Thread: Film scanners?
View Single Post
  #172  
Old April 23rd 17, 05:32 AM posted to rec.photo.digital
Ron C
external usenet poster
 
Posts: 415
Default Film scanners?

On 4/22/2017 11:38 PM, Savageduck wrote:
On 2017-04-23 02:43:30 +0000, Bill W said:

On Sat, 22 Apr 2017 08:26:21 -0700, Savageduck
wrote:

On 2017-04-22 08:21:29 +0000, Eric Stevens
said:

I wouldn't argue with any of that but this guy is saying that it must
be done in the camera or not at all.

Agreed.
That is a silly position to take, particularly since most shooters are
looking for more than SOOC Acros simulations. I am.

My point is that if the processor in the camera can do it then a
processor outside the camera can do it
equally well.

One would think that, and I have been trying to do that ever since the
Fuji PR machine hyped the Acros in-camera simulation with the release
of the X-Pro2, with the X-Trans III sensor and the new X-Processor Pro.
My X-E2 does not have said processor, so I was only able to work with
the in-camera Acros simulation once I got my X-T2. Up until then I got
some pretty good results using Exposure X2, Tonality Pro, and NIK
Silver Efex Pro.

The Camera profiles in LR are camera specific, so Acros was not
available for the X-E2 in LR. It is for the X-T2, so I have only been
able to make that comparison over the last 10 days. The best of the
third party emulations has been Exposure X2. However, when compared
with the X-T2 SOOC rendering, there is a palpable difference which I
have not been able to match, and even though I say so myself, I am not
totally incapable when it comes to working with the software available
to me. While the differences are subtle, I have yet to manage a tweak
in any software to match the X-T2 SOOC in-camera Acros.

I am sure that nospam, and perhaps even you could show me just how to
do it, but I don't see you buying an X-series camera anytime soon, and
I don't anticipate nospam showing us any results.

All it needs is the right programming.

...and that might well be where the difference lies since Fujifilm is
using a proprietary processor and proprietary algorithm. It is one of
their films after all.


Yes, but... I'm not arguing about this, but does the in camera
emulation give you the results you like more than the others, or is it
really more accurate? If you believe it's more accurate, what is there
left to base that on? I have to assume you are not doing side by side
comparisons, and are working from memory. Even if you have more or
less recent film prints, they would be samples, and not the final word
on what those film prints all looked like.


Actually I can make side-by-side comparisons for the various digital
images. It has nothing to do with accuracy, and this for me is only with
regard to the in-camera Acros simulation. As far as comparing with
prints from actual Fujifilm Neopan 100 Acros film goes, I have never
used it. Back in B&W film days I was a Tri-X shooter.

When it comes to making the digital comparison there is a quality and
character to the SOOC image in tone and grain (and you can set three
levels of grain which is reactive to the camera's exposure settings) a
quality which is not present in the third party simulations. As I have
said, I can get pretty close with Exposure X2, but it is not quite there.

I can understand that logic and common sense says that if Acros can be
simulated by one processor, it should be a simple matter to replicate
that with third party software on an external computer. I own and use
most of the software capable of the task, and quiite simply I have not
been able to match the job done in-camera.

I would eventually like somebody in this group, other than the usual
opinionators who don't use the Fujifilm cameras, or appropriate
software, to see for themselves, and report back, to be able to voice
their opinion. I am probably flogging a dead horse here since I don't
believe there is another owner of an X-Pro2, X-T2, X-T20, or X100F in
this group yet.


Probably worth starting a new thread if you want to explore
grain simulation algorithms.
Anyway:
Have you experimented with roll-your-own grain algorithms?
I'm thinking if it can be done in-camera then it can't be all that
complex or processor intensive.
I did a few experiments in PS CS6 and can imagine creating
grain simulation actions. There's a lot of parameter space
to explore, but may be worth the effort to reach your artistic end.
==
Later....
Ron C
--