Cropping and Resolution on 40mp sensors

[JohnNewman]

In a current thread about the expected new X-E5 a contributor wrote:

"Cropping from 40MP to 20MP is exactly 1.4x focal length.



I wouldn't necessarily recommend buying 40MP 'in order' to crop, but it sure does help to have that extra information."

Actually, that's exactly one of the reasons I am considering one of the 40mp bodies, either the X-T50 or X-E5 once I know the specs of the latter. Long story short, I'm looking for a lightweight combo to complement my Sony RX10iv which is a great camera but hardly light or anywhere near jacket pocketable. Either would also allow be to add a pancake prime when wanting to go really light.

In my naivety I thought that a x2 crop in software after shooting would give me a focal length of 100mm and a resolution equivalent to 20mp. As pointed out this sort of crop would be equivalent to a focal length of 50 x 1.4 = 70mm. This still useful but a little more research shows me a 2x crop would mean 10mp resolution which is acceptable for some purposes but maybe not everyone's needs.

Overall I think the 40mp sensor and 16-50 will be a good option for a lightweight kit when I don't want or need the RX10iv and, if cropped, will still give me a decent size file, especially as I no longer print big or enter competitions.

I'm really just posting this for information for anyone considering a camera with higher resolution (whether Fuji or another) and I attach below a couple of results from my research. So if you're deliberating which sensor size (in MP) to select, maybe this will help.

John

p.s happy to be corrected if I've got anything wrong

From Jonas Rask's review of the X100vi clearing showing the drop in resolution when cropping.

From another discussion found on-line

 

[cafe snapper]

You can print a poster from 10mp. It's more than enough for most people.

 

[Mountaingrapher]

The 40Mpx of the X-Trans V sensor is one of the reason I chose my X-T5. It is extremely usefull for my uses cases: mostly travel / landscape, in the deep backcountry and up mountains where weight matters so I can't take 10 lenses.

With the proper lens the results are there:

Shot with a X-T5 and XF 33mm f/1.4.

This photo retains a lot of details, plenty enough for me to do a small print (A4 or similar). It is a heavy crop of this one:

I don't have an IA upscaling/denoising software but I guess it could help with enhancing the crop further?

 

[g7_chord]

This is exactly how I use my very small X-T50. I can put a small lens on it (XF 27mm or XC 15-45mm) and still crop in to compensate for the limited focal length. I have had very successful results this way. I don't do any arithmetic, but just take advantage of the extra cropping room.

 

[g7_chord]

You can print a poster from 10mp. It's more than enough for most people.

Yes, this works because anything that big is normally viewed from a distance. After all, a highway billboard has very limited resolution, but is never viewed up close.

But gallery prints are normally viewed quite closely.

I have had many cameras with 10MP (or less), 12, 16, 20, 24, and 26. The results from my two 40MP cameras are noticeably better -- sharper and with much less grain. And, should it be necessary, I can crop and still have enough resolution for what I'm doing.

I can think of only three disadvantages of higher resolution: Bigger files, more computation time during processing, and higher equipment cost. None of the three outweigh the benefits for me. Of course, another person might reach a different decision.

--
Marc

 

[Truman Prevatt]

I saw a pretty good article comparing a 1.4 teleconverter vs. cropping on a high resolution sensor. The camera was the Nikon Z8 (47 MP) and the lens was the Nikkor 100-400 S series. Not Fujifilm but the same concept holds for evaluating any high resolution sensor.

This a top the line Nikkor lens.

The question address was the which gave better IQ the cropped image at 400 to give an equivalent 560 mm vs the lens with the TC at 560 mm. The conclusions this photographer reached was it was a wash. The cropped version and the and the uncrossed image with the TC were equivalent in IQ - even if the resolution of the cropped image was less.

That makes sense - even if the TC is high quality, it is additional optical elements in the optical path.

Multiple cameras take advantage of this, the Leica Q2/Q3 come to mind with their options to at the press of a button turn a 28 mm lens into a 35 mm lents, etc. The Q2/Q3 in this corp mode also limit the filtering in the frame of the selected focal length. Recently Fujifilm dropped the GFX100RF. The whole philosophy of that camera was to use the GFX100S II sensor for a camera that based the concept of supporting the shooting in not native aspect rations and shooting with a "digital teleconverter."

The GFX100RF also like the Leica provides in finder frame lines as an option. Unlike my Q2 when one shoots in one of these modes, the frame lines appear in the raw. You can chose to ignore the crop or aspect ratio or change it in the post. I didn't use in much in my Q2M. However, I am finding it quite handy in the 100RF.

When I bought the XH2 for my wife and she was complaint that she needed two, sigh that woman is going to drive me crazy, I went with the 40 MP sensor because of the ability to crop. While I have been doing photography for now onto 55 years and somewhat old school and prefer to compose in the viewfinder and not on the computer, the new high resolution cameras are giving us a lot more options.

 

[JohnS59]

In a current thread about the expected new X-E5 a contributor wrote:

"Cropping from 40MP to 20MP is exactly 1.4x focal length.

I wouldn't necessarily recommend buying 40MP 'in order' to crop, but it sure does help to have that extra information."

Actually, that's exactly one of the reasons I am considering one of the 40mp bodies, either the X-T50 or X-E5 once I know the specs of the latter. Long story short, I'm looking for a lightweight combo to complement my Sony RX10iv which is a great camera but hardly light or anywhere near jacket pocketable. Either would also allow be to add a pancake prime when wanting to go really light.

In my naivety I thought that a x2 crop in software after shooting would give me a focal length of 100mm and a resolution equivalent to 20mp. As pointed out this sort of crop would be equivalent to a focal length of 50 x 1.4 = 70mm. This still useful but a little more research shows me a 2x crop would mean 10mp resolution which is acceptable for some purposes but maybe not everyone's needs.

Overall I think the 40mp sensor and 16-50 will be a good option for a lightweight kit when I don't want or need the RX10iv and, if cropped, will still give me a decent size file, especially as I no longer print big or enter competitions.

I'm really just posting this for information for anyone considering a camera with higher resolution (whether Fuji or another) and I attach below a couple of results from my research. So if you're deliberating which sensor size (in MP) to select, maybe this will help.

John

p.s happy to be corrected if I've got anything wrong

From Jonas Rask's review of the X100vi clearing showing the drop in resolution when cropping.

From another discussion found on-line

The "cropping potential" depends not only on the sensor but also on the lens. The sharper the lens the better the result for a given sensor. In your case lens and sensor are not a variable but this is still something to taken into account.

As anyone, I sometimes have to crop and I see quite large differences between lenses. I did not perform systematic tests, but a few lenses stand out, like the 60mm f/2.4 macro, 80mm f/2.8 macro and the 50-140mm f/2.8 zoom. Probably all primes too, especially the more recent ones.

In general, zoom lenses don't have the highest potential, the 50-140mm f/2.8 is an exception. This lens performs also very well with a 1.4 TC. That's no coincidence of course

BTW, it doesn't mean you can't still get very good results with a good standard zoom like the 16-50 when not going too extreme. My experience is also some what limited, other people can have other preferences of course,

 

[Morris0]

In a current thread about the expected new X-E5 a contributor wrote:

"Cropping from 40MP to 20MP is exactly 1.4x focal length.

I wouldn't necessarily recommend buying 40MP 'in order' to crop, but it sure does help to have that extra information."

Actually, that's exactly one of the reasons I am considering one of the 40mp bodies, either the X-T50 or X-E5 once I know the specs of the latter. Long story short, I'm looking for a lightweight combo to complement my Sony RX10iv which is a great camera but hardly light or anywhere near jacket pocketable. Either would also allow be to add a pancake prime when wanting to go really light.

In my naivety I thought that a x2 crop in software after shooting would give me a focal length of 100mm and a resolution equivalent to 20mp. As pointed out this sort of crop would be equivalent to a focal length of 50 x 1.4 = 70mm. This still useful but a little more research shows me a 2x crop would mean 10mp resolution which is acceptable for some purposes but maybe not everyone's needs.

Overall I think the 40mp sensor and 16-50 will be a good option for a lightweight kit when I don't want or need the RX10iv and, if cropped, will still give me a decent size file, especially as I no longer print big or enter competitions.

I'm really just posting this for information for anyone considering a camera with higher resolution (whether Fuji or another) and I attach below a couple of results from my research. So if you're deliberating which sensor size (in MP) to select, maybe this will help.

John

p.s happy to be corrected if I've got anything wrong

From Jonas Rask's review of the X100vi clearing showing the drop in resolution when cropping.

From another discussion found on-line

The "cropping potential" depends not only on the sensor but also on the lens. The sharper the lens the better the result for a given sensor. In your case lens and sensor are not a variable but this is still something to taken into account.

As anyone, I sometimes have to crop and I see quite large differences between lenses. I did not perform systematic tests, but a few lenses stand out, like the 60mm f/2.4 macro, 80mm f/2.8 macro and the 50-140mm f/2.8 zoom. Probably all primes too, especially the more recent ones.

In general, zoom lenses don't have the highest potential, the 50-140mm f/2.8 is an exception. This lens performs also very well with a 1.4 TC. That's no coincidence of course

BTW, it doesn't mean you can't still get very good results with a good standard zoom like the 16-50 when not going too extreme. My experience is also some what limited, other people can have other preferences of course,

You are correct though you left out the most important factor, the Photographer. Till you posted this I was going to answer: Cropping is the best way to find the mistakes you made.

Morris

 

[JohnS59]

In a current thread about the expected new X-E5 a contributor wrote:

"Cropping from 40MP to 20MP is exactly 1.4x focal length.

I wouldn't necessarily recommend buying 40MP 'in order' to crop, but it sure does help to have that extra information."

Actually, that's exactly one of the reasons I am considering one of the 40mp bodies, either the X-T50 or X-E5 once I know the specs of the latter. Long story short, I'm looking for a lightweight combo to complement my Sony RX10iv which is a great camera but hardly light or anywhere near jacket pocketable. Either would also allow be to add a pancake prime when wanting to go really light.

In my naivety I thought that a x2 crop in software after shooting would give me a focal length of 100mm and a resolution equivalent to 20mp. As pointed out this sort of crop would be equivalent to a focal length of 50 x 1.4 = 70mm. This still useful but a little more research shows me a 2x crop would mean 10mp resolution which is acceptable for some purposes but maybe not everyone's needs.

Overall I think the 40mp sensor and 16-50 will be a good option for a lightweight kit when I don't want or need the RX10iv and, if cropped, will still give me a decent size file, especially as I no longer print big or enter competitions.

I'm really just posting this for information for anyone considering a camera with higher resolution (whether Fuji or another) and I attach below a couple of results from my research. So if you're deliberating which sensor size (in MP) to select, maybe this will help.

John

p.s happy to be corrected if I've got anything wrong

From Jonas Rask's review of the X100vi clearing showing the drop in resolution when cropping.

From another discussion found on-line

The "cropping potential" depends not only on the sensor but also on the lens. The sharper the lens the better the result for a given sensor. In your case lens and sensor are not a variable but this is still something to taken into account.

As anyone, I sometimes have to crop and I see quite large differences between lenses. I did not perform systematic tests, but a few lenses stand out, like the 60mm f/2.4 macro, 80mm f/2.8 macro and the 50-140mm f/2.8 zoom. Probably all primes too, especially the more recent ones.

In general, zoom lenses don't have the highest potential, the 50-140mm f/2.8 is an exception. This lens performs also very well with a 1.4 TC. That's no coincidence of course

BTW, it doesn't mean you can't still get very good results with a good standard zoom like the 16-50 when not going too extreme. My experience is also some what limited, other people can have other preferences of course,

You are correct though you left out the most important factor, the Photographer. Till you posted this I was going to answer: Cropping is the best way to find the mistakes you made.

Morris

Yes, if you first think: excellent composition and then later decide to crop in post processing. You can also crop in your mind before if you can't get close enough or the lens is not long enough. That's no mistake. The first option is the way it is done most I think.

BTW, I seldom read or heared people saying: "nice image, but it is cropped too much. On the other hand, "nice image but you over sharpenend it" is more common.

So, even with modest sensors of today, e.g. 24 MP, there is a lot of potential and it is not that critical at all. Only if you go always for the best possible resolution it may be another story. That is a minority.

 

[a_c_skinner]

Exactly. Don't think of it as cropping but as increasing the magnification for display of part of the image so every cause of unsharpness (sensor, lens, technique) will be enlarged and more obvious.

 

[bastibe]

A while ago, I compared to 1.4x TC and 2.0x TC with cropping, on the 70-300/X-T3. I found the benefits of the TC over cropping very marginal, and not worth the price of AF speed and hassle.

Interestingly, the TC faired much better at ~200mm, where my 70-300 is sharpest, than at 300mm, where it gets a bit soft. I would go so far as to say that it was slightly better at 200x1.4, than at native 300. I also regularly use my Ricoh GR's crop modes, which yield 16 MP 23mm, and 8 MP 27mm images.

So if you have a sharp lens, cropping (whether in-camera, in post, or optically) can yield very good results. My 23mm f/1.4 WR, for instance, can easily crop to 35mm and beyond. But beware that a soft lens, such as my 35 f/1.4, falls apart pretty quickly if you crop in.

 

[JohnS59]

A while ago, I compared to 1.4x TC and 2.0x TC with cropping, on the 70-300/X-T3. I found the benefits of the TC over cropping very marginal, and not worth the price of AF speed and hassle.

Interestingly, the TC faired much better at ~200mm, where my 70-300 is sharpest, than at 300mm, where it gets a bit soft. I would go so far as to say that it was slightly better at 200x1.4, than at native 300. I also regularly use my Ricoh GR's crop modes, which yield 16 MP 23mm, and 8 MP 27mm images.

So if you have a sharp lens, cropping (whether in-camera, in post, or optically) can yield very good results. My 23mm f/1.4 WR, for instance, can easily crop to 35mm and beyond. But beware that a soft lens, such as my 35 f/1.4, falls apart pretty quickly if you crop in.

Yes, a sharp lens helps a lot. The point is most people don't know which one really is because the don't need the potential. There seem to be not that much difference between their lenses with normal use and print sizes. The all look good these days.

When cropping (or pixel peeping of course) it shows.

BTW, by definition you crop only a part of the picture. Outer corners, which are the weakest, are often not part of it!

With the high res sensors and good quality lenses of these days I would hesitate to buy a TC. Of you often need the extra reach it is better to buy a lens with it, if not that often why not crop? But be sure the lens is good at the largest focal distance. Zoom lenses tend to be a little soft there but there are exceptions. E.g. the 50-140mm f/2.8 I mentioned.

The 200mm f/2.0 comes with a TC and even using that you still can crop the hell out of it!

 

[jhorse]

Useful post - thanks. While I cannot comment specifically on a XE5 + 16-50 combo as a zoom, I can offer a cropping thought for a 40MP sensor from my experience of using the X100VI.

I am content to use the cropping mode on this camera and in my very unscientific tests I cannot really see a discernible reduction in image quality for normal viewing. Whether cropping in camera or in post is better is a moot point.

Here is a table I drafted for myself as an aide memoire when I use the digital crop. As I have the wide and tel converter lenses for the camera they are included.:


The mm numbers in brackets are full frame equivalent.

I have the 16-50f4.8 (on a XT5) and see no reason why one cannot crop to a about x1.4 or 70mm.

Hope this helps.

--
J.

Jules7

Jules' galleries; enjoy.

jules7.smugmug.com

 

[RWN Photo]

Who cares about the other guy?

If you take this as written:

"I wouldn't necessarily recommend buying 40MP 'in order' to crop, but it sure does help to have that extra information."

To me, it says that they wouldn't recommend cropping every image from 40mp. But you certainly can. Real photographers, or passionate purists, are going to say you should not do it. They will say you need to get it right in post. And so on. There is truth to that, there is an art to that, and cropping every image is pretty much "silly".

But if all you do is post on Instagram, and you are pretty much cropping every single image, to get the only Instagram rations allowed (2x3 isn't one), and then you downsize the heck out of it too.

Now you're talking about taking your 40mp camera and using it eventually as a 2mp one.

 

[Morris0]

In a current thread about the expected new X-E5 a contributor wrote:

"Cropping from 40MP to 20MP is exactly 1.4x focal length.

I wouldn't necessarily recommend buying 40MP 'in order' to crop, but it sure does help to have that extra information."

Actually, that's exactly one of the reasons I am considering one of the 40mp bodies, either the X-T50 or X-E5 once I know the specs of the latter. Long story short, I'm looking for a lightweight combo to complement my Sony RX10iv which is a great camera but hardly light or anywhere near jacket pocketable. Either would also allow be to add a pancake prime when wanting to go really light.

In my naivety I thought that a x2 crop in software after shooting would give me a focal length of 100mm and a resolution equivalent to 20mp. As pointed out this sort of crop would be equivalent to a focal length of 50 x 1.4 = 70mm. This still useful but a little more research shows me a 2x crop would mean 10mp resolution which is acceptable for some purposes but maybe not everyone's needs.

Overall I think the 40mp sensor and 16-50 will be a good option for a lightweight kit when I don't want or need the RX10iv and, if cropped, will still give me a decent size file, especially as I no longer print big or enter competitions.

I'm really just posting this for information for anyone considering a camera with higher resolution (whether Fuji or another) and I attach below a couple of results from my research. So if you're deliberating which sensor size (in MP) to select, maybe this will help.

John

p.s happy to be corrected if I've got anything wrong

From Jonas Rask's review of the X100vi clearing showing the drop in resolution when cropping.

From another discussion found on-line

The "cropping potential" depends not only on the sensor but also on the lens. The sharper the lens the better the result for a given sensor. In your case lens and sensor are not a variable but this is still something to taken into account.

As anyone, I sometimes have to crop and I see quite large differences between lenses. I did not perform systematic tests, but a few lenses stand out, like the 60mm f/2.4 macro, 80mm f/2.8 macro and the 50-140mm f/2.8 zoom. Probably all primes too, especially the more recent ones.

In general, zoom lenses don't have the highest potential, the 50-140mm f/2.8 is an exception. This lens performs also very well with a 1.4 TC. That's no coincidence of course

BTW, it doesn't mean you can't still get very good results with a good standard zoom like the 16-50 when not going too extreme. My experience is also some what limited, other people can have other preferences of course,

You are correct though you left out the most important factor, the Photographer. Till you posted this I was going to answer: Cropping is the best way to find the mistakes you made.

Morris

Yes, if you first think: excellent composition and then later decide to crop in post processing. You can also crop in your mind before if you can't get close enough or the lens is not long enough. That's no mistake. The first option is the way it is done most I think.

BTW, I seldom read or heared people saying: "nice image, but it is cropped too much. On the other hand, "nice image but you over sharpenend it" is more common.

So, even with modest sensors of today, e.g. 24 MP, there is a lot of potential and it is not that critical at all. Only if you go always for the best possible resolution it may be another story. That is a minority.

That's not what I'm talking about. As you magnify an image with a crop or optically, your mistakes are magnified. This includes too slow a shutter speed, bad pan, failing to squeeze off the photo. When one crops and sees these errors, they probably don't display the image.

Morris

 

[Truman Prevatt]

Exactly. Don't think of it as cropping but as increasing the magnification for display of part of the image so every cause of unsharpness (sensor, lens, technique) will be enlarged and more obvious.

In the film days it was denoted by "enlargement ratio." The higher the enlargement ratio between negative to print, the more imperfections in the negative. The higher the enlargement ratio the more the impact the film grain on the final print.

For example an 8x10 of a 35 mm negative represented an enlargement ratio of 8, of a 4x5 sheet of film it was 2. For a 8x10 sheet of film it was a contact print. Of course that also means the 35 mm negative is magnified 4 times more than the 4x5 and 8 times more than the 8x10. No wonder the dramatic prints of iconic landscapes were taken using at large format camera, 4x5 (Bret Weston), 8x10 (Ansel Adams, Bruce Barnbaum, etc.) , 11x14 (Edward Weston), 12x20 (Clyde Butcher).

One can't produce a wall 81 x 130 cm print of Oxbow Bend that looks like you are standing there that beautiful morning using 35 mm film.

Oxbow Bend 63

Limited-edition photograph developed from Clyde Butcher original large-format film.

clydebutcher.com

The more things change - the more they remain the same. Today we just attribute different words to the equivalent concept.

 

[Bill Ferris]

Something that usually gets ignored in these discussions but is good to keep in mind, is that cropping makes noise more prominent. It's not as big an issue when working with a strong exposure, but as the available light diminishes and the resulting image moves closer to one's noise tolerance threshold, cropping may push the resulting image over the line.

 

[jhorse]

Something that usually gets ignored in these discussions but is good to keep in mind, is that cropping makes noise more prominent. It's not as big an issue when working with a strong exposure, but as the available light diminishes and the resulting image moves closer to one's noise tolerance threshold, cropping may push the resulting image over the line.

Good point. I assume this applies regardless as to whether one uses an in-camera crop mode, such as I sometimes use on a X100VI, or in post.

 

[Bill Ferris]

Something that usually gets ignored in these discussions but is good to keep in mind, is that cropping makes noise more prominent. It's not as big an issue when working with a strong exposure, but as the available light diminishes and the resulting image moves closer to one's noise tolerance threshold, cropping may push the resulting image over the line.

Good point. I assume this applies regardless as to whether one uses an in-camera crop mode, such as I sometimes use on a X100VI, or in post.

It does. Cropping effectively reduces the sensor surface area used to capture light. As a result, less total light energy is used to make the photo and that makes noise more visible.

The stronger the exposure one works with, the less likely it is that noise will become obvious & annoying. But in low light, or when using an f-stop or shutter speed that significantly limits how much light is projected on the sensor, it's something to be mindful of.

 

[JohnS59]

Something that usually gets ignored in these discussions but is good to keep in mind, is that cropping makes noise more prominent. It's not as big an issue when working with a strong exposure, but as the available light diminishes and the resulting image moves closer to one's noise tolerance threshold, cropping may push the resulting image over the line.

Cropping digitally also results in less dynamic range I believe (maybe related). This was discussed in a thread about the new GFX100RF which has a dedicated cropping function as it's hallmark.