 |
| CSX Railroad Brakeman. This is a 40mp image. If it were 20mp, no one of could tell the difference. (click to enlarge) Fujifilm X-T5; 16-55mm f/2.8 lens @ 23mm; 1/500th sec. @ f/5.6; ISO 640 |
I'm of the opinion that for the vast majority of us photographers, 20-24mp are enough (need vs. want) for almost everything we want to do with our images. There are exceptions, of course.
Several years ago, as an experiment mainly for my own edification, I had two 24" X 30" (61cm X 76cm) custom prints made of an identical scene using both a 16mp Olympus E-M1 and a 36mp Nikon D800E. The photographs were made in a combination of bright sunlight and deep shade at base ISOs of a subject with lots of nuanced detail so if any of the detail was not visible or discernable, it would have been easy to tell. I literally could not see a difference between the two prints even examining them up very closely over and over again. No one I showed the prints to, including some long time photographers, could tell any difference in minute detail, sharpness, color rendition, etc. They were about as identical as two prints could be. I had to refer to the file names printed on the backs of the prints to know which was which. That experiment was just one example of comparing sensor resolutions and a sample of one doesn't prove much, however. I'm sure if they were made at very high ISOs in poor lighting conditions differences may have been found but the experiment was mostly about how much difference in detail could be seen when using 16mp versus 36mp and the answer in this case was none.
Don't get me wrong, I love diving deeply into one of my 46mp Nikon Z8 images or one of my 40mp Fujifilm X-T5 images and examining the small details at 100% on my 27 inch monitor. There is a great deal of satisfaction in that for me. It is nice to be able to do that but in almost all cases not necessary to have those extra pixels that are really going to waste (we're paying a lot of money for those pixels we discard in every photo when publishing to the web!). But this got me wondering why the camera companies have been pretty much stuck at the pixel counts in the latest digital mirrorless cameras that were also in their predecessors?
I wonder why Nikon has not produced a camera with more than 46mp? They've had that same pixel count since 2017 with the introduction of the D850. They kept that pixel count into the mirrorless era with the Z7, Z7II, Z8 and Z9. Why haven't they gone to sensors with more pixels? I wonder why Canon has only produced their latest and greatest cameras with the same 45mp as their last generation? I wonder why OM Systems has not created their newest stacked, back-side illuminated cameras with more than 20mp? It seems only Fujifilm has given us their latest generation of cameras with 40mp up from 26mp.
Now consider Sony and Leica. Both have both given us cameras with 60-61mp. Do these other companies not have access to the higher megapixel sensors (I'm sure they do)? Engineering-wise, can't they do it (I'm sure they can)? Are the megapixel wars over (maybe, but I wouldn't bet on it)? Market forces? Or is it something else altogether?
Could it possibly be that Nikon’s and Canon’s lenses can’t resolve more than about 45-50mp? Is it possible that OM/Olympus lenses can’t resolve more than 20-25mp? As far as I am aware, only Fujifilm came out and actually told its customers that their older lenses would only resolve up to 32mp and the newer ones are the only ones that will resolve the full 40mp of their latest generation of cameras—the X100V, X-T5, X-T50 and X-H2. However, we don't know if their newest lenses will resolve more than 40mp. At least they told us, gave us some information we, as customers and potential buyers, can use—valuable information, in fact, that we need before purchasing gear. Why haven't the other manufacturers told us anything about if their lenses can fully resolve their current 45-46mp or resolve even more? We are assuming they can. But can they really? I don't think not telling us is a good business practice and certainly doesn't serve their customers' needs.
As I said, I have no answers, I was just wondering why these major manufactures have not produced cameras with more pixels than their last generation of cameras.
Of course, this begs many more questions in my mind such as will future upgrades not be more pixels but be by way of more and better computational features? Will future sensors (along with image processors) keep the same pixel count but have lower noise and more dynamic range? Will more traditional features be improved or refined? Will buffers grow? Will the cameras start coming with a good amount of internal memory? As I said, I have no idea.
Just food for thought as to where cameras may or may not be going in the future. Maybe in my retirement I just have too much time on my hands! That said, I just thought I would throw this out there for you to think about as well. LOL
Thanks for looking. Enjoy!
Dennis A. Mook
All content on this blog is © 2013-2024 Dennis A. Mook. All Rights Reserved. Feel free to point to this blog from your website with full attribution. Permission may be granted for commercial use. Please contact Mr. Mook to discuss permission to reproduce the blog posts and/or images.
Thank you, Dennis, for sharing your thoughts and comparisons between camera resolutions and lenses. I often wonder why I always shoot at the highest resolution possible (e.g., with my X-T5 or X100V). I always choose Large and Fine settings (Dimensions: 6864x5152 = 35.36 MP; File size: 13.60 MB) instead of Medium or Small and Normal (Dimensions: 3888x2592 = 10.08 MP; File size: 2.57 MB). Yet, I’ve never printed any of my photos and mainly view them on a 27-inch screen. The only thing I do regularly is crop the image during post-processing.
ReplyDeleteI’m curious how I can break the habit of always shooting at the highest possible resolution. The advantages of shooting smaller would be less storage usage and faster file handling. How do you handle this temptation? Martin
Martin, thank you for your comment and question. My advice is to always buy a camera that you “want” to use everyday, no matter what the sensor’s resolution. If you buy one that has the resolution you think you’ll need but is not one which you will want to pick and enjoy using, chances are you won’t pick it up and also you won’t enjoy your photographic experiences as much either. Cameras must feel good in your hand, have menus that are easy for your brain to “click’ with, have buttons and dials that feel good and also make sense as to their placement as well as have the features that allow you to make the photos you want to make. Just a few years ago professional photographers were using 12mp cameras and raving about them. So, resolution and the number of pixels on a sensor, in my opinion, should not be a primary concern when choosing a camera, unless you need all those pixels for specific reasons.
DeleteIt is natural to set your camera to use the full resolution of the sensor. It is an insurance policy. Ninety-nine out of hundred times you may not need the extra large files, but you will have the pixels for that one time you need them. You may need to crop heavily or you may want to make a very large print (for yourself or someone else) to be proudly displayed on a wall.
From your comment it sounds as though you are shooting JPEG files and they are considerably smaller than raw files. Storage media is no longer expensive and the older spinning hard drives, SSDs, memory cards, etc. go on sale almost every week. Use the full resolution of your X-T5. You can always edit the full size image file, export it as a smaller JPEG from Lightroom and have it automatically re-import into Lightroom in the same folder. If you are then happy with the edited smaller file and need the storage space, you can then delete the original large file. I hope this helps. ~Dennis
I think this could be a matter of how accurate your yardstick might be. Always consider your final product when in production. In this case, display needs whether a particular print size or for electronic display.
ReplyDeleteI am suspicious of the specifications the manufacturers present. How difficult would it be for them to upsize the resolution in the camera software?
As a test, I went out with camera on a walk one day, photographing typical subjects in a small town. Later, I duplicated all the files. The originals were processed/converted as usual. The duplicates were processed/converted at 1.5 times the usual horizontal resolution. I couldn't see any difference in my standard 4K display or in test 19" prints. But, I could say my OM was functioning as a 45 mp camera.
There is no real check on the manufacturers, so I suspect chicanery does exist.
Long ago, I lost faith in the chart tests on the internet because the test chart is not representative of the real world subjects I photograph. Most lenses are designed with field curvature. It always has been that way. That is why, way back, they told photographers of groups to have the people at the ends stand slightly closer to the camera. That helped bring those people into focus.
My grandmother used to say "The proof of the pudding is in the taste" and I think that applies to most things photographic as well. Believe what you see, not so much what you read, including this. Make your own tests.
I was dismayed at the noise level of my 46mp full frame sensor. It is not much different at ISO 8000 and up than the m43 OM sensor. There is a significant improvement in noise using the Z7 at the Medium setting when you expect to need high ISO's. If you are unlikely to downsize the image for final output, then there is something to be gained in that particular situation. In general, I use the Large setting on my Z7.
I do not find Adobe to be as proficient in downsizing files as upsizing. Adobe is very good at upsizing. This may vary with a particular image and luminance range. It seems to matter with the Z7 when I frequently have to reduce the resolution to fit the output.
As a follow up- I knew I read it here somewhere. Aug. 19, 2015 there was an excellent post on this blog about lens testing by chart and curvature of field.
ReplyDelete