Vuescan: no more streaks on my SE5400!!

  • Thread starter Thread starter Fernando
  • Start date Start date
F

Fernando

I just finished testing a special version of Vuescan on my 5400 (the
3rd in a row that Ed Hamrick sent me), and I'm very happy to say that
the streaks are GONE! :D

I'm too tired now for a report, but it has been great to finally
appreciate my usual test slide without the green lines! :)

I hope Ed will publish the new version soon, and I sure hope he will
listen to the other reports too (shadows "dynamic range", IR
cleaning).

Fernando
 
Fernando said:
I just finished testing a special version of Vuescan on my 5400 (the
3rd in a row that Ed Hamrick sent me), and I'm very happy to say that
the streaks are GONE! :D

I'm too tired now for a report, but it has been great to finally
appreciate my usual test slide without the green lines! :)

I hope Ed will publish the new version soon, and I sure hope he will
listen to the other reports too (shadows "dynamic range", IR
cleaning).

Fernando
Click to expand...

Thats good to hear Fernando. You've been working on this for some time and
it must be good to see the effort pay off.

I'll drimk to less "noise" from now on.

Mark
 
I just finished testing a special version of Vuescan on my 5400 (the
3rd in a row that Ed Hamrick sent me), and I'm very happy to say that
the streaks are GONE! :D

I'm too tired now for a report, but it has been great to finally
appreciate my usual test slide without the green lines! :)

I hope Ed will publish the new version soon, and I sure hope he will
listen to the other reports too (shadows "dynamic range", IR
cleaning).

Fernando
Click to expand...
Perhaps some of those who just keep bashing Ed will now apologize...
How many other software developers keep trying to fix a problem for
an "obsolete" model.
 
Robert Feinman said:
How many other software developers keep trying to fix a problem for
an "obsolete" model.
Click to expand...

Speaking of which: is anyone aware if ICE has found its way back into
the current OSX version of Silverfast for the Nikon LS30/2000?

Ralf
 
Fernando said:
I just finished testing a special version of Vuescan on my 5400 (the
3rd in a row that Ed Hamrick sent me), and I'm very happy to say that
the streaks are GONE! :D
Click to expand...
Let's hope that this is a general fix and not just an increase in
tolerance to include your scanner. ;-)
 
Robert Feinman said:
Perhaps some of those who just keep bashing Ed will now apologize...
How many other software developers keep trying to fix a problem for
an "obsolete" model.
Click to expand...

"Obsolescence" is not synonymous with "superceded".

For example, Win95 is superceded and obsolete, Win98/SE is superceded
but not yet obsolete (it was originally scheduled for obsolescence on
16th January 2004 but has been extended to 30th June 2006), while WinXP
is neither superceded nor obsolete.

The SE5400 may have been superceded by a Mk.II version (which may or may
not exhibit the same streak problem there aren't enough in the user base
yet to know) but it is not yet an obsolete product and Minolta will
continue to support it for some considerable time yet.

In the EU, and probably also in the US, there are laws governing the
time that a product must be supported after being discontinued - and
that period is a lot longer than 2 and a half weeks!
 
Perhaps some of those who just keep bashing Ed will now apologize...
Click to expand...

Now that's bashing! Contrast that to these objective facts:

If anyone should apologize it's the author for *misrepresenting* the
software for *two years* and refusing to reimburse *defrauded*
customers.

Secondly, he should then also apologize for demonstrated incompetence
and not being up to the task. Instead, depending on users to provide
him with *code* to fix his own long standing bugs!

Thirdly, will he now do the right thing (no danger of that...) and
pass on a proportional part of his profits to Fernando and, most of
all (!) Kennedy for showing him how elementary the solution was and
"how it's done"?

But most importantly, jubilant VueScan fans rushing out to celebrate
would be well advised to keep those Champaign bottles on ice.

Judging by VueScam's history this may very well be (and indeed
probably is) just another case of "hope over reality". It wouldn't be
the first time that after a lot of loud fanfare the bug(s) reappear.

And, of course, this is only *one* bug! What about all the others?

Don.

P.S. So, the answer to your question is: Nope! ;o)
 
Indeed!

But some hardcore Vuescan fans are notorious for playing fast and lose
with facts. Actually, they take no notice of facts and get quite upset
when facts are used.

Years ago when I moonlighted as a freelance journo my editor used to
joke "Don't let the facts get in the way of a good story". ;o)

Don.
 
I just finished testing a special version of Vuescan on my 5400 (the
3rd in a row that Ed Hamrick sent me), and I'm very happy to say that
the streaks are GONE! :D
Click to expand...

Do you still have to calibrate the scanner first using Minolta
software?

Don.
 
Let's hope that this is a general fix and not just an increase in
tolerance to include your scanner. ;-)
Click to expand...

Bingo!

It which case this increase is tolerance is bound to pop up elsewhere
(even on his scanner) with... shall we say... unintended side
effects.

Trying to fix VueScan is like trying to squeeze a balloon. Something
always pops up! ;o)

Don.
 
Do you still have to calibrate the scanner first using Minolta
software?
Click to expand...

No, no more. :)
This also means I can finally use the SE5400 with Linux, if I need to
(there are still issues with IR cleaning and shadow linearity), since
there's no Minolta software for Linux...

Fernando
 
Let's hope that this is a general fix and not just an increase in
tolerance to include your scanner. ;-)
Click to expand...

Reporting what Ed Hamrick told me: there is a lot of variation among
the sensors Minolta used for the SE5400 (let's say those sensors are
not first class? ;-) ).
The unit he had was successfully working even with older Vuescan
versions; but some units had more ample per-cell variations; moreover,
he initially did the calibration with CCD exposures that were not long
enough to properly discriminate the "hot pixels" of some sensors
(including mine), that start to present visible streaks only when
appreciably long exposure times are used.

So, I think this is a "proper" fix, not a quick hack... :)

Fernando
 
Secondly, he should then also apologize for demonstrated incompetence
and not being up to the task. Instead, depending on users to provide
him with *code* to fix his own long standing bugs!
Click to expand...

Just to put this straight: I did not provide Ed Hamrick any code, I
just explained (and showed with picture samples) that I was having
good results with a dark frame subtraction.
But he did not use d.f.s. (as far as I know): he just re-thinked
Vuescan calibration routine.

This is not to say that I'm happy and dancing (I had to wait quite a
bit of time for this, and I submitted quite a bunch of bug reports),
but the fact that a producer actually fixed a reported issue, is much
more the exception than the rule, nowadays (as sad as it is).
This time he has been proactive, helpful and "solid". I appreciated
this.
Judging by VueScam's history this may very well be (and indeed
probably is) just another case of "hope over reality". It wouldn't be
the first time that after a lot of loud fanfare the bug(s) reappear.
Click to expand...

You're right that some bugs have returned over time. I hope this one
will not, and that..
And, of course, this is only *one* bug! What about all the others?
Click to expand...

.... all the others will be fixed, too, one day or another. :-)

I'll keep the bug reports flowing... and urge all the other users to
do the same!
It's the only way.

Fernando
 
Don said:
Thirdly, will he now do the right thing (no danger of that...) and
pass on a proportional part of his profits to Fernando and, most of
all (!) Kennedy for showing him how elementary the solution was and
"how it's done"?
Click to expand...
Oh come on, Don! What evidence is there that Ed even used a similar
approach to Fernando's? As I said before, what I proposed (and Fernando
implemented) was a work around, not a solution.
 
Fernando said:
Reporting what Ed Hamrick told me: there is a lot of variation among
the sensors Minolta used for the SE5400 (let's say those sensors are
not first class? ;-) ).
Click to expand...

Err - none of the CCDs that go into consumer scanners are first class,
so it's a bit of a surprise that Ed expected they might be. ;-)
The unit he had was successfully working even with older Vuescan
versions;
Click to expand...

Something I said a while ago - he obviously thought the problem was
licked when he first announced it.
but some units had more ample per-cell variations; moreover,
he initially did the calibration with CCD exposures that were not long
enough to properly discriminate the "hot pixels" of some sensors
(including mine), that start to present visible streaks only when
appreciably long exposure times are used.
Click to expand...
That doesn't quite ring true to me. Excessively hot cells (high dark
current) will show up in short exposures more than other cells. You
increase the exposure to detect the less variation in the better cells.
 
I just finished testing a special version of Vuescan on my 5400 (the
3rd in a row that Ed Hamrick sent me), and I'm very happy to say that
the streaks are GONE! :D

I'm too tired now for a report, but it has been great to finally
appreciate my usual test slide without the green lines! :)

I hope Ed will publish the new version soon, and I sure hope he will
listen to the other reports too (shadows "dynamic range", IR
cleaning).
Click to expand...
Good for you Fernando. Now, let's hope it makes it through to the
latest version and doesn't reappear later in another version after
something else is "fixed". :)

--

Hecate - The Real One
(e-mail address removed)
Fashion: Buying things you don't need, with money
you don't have, to impress people you don't like...
 
Reporting what Ed Hamrick told me: there is a lot of variation among
the sensors Minolta used for the SE5400 (let's say those sensors are
not first class? ;-) ).
Click to expand...


We have a saying in the UK:

A poor workman blames his tools :)

--

Hecate - The Real One
(e-mail address removed)
Fashion: Buying things you don't need, with money
you don't have, to impress people you don't like...
 
... all the others will be fixed, too, one day or another. :-)
Click to expand...

At this rate it will take 800 years... But only if users provide the
author with solutions/workarounds!

VueScan boasts supporting 400 scanners. If we are extremely generous
(to the point of irrationality) and assume *only* 1 bug per scanner:

400 bugs * 2 years per bug = 800 years...

But only *if* the users provide the solutions/workarounds!

Good luck with your wait! ;o)

Don.
 
Oh come on, Don! What evidence is there that Ed even used a similar
approach to Fernando's? As I said before, what I proposed (and Fernando
implemented) was a work around, not a solution.
Click to expand...

Well, whatever it was, it did something the author was unable to do
for two years. So, pending messages to the contrary, I think we can
agree on that much.

And then following Fernando's emails a version suddenly appears (after
two years of false starts!) which seems to actually address it.

I think a reasonable conclusion is, that's more than a coincidence.
Indeed, it clearly points to a definitive causal link between the two.

Circumstantial? Most definitely! Probable? Very much so! To the point
of near certainty.

Don.
 
Back
Top