ðòïåëôù 


  áòèé÷ 


Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 

  óôáôøé 


  ðåòóïîáìøîïå 


  ðòïçòáííù 



ðéûéôå
ðéóøíá












     áòèé÷ :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: filmscanners: Canon's scanner



Alex--

Last night I compared scans of the face on Kodak's Q-60 Ektachrome 
transparency targets from the LS4000, SS4000, FS4000 and a Howtek 
drum scanner and a ScanMaker 8700 flatbed Scanner and felt that:

o Of the 35mm film scanners the Canon had the finest grain pattern, 
very much like the drum scanner's.

o The Nikon had the best shadow detail:  The Canon clealy blocked up 
before either the Polariod or Nikon.

o The film grain of the Nikon was the most distinct of all scanners, 
yet when viewed from greater than 4 feet nonetheless seemed slightly 
sharper than the Polaroid (don't remember how it compared with the 
Canon).

o The ScanMaker 8700 scan looked just as good as the drum scan in 
terms of grain, sharpness, color balance, resolving hairs (such as 
the eyelashes), but the shadows blocked up and were filled with green 
noise much worse than any of the scanners.


Technical details:

o The monitor is a 22", Mitsubishi 2040u calibrated and profiled with 
PhotoCal and the Monitor Spyder.

o The Nikon scan was made by me and output with the Wide Gamut 
profile.  This profile is the best I've found of the ones Nikon 
supports for pulling out shadow detail. It makes a substantial 
difference over, say, Adobe RGB.

o The Canon scan was taken from the Hively website, 
http://www.hively.com/canoscan/.

o I had two Polariod scans: one from the VueScan website 
(http://www.hamrick.com), the other from Tony Sleep's website.

o The Howtek drum scan was taken from Tony Sleep's website.  Note 
that he had to sample it down to make it match the size of his other 
scans, so a significant amount of detail must have been lost which 
would explain why the ScanMaker 8700's scan had as much detail 
(except in the shadows).  It was also made from a 4 x 5 target, which 
inherently has lots of detail that the 35mm targets don't.

o The ScanMaker 8700 scan was made from a 4 x 5 target by me at it's 
maximum optical resolution of 1200 dpi.


Comments:

o Shadow detail is very important to me, and I'm finding that using 
the analog gain control and outputting to the Wide Gamut color space 
my Nikon LS4000ED is doing quite well on that account.  Clearly way 
better than the Canon.  Possibly better than the Polaroid with dark 
slides where the analog gain can be bumped up without blowing out the 
highlights.

o The sharpness and fineness of grain of the Canon is beautiful, and 
note that Norman Koren (www.normankoren.com) who's extremely fussy 
about sharpness (I know him :-) seems pleased with his.

o The GEM grain reduction feature on the Nikon works VERY well, and I 
find that GEM scans sharpen up very nicely whereas straight Nikon 
scans accentuate the grain so much that you can only sharpen them a 
tiny bit.

o I have noticed no increase in shadow detail nor reduction in shadow 
noise using multisampling (I've tried up to 16x) on the Nikon.


Conclusions:

I want a drum scanner!  Short of that it seems to depend on what 
tradeoffs you want to make between cost, shadow detail, etc.

Good luck,

--Bill



At 4:21 PM +0200 3-10-01, Alex Z wrote:
>The view things I'm still concerning about thinking about Canon FS4000US:
>1. Somewhat reduced Dynamic range (seems to be 3.4) comparative to the
>rivals (manufacturer claimed 3.6 for SS4000 and 4.2 for CoolScan 4000ED).


>2. The reviews gave me an impression that it even better suited for
>negatives rather then slides (there is opposite problem with rest of 
>the scanners).


>3. Lack of multipassing capability like of Minolta Elite and Elite II.

-- 

======================================================================
Bill Fernandez  *  User Interface Architect  *  Bill Fernandez Design

(505) 346-3080  *  bill@billfernandez.com  *  http://billfernandez.com
======================================================================




 




Copyright © Lexa Software, 1996-2009.