iiiNelson
Some solid in depth point there! Thanx for your contribution, learned a few things here!
About 8k coverage in terms of mtf. I always find this to be a cumbersome topic, in the following topic I already went a bit in depth with the issues of resolution, and why resolution doesnt seem like it is.
https://gafpaforum.com/d/3-kinefinity-mavo-edge-8k-footage-vs-4k-over-sampled-footage/2
-
-
-
But lens wise, i think it’s interesting to state, that you can already see a huge difference in optical resolution on a Arri Alexa classic, even though this camera has quite a thick olpf and is 2k after its sublime oversample algorithm, you can still see which lenses are somewhat sharp and which ones are not. To me sharpness is a join definition for contrast and resolving power, and contrast can be divided in macro and micro. For instance a Signature prime has great macro contrast, low micro contrast (gentle on the skintones), but yet high resolving power, quite a unique combination. What i explained in the other topic is that 8k is more or less a capturing resolution which is preferably much higher than master resolution (nyquist), due to sampling at high resolution we can leave very high frequencies of the lens more untouched but with the same aliasing breaking point as on a 4k capture camera. We can do this by inserting a lighter less dense olpf in the 8k camera, and a stronger one on the 4k camera. the aliasing will happen on the same level, yet the 8k capture camera will be able to resolve more lines before it will aliase as the 4k camera with the denser olpf. Yet 8k is just sampling and in order to display this 8k in its honest resolution we have to diminish it with the Nyquist theorem and also bayer theorem. We will then perceive it at the perfect resolution with no digital artefacts like spatial issues or color issues. 8k is not sharper like some people believe, 8k is smoother. Because when you use a camera pixel to pixel at tits sample capturing resolution, you will get lots of edge issues and enhancement issues due to extrapolating debayer algorithms and sample resolutions past its nyquist limits. capturing at 4X the intended master resolution will by far lead to a cleaner and better presentation of the lens capabilities, and in this case 8k to 4k is still way past what our retinas can see at a normal viewing distance. To me vintage lenses look better on a 8k sampling camera because its harder for a camera to display smoothness (OOF out of focus), and other subtile lens effects, than plain crisp contrast sharpness (mtf). In that light leica lenses designed for mirrorless film cameras without digital sensors , will look better on a 8k camera with light olpf, than on a 4k camera with dense camera, due to the inclining rays which are projected under a steep angle, and will be less refracted in the 8k sample camera, due to lighter olpf. So long story short 8k sampling doesn’t lead to sharper images, but better images, more smoother images. a 1k camera can look very sharp, and can be perceived enormously sharp. And some people think about resolutions as details, but its instead all about smoothness. A sharp lens like a master prime will look sharp on a standard definition camera and on a 8k camera. a 9600 0 audio recording mastered at 441100 will sound better than a 44100 audio recording mastered at 44100. etc. Our eyes are much more reactive to smoothness than details, I have a nice library about how our eyes and brain transform visual data, and what stands out, is that once you start to ramp up resolution is that the first thing that bottlenecks is detail perceiving, the last thing that bottlenecks is perceiving smoothness, like bokeh, and fall off curves from sharp to OOF.
SO to state that a carl zeiss contax lens is less ideal for a 8k camera, and to get to its potential is not a good way of looking at it (in my humble opinion). If you don’t like contax zeiss, you probably dont like it on 1k, not on 2k not on 3k etc. you just don’t like the characteristics of the lens, but if you DO like the characteristics of the lens you probably love to see the lens digitally sampled at the highest resolution possible, and oversampled according to the oversample rules. (nyquist and bayer).