Should there be a 3rd scanner bake-off?

[jim.hutchison]

I'm wondering if enough new scanning technology has emerged since last
year to justify another scanner bake-off.

Replies more than welcome...


--jim h


http://www.jamesphotography.ca

 

[-]

I'm wondering if enough new scanning technology has emerged since last

year to justify another scanner bake-off.

There are a few new models but the technology hasn't changed that much. PMA
is this month, so you might want to wait until after those announcements.

Doug

 

[Philip Homburg]

I'm wondering if enough new scanning technology has emerged since last
year to justify another scanner bake-off.

Replies more than welcome...

What are the areas that are not tested yet? I can think of a couple of
areas:
- color accuracy.
- Dmax and scanner noise
- Scanning time

 

[Roger S.]

Hi Jim,
By all means, yes! I'd like a slide that tests Dmax where scanners are
rated on their ability to control shadows noise and preserve highlight
and shadow detail. Color is something that can be corrected, but
missing data cannot.

Maybe test the MF ability of flatbeds as well?

Roger

 

[Guy Jordan]

I'm wondering if enough new scanning technology has emerged since last
year to justify another scanner bake-off.

Replies more than welcome...


--jim h


http://www.jamesphotography.ca

I'd love to see a comparison of techniques rather than just the scanner
model. Workflow and the individuals skill at executing those steps must
play a part in the scan qualities I've seen in your bake-off results.
Scanning software, computer hardware, scanner revision also might play a
part too but I'd really like to know what skills are universally best
practice. I'd also love to know what variations on the best practice
theme are likely to offer improvements to my own scanning.

I'd also be willing to participate if a rank amateur with a second hand
1800f is welcome.

-Guy

 

[Djon]

Guy's right. Previous bakeoff proved that all sorts of elderly scanners
do beautiful work in good hands. John

 

[jim.hutchison]

I'm wondering if enough new scanning technology has emerged since last
year to justify another scanner bake-off.

Replies more than welcome...


--jim h


http://www.jamesphotography.ca

Guys,

If I do another bake-off, I won't be able to measure some of the
things you're asking, such as d-max... unless someone knows how to do
that. I certainly can ask participants to select a standard color
profile, and not to change the image gamma, or sharpen it...

Color accuracy is another area of many variables. I would have to ask
everyone to calibrate their systems, which not everyone is willing to
do. It costs money to do properly.

Rather, I'm looking at using Imatest again instead of a judging panel,
but this time have more specific requirements of the scanner operator.
Sharpness is something I can test scientifically, and it is known that
ALL aspects of scanning are influenced by operator skill, so I'll try
to nail it down as tight as possible.

Requirements would be:

-manual sharpening
-Adobe 1998 color profile
-sharpening not allowed
-brightness/contrast/gamma adjustments not allowed
-8 bit color depth
-single-pass


If possible, I'll also include a microscope image of the target slide.


Please chime in if you have any suggestions.

 

[Roger S.]

Hi Jim,
Would an approach like this work for DMAX? I don't really care about
absolute values but rather the relative performance of scanners. A
subjective review of scanners would give useful information if Imatest
doesn't work for this.

http://www.dpreview.com/reviews/nikond200/page22.asp

Personally, image sharpness and resolution aren't major areas of
concern. I don't look at my scans and wish they had more resolution-
generally, what I have is overkill for my purposes (small prints and
on-screen display). Photokit sharpener makes my prints quite a bit
sharper than optical ones ever were. Dynamic range and noise for slide
scans are a different story.

 

[jim.hutchison]

Hi Jim,
Would an approach like this work for DMAX? I don't really care about
absolute values but rather the relative performance of scanners. A
subjective review of scanners would give useful information if Imatest
doesn't work for this.

http://www.dpreview.com/reviews/nikond200/page22.asp

Personally, image sharpness and resolution aren't major areas of
concern. I don't look at my scans and wish they had more resolution-
generally, what I have is overkill for my purposes (small prints and
on-screen display). Photokit sharpener makes my prints quite a bit
sharper than optical ones ever were. Dynamic range and noise for slide
scans are a different story.

A very interesting article... something to chew on. The first couple
things that come to mind are:

1) Sourcing a target image to shoot (calibrated Stouffer Step Wedge).
I suppose I could create one in PS, but then the question of printer
accuracy springs to mind.
2) The accuracy of the exposure latitude as captured by the film
media... Film choice would be crucial here. Perhaps Provia or even
Astia would suit best, but that's just an educated guess.

I appreciate your response; keep your eye on this NG for updates.

--jim

 

[Roger S.]

Would shooting this on Provia or Astia do the trick?

http://www.stouffer.net/Stoufferhome1.htm

I think we could use a decent quality drum scan of the wedge as a
baseline- say if there's no info beyond step 27 out of 31 on the drum
scan, we can then see how close the other scanners come to matching it
but not penalize them for the limited Drange of the film. It will take
careful exposure on your part, however.

DPReview talks a bit more about the methodology here (they are using a
transmissive wedge)
http://www.dpreview.com/reviews/canoneos5d/page22.asp

 

[Roger S.]

Imatest does seem to support this:
http://www.imatest.com/docs/q13.html

 

[jim.hutchison]

Imatest does seem to support this:
http://www.imatest.com/docs/q13.html

I created a couple different wedges in Photoshop, one with 1%
luminance increments, another with 5% increments (20 steps). There is
very slight - but perceivable - difference between 95 and 100 when
printed on my Epson R200. It will be interesting to see if slide film
can pick that up.

Thanks for those links; Norman's Imatest site is chock full of useful
data. He'll be donating a copy of Imtatest Light for the bake-off.

--jim

 

[Roger S.]

I created a couple different wedges in Photoshop, one with 1%

luminance increments, another with 5% increments (20 steps). There is
very slight - but perceivable - difference between 95 and 100 when
printed on my Epson R200. It will be interesting to see if slide film
can pick that up.

Hi Jim,
Can't slide film record a far greater range of tones than you can print
on ordinary paper (Dmax 2 or so)? I would think that a real scene with
a light source going into shadow (say a mid-day shot) would have a far
greater range than a print. I think this is why the test I linked to
used a backlit transparency. I use the Epson R220 with MIS inks, by
the way.