Reach vs Megapixels

[telscossie]

Hi i wonder if i can have peoples views on my quandary at the moment i use a 1dx 18 megapixels + 7dii 20 megapixels which i use for bird /Wildlife photography on my 500mm f4 is version one and sometimes with a 1.4xiii extender but i still find my shots are not close enough so my question is would i benefit from having a camera with larger megapixels so i could crop in more or getting a 600mm f4 ii lens ? any comments are most welcome. Thanks

 

[Harlequin565]

How big do you print? That's the main determining factor in terms of the number of pickles you need. How much do you crop in already?

The 7d2 outputs 5400 x 3600. If you're talking web use you could easily crop down to 2700 x 1800 (half the height & width which is a quarter of the area) and that will still give you magazine quality 6x9 prints as well as "good enough" for the majority of "normal" screens.

 

[snerkler]

In terms of outright IQ you’re always better off having the extra reach of the lens, any kind of cropping will ‘degrade’ the overall IQ. However there are a lot of variables, and the one that’s difficult to quantify is ‘acceptance’ of the end user. Some may crop 2x or more and be perfectly happy with the resulting image and some not.

With regards to cameras with high mp count whilst this gives you the ability to crop more it’s not necessarily going to give you as good an image as having the extra reach in the first place and will depend a lot on the resolving power of the lens. If you have a lens capable of resolving nearly all the mp and you crop then the image will still look good, but have a lens that can’t resolve well and it won’t crop so well.

 

[sk66]

A longer/sharper/faster (or at least equally as sharp/fast) lens will give you some benefit... if used appropriately; but probably not as much as you would hope. Otherwise, it's really just a big circle of tradeoffs with no clear "best choice." IDT there's really a significant difference between 500/4 and 600/4; unless your 500/4 is a poor performer.

The one thing that will always help to a large degree is getting closer (responsibly).

 

[sk66]

In terms of outright IQ you’re always better off having the extra reach of the lens, any kind of cropping will ‘degrade’ the overall IQ.

Using a longer lens is really just another version of cropping... so the longer a lens is, the better it has to be.

 

[jakeblu]

That's the main determining factor in terms of the number of pickles you need.

Would that be Branston or Piccalili?

 

[AndrewFlannigan]

Would that be Branston or Piccalili?

Gherkins. The gherkin count integrates the lines per millimetre of the lens with the number of receptors in the sensor to provide the overall pickle rating of the camera.

 

[GyRob]

I had your setup then went to the 1DX MKII and 600f4 MKII just to get that little extra reach even then I often had the 1.4tc mkIII on it worked very well but times have changed Mirrow less has come of age I sold all the above and now use an a9mkII and FE 600F4 and 1.4tc .

you might want to check that out before you buy even more into Canon the AF blows the 1dx out the water imho I don't regret it for a second .

Yes the R5 / R6 are on the way and look very good but you need an adapter for EF lenses so i don't think they will be as good as native lenses , just my thoughts .

Rob.

Edit I did buy the a7r4 for the MPs but it's not as fast as the a9mkII so i sold it i want speed / AF over MPs .

 

[Bobsyeruncle]

I had your setup then went to the 1DX MKII and 600f4 MKII just to get that little extra reach even then I often had the 1.4tc mkIII on it worked very well but times have changed Mirrow less has come of age I sold all the above and now use an a9mkII and FE 600F4 and 1.4tc .

you might want to check that out before you buy even more into Canon the AF blows the 1dx out the water imho I don't regret it for a second .

Yes the R5 / R6 are on the way and look very good but you need an adapter for EF lenses so i don't think they will be as good as native lenses , just my thoughts .

Rob.

Your new kit doesn't give you any more reach though.

 

[snerkler]

Using a longer lens is really just another version of cropping... so the longer a lens is, the better it has to be.

How so? The optics are filling the full image circle and so less enlarging needs to be done vs cropping? I get atmospheric aberrations come into play but are you saying if you have two optically identical lenses that a 500mm cropped to a 600mm reach is going to look as good as a 600mm uncropped?

 

[GyRob]

Your new kit doesn't give you any more reach though.

Well 4 mp more but almost every shot in focus and that's worth more than MPs .

Rob.

 

[AndrewFlannigan]

are you saying if you have two optically identical lenses that a 500mm cropped to a 600mm reach is going to look as good as a 600mm uncropped?

That will depend on the sensors, the image stabilisation and the physical steadiness of the setups being used. If your question refers to testing both lenses on the same rig in the same place at the same time you would eliminate those issues to a degree but there might be environmental differences due to the time between the two tests. Much depends on what degree of detail you are testing.

 

[sk66]

Using a longer lens is really just another version of cropping... so the longer a lens is, the better it has to be.

How so? The optics are filling the full image circle and so less enlarging needs to be done vs cropping? I get atmospheric aberrations come into play but are you saying if you have two optically identical lenses that a 500mm cropped to a 600mm reach is going to look as good as a 600mm uncropped?

All lenses see the same light/scene; the same as your eyes do. So if a detail is x-size relative due to distance, it is the same relative size entering every lens (discounting the negligible differences in a lens' physical length). What a lens does is "crop" the scene you see down to it's FOV, and then it magnifies it after the light enters the lens. And in order for the image to hold up with the greater magnification, the lens has to be better optically. It is also spreading the light out more, so it needs a larger entrance pupil (same f#) in order to collect more light and equalize the exposure.

Optically identical is very hard to achieve. But let's assume a 500 and a 600 with equivalent MTF (sharpness) ratings... In fact, the 600mm is sharper; because in order to conduct the MTF test the target must be at a greater distance so that the lines cover the same area of the sensor (lines/mm)... it is resolving details which are relatively smaller. In this case, if used from the same distance then the 600 will look better. Because it is able to resolve relatively smaller details. And if you crop the 500mm image you are discarding image area/light, which is going to be additionally detrimental.

If instead we compare lenses that can resolve the same size detail, at the same distance, and to the same level of sharpness, then the greater magnification is going to make any flaws more apparent; and increase diffraction. This is somewhat typical of the large telephoto primes where the max aperture/entrance pupil (F#) reduces in size as FL increases... that's because they are essentially the same basic design/lens (optically the same), only with stronger telephoto elements at the exit pupil (lenses of same generation/technology). In this case the images may look extremely similar, except that you discarded light for the cropped image; assuming the same f#/exposure was used and you didn't take advantage of the shorter lens's larger aperture/lower diffraction. This is essentially the same as adding a TC (increasing the telephoto magnification).
If instead you did take advantage of the shorter lens' larger aperture/lower diffraction then the differences become even less and the 500 cropped could actually look better.

Assuming neither lens can resolve to the level of the sensor (very probable), then the MP's are not much of a factor.

This is why getting closer is much better than using a longer lens... the details are relatively larger entering the lens and there is a much lower requirement from the lens (plus environmental/technique considerations).

 

[Harlequin565]

Well written explanation

Really useful and well explained information. Thanks.

 

[nandbytes]

As alluded to above nothing really beats physically getting closer to the subject.

I don't see reach vs. mp as separate things but rather mp being part of "reach".

I personally see reach as a FoV (35mm/FF equivalent) at certain amount resolution. So for example with my A7RIV+200-600mm, I can get 600mm FoV at 61mp of "reach" (i.e. FF equivalent) OR 900mm FoV at 26mp "reach" (i.e. APS-C equivalent) OR 1200mm FoV at 15mp of "reach" (i.e. M43 equivalent). Of course cropping especially at higher ISOs will show more noise and cause loss in IQ (just the same as you'd see had you used a smaller sensor).

and how much "reach" you need depends on how big you print, how close you can get to your subject etc. But as its commonly noted you can never have enough reach

 

[sk66]

I don't see reach vs. mp as separate things but rather mp being part of "reach".

With modern sensors MP's is largely a red herring... there's no lens that can resolve to 61MP on the Sony A7IV sensor. There are a few Ziess lenses that can resolve to ~ 50MP (e.g. the Loxia 21/2.8 @ f/2.8). But the big Sony lenses (300mm +) are more like 20-30MP (which is actually way more than you are likely to ever need).

And the 300ppi for printing is also mostly nonsense. What actually matters is the physical size of negative you have (sensor area/area remaining), and the quality of the image on that negative (sharpness/resolution/noise/grain)... that's what's always mattered; and digital didn't really change anything (until you get to pixelation).

 

[nandbytes]

With modern sensors MP's is largely a red herring... there's no lens that can resolve to 61MP on the Sony A7IV sensor. There are a few Ziess lenses that can resolve to ~ 50MP (e.g. the Loxia 21/2.8 @ f/2.8). But the big Sony lenses (300mm +) are more like 20-30MP (which is actually way more than you are likely to ever need).

And the 300ppi for printing is also mostly nonsense. What actually matters is the physical size of negative you have (sensor area/area remaining), and the quality of the image on that negative (sharpness/resolution/noise/grain)... that's what's always mattered; and digital didn't really change anything (until you get to pixelation).

while I don't disagree, you almost always get an increase in quality i.e. pixels/details with increase in resolution even with bad lenses. Improve the optics or increase the digital resolution and you will end up with more details (of course largely depending on the optics and sensor). Increase both and you will be doing even better.

You mention loxia and I am not sure what your experience with Sony lenses are but there are sharper lenses than loxia 21 these days.
for eg: https://www.lensrentals.com/blog/2019/03/sony-fe-135mm-f1-8-gm-early-mtf-results/
So as I see it just as sensor tech, optics are improving and there is a fair amount of modern optics/lenses now that can make use of these high res sensors.

All the theory aside, I have been shooting with 24mp, 42mp and now 61mp. there is no way I'd go back to 24mp. The difference between 42mp and 61mp is not huge but still helps with getting a little extra "reach". tbh I am not sure upgrade to A7RIV was worth it but its done now lol.

As for printing that's another thread and half, let's keep that debate to another day in interest of not derailing OP's thread. But again I don't disagree with you for the most part.

 

[sk66]

while I don't disagree, you almost always get an increase in quality i.e. pixels/details with increase in resolution even with bad lenses.

Yes, this is true... due to increased oversampling. But the gains in terms of resolution are typically quite small compared to the increase in MP's (often only around 10%). That said, there are also numerous other (potential) gains of perhaps more significance due to higher MP sensors (e.g. color accuracy, moire suppression, etc).

Interesting about that 135; it's at 55MP on the A7RIV... really very impressive. In any reasonably normal situation, that's much more resolution than a human can see.

These days I shoot with the 46MP D850 and the 20MP D5; they are different tools for different situations, and there are times where the 20MP D5 is the better choice in terms if IQ... or at least no worse.

 

[snerkler]

All lenses see the same light/scene; the same as your eyes do. So if a detail is x-size relative due to distance, it is the same relative size entering every lens (discounting the negligible differences in a lens' physical length). What a lens does is "crop" the scene you see down to it's FOV, and then it magnifies it after the light enters the lens. And in order for the image to hold up with the greater magnification, the lens has to be better optically. It is also spreading the light out more, so it needs a larger entrance pupil (same f#) in order to collect more light and equalize the exposure.

Optically identical is very hard to achieve. But let's assume a 500 and a 600 with equivalent MTF (sharpness) ratings... In fact, the 600mm is sharper; because in order to conduct the MTF test the target must be at a greater distance so that the lines cover the same area of the sensor (lines/mm)... it is resolving details which are relatively smaller. In this case, if used from the same distance then the 600 will look better. Because it is able to resolve relatively smaller details. And if you crop the 500mm image you are discarding image area/light, which is going to be additionally detrimental.

If instead we compare lenses that can resolve the same size detail, at the same distance, and to the same level of sharpness, then the greater magnification is going to make any flaws more apparent; and increase diffraction. This is somewhat typical of the large telephoto primes where the max aperture/entrance pupil (F#) reduces in size as FL increases... that's because they are essentially the same basic design/lens (optically the same), only with stronger telephoto elements at the exit pupil (lenses of same generation/technology). In this case the images may look extremely similar, except that you discarded light for the cropped image; assuming the same f#/exposure was used and you didn't take advantage of the shorter lens's larger aperture/lower diffraction. This is essentially the same as adding a TC (increasing the telephoto magnification).
If instead you did take advantage of the shorter lens' larger aperture/lower diffraction then the differences become even less and the 500 cropped could actually look better.

Assuming neither lens can resolve to the level of the sensor (very probable), then the MP's are not much of a factor.

This is why getting closer is much better than using a longer lens... the details are relatively larger entering the lens and there is a much lower requirement from the lens (plus environmental/technique considerations).

Interesting, thanks for the explanation. So cropping in post and having a longer lens will likely yield similar quality images then (assuming similar lenses etc)?

 

[nandbytes]

Interesting, thanks for the explanation. So cropping in post and having a longer lens will likely yield similar quality images then (assuming similar lenses etc)?

Depends on the lenses of course. I remember back in the a-mount days showing how little difference there was between using the 70-400mmG and then newly released tamron 150-600mm (which was first of its kind). But as I noted above just as sensors, optics have also improved since then.

A sure way to know is shoot two lens and do the test yourself.

 

[sk66]

Interesting, thanks for the explanation. So cropping in post and having a longer lens will likely yield similar quality images then (assuming similar lenses etc)?

Yes, it often happens... I call it "the equivalence circle," where every choice has an equal tradeoff.
E.g. choosing between a 400/2.8, 600/4, and 800/5.6... each longer FL costs you light, exactly as if you had cropped the shorter lens' image to equivalent (or added a TC/used a crop body).
And each increase in FL/magnification causes greater diffraction and reduced DOF, which reduces the ability to record finer details with the same sharpness/clarity/contrast... which is exactly what happens (perceptually) when you enlarge a smaller image/negative to the same size.
And each increase in FL/magnification puts an increased demand on technique required to record an acceptably sharp image... which also happens when you enlarge a smaller image/negative to the same size.
And around it goes...

There are certainly situations where it doesn't work out quite that way... but then you're probably not looking at very similar lenses/designs. E.g. a 600/4 will stomp all over a 70-200/4... and the latest generation 600/4FL will probably stomp all over any of the big primes from the late 80's early 90's.

 

[LongLensPhotography]

I've sort of been there and while my resolution was going back to what I know best (landscapes and commercial) I made some clear conclusions that you may find useful.

600mm f/4 lens is a beast. 500mm is a lot more user friendly. While mkII is of course lighter than mkI, it is still ridiculous. At 600mm I still struggled to get near enough anything, and if I did the little 400mm was actually preferable for ease and speed of use. I think the key is managing to get close enough, or even 800mm won't help. You have to first make the most of the existing focal length any way possible. I think there is a reason why canon introduced RF 800mm f/11 (it takes TC too!); that's for wildlife losers like me and I fear it wouldn't help much either.

1DX is a sort of camera that really wants to fill at least 60-70% of the frame or you start losing too much. With something like 45MP EOS R5 or 50MP 5Ds you would have far more flexibility to crop in much closer occasionally, but you may not wish to in case focus is less than perfect across the subject which it will always be on one of these lenses, and then there are noise issues. With 7DII you are already starting at the worst case scenario; it's very dated sensor. For ISO100 maybe it's not too bad, but I'm guessing you may be closer to 1600.

 

[Cheffievrs]

Surely fieldcraft needs to be considered no.? My thoughts are you have all the tools you need kit wise, just learn to get closer to your subject..

 

[GyRob]

Surely fieldcraft needs to be considered no.? My thoughts are you have all the tools you need kit wise, just learn to get closer to your subject..

Yes i agree fieldcraft does help a lot and i always try to hide, with the OPs kit once he fires the shutter the bird / animal is well on the way to turning the other way more so with the 1dx Silent electronic shutter really helps as often the bird say an owl just keeps coming until it sees you move by then you can have lots of shots .

Rob.