Possible 7D mkII Spec.

Started 6 months ago | Discussion thread
40d_dane
Senior MemberPosts: 1,004
Like?
Re: Possible 7D mkII Spec.
In reply to hjulenissen, 5 months ago

hjulenissen wrote:

40d_dane wrote:

The returns have already diminished going from the 40D to the 7D. As far as I'm concerned there really isn't much reason to go beyond 18MP. Since you also mention 36 MP and 200 MP, I have to assume you really don't know what you are talking about.

It will be easier to take your posts seriously if you avoid such assumptions about me in the future.

Well, sorry, suggesting that a 200MP APS-C sensor would make much sense doesn't really add any credibility to your posts.

I object to the claim that a 25 MP camera "requires" more sharpening than an 18 MP camera. It does if you evaluate the end-result on a per-pixel basis, but I reject that criteria as a relevant criteria for comparing the quality of two cameras with different pixel counts. I printed my 350D and 7D images at equal size, and the 7D image did not need "more sharpening" to look pleasing, on the contrary.

"Oversampling" should (in itself) be seen as a good thing: even if the sensor "outresolves" the lense or the camera stand, the high pixel density means that moire, color pattern/bayer artifacts, etc are that much easier to avoid. Sampling/recreating a smooth signal (something that is blurry relative to the sensel pitch) is that much easier, well supported by Nyquist theory, etc.

Well, you forget to mention that oversampling is called oversampling because you sample your signal *more* than required to reconstruct the original signal accurately. If you are truly oversampling all you do is adding data that you really don't need.

Well, audio ADC/DACs that employ some form of oversampling are the most accurate ADC/DACs around. So while in a theoretical world, you might call it "adding data that you really dont need", in practice, it tends to be needed for optimal results. Of course, not everyone needs optimal results (and the use of oversampling can lead to other undesireable trade-offs).

And how many times oversampling does audio ADC/DAC's employ ? A lot more than a factor of two. And don't venture too far down the audio path. A lot of audiophile stuff is based on superstition rather than science.

The core objective of oversampling is to move the filtering operation prescribed by Nyquist from one domain where it may be hard/expensive to do it right, to a domain where it is easier. Making good optical lowpass filters is hard. By moving the aliasing frequency of the sensor into a range where the lense++ can be assumed to have very low contrast, you know that there will be no aliasing (since there is no signal to alias).

If you are doing oversampling, all you are doing is trying to make a better signal and/or lower the cost. How can that be a bad thing for a product?

Your haven't thought this well through.

The question then only is if an increase in pixel density leads to visibly worse noise, DR, etc as seen on a fixed size display or print, or lower framerate, worse battery life etc. Is the D800 worse than the 5Dmk3 in those respects?

The D800 has a better sensor than the 5DmkIII below ISO 800. Above the reverse is true. And if you work out the pixel density of the D800 it's about 14 MP on (Canon) APS-C. So the 7D sensor has a more dense sensor than the D800. To really benefit from the full D800 resolution you have to put your camera on a tripod. How many actions shooters do that ? Some Nikon shooters have returned a D800 in favor of a D4 in disappointment over the IQ of the D800 in practical shooting.

Don't forget this debate is about the next incarnation of an action shooters camera. Not a landscape or what ever else camera.

The D800 edge at ISO 100 is worth something. The 5Dmk3 edge at higher ISOs seems to be small enough so as to really only affect measurements,

I don't care about image sensor performance at ISO 100. I never shoot below ISO 400 because I don't have the benefit of sufficient light for doing so. I care much more about DR and high ISO performance.

Having to use a stand to benefit from that resolutoin is of course nonsense. It is a matter of exposure time, stabilization and light (natural or artificial). If you are shooting a moving bird at dusk handheld, the benefit will of course be a lot smaller than in more ideal conditions.

The benefit will be non-existent - most likely it will be a disadvantage.

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow