Digital equal to Film? Lets Hear your opinion!

Started Jun 30, 2008 | Discussions thread
GordonBGood
Veteran MemberPosts: 6,276
Like?
Film has more Dynamic Range than you think...
In reply to Daniel Lee Taylor, Jul 2, 2008

Although it isn't really the subject of the Gary Mercer's OP, which was more concerned with resolution, as follows:

Daniel, I happen to agree with the point you made on this post to which I am replying, but there are quite a few misconceptions on what the Dynamic Range (DR) of film really is from the links that you have made in two of your posts to this thread and also spt_gb's related post which I have included below.

Daniel Lee Taylor wrote:

http://www.fujifilmusa.com/shared/bin/AF3-0221E2Velvia50PIB.pdf

At best Velvia 50 can hold 6 stops, and that's stretching it. Many of
the best landscape photographs in the world were made on, and
continue to be made on, this film which has less DR than even an old
DSLR. None of the pros using it care. They know its exposure
characteristics and they make the best exposures for the scene and
medium.

from spt_gb's related post:

spt_gb wrote:

How about from Fuji's own data sheet? From fully saturated to the
point that colour response diverges is only about 1.75 log H, or 5.8
stops.

I used to look at the above linked Fuji Velvia 50 data sheet and the characteristic curve and come to the same generally accepted conclusion: that Velvia 50 is only usable to about 6 stops of DR. Then, I received the following Velvia 50 target for such uses as calibrating scanners as scanned without calibration on a fairly high end scanner to a gamma type JPEG and here reduced in size to be readily viewed in this post:

Note that this target is calibrated to cover a DR range in steps from dMax to dMin and that the darkest steps are for the steps labelled 17 onwards to stop 22 represent luminances that are about 4.9, 5.5, 6.2, 6.8, 7.4, 8.6, and for dMax 10.6 stops below dMin, respectively. If you download this image and bring it into an editor such as Photoshop, you will see that there are still distinct differences in level all the way to step 22 in spite of the JPEG format and the severe downsampling and compression used. If you use the said editor to boost brightness and contrast so as to see these differences clearly, you will see that there is a slight colour shift toward red as the tones get darker, but one has to apply some extreme processing to begin to see this.

When one does some numerical analysis, one sees why this is so slight in "real world" terms and why we can actually use about 8 or 9 stops of DR from Velvia. Notice from the characteristic curve for Velvia 50 from your link that red "poops out" at about log density 3.0 below dMin where as the best green doesn't "poop out" until log density about 3.5 below dMin. In real numbers, that means that red doesn't start to saturate until the red is 1/1000 or 0.001 of its bright level and green doesn't saturate until about 1/3000 or 0.00033. Note that there still is red response below this level, just that the response starts to diverge. If one actually uses the full available range, they could pull out a usable eight or nine stops of DR from Velvia, noting that the slight colour divergance at the dark end could be calibrated out of a scan (which is the purpose of targets such as this one).

So slide film has a usable 8 to 9 stops of DR, which we agree to be quite acceptable for most uses, and is approximately the extremely satisfactor DR range of most good digital cameras, especially DSLR's.

Now, on to looking at colour negative film:

From another Daniel Lee Taylor post:

http://www.clarkvision.com/imagedetail/dynamicrange2/

Not all neg films have a wide dynamic range. I wonder if you even
shoot the ones that do.

This is an often quoted reference showing the limited DR of colour negative film. Recently, I have had reason to question the charts from the above link and Ed Leys, Jay Turberville, and I (all contributors to this forum) actually obtained Kodak Gold 200 film and used Jay's Stouffer step wedge setup to test the DR of this film, scanning with Ed's Minolta Elite 5400 scanner with careful technique. Although we haven't completed our analysis to the point where we are ready to publish to the Internet, suffice it to say that we don't know what technique or scanner software Roger Clark used but we in now way see this colour negative film llmited to the about 6 stop his charts show (and also much more than the about 6.5 stops of DR he shows for Velvia 50 as per the above analysis). We get a DR for this negative film of at least 10 or 11 stops, and can see with further work that some colour negative films can likely be pushed to capture up to 12 or 13 stops of DR, although with decreasing image qualtity as to the relative strength of the grain appearance.

In summary, the results of our analysis will likely show that average slide film has a usable DR of about 8 stops plus or minus one and average colour negative films have a usable range if pushed of about 12 stops plus or minus one. Further, from these tests it is my observation that although film has a theoretical higher resolution as to support larger prints, the magnitude of the film grain means that these film prints reduced to about the largest size supported by digital cameras will produce roughly the same image quality as the pure digital image.

Thus, colour negative film likely still has a DR advantage compared to digital whose useable "pushed" range with the best technique and best cameras is probably limited to something a little over 10 stops.

Regards, GordonBGood

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow