Big Megapixel Camera in 2014

[Apop]

Come one , no need for insults....

why shouldn't people in the food business be able to get in focus shots btw???



OT: i still hope to learn to shoot with my current camera!, it seems more advanced then the d800, but also less forgiving when used with the my current lens

(getting the shot to fill the frame , of moving subjects vs just getting the shot and crop afterwards)

[jrista]

As far as numbers of megapixels and frame rate are concerned... the real bottleneck is not reading the sensor, it's processing the data afterwards.

For example, we pick on a 60D with a digic4 processor and an 18Mp sensor. It is capable of reading that sensor 60 times per second, but can only process 5.9 frames per second.

And I know this will be controversial, but this is one of those cases where a mirrorless design is superior.... no mirror to shake the camera or to slow down the burst. Even a gopro can read it's 11Mp sensor 240 times per second.... Every design decision has it's plusses and minuses, nothing is perfect for everyone.

You just have to question the usefulness of reading still photos 240 times per second. That is not only video, that is high speed video. Unless you have the literal intention of playing that "video" back on a 240Hz screen at its native rate, the usefulness of such a frame rate for a still camera actually eludes me. Same goes for 60fps. It is one thing to look at 12, 14, 16, maybe 20 fps for a still camera. But at 60fps, and worse at 240fps, one momentary press of the shutter will capture a ton of frames.

Assuming we had image processors that could keep up, a 1/10th second shutter press in AI Servo with a 60fps DSLR would mean you get 6 frames. SIX FRAMES for one tenth of a second. That might get you the best shot up to the moment where an Osprey flexes out its claws in preparation for grabbing a fish....but it would still miss the moment where the fish is in-talon, the moment after it is pulled out of the water. If you hold the shutter down for a full second to get the entire sequence of action...that is SIXTY FRAMES!! I don't even want to think about the issue if we actually had a 240fps frame rate (and an image processing chip that could keep up).

There are bounding limits of practicality here. Not only is there a limit on how LOW of a frame rate you can really have to get enough frames to capture the right moment of action, but there is also a limit on how HIGH of a frame rate is really practical. I'd say a range between 8fps to 20fps is probably it. I don't think I'd want to wade through more than 20 frames per second at the most, and 12-16 is probably really the sweet spot. I'd also venture that the higher the frame rate, the lower you would really want your sensor resolution to be. Lacl of sufficient memory card and disk space would become a huge problem with ultra high frame rates on high or big megapixel sensors. I like 18mp...it's a good number, especially in FF format. Maybe 22, but you'll start having space issues at that point. Even if ultra high stills frame rates are possible, it blows the bounds of practicality.

[Don Haines]

As far as numbers of megapixels and frame rate are concerned... the real bottleneck is not reading the sensor, it's processing the data afterwards.

For example, we pick on a 60D with a digic4 processor and an 18Mp sensor. It is capable of reading that sensor 60 times per second, but can only process 5.9 frames per second.

And I know this will be controversial, but this is one of those cases where a mirrorless design is superior.... no mirror to shake the camera or to slow down the burst. Even a gopro can read it's 11Mp sensor 240 times per second.... Every design decision has it's plusses and minuses, nothing is perfect for everyone.

You just have to question the usefulness of reading still photos 240 times per second. That is not only video, that is high speed video. Unless you have the literal intention of playing that "video" back on a 240Hz screen at its native rate, the usefulness of such a frame rate for a still camera actually eludes me. Same goes for 60fps. It is one thing to look at 12, 14, 16, maybe 20 fps for a still camera. But at 60fps, and worse at 240fps, one momentary press of the shutter will capture a ton of frames.

Assuming we had image processors that could keep up, a 1/10th second shutter press in AI Servo with a 60fps DSLR would mean you get 6 frames. SIX FRAMES for one tenth of a second. That might get you the best shot up to the moment where an Osprey flexes out its claws in preparation for grabbing a fish....but it would still miss the moment where the fish is in-talon, the moment after it is pulled out of the water. If you hold the shutter down for a full second to get the entire sequence of action...that is SIXTY FRAMES!! I don't even want to think about the issue if we actually had a 240fps frame rate (and an image processing chip that could keep up).

There are bounding limits of practicality here. Not only is there a limit on how LOW of a frame rate you can really have to get enough frames to capture the right moment of action, but there is also a limit on how HIGH of a frame rate is really practical. I'd say a range between 8fps to 20fps is probably it. I don't think I'd want to wade through more than 20 frames per second at the most, and 12-16 is probably really the sweet spot. I'd also venture that the higher the frame rate, the lower you would really want your sensor resolution to be. Lacl of sufficient memory card and disk space would become a huge problem with ultra high frame rates on high or big megapixel sensors. I like 18mp...it's a good number, especially in FF format. Maybe 22, but you'll start having space issues at that point. Even if ultra high stills frame rates are possible, it blows the bounds of practicality.

Exactly!

It's a number, but it is a useless number. Cameras are systems, and the fact that a 60D can read it's sensor 60 times per second would imply that a 1DX should be even faster.....  A single number like that is so out of context as to be misleading.... it forgets about shutter delay, processing time, storage time, and probably 100 other things that I am not aware of.

We get so obsessed about a particular component we forget about the rest of the camera

[pedro]

by jirista: "My 7D is a pretty great camera. It has a high frame rate, a good AF system that does a great job, but can bail out and get sketchy at any moment. With quality L-series glass it pounds out the sharp like you wouldn't believe. But...the 5D III, despite its SLOWER frame rate, still out-does it because of the better AF system. The larger pixels on the 5D III result in much better noise characteristics, allowing usable ISO to jump from around ISO 1600-2500 on the 7D to as high as 12800 on the 5D III, while concurrently making ISO 1600 WORLDS BETTER."

Coming from a 30D the 5D3 just blew me away. 12800 are a no-brainer. I'll even go as high as 25600 and sometimes 51200 (required in very dark rooms, exposing well to the right, final output color or b/w) I am playing with it since my cat-photograph looked pretty usable  even without NR...


Shooting my Cat at ISO 51k by Peter Hauri, on Flickr
here's a 25k lowlight shot


Z96A3292bNoNRBWcropDEFKLEIN by Peter Hauri, on Flickr

[dilbert]

For the 5DC, 7D and rebel-crowd, I'm sure the 5D3 is a great upgrade.

...and for the 5DII crowd who shoot things that move. 

You're telling me that using a 5DIII with the same shutter speed and aperture as a 5DII will result in better photos of "action"?? I'd like to know how you work that magic. Do things automatically go slower when they sense a 5DIII is taking a photo of them?

Thanks, now I understand you.  Your dislike of the 5DIII (or maybe Canon, in general) has completely eliminated any objectivity you may have had.  This assertion is even more ludicrous that others you've made.   If you honestly believe that the better AF system of the 5DIII cannot yield better images of moving subjects than the 5DII, then I feel comfortable dismissing your arguments on the subject as biased to the point of irrelevance.

Pity that you didn't understand what I wrote.

What makes a sharp photo?
Shutter speed + lens + focus

If the speed of the object is such that the shutter speed isn't enough to freeze the object then it don't matter how good the focus is, you're still left with a blurry image.

Pity that you don't understand what you wrote. 

What makes a sharp photo?
Shutter speed + lens + focus

Since the 5DII and 5DIII can obviously be set to the same shutter speed with the same lens, the difference is the AF system.  "Focus" - your word. So...what you wrote suggests that you think the 5DIII's AF is not better than the 5DII's.

I didn't say that the 5DIII's AF wasn't better than the 5DII's. However the way you and everyone else talks about the AF, you could be forgiven for thinking that AF was the only factor in obtaining good action shots.

There is a certain range of subject motion for which the newer AF does make an improvement but it isn't complete.

I suppose I shouldn't be surprised...you also argued that the Canon DIGISUPER 75 broadcast TV lens was a camera, and you were way off base there, too (despite the fact that you refused to admit your mistake).

Well given that it was the only marking that could be seen on the entire device was on a lens that was not easily discernible as being a lens, one could easily be forgiven for mistaking the entire thing to being a camera.

[dilbert]

...
I like the Osprey example....

I wish to take a picture of an Osprey picking a fish out of the river.... this happens pretty fast, about 2 or 3 tenths of a second. This is faster than my reaction time so I have to anticipate the action and hope that my picture is taken at the right time... and with the speed things are moving I only get one shot.

Beside me is another photographer with a 1Dx set to take a burst at 10 frames per second. The other photographer starts thier burst before the Osprey hits the water and it continues until the fish is well clear. My picture is of the claws going into the water.. my timing is off. The other photographer has a string of twenty or so pictures from the Osprey approaching the water, through the grab, and as the catch clears the water. I have a great shot. The other photographer has twenty great shots, one or two of which show the intended image.

That is an indisputable reason for the use of bursts and how they can improve composition.

You might as well take a high resolution video camera...

And indeed that is what various people used to do to get good action shots of animals: they used film to capture lots of frames to get the one they wanted and blow that up large.

Sure, it may be the only way to "get the shot" but from a photography perspective it's kind of like cheating because there's no skill required on the photographer's part in knowing when to snap or compose.

[LetTheRightLensIn]

A high frame rate can correlate with sharpness, when a subject is periodically moving and you want a frame where it is not. In the case of bird photography specifically, birds have periods of fast action, interleaved with moments of stillness. You need at least 6fps, better 8-10fps, such that in any burst of frames you get one where the subject is perfectly still. Even at high shutter speeds like 1/1250th or 1/1600th and a perfectly stable camera, the fast motion of a bird can still cause blurring. The 5D III is certainly better in this regard than its predecessor. Obviously there are even better options, but that does not invalidate the fact that the 5D III is better than its predecessor, by 54%.

Very true and this also applies to trying to get away with natural light macro shots of bugs.

Sure, if you were a professional sports shooter. But there are also wildlife shooters. Bird photographers. Arial photographers (some who use the 1D line, some who were using the 7D line but have been all too happy to move to the 5D III). Sports is not the only source of high speed action. I chose my terms carefully. The 5D III, as a general purpose camera, is ideal in an extremely broad set of circumstances. The 1D X is obviously the better choice for the highest speed action, but the 5D III does the job extremely well well when money is an issue.

It is and the 6fps a big key. Some people shoot everything from landscapes to wildlife to action and it can be nice to have one camera that can manage all of that at least decently.

Again, your missing the point. The notion that IQ is purely dependent upon the sensor is a fallacy. Getting a quality image means getting the little aspects of each key thing within the image correct. And that means all of the little aspects correct...not just exposure, not just sharpness, but composition, subject position, pose, head angle, and eye contact. More frames per second, larger frame buffer, better AF system, etc. all lead to better IQ. The sensor is certainly the most important factor in getting each and every individual pixel that is recorded perfect, but if you don't record the right thing...well, it doesn't matter how good your pixels are. The wrong frame is the wrong frame. Bad focus is bad focus. The best pixels in the world won't give you enough post-process editing latitude to fix those issues.

IQ is dependent upon multiple factors. Sensor is only one of many. As I said previously (not sure which thread, maybe this one), if one were to rank the most important factors in IQ, I'd say the best rankings would be: Sensor is #3. Frame rate is #2. AF system is #1! Sensor only matters if the other two factors do their job, *lock focus* and capture the *right* frame.

It depends what you are shooting, if it is landscapes then sensor is far and away #1. The 5D3 sensor was a bit of a let down after 3.5 years. The rest of the camera is fantastic but the sensor is bit outdated feeling, basically even if the 5D4 arrives in 2 years you are talking 5.5 years with the same sensor for many types of shooting although somewhat better for high iso shooting in some cases and a good 7.5 since they'd done anything to improve low ISO in any fashion at all.

It's a very good camera though. (It could have been earth-shattering for both stills and video if they were less conservative and hadn't gone into king of the hill hahaha mode.)

[jrista]

I didn't say that the 5DIII's AF wasn't better than the 5DII's. However the way you and everyone else talks about the AF, you could be forgiven for thinking that AF was the only factor in obtaining good action shots.

There is a certain range of subject motion for which the newer AF does make an improvement but it isn't complete.

It is actually not everyone. It is really actually a smaller group than those who are complaining about the 5D III sensor. A MUCH smaller group. Which is why we are so vocal. The notion that sensor is the most important factor, the sole factor, in achieving IQ is fundamentally flawed. It is a fallacy. No, the 5D III does not have the best sensor on planet earth. However, despite that, the 5D III is still immensely capable of taking really freaking awesome photos with excellent IQ in an extremely broad set of circumstances for a very broad set of photographic types.

The D800, the D3200, hell even the D5200 may all have sensors that kick the crap out of the 5D III's sensor at low ISO, but their range of usefulness is more limited. The D3200 and D5200 are lower end, and crop, so their usefulness is intrinsically limited. The D800 is excellent for some types of photography, but it is way overpowered or way too much for a lot of other types of photography. For the narrower band of photography types the D800 is great for, it can produce better images IF you use ISO 100-400, however in pretty much any other case, the 5D III should be just as capable if not more capable of getting a better, sharper shot with the right composition and ideal subject pose/posture/head angle/eye contact/etc.

[jrista]

A high frame rate can correlate with sharpness, when a subject is periodically moving and you want a frame where it is not. In the case of bird photography specifically, birds have periods of fast action, interleaved with moments of stillness. You need at least 6fps, better 8-10fps, such that in any burst of frames you get one where the subject is perfectly still. Even at high shutter speeds like 1/1250th or 1/1600th and a perfectly stable camera, the fast motion of a bird can still cause blurring. The 5D III is certainly better in this regard than its predecessor. Obviously there are even better options, but that does not invalidate the fact that the 5D III is better than its predecessor, by 54%.

Very true and this also applies to trying to get away with natural light macro shots of bugs.

Ah, good point. Macro is a great example of a more niche use of higher frame rate.

Sure, if you were a professional sports shooter. But there are also wildlife shooters. Bird photographers. Arial photographers (some who use the 1D line, some who were using the 7D line but have been all too happy to move to the 5D III). Sports is not the only source of high speed action. I chose my terms carefully. The 5D III, as a general purpose camera, is ideal in an extremely broad set of circumstances. The 1D X is obviously the better choice for the highest speed action, but the 5D III does the job extremely well well when money is an issue.

It is and the 6fps a big key. Some people shoot everything from landscapes to wildlife to action and it can be nice to have one camera that can manage all of that at least decently.

Exactly. The higher frame rate and better AF system really make the 5D III an ideal general purpose FF camera with extremely broad applicability.

Again, your missing the point. The notion that IQ is purely dependent upon the sensor is a fallacy. Getting a quality image means getting the little aspects of each key thing within the image correct. And that means all of the little aspects correct...not just exposure, not just sharpness, but composition, subject position, pose, head angle, and eye contact. More frames per second, larger frame buffer, better AF system, etc. all lead to better IQ. The sensor is certainly the most important factor in getting each and every individual pixel that is recorded perfect, but if you don't record the right thing...well, it doesn't matter how good your pixels are. The wrong frame is the wrong frame. Bad focus is bad focus. The best pixels in the world won't give you enough post-process editing latitude to fix those issues.

IQ is dependent upon multiple factors. Sensor is only one of many. As I said previously (not sure which thread, maybe this one), if one were to rank the most important factors in IQ, I'd say the best rankings would be: Sensor is #3. Frame rate is #2. AF system is #1! Sensor only matters if the other two factors do their job, *lock focus* and capture the *right* frame.

It depends what you are shooting, if it is landscapes then sensor is far and away #1.

True, there are some kinds of photography where sensor is #1. I'd say at least landscape and studio photography, although a lot of studio photography can benefit from a highly capable AF system and even frame rate as well. I guess those who shoot very high contrast scenes that might actually have a strong argument for needing 14 stops of DR (which again includes landscape photography, but may also include things like street photography) could say sensor is #1 as well.

[LetTheRightLensIn]

I didn't say that the 5DIII's AF wasn't better than the 5DII's. However the way you and everyone else talks about the AF, you could be forgiven for thinking that AF was the only factor in obtaining good action shots.

There is a certain range of subject motion for which the newer AF does make an improvement but it isn't complete.

It is actually not everyone. It is really actually a smaller group than those who are complaining about the 5D III sensor. A MUCH smaller group. Which is why we are so vocal. The notion that sensor is the most important factor, the sole factor, in achieving IQ is fundamentally flawed. It is a fallacy. No, the 5D III does not have the best sensor on planet earth. However, despite that, the 5D III is still immensely capable of taking really freaking awesome photos with excellent IQ in an extremely broad set of circumstances for a very broad set of photographic types.

The D800, the D3200, hell even the D5200 may all have sensors that kick the crap out of the 5D III's sensor at low ISO, but their range of usefulness is more limited. The D3200 and D5200 are lower end, and crop, so their usefulness is intrinsically limited. The D800 is excellent for some types of photography, but it is way overpowered or way too much for a lot of other types of photography. For the narrower band of photography types the D800 is great for, it can produce better images IF you use ISO 100-400, however in pretty much any other case, the 5D III should be just as capable if not more capable of getting a better, sharper shot with the right composition and ideal subject pose/posture/head angle/eye contact/etc.

It all depends. It is also worthwhile to point out that many a Canon user used to make it all about the sensor back in the days when the Canon sensors ruled but the Nikon bodies were always more fully featured and some of the same now suddenly say that the sensor is the most minor point  .

The D800 is pretty versatile too in that it does have 5 and 6fps crop modes that coupled with the high sensor density one might say it may actually be better for wildlife. You do need the grip to get to 6fps though. For something like indoor basketball without strobes it would be worse for sure though since you'd either have to crop it and lose SNR to get 6fps or get good SNR but be stuck at 4fps.

But yeah the 5D3 is pretty well rounded and it doesn't have the aliasing/moire the D800 video suffers from even if the 5D3 video is kinda weak for micro-contrast.

[jrista]

Sure, it may be the only way to "get the shot" but from a photography perspective it's kind of like cheating because there's no skill required on the photographer's part in knowing when to snap or compose.

That is so unbelievably naive I don't even know where to start!!! Go read Art Moriss' blog, or any one of the worlds most renown bird photographers. MOST will tell you they are not good at BIF (Birds in Flight), because it takes a tremendous amount of detail-oriented SKILL. Good BIF photography requires dedicated practice to get right, as there are so many factors to consider, all of which must be continually addressed in real time!

It is not simply a matter of point, frame, and shoot. Achieving and maintaining correct exposure, particularly of birds, while they are in flight against a constantly changing background, is one of the most challenging things in photography. You've just shown your excessive naivete with the claim you've made above. Snap and compose my ***. You just insulted the entire community of bird photographers, particularly the experts at Bird in Flight photography.

[LetTheRightLensIn]

Sure, it may be the only way to "get the shot" but from a photography perspective it's kind of like cheating because there's no skill required on the photographer's part in knowing when to snap or compose.

That is so unbelievably naive I don't even know where to start!!! Go read Art Moriss' blog, or any one of the worlds most renown bird photographers. MOST will tell you they are not good at BIF (Birds in Flight), because it takes a tremendous amount of detail-oriented SKILL. Good BIF photography requires dedicated practice to get right, as there are so many factors to consider, all of which must be continually addressed in real time!

It is not simply a matter of point, frame, and shoot. Achieving and maintaining correct exposure, particularly of birds, while they are in flight against a constantly changing background, is one of the most challenging things in photography. You've just shown your excessive naivete with the claim you've made above. Snap and compose my ***. You just insulted the entire community of bird photographers, particularly the experts at Bird in Flight photography.

Same for sports too. Even with tons of fps you still need a lot of skill. A newbie with 30fps will get much worse sports shots than a talented sports shooter with 4fps.

[jrista]

I didn't say that the 5DIII's AF wasn't better than the 5DII's. However the way you and everyone else talks about the AF, you could be forgiven for thinking that AF was the only factor in obtaining good action shots.

There is a certain range of subject motion for which the newer AF does make an improvement but it isn't complete.

It is actually not everyone. It is really actually a smaller group than those who are complaining about the 5D III sensor. A MUCH smaller group. Which is why we are so vocal. The notion that sensor is the most important factor, the sole factor, in achieving IQ is fundamentally flawed. It is a fallacy. No, the 5D III does not have the best sensor on planet earth. However, despite that, the 5D III is still immensely capable of taking really freaking awesome photos with excellent IQ in an extremely broad set of circumstances for a very broad set of photographic types.

The D800, the D3200, hell even the D5200 may all have sensors that kick the crap out of the 5D III's sensor at low ISO, but their range of usefulness is more limited. The D3200 and D5200 are lower end, and crop, so their usefulness is intrinsically limited. The D800 is excellent for some types of photography, but it is way overpowered or way too much for a lot of other types of photography. For the narrower band of photography types the D800 is great for, it can produce better images IF you use ISO 100-400, however in pretty much any other case, the 5D III should be just as capable if not more capable of getting a better, sharper shot with the right composition and ideal subject pose/posture/head angle/eye contact/etc.

It all depends. It is also worthwhile to point out that many a Canon user used to make it all about the sensor back in the days when the Canon sensors ruled but the Nikon bodies were always more fully featured and some of the same now suddenly say that the sensor is the most minor point  .

Ah, touché, touché. Great point. I was a newly minted toddler of a photographer back when the 5D II was first released, but I remember the elation of the Canon community at that little gem. Heh, seems so odd, thinking back...that camera WAS the D800 of its time, and everyone raved about it. So funny how times change, but things really stay the same...it's apparently always about the sensor.

[jrista]

Sure, it may be the only way to "get the shot" but from a photography perspective it's kind of like cheating because there's no skill required on the photographer's part in knowing when to snap or compose.

That is so unbelievably naive I don't even know where to start!!! Go read Art Moriss' blog, or any one of the worlds most renown bird photographers. MOST will tell you they are not good at BIF (Birds in Flight), because it takes a tremendous amount of detail-oriented SKILL. Good BIF photography requires dedicated practice to get right, as there are so many factors to consider, all of which must be continually addressed in real time!

It is not simply a matter of point, frame, and shoot. Achieving and maintaining correct exposure, particularly of birds, while they are in flight against a constantly changing background, is one of the most challenging things in photography. You've just shown your excessive naivete with the claim you've made above. Snap and compose my ***. You just insulted the entire community of bird photographers, particularly the experts at Bird in Flight photography.

Same for sports too. Even with tons of fps you still need a lot of skill. A newbie with 30fps will get much worse sports shots than a talented sports shooter with 4fps.

And the talented sports shooter with 4fps will miss the "perfect moment" far more often than the talented photographer with 12fps.

[Don Haines]

...
I like the Osprey example....

I wish to take a picture of an Osprey picking a fish out of the river.... this happens pretty fast, about 2 or 3 tenths of a second. This is faster than my reaction time so I have to anticipate the action and hope that my picture is taken at the right time... and with the speed things are moving I only get one shot.

Beside me is another photographer with a 1Dx set to take a burst at 10 frames per second. The other photographer starts thier burst before the Osprey hits the water and it continues until the fish is well clear. My picture is of the claws going into the water.. my timing is off. The other photographer has a string of twenty or so pictures from the Osprey approaching the water, through the grab, and as the catch clears the water. I have a great shot. The other photographer has twenty great shots, one or two of which show the intended image.

That is an indisputable reason for the use of bursts and how they can improve composition.

You might as well take a high resolution video camera...

And indeed that is what various people used to do to get good action shots of animals: they used film to capture lots of frames to get the one they wanted and blow that up large.

Sure, it may be the only way to "get the shot" but from a photography perspective it's kind of like cheating because there's no skill required on the photographer's part in knowing when to snap or compose.

I sincerely hope that you take all your photos in manual mode using manual focus and don't bother to look at exposure displays or anything else, because that would be like cheating.

An exposure meter is a tool, developed to help you expose better.
Autofocus is a tool, designed to help you focus better.
Lightroom is a tool, designed to help you modify images.
Shooting RAW is a tool, designed to give you better data for editing.
display screens are a tool, designed to help you understand what your camera is set to and doing.
I could keep going here all night......
Burst mode is a tool, to help you in times where action happens too fast for you to react.

These are all tools to enhance your abilities and creativity. Only a fool ignores good tools.