slides vs prints dilemma

[Scubatooth]

-hh

i would have to disagree on the 600 mp of pro film(s) as even the highest quality scanners are only 8000 DPI (or about 61.4 mpx 6400x9600 scan)and those are pushing the limits the 61.4 mpx is about the max even for medium format.

and the megapixel to quality ratio that alot of people talk about is completely wrong because i have 24x30 prints from a 4 mp cam that look better the film prints(and the same could be said the other way around), in the final analyse its partially the camera but mostly the photographer


also there are recent test of prints from Canon 1d, 1ds, nikon D1h & x (note that these arent consumer cameras the are Pro level cameras and range from 3-5 mp except for the 1ds which is 11mp)compared with some 35 mm bodies (both C & N) and medium format bodies and some of the prints were able to be tell that they were digital and then there were some film prints that the judges thought were digital were and were 35mm or medium so it evened out (BTW the judges were editors for National Geographic so you know they can tell good from bad)

FWIW

 

[segal3]

-hh once bubbled...

The absolute resolution potential of 35mm (best everything) has been estimated at ~600 MP. Being more realistic with merely going to "Pro" class equipment, this drops, but to no less than 100MP.

I very much doubt this. I have a friend in the photography industry whose employer owns a 48MP CCD backing that attaches to a lens setup. It costs $125k and came with a Dual 1gz G4 PowerMac, that it directly dumps the image to. The friend says the image is indistinguishable from regular film (this is larger film, bigger than 35mm) - I'll ask him tonight for exact specs.

~Matt Segal

 

[James connell]

segal3 once bubbled...
It costs $125k and came with a Dual 1gz G4 PowerMac, that it directly dumps the image to. The friend says the image is indistinguishable from regular film

Anybody who'll spend that much for that setup HAS to convince Themselves it's as good as film!!! that doesn't make it true however.

 

[segal3]

James connell once bubbled...


Anybody who'll spend that much for that setup HAS to convince Themselves it's as good as film!!! that doesn't make it true however.

He didn't spend on it. Like I said, his employer did. I would assume his employer also has an idea of what he's doing, simply because of the money he makes to afford such things. They use it to shoot photos for billboards, adverts. on sides of buses, etc.

Edit - my friend just gave me a little more info on it. The size of the image file (RAW) that it dumps is 236mb. Depth index of 0.11 . The CCD is made by Leaf, and the camera front is by Haskuban.

~Matt Segal

 

[James connell]

TheAvatar once bubbled...
I am a film snob, but 600MP seems like a very high number for 35mm.

So this equation would call for a 30.86MP Foveon style sensor or a 92.57MP bayer pattern sensor.

depends on the film!
i can say that when foveon technolgy gets to ~10MP and full frame 35MM, i'll give up on film. till then film has it over digital all the way.

 

[TheAvatar]

James connell once bubbled...


depends on the film!
i can say that when foveon technolgy gets to ~10MP and full frame 35MM, i'll give up on film. till then film has it over digital all the way.

I was just throwin numbers and doin ideal film/lens/exposure. Just a guess. I am waiting for 16MP foveon... although for the price... the 300D is tempting as a snapshot camera when I'm not going B&W, slide, or 4x5.

 

[-hh]

Scubatooth once bubbled...
-hh

i would have to disagree on the 600 mp of pro film(s) as even the highest quality scanners are only 8000 DPI (or about 61.4 mpx 6400x9600 scan)and those are pushing the limits the 61.4 mpx is about the max even for medium format.

I have a correction to make: I got the units incorrect on the numbers: its not 600MP, but 600lpmm (lines per millimeter). Ditto 100MP -> 100lpmm.

In theory you can go (24mm*600lpmm)*(35mm*600lpmm) to start to estimate its MP equivalancy. If you run this math, you get ~296.6MP before you add any color depth bits, and this is where it starts to run into problems: since the original test consisted of a series of black/white lines (max contrast), the resolvability of lesser contrasts (colors) is untested and thus, unknown. There's probably some good questions on film grain too.

What I personally think is a pragmatically more effective number is the 100lpmm value, times 24 bit depth (3 channels @ 8 bits each), and this works out to ~200MP equivalent.

...the megapixel to quality ratio that alot of people talk about is completely wrong because i have 24x30 prints from a 4 mp cam that look better the film prints(and the same could be said the other way around), in the final analyse its partially the camera but mostly the photographer

There's also the eye of the viewer to consider too: I have an 8x10 print hanging in our dining room that I hate. It was taken with ISO 400 film over 15 years ago and its "incredibly grainy", which is why I hate it. But the reason its hanging up is because my wife doesn't think it has any grain at all and she absolutely loves the image...from the day that I had it made, she's insisted that we display it. (Argh!).

For example, this pic: http://tinyurl.com/lzm5 has "way too much" grain as far as I'm concerned, but it might not bother other people.



-hh

 

[-hh]

James connell once bubbled...


Anybody who'll spend that much for that setup HAS to convince Themselves it's as good as film!!! that doesn't make it true however.

Sometimes its merely the right tool for the job. At work, I picked up a Vision Research Phantom 7 digital camera system last year, which is a sweet specialty camera...its nickname is "The King of Speed": take a peek: http://www.visiblesolutions.com/phantomv7.html


-hh

 

[Scubatooth]

-hh once bubbled...



I have a correction to make: I got the units incorrect on the numbers: its not 600MP, but 600lpmm (lines per millimeter). Ditto 100MP -> 100lpmm.

In theory you can go (24mm*600lpmm)*(35mm*600lpmm) to start to estimate its MP equivalancy. If you run this math, you get ~296.6MP before you add any color depth bits, and this is where it starts to run into problems: since the original test consisted of a series of black/white lines (max contrast), the resolvability of lesser contrasts (colors) is untested and thus, unknown. There's probably some good questions on film grain too.

What I personally think is a pragmatically more effective number is the 100lpmm value, times 24 bit depth (3 channels @ 8 bits each), and this works out to ~200MP equivalent.




-hh

i dont think even the best pro lenses resolve that high and i believe the lens test charts dont even go that high. and i thought that most lenses (Pro) only went up to maybe 100-150 lpmm

 

[-hh]

Scubatooth once bubbled...


i dont think even the best pro lenses resolve that high and i believe the lens test charts dont even go that high. and i thought that most lenses (Pro) only went up to maybe 100-150 lpmm

You're right - the best Pro stuff typically is 100-150 lpmm.

However, the "600 lpmm" is real: it's a rounding off of testing done on the 100mm f2 Kinoptik, which is the lens that holds the record for highest measured resolution delivered by any 35mm lens.

In the test, Perkin-Elmer got up to 622 lines per millimeter before giving up. Which means its max resoution potential is actually somewhere above 622 lpmm, but by how much, we don't know.


-hh