Thank you Dan. I can only imagine how MF (4x5 or larger) transparency images look like in print! Please post your samples, once processed.
Interestingly, I shot few rolls of Fuji Velvia side by side with digital just to be able to compare them. Although word 'sterile' come to mind when viewing digital images, an old paraphrased conclusion from comparing audio media (vinyl vs. CDs) comes to mind; film is like sex where digital is pornography. (I believe it was Michael Framer of The Absolute Sound)
When it comes to colors, digital seems a bit castrated. There are some digital cameras that produce really good colors, but they are very small minority.
On other hand with film, you can put it into almost anything, use any crappy lens and at least colors will look good.
Comparsion of CD vs vinyl is pretty accurate. And like with film, you dont need world best equipment to enjoy it (tho it helps quite a bit .
I always wondered how would hyper-contrasty lens like Voigt 40/2 work on film. It seems that rather nicely actually. Wonder if all Voigts are so good on film..
Czech guy which shot strictly with Hasselblad and Velvia 50. These days hes doing some astrophotography with modded 5DMK2. I think colors in most of his film pics are pretty good.. He was sorta inspirational for me.
Mescalamba wrote: When it comes to colors, digital seems a bit castrated
Well said! There is something about that noise, grains, saturation, etc. By the same token movie going experience (some producers/directors still refuse to shoot in any format but film) has some embodied and unexplained viewing pleasure.
I would say that "issues" with colors comes from digital making everything from:
1) interpolated colors (exception is Foveon)
2) only RGB (certain films had 8 layers of colors, not sure how much current Velvia does have or how exactly it works, maybe less, maybe different, but still seems to be good)
3) colors are saturated different way, even very saturated films seem to have sorta "natural" saturation, over-saturated digital looks like exactly what it is - over-saturated
When printed (especially if its old way printing, I think its called contact printing in English?) difference is much bigger than when viewed on display.
Grain is another matter, in most films it doesnt matter, its just sorta part of it. Noise in most digitals looks just like something that shouldnt be there at all (expections are, but about as rare as those with good colors).
I think digital could reach something similar, if colors were non-interpolated and made from RGBCMY. Which is currently most likely impossible, so I would say maybe 80 mpix digital back with special CFA, could do that trick. I would probably try to merge 6 color pixels into one, to prevent interpolation. Which would result in 13 mpix image, most likely crazy sharp and with good colors.
Tho someone would really need to be pretty crazy to do that.. Wonder if it would really work. Most likely it would hit color limit with most monitors and printers anyway. Thats another issues too..
http://www.horolezec.cz/gallery.htm
Czech guy which shot strictly with Hasselblad and Velvia 50.
Wow, impressive set of images, thanks. I can only imagine how much resources it took to get there, how much energy it took to get up there and finally, $$$ support must be fairly substantial even plan and to get there!
Being an engineer I have always wanted an explanation for the subtle.
A photographer friend (also an engineer) has described one of the film qualities as "progression of color".
Where digital is taken by the camera as rows & columns of color info.
When you scan film, you really don't want to scan at the particle level
But I wonder if you scan at a dpi that is multiples of the particle size (say 4/8/16-more particles per), that you get a pseudo-random color transition quality from the scanner's digitization. This would give the more pleasing color transitions along with the detail. Optical printing is another matter and would be better yet.
These are all just musings, so don't be too harsh on me for searching for an explanation
Mescalamba wrote:3) colors are saturated different way, even very saturated films seem to have sorta "natural" saturation, over-saturated digital looks like exactly what it is - over-saturated
That's the unexplained beauty of film!
When resizing/cropping my film scans I had to compare them to digital takes from the same day in disbelieve that film needed nothing in PP (I am sure developing/scanning services have their own addition to the final scan, however)
With scans you can do all sorts of tweakes, but developing itself is simply in doing it right and when its right theres no need to do more with it. When film is perfectly exposed and developed, its pretty much finished. All other things like color saturation, contrast curve, grain is given by emulsion so no other work is needed.
Compared to digital, its fairly easy, but if something goes wrong in development, then its really bad. In digital, you just reset to defaults and do it again.
But yea in lots of ways, film is easier, cause when you do everything right, it will be right in results. I dont find much fun in shooting digital picture in certain way and painfully dig those data and shape file to look how I wanted it to look at first place. Not mentioning it rarely looks like I saw it and wanted it to look. Im most of the time pretty close but its almost never "it".
I should stop writing, get myself decent film camera and shoot while there is still film left.
dswiger wrote:
Being an engineer I have always wanted an explanation for the subtle.
A photographer friend (also an engineer) has described one of the film qualities as "progression of color".
Where digital is taken by the camera as rows & columns of color info.
When you scan film, you really don't want to scan at the particle level
But I wonder if you scan at a dpi that is multiples of the particle size (say 4/8/16-more particles per), that you get a pseudo-random color transition quality from the scanner's digitization. This would give the more pleasing color transitions along with the detail. Optical printing is another matter and would be better yet.
These are all just musings, so don't be too harsh on me for searching for an explanation
Mescalamba wrote:
That would need to be insanely high resolution. But, if it was downsized right at HW level, it actually might do what you said.. Interesting idea.
Yea the scalar part was a WAG. Could be from 10 to 1000 times what I said.
Found this document on film particle size. Need to digest this vs scanning pixes?? http://cool.conservation-us.org/coolaic/sg/emg/library/pdf/vitale/2007-04-vitale-filmgrain_resolution.pdf
Recently picked up an RB67, and this week I finally had the chance to test it out properly on a shoot. FP4+ home developed, scanned with a Canon 9000f
I'm mainly from the digital generation, on film, I only played a bit with cheap 35mm rangefinders before. First time shooting medium format, and first time doing an entire shoot on film. It was a bit strange, not being able to chimp and look at fancy histograms. Kinda liked it though, certainly gonna do this more often
.
