Comparision of 4 Films in Diafine

Had a bit to time free this afternoon so I decided to look at grain and sharpness of 4 different films in Diafine. This was a very, very basic test but the results are very interesting. The images and stuff are on my web site...a direct link is here:

http://figitalrevolution.com/2008/04/15/sharpness-and-graindiafine-processed-films-compared/

I'd love to hear your thoughts here or there. I will add to the mix over the next few days Plux-X in Diafine.

Cheers-
Stephen Schaub

Stephen

Personally, I'd find the results considerably more interesting if there had been no noise reduction applied at all. The addition of a software process (which is likely not consistently applied across each film - and even if it is applied with the same "levels" it will have different effects on different grain structure) which intereferes with grain, actuance and sharpness simply makes a meaningful sharpness and grain comparison impossible. Why not simply show the native scans with levels adjusted - surely that would be far more interesting?

Not really (IMHO)...I am more interested in final results as it is reflected in my workflow and as such each film was "optomized" for as good as it could be...the balance between lost of grain/ noise and sharpness. The variables of type of scanner, type of mounting and personal workflow makes a "straight" comparrision just as individual in nature as the results I posted.


Best-
Stephen

Figital Revolution said:

Not really (IMHO)...I am more interested in final results as it is reflected in my workflow and as such each film was "optomized" for as good as it could be...the balance between lost of grain/ noise and sharpness. The variables of type of scanner, type of mounting and personal workflow makes a "straight" comparrision just as individual in nature as the results I posted.


Best-
Stephen

Click to expand...

OK... While I understand that your entire "workflow" is what you're interested in, it does make an isolated comparison of these films a lot less valuable - for example, my own tests done on a drum scanner using your development suggestions and IE have produced very different results from your published examples, up until the application of Noise Ninja (where screen examples frankly all look pretty similar after the application of NN in varying doses).

I thought the results of the 3 films I tested according to your suggestions in Diafine were way inferior to the results I had produced with the same films using different approaches and chemistry - however, after applying NN, they all look pretty much the same. While I use noise reduction software in my workflow, it is generally close to final output and very dependant on final output size, material etc. I think it would be far more relevant to show the film combinations without it. I understand that you have a different scanner, scanning method etc compared to others, but why not remove an easily "removable varibale" when making these comparisons?

In short, I think there's a huge amount of value to a lot of folks in doing this work and publishing it, but I do think that it's value is undermined by introducing "hidden" variables which are unneccesary.

I understand your point (I really do) but as I indicated the noise reduction in my post is present I don't see your point of a "hidden" variable as all variables are on the table. I also feel that the scanner and scanning software are a HUGE variable, much larger than NN (esp when you consider the range of scanners made today). I also feel that using NN with a careful approach there really is a big difference in grain from film to film...using the default NN is not my technique as it kills (IMHO) the look of film. I hope that my posts on FR get others, as you have indicated you have, to try these "solutions" for themselves (which is always best) and see if it works for them and their needs. I never personally trust anything I see on my monitor from the web site as Gospel- but it does give one an idea and something to perhaps consider and try. Great conversation!

Best-
Stephen Schaub

Figital Revolution said:

I understand your point (I really do) but as I indicated the noise reduction in my post is present I don't see your point of a "hidden" variable as all variables are on the table. I also feel that the scanner and scanning software are a HUGE variable, much larger than NN (esp when you consider the range of scanners made today). I also feel that using NN with a careful approach there really is a big difference in grain from film to film...using the default NN is not my technique as it kills (IMHO) the look of film. I hope that my posts on FR get others, as you have indicated you have, to try these "solutions" for themselves (which is always best) and see if it works for them and their needs. I never personally trust anything I see on my monitor from the web site as Gospel- but it does give one an idea and something to perhaps consider and try. Great conversation!

Best-
Stephen Schaub

Click to expand...

To the extent possible we should eliminate, IMO, as many variables as possible in comparing grain and sharpness in films. I agree that scanners, and even scanning technique, are big variables in the perception of sharpness and grain. But why add another variable, i.e. software that impacts grain, to the equation.?

For what it is worth, I agree with Don (and by extension Marcus Aurelius) in that you should take Noise Ninja, or any grain reduction software, out of the equation. Then we only have the issues of film, developer and scanner to consider. That is complicated enough in itself.

Consider it a matter of first principles. If you add fourth principles over first, second and third principles the result may be chaos in understanding. The point that you make then becomes relevant only for your specific work flow.

Sandy King

I understand both of your points regarding the use noise ninja but my tests were for end results...that is to say what is possible from each combo in the best case using my workflow as described in an earlier post on FR. I really could not care what a negative looks like just scanned as I never end my process with just scanning. If either of you would like to run a test based on your suggest testing procedures than that is great and would fit your testing needs but my tests as I have indicated had a different end goal than what you both seem to be looking for.

Best-
Stephen Schaub

Ok, maybe on this Stephen, I'm trying to be a believer, but how about a scene which has a little more challenging contrast than a coin collection? It's had to say what's really going on in the real world shadows (or highlights for that matter). Can we see the unadjusted histograms from the scans? For B&W I usually lock my scanner in at 1.0 or 1.1 exposure using view Vuescan which pretty much tells me the tale of expansion and compression. Also, are not lenses going to play a part in the final equation if say a shooter tends toward using a lot of wide angles outdoors which tend to yield larger SBRs?

mkochsch said:

Also, are not lenses going to play a part in the final equation if say a shooter tends toward using a lot of wide angles outdoors which tend to yield larger SBRs?

Click to expand...

Michael,

I understand what you mean about wide angle lenses, but the lens does not determine the SBR of a particular scene.

Don Bryant

Yeah Don I could have said that better of left out that last line. ~m

Here is a quick link to an article I posted a few days ago on FR...shows Acros in Diafine with sample pics! I am now about to test Plus X...more on that in a few day.

http://figitalrevolution.com/2008/04/10/fuji-acros-100-and-diafine-developer-part-2/

Cheers-
Stephen