When you think of it, Ice Age - The Meltdown could be a pretty good metaphor.Originally Posted by amilne
When you think of it, Ice Age - The Meltdown could be a pretty good metaphor.
I remind you this is not my thesis. This is a quote.
Yes, a quote. Sorry about that.
Bruce Watson
The person I would think of with this, would be Edward Burtynsky. He prints from an extremely good enlarger, and has done some (LightJet - I think) digital intermediate prints. Still RA-4 prints in both cases, yet telling the difference between prints at an exhibit is practically impossible.
There might be a misconception that someone could see dots in an RA-4 print done with an LED or Laser based printer. Even with a loupe, there are no apparant dots. This is still continuous tone. Compare that to inkjet, which at best attempts to mimic continuous tone prints.
Anyway, back to the scanners. The print size, original film size, and scanning choice, all conspire to create limits. In smaller prints, no limits might be reached, at least in resolution comparison. Larger prints are where resolution limits might be more important, but only if the level of detail in an image, and the ability to show detail in the print, are not also limited.
A better indicator of limits in scanning, especially comparing low to mid range with high end, is DMax. The ability to resolve shadows details has a loose relationship to the price of the scanner. While consumer scanners claim really high numbers, most are hitting closer to 3.0 to 3.1 DMax. If you shoot night scenes, or high contrast scenes with lots of shadows, you might not capture those details in the scan; at least not without blowing out the highlights entirely. If you only shoot evenly lit scenes, then you might wonder what all the fuss is about.
Transparency films can need near 3.6 to 3.9 DMax from a scanner (actual, not marketing claims) in order to capture shadow details. That does not mean every transparency would need that capability, though you might find in practice that at some point, and with some images, the colour information is what you will be missing . . . and not the resolution.
Ciao!
Gordon Moat
A G Studio
There are two ways to judge a photographic print.
Objectively and subjectively.
Subjectively there can be no definitive answer because it boils down to what people like and therefore I see no reason to debate it.
Objectively, I have seen no reproduceabe data (and I have looked a lot) that any digital process can get as much information to paper as a contact print and even most (if not all) enlarging processes.
The flaw in the data that is presented is that they use a scanner as there measuring instrument so are really measuring the scanner and not the photographic process.
To compair digital and film you must have either a good microscope are be willing to spring for the cost of a drum scan as I have.
In my experiements a 12,000 dpi drum scanner has about 150% of the resoulation needed to test fine grain film.
Try 2.4 to 2.6.
I haven't seen a consumer scanner (or even an Imacon, for that matter) that is better than 2.65.
---Michael
Could you explain how you made this determination?
I am attaching a small .jpeg image of a Stouffer step wedge, scanned with an Epson 4990. The Dmax of Step 21 is log 3.07. It certainly appears to me that this scanner is capable of capturing density up to at least 3.07. I figure you must have made these tests before yourself so I am trying to figure out why your conclusion is that consumer scanners can not capture density over 2.6?
The step wedge was scanned with SilverFast, with the histogram input and output values set so no clipping would take place in either the highlights or shadows. No other modification was made of the scan other than to reduce it for posting.
Sandy King
Lasersoft Imaging have a link on their website to tests done independently of their multi-exposure feature. If you look through those tests, and see the scans without multi-exposure, then you find near what I posted as results. This is a feature that only works on a few low to mid level scanners, and unfortunately greatly increases scan times.
Canon 8600F - 2.89
Epson 4990 - 3.11
Epson V700 - 3.10
Nikon LS5000 - 3.53
There have been a few earlier flatbed scanners that hit closer to 3.4 DMax, though without much resolution capability. It seems that getting a true 3.6 DMax or higher is expensive, even when buying refurbished (unless you get lucky).
Ciao!
Gordon Moat
A G Studio
Last edited by Gordon Moat; 19-Jul-2007 at 14:24. Reason: clarity, information, grammar
Sandy,
Interesting results. Using a Stouffer Step Wedge we have never come close to 3.07 but I have asked Michael to take a look at your post.
Gordon, bothof your posts are hitting the nail on the head. You aren't going to get the DMax or the resolution with a really cheap scanner. IMO, you aren't going to get there with anything approaching total satisfaction with any of the consumer scanners either. Now is an enlarger better? Maybe better than a cheap scanner but not if it is a cheap enlarger ... and, as I have mentioned before, if you are spending or have spent the money on a top-of-the-line enlarger with the best optics then you are well over the cost of the consumer scanners and are not going to outperform them by all that much ... but you will outperform them in some respects. YMMV but my experience has been that the high end scanners outperform the best enlarger/optics setup when you start to go to 4x and larger.
BTW for the high end scanners we are talking DMax of at least 3.6 for some of the older less capable ones and ~4.0 or better for most.
Ok, Gordon, but you have to be clear that you are doing a multi-exposure trick to possibly achieve that result. This is no different that HDR shots with a digital camera.
Unfortunately, the multi-exposure approach will cause the sharpness of most of these scanners to be negatively impacted (sometimes severely), so this is not a win-win kind of change.
Plus, considering the terribly poor contrast that the scanners experience near their limit (and considering the high noise also experienced) I am somewhat doubtful that there is meaningful information bering pulled out of the film at that high density.
Not that the approach doesn't improve the scans, but I am skeptical that the improvements aren't due to better data processing and in particular, noise reduction, rather than meaningful information gathering.
---Michael
Posting Permissions
Reply With Quote
Bookmarks