead use a 600mm lens the DSLR would certainly win: This would come out to a field of view that is wider by a factor of 1.3, which in turn means cropping to equal field of view results in 11.8MP. Obviously the 600mm is significantly heavier and more expensive.

std_photo
  • 39
  • 1
  • The sensor resolution is irrelevant in this comparison... at 215mm f/6.3 the SX50 will record an average of .35 MP even with a perfect lens; and the superzoom lens is certainly not perfect. – Steven Kersting Jul 09 '20 at 14:24
  • Cheap 70-300mm or 75-300mm lenses for APS-C cameras are also far from perfect, particularly at the long end, where they tend to be their worst. – Michael C Jul 11 '20 at 14:11
  • @std_photo is correct. If this were a full sensor to full sensor comparison, the DSLR would be better, but that's not what's happening. Only a small part of the image is of importance, the same thing happens with astro photograpy. The little 1/2.5 sensor gets all of it's sensor elements used (12 million) on the pertinent part of the picture (the Bird, or Moon for example). The larger DSLR sensor only has a small fraction of it's sensor elements used on the target (2.5 million). It's like using a wide angle lens then cropping. – user10216038 Jul 11 '20 at 22:35
1

We can not answer that over the internet.

Seems like you need to get a DSLR and a 300mm lens and Do a test shoot.

Find a stationary subject set up your tripod and do a test shoot with both cameras. I.E.carry out your hypothetical situation you described.

How do you compute this?

That is up to you and your criteria, compare the images with your eyes and on your computer to see how they differ.

Of coarse those test results will only apply to the two specific camera/lens combos in that test.

Alaska Man
  • 3,608
  • 10
  • 17
  • Sorry, I did not realize the specific camera would make that much difference. The superzoom camera I own is a Canon SX50. If I had a DSLR I would do a test shoot, but alas I do not. The DSLR I am considering is a Nikon D7500 with a 70-300 mm VR lens. Thanks for your help. – John Long Jul 09 '20 at 01:47
  • @JohnLong for just a few bucks (in comparison to purchase price) you can rent for a couple of days to test it out. You could even rent a fixed focal length 300mm and the zoom to compare image quality. There are a number of places online where you can rent equipment - an internet search for "camera rental" will probably turn them all up quite quickly. – FreeMan Jul 10 '20 at 17:08
  • @FreeMan Rentals are easier and more affordable in some places than others. – Michael C Jul 11 '20 at 14:00
  • Quite true, @MichaelC. I took a very US-centric view of the issue. Mea Culpa – FreeMan Jul 11 '20 at 14:31
  • @FreeMan Even in the U.S., if one has to have the rental shipped both ways, the shipping charges can significantly increase the cost of renting lower end products such as the D7500 and 70-300VR. At lensrentals its $42 + $17 for that body + lens, and another $50, maybe more depending on how far you live from Memphis or L.A., for shipping. – Michael C Jul 12 '20 at 03:40
  • I've only done a lens rental for a week for some under-lights outdoor sports shooting, so the shipping was, I felt, reasonable compared to the lens rental itself, @MichaelC. Of course, everyone's tolerance and definition of "reasonable" will be different. All things considered, though, under $150 to rent & ship is still cheaper than a purchase, I'd think. YMMV. – FreeMan Jul 12 '20 at 16:38
  • That depends. There are 70-300mm f/4-5.6 zoom lenses available for less than twice that. – Michael C Jul 12 '20 at 22:48
1

This is not a simple question to answer as there are many variables; but the simple answer is that the probability of recording a better image exists with the APS camera.

With the SX50 you have a 5.6x crop of a 215mm lens' image circle. And with the APS you have a 4x crop of a 300mm lens' image circle; which means the APS retains more of the light from the scene/image... advantage to the APS.

At 215mm the SX50 lens is at f/6.3 and projects details at a minimum average size of approx. 9microns, which is approximately 6x the size of a pixel on that sensor (1.5um)... but that's only if the lens were optically perfect, and that lens most certainly is not. Best case (theoretical) scenario is ~ .35MP average of recorded resolution w/ a lower contrast level... reality is probably notably worse.

With the APS there is the potential of using a faster and better lens, lets say a 300/2.8. At f/2.8 a lens projects details at an average minimum size of 3.8 microns. Or say a 300/4 which will project details at an average minimum size of 5.4 microns. Both would fit within the size of a pixel of most APS cameras (or w/in the diffraction limit of 2x). E.g. the D7500 has 4.2 micron pixels and ~ 8 micron diffraction limit. Best case (theoretical) scenario, ~ 2.5MP recorded resolution with a much higher level of contrast... advantage to the APS again.

Edit: even the 70-300vr would be at f/5.6-300mm and could project details at an average minimum of 7.5 microns, less that the D7500 diffraction limit (again theoretical, the lens probably isn't that good wide open).

Then there is the consideration of motion (camera shake/subject/atmospheric/etc)... and again, the larger pixel size on the larger APS sensor gives a potential increase in actual recorded resolution because a detail can move much farther before crossing an equivalent number of pixels.

Steven Kersting
  • 17,087
  • 1
  • 11
  • 33
  • APS-C is already cropped 1.5X compared to "35mm equivalent" FF cameras. That is, to get a 1200mm FF field of view with a Nikon APS-C camera, one only need use an 800mm lens, compared to the 215mm lens of the 5.6X crop superzoom. In other words, one must only crop a 300mm lens on an APS-C camera by a linear factor of 2.67X to get the same FoV as the 5.6X crop camera with a 215mm lens. – Michael C Jul 11 '20 at 14:07
  • @Steven Kersting - You're going down the the wrong path with diffraction limits. It's pixels on target that dominate. An easy comparison test is to shoot the Moon. The little super zoom blew away a 300mm on a Nikon D7200. Try it yourself. – user10216038 Jul 11 '20 at 22:54
  • That's probably because cheap 70-300mm lenses are crap at 300mm. I do rather well with a 70-200/2.8 + 2X extender (or even just the bare lens with the stink cropped out of it) on a 7D Mark II. I've yet to see a superzoom that can equal the results from that combination, several examples of which are included in my answers here. – Michael C Jul 11 '20 at 23:27
  • @MichaelC, the 2.67x + 1.5x APS crops equal the 4x crop I stated. What is relevant here is how much of the original image circle is discarded... it doesn't matter where/how the cropping occurs. – Steven Kersting Jul 12 '20 at 14:17
  • @user10216038, you cannot record something the lens cannot resolve/project regardless of the any sensor characteristic (MPs). – Steven Kersting Jul 12 '20 at 14:18
  • @MichaelC - Your assumption of a poor 300mm is not germane, and you 2x extender on your 200 makes it a 400 (which I'm sure you know). This whole argument is easily tested in reality, and I've done so. Here's a reference that may help: https://www.discoverdigitalphotography.com/2013/how-does-the-zoom-of-a-superzoom-camera-compare-to-a-dslr/ – user10216038 Jul 12 '20 at 15:36
  • @user10216038 My "assumption" may or may not be correct, but it is certainly germane since it is a distinct possibility. The results using the bare 200mm and cropping twice as much are still very good as long as one doesn't try to view them larger than 100% (1 imagepixel + 1 screen pixel). – Michael C Jul 12 '20 at 22:52
  • @Steven Kersting - You said, "... the larger pixel size of the APS sensor gives a potential increase in actual recorded resolution ..." . I'll refer you to https://photographylife.com/what-is-diffraction-in-photography and "... a high-resolution sensor will always capture more detail than a low-resolution sensor of the same size. More pixels will never lead to lower detail, even at the tiniest of apertures. ..." – user10216038 Jul 18 '20 at 14:43
  • @user10216038, that statement is true when comparing different sensor resolutions when the sensor is the same size. If the lens is projecting an image of 8MP it doesn't really matter if the sensor is 12MP or 24MP, both will record the same 8MP image. And there is often some small increase in recorded resolution with the higher MP sensor due to the increase in oversampling. But if the lens is only able to project 10 dots/mm (for example) that is going to result in fewer dots fitting on a smaller sensor... i.e. a lower recorded resolution. – Steven Kersting Jul 18 '20 at 16:20
1

I have the Nikon P900 with its 2000 mm effective focal length generated by a 357.5 mm f/6.5 lens on a 1/2.33" sensor. I use it handheld mostly for birds, trying to get shots good enough for ID on iNaturalist. I have tried a Canon 7D with the 100-400 lens for comparison, which works out to a 640 mm effective focal length. I prefer the Nikon for exactly the reasons you are asking about. That said, the images often do not have the "snap" you can get from a nice DSLR photo if you are close enough. Getting close enough can be hard work. The other downsides to the Nikon are that it is probably slower to focus and does not have a usable manual focus. If you have a bird with some branches in front the camera will often focus on the branches. Spot focusing will sometimes let you focus correctly, and sometimes you will get lucky. But nothing competes with the reach. The lenses are not too different in focal length or f/stop. I want a zoom sensor behind the lens, so I can use big pixels when I am close.

I did a direct comparison between a Canon 7D Mk I with the 100-400 zoom (first generation) and the P900 in a range of distances in good sunlight. Roughly, I preferred the Canon when the "right" effective focal length (so the bird fills the frame) was about 1200mm or less. That results in a crop to about 1/2 of the frame in each direction. The better lens and lower noise overcame the fewer pixels on target. When the "right" focal length was longer, the pixels on target from the Nikon won out because the crop of the Canon was more severe. The further from 1200mm, the clearer the difference was.

Ross Millikan
  • 1,924
  • 1
  • 11
  • 11