Film vs. Digital

[David Luttmann]

WG,

Unfortunately, extrapolating doesn't work. Because of differences in sharpness and film flatness between 35mm & MF gear, as well as scanning tolerances, I have found that 22MP backs exceed what is achievable with 6x7 MF gear. And don't get me wrong....I own an RB67. I compared a while ago scans shot on Astia F vs a 22Mp back from Imacon. While it is possible that the film holds more resolution.....it's not usable as at magnifications large enough to see it, you become overwhelmed in grain. The human eye responds quite well to accutance with regards to perceived detail & sharpness. You are able to increase accutance from a digital source through USM to a much greater extent than film because of the fact that digital sources lack noise. ...which would be increased far more using USM from a film source.

That is why you'll find a 48MP betterlight back easily equals 4x5 film. It is also why on a 16x24 print, an 8MP source exceeds 35mm film when many people throw around numbers like 10,12, even 25MP. Give it a try yourself. Print out some 8x10 crops from a 16x20 film & digital source. I've found whenever I do this for people, they are really surprised how well a digital source can do.

Of course, if all you're doing is 16x20, you'll do well with any of 4x5, MF 6x7, or a 12MP digital source.

Best regards,

[Emre Yildirim]

Roger Clark has a page dedicated to this topic here:

http://www.clarkvision.com/imagedetail/film.vs.digital.summary1.html

According to his findings:

35mm Provia 100 - 7MP (aparent image quality = 7MP)

35mm Velvia 50 - 16MP (apparent image quality = 14MB)

8MP DSLR at ISO 100 - 8MP (aparent Image quality = 38MP)

35mm vs. 8mp DSLR, Digital wins here.

6x4.5 Velvia 50 - 50MP (aparent image quality 45MP)

6x7 Velvia 50 - 78MP (aparent image quality 70MP)

Theoretical 39MP digital medium format - 39MP (aparent image quality = 185MP)

Film MF vs. Digital MF, Digital wins here again.

4x5 Velvia 50 - 240MP (apparent image quality = 220MP)

8x10 Velvia 50 - 960MP (aparent image quality = 860MP)

4x5 film vs. 39MP Digital MF, Film wins.

In the end, I'll probably stick to film for resolution. I have a 20D that I use for snapshots, but I never actually made prints from it. I have a friend who makes 8x10 prints from his 5MP point & shoot camera, and they look very decent.

[Emre Yildirim]

Oops, looks like I don't know how to spell "apparent". Sorry I just woke up

[Ron Marshall]

Here is a link to a page on Norman Koren's site which lists the effective resolutions of most of the current DSLRs (scroll to the bottom of the page).

http://www.normankoren.com/Tutorials/MTF7.html

[David Luttmann]

Emre,

I can tell you right away that it takes far less than 78MP to equal 6x7 Velvia, as well as far less than 240MP to equal 4x5. He is basing these figures on theory rather than real world testing.....which unfortunately is what many people do.

I was really surprised when I found 4x5 exceed by a multi shot 22MP back.....which gave the equivalent of an 88MP image. The 4x5 was left far behind in resolution, noise, and color accuracy......but at $23,000......I'm not rushing out to buy one!

[Bill_1856]

I told you so.

[Hans Berkhout]

"Adding some digital in the future". Why?? That's the question you need to answer first. Only consider digital if it offers you something you need, and you can't get it with your current equipment- whatever that is.

[Matthew Cromer]

Some of you guys need your eyes checked. . .

[Emre Yildirim]

Dave,

I'm still a little skeptical about this. I've seen shots from a 16.7MP 1Ds MkII and compared them to 4800dpi scans from my Epson 4990. The 4x5 scans were far suprior to the digital capture. Maybe the extra 6MP makes a difference with the 22MP back, but as you said, I don't think it would be worth the money.

Another [cheaper] alternative is 8x10. I doubt any digital back can match a 4000 dpi scan of an 8x10 Astia slide.

[Dan_4341]

Here's what my (very limited in scope) tests comparing film to digital showed me:

1. Compared a 13x19 print made at Costco from Fuji Reala 35mm shot with a Rollei AFM35 P&S camera (with an excellent lens for a P&S), to same size print made from a digital file shot with a Nikon D50 6mp RAW (converted for printing to TIFF) using 18-70mm Nikon lens, camera set at ASA 200 (the lowest available if I remember correctly). The film shot showed a fair amount of grain. The digital showed none. Costco did a lousy job of color on the film shot - color balance way off. The digital was great. However, there was a marked lack of detail and lack of texture to the digital shot. Nice and smooth but pathetic lack of detail in comparison to the Fuji Reala shot.

2. Compared a 35mm Provia F slide taken from an Olympus stylus 35mm P&S camera to a Canon SD500 7mp digital. View slide with 20x magnifier, and compared the detail observed to 100% and 200% (for easier viewing as a double check) on a 17" monitor for the digital file. The slide had a tremendous amount of more detail than the digital file. Of course, it had film grain as well, while the digital had a very small amount of noise.

3. Printed some digital files I found at various camera review sites and manufacturer's sites on the internet - large files made from a Canon 5d, a 22mp back (forgot which ones), etc. Printed at 13x19" at Costco. Again, no grain, very little noise, sometimes some digital artifacts (maybe jpeg detail shadows or whatever they are called, sometimes banding in large areas of sky). Very nice all of them, but in comparison to what I am used to seeing from 6x7 and 4x5, .. no cigar for digital. Yes my nose is pressed against the print !

4. Had a 6x9 color neg scanned commercially at 3000 dpi from a high end Nikon scanner, and printed to about 13x19" at Costco (on photo type paper - not inket; the same as my other comparison shots). Compared it to an analog print made some years before at a pro lab. No comparison - the analog print blew away the one made from the scanned file. Much more detail on the analog print, and smoother, and less apparent grain. I suspect that I would need an expensive drum scan for more detail and more expert operators dealing with something that doesn't make the film grain more pronounced than the analog print.

In my opinion (which reflects my priorites, tastes, possible shortcomings in observations, etc.), while a 13x19" print from a good 6mp SLR shows no grain/no or little noise and therefore presents a 'smoother' picture than a 35mm shot taken with a very good color film, it has much less detail and texture. So for me, a 13x19 is much too big a print to be made from a 6mp camera.
At some point in increasing the megapixels - maybe 8, maybe 10, I would say digital would equal 35mm film. For portraits, most people probably prefer smooth skin tones - so no film grain and lack of detail would favor a 6mp DSLR. For anything else I think film would be favored. For me, my guess is that I wouldn't consider digital to equal 35mm film unless the digital had at least 10 mp.

A 4x5 pro, Jack Dykinga, has written at least a couple of articles on the subject, and my interpretaion of what he has said is that it might take about a 50mp camera to equal the quality from a 4x5 RVP transparency. Since a 4x5 trannie has about 13 times the area of a 35mm trannie, than puts the 35mm/digital equivalent to about 4mp. Apparently Jack is allowing for the fact that 4x5 film is not held as flat as 35mm, and the 4x5 lenses aren't as sharp, etc.

It will be awhile before that type of digital becomes affordable. In addition, some LF folks , like me, are 'moving up' to larger size camera, like 5x7 and 8x10, even for color. Not very practical, but any view camera, even 4x5, is not very 'practical' or maybe 'convenient' is a better word. I dearly wish high end digital was more refined and affordable, because it's so darn convenient. Some day it will be. For many, it's already good enough and affordable enough. But, not me :-)