[Wired]

IMO, 50mp cameras exist for the select few professionals who need it for billboards and whatnot, and is then sold to everyone else who thinks they need it to fund more research for important things like clean high ISOs and dynamic range.

Not really. There is something about perceived viewing distance. The further away you are from the image, the less dpi/mp it can be to still look good. I took an image of a car dealership on a Pentax Q. I posted it to Flickr, and someone from the dealership contacted me asking if they could use it in advertising, so I sold them the image. They used it on a digital billboard on one of the busiest intersections in town for 8 months and it looked great. Yet it came from a Q.

I'm sure if someone were to get within 5 feet it would look like poo. but perceived viewing distance from a pedestrian or car traveling by, it was sharp enough.

[RonHendriks1966]

No , 5 Mpixels is not enough. According to the resolution of your own eyes and the viewing distance at least the diameter of the print, the right number seems to be 7 mega pixels (Print Resolution Calculator - Points in Focus Photography).

wow thanks, well, I was close with my pixelnumber.

[ScooterMaxi Jim]

Yes it does. The optical bandwidth is measured in lines per mm for a given contrast value, there is also an attenuation of the spacial bandwidth due to diffraction, so even a lens built with perfect glass material has a limited resolving power, that's why is you close down the aperture the optical imperfection diminishes but the diffraction increases, independent from the sensor. Unfortunately, there are limits to everything.

---------- Post added 03-25-15 at 08:30 PM ----------


Yes, indeed, the lens is the same, but when you had the K100D, you zoomed your image at 100% on you LCD and "waooow, this lens is sharp", a number of years later, you buy the K-3, you capture a photo with the same lens, you load it on your computer and zoom it at 100% on you LCD, and realize, beurk, it's blurred, this lens is not as good as it used to be ;-)

No , 5 Mpixels is not enough. According to the resolution of your own eyes and the viewing distance at least the diameter of the print, the right number seems to be 7 mega pixels (Print Resolution Calculator - Points in Focus Photography).

Empirically, you are right. Empirically, the Nyquist limit should be the end of the discussion. However, my eyes tell me something very different than the science. If what you say is right, I should be fine with my good, old Canon 20D (8 megapixels). In fact, I was fine with it for various reasons until the K-3 (which finally matched and surpassed the 20D AF performance - so many years later). All I can say is that really good lenses get better with the increased resolution - well beyond what Nyquist theorizes. The m50mm f/1.4 does so much better on the K-3 than on the previous dSLRs. Mount it on a Q - and the resolution and rendering are amazing shooting at f/2.4 or 2.8. That lens looked like garbage on the *istD.

Out of thousands of photos taken of this dog over the years - most taken with modern lenses and only a few taken with small sensor cameras - none are meaningfully optically superior to this one (Q / m50):
Remembering Max - James Robins - Powered by Phanfare

[reeftool]

I have no desire right now for a 50 MP camera of even a 36. The file sizes are huge. I was getting good crops from my K10D. I print a lot of shots and many others do also but I think far too many people are staring at 100% crops on very large Hi Res monitors. A lot of people aren't looking at prints. I have a 23 inch 1920 x 1080 monitor. I have had shots with slight motion blur or slightly missed focus that were obviously soft looking on my monitor but looked super on a 4x6 or even a 5x7 print . Back in the days of film, high resolution meant getting a sharp 8x10 print. That's tablet sized today and even the tablets are getting bigger. The 50 MP sensors are for large, super high resolution monitors, not prints

There certainly are reasons for more resolution and pixels on sensors but not for what most of us are shooting. The files are too big to upload for general online viewing anyhow.

[Nicolas06]

If you want to crop heavily, everything (lens, technique, conditions) must be more than up to scratch, and if just one thing is missing (e.g. low light, hence can't stop down as required) it's a no-go.

This is key, in particular the last part. If the light is not good enough, you'll not get much resolution out of your picture.

From experience doing this, quite often you can crop heavily in the 35-135mm range if you have prime or a great quality zoom provided the light is good. For the things I tend to shoot and the lighting condition I tend to have even without thinking too much of it, it is fine most of the time.

If you continuously shoot in low light or with very long lenses that might be entirely different. An heavy crop of 135mm give you 300mm still something that may require only 1/100 or 1/200s with SR and that will be sharp in good light. If you crop a 400mm lense, that may get your to 800mm equivalent. If you need to push the isos a bit too much, you'll be more limited in your cropping.

[Nicolas06]

Yes, indeed, the lens is the same, but when you had the K100D, you zoomed your image at 100% on you LCD and "waooow, this lens is sharp", a number of years later, you buy the K-3, you capture a photo with the same lens, you load it on your computer and zoom it at 100% on you LCD, and realize, beurk, it's blurred, this lens is not as good as it used to be ;-)

No , 5 Mpixels is not enough. According to the resolution of your own eyes and the viewing distance at least the diameter of the print, the right number seems to be 7 mega pixels (Print Resolution Calculator - Points in Focus Photography).

Forward to my own experience with K5 and K3:

K5: this is blured.
K3: this is tack sharp.

This works with DA21 (center), DA35 f/2.4, FA50, DA50-135. FA77 and DA50-135 are also talk sharp on K3, much more at 100% than the image where at 100% on K5... I mean the K5 has a low pass filter to blur the image and this is very visible at 100%. I know it because I got the 2 bodies.

It may not work with a 55-300 but hey even a 18-135 is very sharp in the center for example closed down to f/8 at 24mm.

Sometime I think the point here is to speak theory, explain how you would not gain anything, how it is not possible blablalba...

And ignore completely all the people that experience the thing everyday.

Instead of saying how a lense should or should not behave on a K3 TRY IT !

[Nicolas06]

Out of thousands of photos taken of this dog over the years - most taken with modern lenses and only a few taken with small sensor cameras - none are meaningfully optically superior to this one (Q / m50):
Remembering Max - James Robins - Powered by Phanfare

There not much more than the websize display on the first image. At 100% is it not sharp, at 50% neither. It is sharp at 27% on my screen that is matching the normal display size... From a fast computation on this example there no much more detail on this Q with the 50mm than you would get from a 16MP crop reduced to say 1 to account the sensor size difference. I say the low pass filter would make the K5 image look more blury through.

While I see difference between K3 and K5 it is really complicated to conclude on this case with the Q.

I suppose through that some Q lenses or in some condition you can get 100% sharp results on the Q meaning we have lot of margin in MP.

[Dlanor Sekao]

If we make a photobook large size square, 30x30cm, that is almost 12x12 inch one only needs 5 megapixels for good quality. Yes it gets a little better at 8 megapixel and even just another notch better at 12 megapixel, but I doubt you will see this on a regular base.


So why do some people think that 50 megapixels for the next FF camera is a great idea?

I don't at all ! I wont even consider it if its over 36mp and will pass it right on by.

[ScooterMaxi Jim]

There not much more than the websize display on the first image. At 100% is it not sharp, at 50% neither. It is sharp at 27% on my screen that is matching the normal display size... From a fast computation on this example there no much more detail on this Q with the 50mm than you would get from a 16MP crop reduced to say 1 to account the sensor size difference. I say the low pass filter would make the K5 image look more blury through.

While I see difference between K3 and K5 it is really complicated to conclude on this case with the Q.

I suppose through that some Q lenses or in some condition you can get 100% sharp results on the Q meaning we have lot of margin in MP.

It probably depends where you are looking because DoF is very low on the shot; if the goal was ultimate sharpness on that shot it would have been dialed down to around f/4 (but the shot would have been less interesting and had more noise due to lower light brought in). Having printed the image significantly enlarged, it is quite a bit sharper than what you state. Possibly a monitor problem - is it old or poorly aligned? In any event, what you downloaded and what was originally processed to TIF are not identical. My essential points are that lenses don't just stop getting sharper once you exceed the Nyquist limit, and this lens performs better on newer, higher resolution (low AA or no AA filter) sensors.

I agree that the difference between K-3 and original K5 is easier to compare due to their comparable sizes - and the sharpness differences are significant (far less significant difference on the newer 16mp sensors with low AA or none).

Keep in mind, as well, that sharpness as we relate it to photography is human perception, not a scientific absolute. What you need to shoot portraits (human or animal) is far less critical than the sharpness needed to convey excellence in a landscape. Pretty much no optic will perform adequately on the Q for landscape work.

[Nicolas06]

It probably depends where you are looking because DoF is very low on the shot; if the goal was ultimate sharpness on that shot it would have been dialed down to around f/4 (but the shot would have been less interesting and had more noise due to lower light brought in). Having printed the image significantly enlarged, it is quite a bit sharper than what you state. Possibly a monitor problem - is it old or poorly aligned? In any event, what you downloaded and what was originally processed to TIF are not identical. My essential points are that lenses don't just stop getting sharper once you exceed the Nyquist limit, and this lens performs better on newer, higher resolution (low AA or no AA filter) sensors.

I agree that the difference between K-3 and original K5 is easier to compare due to their comparable sizes - and the sharpness differences are significant (far less significant difference on the newer 16mp sensors with low AA or none).

Keep in mind, as well, that sharpness as we relate it to photography is human perception, not a scientific absolute. What you need to shoot portraits (human or animal) is far less critical than the sharpness needed to convey excellence in a landscape. Pretty much no optic will perform adequately on the Q for landscape work.

This is just I was expecting better from the lense... But as I understand the apperture was the best of the lense and this explain a lot to me. The fast 50mm are not that sharp at large appertures but do offer great bokeh.

As a monitor problem... A problem that would affect only this image, I'am sorry, this look very unlikely to me.

I mean the picture is good enoug to print 12"x8" and have something pleasant. The type of picture and the apperture make it not so sharp overall that is not an issue.

As if the photo is great, for sure it is We should not think too much of sharpness but use it as a tool. Here you didn't need crop, huge print size or whatever special, so that's perfect. And I think that the Ron point.

The final image once cropped/edited can likely do with 5MP, some say 7, DxO consider 8MP the standard (for 300dpi A4 approximarely 12"x8" prints) in most uses.

As what you need to get theses great 5-7MP, it may require a 24 or 36MP sensor for heavy cropping if you want to stay light and avoid long/heavy tele, if your are after wildlife or if you are after macro. Even for landscape, finding new framing inside the initial image and still get a good quality shoot that can print large enough or the "keep light" strategy is interresting to me. With K5 one shoot panos for example where often a bit limited in term of resolution, with K3 it is ok. All in all it is a TC that doesn't have any optical aberation.

For portraiture if you are not a paparazi, I don't think you should care too much even through it mean if your are using prime like me you can have less prime overall to cover the range and can concentrate more on getting the best one and the one you like the more.

So maybe because I like the possibility for cropping and still have great sharpness in my shoot I can see use for 24MP... Would I see use of 50MP? Maybe but that's not at all critical to me. Putting it inside an FF anyway mean that it goes against my need for staying small/light. Still I see a few usage where it could be of use and of course the more you get, the more limited are the case where you need the extra.

[RonHendriks1966]

Still I see a few usage where it could be of use and of course the more you get, the more limited are the case where you need the extra.

We have a discussion about a New stadium for the soccerclub in Rotterdam overhere. The current stadium is old and seats 50.000 People. The idea is to build a New One seating 70.000 People. The club almost never need this. Just some games a years It's completely sold out and they Could use extra setting up to 60.000. But for international games it is different, not for our national team, but for Champions League. If you want the Champions League final you need the 70.000 seating. So for that One evening in may 2023 we might build that oversized stadium. To my my feelings the 50 megapixels is the same.

[Nicolas06]

We have a discussion about a New stadium for the soccerclub in Rotterdam overhere. The current stadium is old and seats 50.000 People. The idea is to build a New One seating 70.000 People. The club almost never need this. Just some games a years It's completely sold out and they Could use extra setting up to 60.000. But for international games it is different, not for our national team, but for Champions League. If you want the Champions League final you need the 70.000 seating. So for that One evening in may 2023 we might build that oversized stadium. To my my feelings the 50 megapixels is the same.

All the options we have in camera almost we can do the same without. Some shoot the olympics with large format camera and film. They may make more money on their shoot for this very reason than if they used the latest canikon body. Still we think we need a camera that can bump the shadow, jump in isos, do very fast burst mode, allow for multiple exposure, can be tethered etc etc. Not long ago just thinking of a digital camera what seen as completely ridiculous by most photographers.

All of this in the end each time serve a few people really why the other don't really use it. So when you buy the gear you should see if there something you need (at least 1 feature) that the more basic, less expensive model can't do. You should also see how you want to operate. If you make a good living of your photography the gear is pretty unexpensive in particular if you limit yourself to FF format. It can make sense to have the greatest/latest even if it serve 1 time out of 1000. Because you'll have the better result then. If you make 100K$ a year out of your photos it make lot of sense. If you make 20K$, it is just throwing money out of the windows.

That's the same for your stadium. if you plan to make the worldwide cup, the olympics or simply the old one just burned and 70K seat is the same price as 50K seat, there no reason to not go for 70K.

[reeftool]

Megapixels on sensors are going to increase. That's a given and camera manufacturers for the most part are stuck with what the sensor manufacturers are giving them. There are some downsides to more pixels such as file size which in turns, slows down the burst rate and overall speed of the camera. It may not matter all that much to some but could be extremely important to others. I know it doesn't take long to fill up a hard drive these days. Each year, we all take more and more photos. I have been shooting motorcycle races this winter and have been averaging 1000 shots per event. I filled a 1TB external drive last year and I now have a half filled 3TB drive. I also have internal drives filled on my desktop towers. These are K10D and K5 shots. Imagine 36 or 50 MP files. Start spreading this over a lifetime of shooting.

Elsewhere in the Forum are threads about PP and computers which are just as important in all this as our cameras. The trend of the computer industry are tablets, small form factor desktops like the NUC and Mac Mini, and lightweight laptops. They are pushing devices with flash memory and SSD's and urging users to use the cloud. There's a lot of threads with the " What Laptop?" or "What Computer?" should I buy titles. Yet, the 2 industries are going in opposite directions. Think about where this leaves us a photographers. Just for fun, try uploading full sized RAW files to the cloud some time. This isn't a reasonable option for serious photographers. We're going to need our own servers to store all these hi res photos!

I believe there's a lot of improvement to be made in digital photography and just making sensors with more pixels isn't going to be the answer but that's all they can currently do right now. We need high resolution lossless photos with small file sizes.

[Nicolas06]

Megapixels on sensors are going to increase. That's a given and camera manufacturers for the most part are stuck with what the sensor manufacturers are giving them. There are some downsides to more pixels such as file size which in turns, slows down the burst rate and overall speed of the camera. It may not matter all that much to some but could be extremely important to others. I know it doesn't take long to fill up a hard drive these days. Each year, we all take more and more photos. I have been shooting motorcycle races this winter and have been averaging 1000 shots per event. I filled a 1TB external drive last year and I now have a half filled 3TB drive. I also have internal drives filled on my desktop towers. These are K10D and K5 shots. Imagine 36 or 50 MP files. Start spreading this over a lifetime of shooting.

Elsewhere in the Forum are threads about PP and computers which are just as important in all this as our cameras. The trend of the computer industry are tablets, small form factor desktops like the NUC and Mac Mini, and lightweight laptops. They are pushing devices with flash memory and SSD's and urging users to use the cloud. There's a lot of threads with the " What Laptop?" or "What Computer?" should I buy titles. Yet, the 2 industries are going in opposite directions. Think about where this leaves us a photographers. Just for fun, try uploading full sized RAW files to the cloud some time. This isn't a reasonable option for serious photographers. We're going to need our own servers to store all these hi res photos!

I believe there's a lot of improvement to be made in digital photography and just making sensors with more pixels isn't going to be the answer but that's all they can currently do right now. We need high resolution lossless photos with small file sizes.

I think we speak of opposite needs.

The amator is not going to shoot all the time 1000 a days and in 5, 10, 50 years appreciate to relook at all the photos he would have taken. 1TB filled in 1 year is 30K RAWs or 150K JPEG. This mean in 20 years he would end up with 600K to 3000K photos. I simply don't trust him to appreciate looking at all of them or any of his friend/relative etc to appreciate looking at them too. By keeping no more than 10% of that, he would on the contrary make the experience far better as he would be able to go back to the photos with pleasure... His friend would not be so affraid as to hear of a planned photo session and invent reasons they cannot come.

If you are a pro and are selling image, the storage should be part of the cost, and the cost would be very low. 1TB is 60-70€. Backup to the cloud likely more expensive but necessary to sell the photos online in event like a motocross day so everybody can choose the photos he is interrested into and buy them. What I don't get is why you would keep most of theses photos more than a year anyway, in particular the photos nobody asked. It make no sense from a business point of view.

If you shoot a wedding, it may take more sense to backup and store the photos far longer, but then you need to include this in the price. I suppose you can even sell an hosted website with all the picture presented nicely etc etc for all the familly to access or even public. But this is a service the client would pay for.

But really to me that a pro should have pro gear to manage photos while amators would need to use the easy tools at their disposal it make lot of sense. For most of the people taking photos anyway, they don't even use RAW, a DSLR, they would relook at the pictures a few years from now and if they want they'll discover they have lost already a good share of them anyway.

[Clavius]

Each year, we all take more and more photos. I have been shooting motorcycle races this winter and have been averaging 1000 shots per event. I filled a 1TB external drive last year and I now have a half filled 3TB drive. I also have internal drives filled on my desktop towers. These are K10D and K5 shots. Imagine 36 or 50 MP files. Start spreading this over a lifetime of shooting.

Ooph, I get tired just by reading this. What do you actually do with that amount of photo's? After filtering the first 100 I'd call it a day. With me, it's actually the other way around. The more experienced I become, the less photo's I tend to take, whilst the amount of keepers has gone drastically up.