Canon 'Rep' tells BBC that "fewer megapixels are better"

[craigkg]

I'm awaiting the release of a 400 mp camera so I can sell off all my telephoto lenses.

I'd like a 400MP camera...if it were a 4x5 or 5x7 sensor. As someone that likes to photograph landscapes, digital large format would be the holy grail.

400MP on a APS-C or even full frame 35mm would be so overkill it isn't even funny. Your lens couldn't resolve anywhere near that resolution and you'd be sacrificing dynamic range and noise performance because of the amount of photons per pixel the sensor can capture.

 

[DJL329]

I'm awaiting the release of a 400 mp camera so I can sell off all my telephoto lenses.

400MP on a APS-C or even full frame 35mm would be so overkill it isn't even funny.

It would be so "noisy" that you'd need to wear earplugs!

 

[alipaulphotography]

I'm not sure that I understand the comment about crop sensor cameras being for amateurs only. There are tons of benefits to having them. They are typically faster (great for sports) & they have a much higher pixel density (great for cropping). Both of those are reasons enough to be a number one choice in many situations.

Don't get me wrong, I want a full frame, but even with one I wouldn't always use it. If I were taking pictures at a game I would definitely choose the crop sensor, and if I were shooting wildlife I would probably go crop sensor also. What's the big advantage with having a full frame? More picture? You can get just as much picture with a crop sensor - just take a couple steps back. Now you have a higher resolution image of basically the same thing. Also, don't full frame cameras have issues with the edge of the picture sometimes? They are also typical worse at handling noise right?

Either way, I'm curious to hear the reasons why FF cameras are pro cameras and everything else is for amateurs. By that logic the 1D is basically a Rebel.

Full frame cameras can be prone to vignetting and unsharp corners in poor quality lenses but they make the most a camera can out of an EF lens. Why waste glass and only use the middle of it when you can use it all? The focal length change gives an apparent shallower depth of field as you can then get closer to your subject also. I would hate if my 50mm wasn't actually 50mm...

Another massive reason for full frame is its hands down superior noise handling. No crop sensor compares and noise reduction technologies don't count as they affect the sharpness of your image.

They both have their uses.

 

[awinphoto]

I'm not sure that I understand the comment about crop sensor cameras being for amateurs only. There are tons of benefits to having them. They are typically faster (great for sports) & they have a much higher pixel density (great for cropping). Both of those are reasons enough to be a number one choice in many situations.

Don't get me wrong, I want a full frame, but even with one I wouldn't always use it. If I were taking pictures at a game I would definitely choose the crop sensor, and if I were shooting wildlife I would probably go crop sensor also. What's the big advantage with having a full frame? More picture? You can get just as much picture with a crop sensor - just take a couple steps back. Now you have a higher resolution image of basically the same thing. Also, don't full frame cameras have issues with the edge of the picture sometimes? They are also typical worse at handling noise right?

Either way, I'm curious to hear the reasons why FF cameras are pro cameras and everything else is for amateurs. By that logic the 1D is basically a Rebel.

Full frame cameras can be prone to vignetting and unsharp corners in poor quality lenses but they make the most a camera can out of an EF lens. Why waste glass and only use the middle of it when you can use it all? The focal length change gives an apparent shallower depth of field as you can then get closer to your subject also. I would hate if my 50mm wasn't actually 50mm...

Another massive reason for full frame is its hands down superior noise handling. No crop sensor compares and noise reduction technologies don't count as they affect the sharpness of your image.

They both have their uses.

There are a lot of pro's for either system, however people need to realize there are professionals who shoot with what they have including crop sensors, full frame, film, and beyond... There are situations where a 7D can shine over a 5d... It's using the best gear for the situation and conditions and knowing how to use the gear to it's fullest potential that makes one a pro (and get paid for doing so)...

 

[Hillsilly]

I would hate if my 50mm wasn't actually 50mm...

A 50mm on an APS-C sensored camera is still a 50mm. You need to free your mind from such restrictive thought paradigms!

 

[UncleFester]

Seriously, those who think they "need" 21 or 24MP would have/should have moved to FF by now. All DSLRs should have FF sensors, maximize the lens mount FFS. Leave the crop sensor to the lowest-end Rebel and the so-called "large sensor compacts" .

I think I know what you mean here. My experience this past summer with both a cropped and ff was that the cropped sensor, IQ-wise, had no advantage over the FF cropped. And I certainly couldn't use a converter with the 7D,but the 5DII worked amazingly well with a 2x. And still able to crop at %100.

So, either 1.6 needs to get better, or ff needs to get faster. Until then, I need to use 2 cams, or buy one really expensive one.

 

[Woody]

My experience this past summer with both a cropped and ff was that the cropped sensor, IQ-wise, had no advantage over the FF cropped. And I certainly couldn't use a converter with the 7D,but the 5DII worked amazingly well with a 2x. And still able to crop at %100.

Then, you have not done a proper apple to apple comparison. Quote from:
http://www.juzaphoto.com/eng/articles/canon_7d_review.htm

"Comparing a lens without TC on 7D vs lens with 1.4x TC on FF, the 7D gives slightly better image quality - more detail and less chromatic aberration."

 

[NotABunny]

What's the big advantage with having a full frame? More picture? They are also typical worse at handling noise right?

Where did you here that they are worse at handling noise?

The sole, absolute, and single most important technical aspect of a photo camera is the sensor SIZE. A larger sensor size means that more light is captured for the SAME exposure (obvisouly, it's the entire light capturing system, meaning, the lens diameter must scale). Tough to understand (the relationships between all the relative and absolute factors involved), but it's physics.

This extra light actually reduces the noise in a photo (when scaled to the same given physical size). Of course, it all depends on the used technology, but this is how the current technology scales.

You can get just as much picture with a crop sensor - just take a couple steps back. Now you have a higher resolution image of basically the same thing.

Not really, you have an image with a slightly lower technical quality. (Though it's unlikely to see the difference between a FF and a APSC at low ISO.)

Also, don't full frame cameras have issues with the edge of the picture sometimes?

No. Lenses have that. Besides, after I apply lens correction in post-processing, I add some vignette.

 

[NotABunny]

Then, you have not done a proper apple to apple comparison. Quote from: http://www.juzaphoto.com/eng/articles/canon_7d_review.htm

The problem is that he's comparing pixels, not images; he compares photos coming from sensors of different physical sizes and different resolutions. The apples that he's comparing are made of wax... and consequently of no importance (to humans).

I have however used some of his photos to prove to myself that indeed, the noise levels for 1D4 and D3s are the same per UNIT AREA (square millimeter, for example), not pixel (nor image for that matter).

 

[NotABunny]

Regarding why sensor size matters most, I am separating some details in another post:

This is why astronomy is being done with ginormous telescopes: more light per image / "exposure".

See http://www.galileotelescope.com/choosing_a_telescope.htm

A telescope's aperture relates directly to the two vital aspects of the scope's performance: its light-gathering power (which determines how bright objects viewed in the scope will appear), and its maximum resolving power (how much fine detail it can reveal).

The truth is, any telescope can be made to provide almost any magnification, depending on what eyepiece is used. The factor that limits the highest power that can be used effectively on a given scope is, you may have guessed, its aperture. As magnification is increased, and the image in the scope grows larger, the light gathered by the telescope is spread over a larger area, so the image is dimmed.

Or http://www.howstuffworks.com/telescope1.htm

A telescope's ability to collect light is directly related to the diameter of the lens or mirror -- the aperture -- that is used to gather light. Generally, the larger the aperture, the more light the telescope collects and brings to focus, and the brighter the final image.

Then there is the experience of this guy http://www.waid-observatory.com/article-what_can_I_see.html

I had an eight-inch SCT (Schmidt Cassegrain Telescope) that had very good optics. I then purchased a big 12-inch SCT and thought the heavens would suddenly "open up". Quite frankly, I was disappointed when I looked through the "Big Scope". The difference was there, but it was not like night and day.

What he didn't realize is that the difference was only of sqr( 12 / 8 ), which is just a bit more than 1 stop of light.

 

[oosh]

Though I welcome more MP so I can make larger, more detailed prints I strongly disagree that we need to cram so many extra pixels onto a sensor with each model. We all know the main issue that need to be working on is DYNAMIC RANGE. Give us more. They have the technology to do it, but instead they insist on pumping out more megapixels for each and very model. This is doubly worse with crop sensors, I have a 7D, which I think is a great cam, but man, 18mp is as far as you'd want to go for such a small sensor size. If they slowed down the MP increase on bodies and concentrated on higher DR and better noise handling they'll certainly make a superior product - IQ will be increased because the pixel density isn't increased, and when you think about it, you'll be able to upscale images much bigger because of this. Sooner or later the consumer level market is going to wake up to the MP marketing hype, and will be sick of DR being the same as the previous body and noise only being slightly improved.

I can assume that the main reason people would want and actually use more pixels is for example landscape images or images that are likely to be printed extra large - generally these genres don't require fast frame rate, so if you really, really want more pixels save for medium format - you'll get 2 or 3 more stops of dynamic range.

Better yet, spend less than $2000 on a 4x5 field cam and an epson v700 and get hugely detailed negatives with more resolution than you can poke a stick at - and save your cash!

 

[NormanBates]

I have however used some of his photos to prove to myself that indeed, the noise levels for 1D4 and D3s are the same per UNIT AREA (square millimeter, for example), not pixel (nor image for that matter).

I support that idea too, but I don't know if your results are ok: DXOmark does the same thing, and measured much lower noise in the D3s (nearly one stop better):
http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/629%7C0/(brand)/Canon/(appareil2)/628%7C0/(brand2)/Nikon

 

[UncleFester]

My experience this past summer with both a cropped and ff was that the cropped sensor, IQ-wise, had no advantage over the FF cropped. And I certainly couldn't use a converter with the 7D,but the 5DII worked amazingly well with a 2x. And still able to crop at %100.

Then, you have not done a proper apple to apple comparison. Quote from:
http://www.juzaphoto.com/eng/articles/canon_7d_review.htm

"Comparing a lens without TC on 7D vs lens with 1.4x TC on FF, the 7D gives slightly better image quality - more detail and less chromatic aberration."

I don't photograph book cases or stuffed gorillas and I don't own a Sigma lens. So, that sort of "proper apple to apple comparison" wouldn't help me anyway.

 

[NotABunny]

I have however used some of his photos to prove to myself that indeed, the noise levels for 1D4 and D3s are the same per UNIT AREA (square millimeter, for example), not pixel (nor image for that matter).

I support that idea too, but I don't know if your results are ok: DXOmark does the same thing, and measured much lower noise in the D3s (nearly one stop better):
http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/629%7C0/(brand)/Canon/(appareil2)/628%7C0/(brand2)/Nikon

The only thing I can read on the DXO website is "bla X bla bla, Y bla bla Z bla, X * Y bla, Y* Z blablabla..." Seriously, if they show me the photos to interpret the results with my own eyes, it's good, otherwise it's meaningless. Formulas can get you anything you want. In the end, the conclusion must make a difference for a human.

Anyway, it's likely they compare noise per image (= entire sensor area), so the D3s would have to have 0.76 stops of advantage.

By the way, where have you seen that nearly 1 stop?