Discussion in 'Film and Processing' started by alexandergambino, Oct 30, 2017.
That camera shown by ben is not unlike what is claimed to be the invention of the digital beast:
Steven Sasson and the beast itself (Credit: Steve Kelly/Kodak)
Something that would actually warrant those 2 inch wide camera straps people use to use.
Likely enough, it is monochrome.
Aliasing in current systems wouldn't be so bad, except that it interacts with the Bayer array to generate colors that shouldn't be there.
Small color changes show up much more than a little aliasing.
I have Nikon AN6 straps (2" wide) on all my cameras, in case I want to shoot with a motor drive or my Nikkor 25-50 AIS lens. Also, when I'm shooting with 2 cameras, I want as much rough surface area on the underside of the strap as possible to keep the camera from slipping off my shoulder.
I know I've posted this before, but I think it perfectly makes the point how un-analogue B&W film really is.
It shows a small section of Ilford Delta 3200 compared with an equal area of digital capture. Same subject, same lens and EI/ISO rating used.
BTW - the reason I didn't give the source I had for the picture above of Steven Sasson was that I got it from a Israeli newspaper and the partially Hebrew language URL wouldn't play well with other sites. Here, for the curious, is the Wiki article on him:
Steven Sasson - Wikipedia
3200? ya could have picked a better film to compare.
For both digital and film, you need to low-pass filter when viewing.
Appropriate low-pass filter will smooth out the grain, not completely but much more than shown.
You also have to watch out for grain aliasing when scanning film.
But yes, we know that Delta 3200 isn't a fine grain film.
"3200? ya could have picked a better film to compare"
- The point is that all silver-based commercial B&W film looks like that if you magnify it enough. Any semblance of continuous tone is created by dithered opaque/clear areas, and therefore is effectively digital in nature.
Another point is that both images above were shot at EI 3200, and the same film/sensor area is shown.
"For both digital and film, you need to low-pass filter when viewing."
Errr, no! I don't usually put an LPF over my eyes when looking at photographic prints.
If you mean that some distance is needed when viewing prints or monitor images - true, but what counts is the amount of distance required to give the illusion of smooth tone. I don't think anyone could argue that 35mm film at 3200 ISO, or even at 400 ISO, looks smoother than its digitally-shot equivalent.
(snip, I wrote)
>> "For both digital and film, you need to low-pass filter when viewing."
> Errr, no! I don't usually put an LPF over my eyes when looking at photographic prints.
> If you mean that some distance is needed when viewing prints or monitor images - true, but what counts is the amount of distance required to give the illusion of smooth tone. I don't think anyone could argue that 35mm film at 3200 ISO, or even at 400 ISO, looks smoother than its digitally-shot equivalent.
From sampling theory, you can sample above the Nyquist frequency an appropriately bandwidth limited source, and reconstruct it perfectly, with appropriately bandwidth limited reconstruction filters.
And yes, in many cases, digital and AgBr, the low-pass filter is the limits of the optical system, including our eyes.
We live in a quantum universe, where many physical quantities can only have a finite number of values.
We consider the incoming light as analog, though it is quantized into photons. The CCD array outputs an integer number of electrons to the A/D converter.
Analog magnetic (audio and video) tape magnetizes ferric oxide grains either one way or the other, on the average with a magnetic field approximating the desired value. (The finite grain structure causes noise at higher frequencies, which is normally filtered out. It is also the cause for all the noise reduction systems in use for analog recording.)
Analog vinyl disks are made of atoms, which only approximate the desired continuous signal.
All these analog systems store quantized values. Unlike digital systems, though, the allowed values are continuously distributed, or at least as close to continuously as the universe allows.
My all-time favorite quantization problem, is how many different velocities a baseball pitcher and pitch at in a baseball stadium.
It is a function of the maximum speed, size of the stadium, and mass of the ball. (and is a very large, but not infinite, number)
iinteresting, no MTF curves for Delta 3200. They don't want you to know!
Kodak gives the TMZ curve out to about 140 cycles/mm
Separate names with a comma.