Minolta 5400 - Lines in scans - Please HELP!
Ricky
I recently posted a message regarding a problem I had with a Minolta
5400. The scanner was putting straight lines in the scan direction (ie
lengthways) in every scan I made, so I returned it believing it to be
faulty.
I have received my second SE5400, from a completely different
supplier, and guess what? Same deal. Every scan has these lines! The
lines are only really visible if I turn the gamma up and zoom in to
around 80%, but nevertheless they are there and they are unacceptable.
I have done a few scans over the last couple of days, with both
negatives and slides, and have come to the following conclusions:
1: The lines vary in thickness between 3 - 6 pixels wide
2: The lines are either magenta or a yellow/green colour
3: The lines only seem to be visible in dense areas
4: The lines are much more visible when scanning negatives
5: The lines occur with both DiMAGE scan utility and Vuescan
6: The lines occur regardless of whether ICE/Grain Dissolver is
enabled
7: Letting the scanner warm up before scanning doesn't help, and
neither does re-initializing
I find it hard to believe that I could be unlucky enough to receive
two different scanners from two different suppliers, and both of them
have the same fault. Is it possible something in my computer hardware
is behind all this? Or maybe some kind of electrical interference? Or
am I just really unlucky?
I'll do some more tests over the weekend and hopefully test the
scanner on a friend's computer and see what happens. I'll also try
moving the Firewire card to a different slot further away from the
other cards and see if that helps.
Until then though, I was wondering if anyone out there has suffered
from the same problem, or if anyone has any suggestions as to what
could be causing these lines, as I can't take much more of this, and
my wife is getting fed up with me hunched over my desk grimacing and
cursing!
Many thanks,
Ricky.
Mike Hide
Mike Hide
I think Ed Hanrick mentioned dust particles in the calibration area and
someone else mentioned power supply problems . I cleaned the former ,it made
no difference . I did originally have lateral bands ,quite severe ones I did
solve that problem by dumping the mat that gis applied to the scanner glass
surface .
One thing I have noticed is a small buildup of dust and "smeers" on the
inside of the scanner glass .I have no idea how to clean that surface ..
--
mike hide
"Ricky" <ricky...@hays.com> wrote in message
news:1b3b5b53.03102...@posting.google.com...
Ira Solomon
I've sent you an e-mail message with a jpg.
Another person had line problems and this is his sample.
If this is what you are getting I'll try to dig up his address.
He solved it by removing some sort of accelerator program.
Good Luck
Ira Solomon
JMooreTS
I, too, have had exactly the same problem . . . except it is with Vuescan and
an Epson 1260 flatbed. I even purchased a second scanner thinking my first one
(bought used) was defective. But the same problem occurs intermittently. If I
restart everything it goes away.
John Moore
Bart van der Wolf
> Until then though, I was wondering if anyone out there has suffered
> from the same problem, or if anyone has any suggestions as to what
> could be causing these lines, as I can't take much more of this, and
> my wife is getting fed up with me hunched over my desk grimacing and
> cursing!
Are you using the latest version (VueScan 7.6.64, as I write)?
I used to have issues with the calibration with VueScan, but not with the
Minolta Software. If the Minolta utility also returns "striped" images, it
may be some electronic interference (power source/converter, speakers)
causing it. Try switching off as many devices in the vicinity as you can,
and check if that helps.
Posting a small image sample somewhere might also help in identifying its
cause.
Bart
Ricky
I've posted an example of the lines here:
http://www.photo.net/bboard/uploaded-file?bboard_upload_id=14996684
This is a cropped section of a scan from Provia 100F. I've turned the
gamma up so the lines are easily visible. It was scanned on firewire
connection (lines appear when using USB as well). The image is
'upright' so the lines go from bottom to top in the scan direction.
Is this the kind of thing you guys are seeing?
Thanks,
Ricky.
Kennedy McEwen
<ricky...@hays.com> writes
>Hi Guys,
>
>I've posted an example of the lines here:
>
>http://www.photo.net/bboard/uploaded-file?bboard_upload_id=14996684
>
>This is a cropped section of a scan from Provia 100F. I've turned the
>gamma up so the lines are easily visible. It was scanned on firewire
>connection (lines appear when using USB as well). The image is
>'upright' so the lines go from bottom to top in the scan direction.
>
variances well outside what the calibration can cope with! Appears to
be only one pixel wide in this example. You mentioned cases of several
pixels wide examples - are you sure these are not as a result of
post-processing such as sharpening?
Either way, there is little you can do to fix it, but its a bit
concerning that you have had two examples from independent sources.
Return the scanner and send this image to Minolta with a strongly worded
complaint. This type of problem out of the box is completely
unacceptable.
Good luck third time - or with Nikon!
--
Kennedy
Yes, Socrates himself is particularly missed;
A lovely little thinker, but a bugger when he's pissed.
Python Philosophers (replace 'nospam' with 'kennedym' when replying)
Kennedy McEwen
<reani...@miskatonic.edu> writes
>
>Just a harebrained idea... But I always try to rule this possibility
>out before I start blaming software or the electronics.
>
>Any chance that it might be caused by dust in the optics?
>
Not unless the optical design has a surface exactly on an intermediate
focal plane - which would be an extremely poor design for most
applications, including this one.
As my old optics designer used to say: "Lenses are for looking through,
not looking at!". In other words, a little dust or defect on or in a
lens has negligible effect on the image - it takes a lot of dirt, and
the effect is nothing like what is being experienced here.
tomas
from Visioneer read this NG?) I too already exchange the unit with no
improvement. See my original posting:
http://groups.google.com/groups?selm=vdshkbrv18mk75%40corp.supernews.com
I was told this is normal result from my "cheap" scanner (~$100).
Tomas
"Ricky" <ricky...@hays.com> wrote in message
news:1b3b5b53.03102...@posting.google.com...
Bart van der Wolf
"Kennedy McEwen" <r...@nospam.demon.co.uk> wrote in message
news:37WszYOJWEm$Ew...@kennedym.demon.co.uk...
SNIP
> This looks very much like dirt/dust/defects on the CCD itself, producing
> variances well outside what the calibration can cope with!
I second that. This *looks* like dirt on the sensor or defective sensels
(this would produce identically spaced defects on other scans).
Very unlikely to happen in two independent samples, but apparently possible.
Bart
Stephen Harker
> "Ricky" <ricky...@hays.com> wrote in message
> news:1b3b5b53.03102...@posting.google.com...
> SNIP
> > Until then though, I was wondering if anyone out there has suffered
> > from the same problem, or if anyone has any suggestions as to what
> > could be causing these lines, as I can't take much more of this, and
> > my wife is getting fed up with me hunched over my desk grimacing and
> > cursing!
>
> Are you using the latest version (VueScan 7.6.64, as I write)?
> I used to have issues with the calibration with VueScan, but not with the
> Minolta Software. If the Minolta utility also returns "striped" images, it
> may be some electronic interference (power source/converter, speakers)
> causing it. Try switching off as many devices in the vicinity as you can,
> and check if that helps.
This is good advice. I used to occasionally get striped images from
my ACER ScanWit 2720S, it was rather intermittent but when it occurred
would affect most scans taken at that time. I could not work out the
problem. Then my monitor died and after I bought a replacement the
lines _never_ recurred (at least over the last eight months). If I
had thought of this and tested the monitor I would have avoided a lot
of scanning problems.
--
Stephen Harker Stephen...@spme.monash.edu.au
School of Physics & Materials Engineering
Monash University http://www.ph.adfa.edu.au/s-harker/
Baloney Baffles brains: Eric Frank Russell
Alan Browne
Ricky wrote:
Looks like dirt or partially dead sensors (many causes possible) in the
linear array or a problem in the associated electronics.
Send it back to Minolta with example print or electronic file and
somewhat strong words.
Cheers,
Alan.
Alan Browne
Kennedy McEwen wrote:
> In article <aqdgpvkge8f46gn4e...@4ax.com>, Herbert West
> <reani...@miskatonic.edu> writes
>
>>
>> Just a harebrained idea... But I always try to rule this possibility
>> out before I start blaming software or the electronics.
>>
>> Any chance that it might be caused by dust in the optics?
>>
> Not unless the optical design has a surface exactly on an intermediate
> focal plane - which would be an extremely poor design for most
> applications, including this one.
>
> As my old optics designer used to say: "Lenses are for looking through,
> not looking at!". In other words, a little dust or defect on or in a
> lens has negligible effect on the image - it takes a lot of dirt, and
> the effect is nothing like what is being experienced here.
Unless the dirt is on the surface of the sensor array itself, like dirt
on a negative or slide ... will show.
Alan,
Ricky
been kind enough to make a few suggestions. He suggests disabling
everything that runs at start-up and if I get a scan with no lines to
enable each one, one by one, to see which is causing it. He knows
someone who had a similar problem, which he fixed by removing an
accelerator program.
Failing that, he suggests it may be something in my computer hardware.
I will test the scanner on a friend's computer this weekend. If I
still get lines then I'll arrange with the supplier to send the
scanner back for replacement (I had better be third time lucky!). If,
however, there are no lines then I guess I have the tedious task of
finding out what exactly in my computer hardware is behind the
problem. For info, the scanner is at least two feet away from the
monitor and at least three feet away from the computer.
Thanks again everyone for your advice,
Ricky.
Ricky
> linear array or a problem in the associated electronics.
I've had this and the original scanner plugged into a Belkin
Mastercube surge protector. Is it possible that this has become faulty
and damaged the scanners?
Bart van der Wolf
*Very* unlikely.
Bart
Alan Browne
Ricky wrote:
If it is electronic or sensor related it is most likely a manufacturing
fault at Minolta or one of their suppliers.
The P/S in the 5400 is fairly robust and would likely protect against
most surges, other than a lightning strike on the mains outside your house.
The fact that the stripes are in the same place suggests something is
wrong with the scanning guts (the sensor array, dirt there, or electronics)
Send the critter back.
Cheers,
Alan
Kennedy McEwen
<"Alan Browne"@videotron.canospam> writes
I suspect that this problem is actually dust/dirt/defects on the sensor
itself - but it certainly isn't in the optics!
Kennedy McEwen
<ricky...@hays.com> writes
It's possible, but extremely unlikely and even if it did, it's unlikely
to produce this type of fault. A poor power supply would cause other
parts of the scanner would fail completely long before this sort of
damage, if indeed that is what it is, started to appear on the sensor.
I don't buy the other suggestions about PSU noise or interference
either. This effect is just too regular and too isolated to single
pixels in the scan to be cause by anything other than an optical effect
(dirt/dust) or the sensor itself.
Ricky
> > Mastercube surge protector. Is it possible that this has become faulty
> > and damaged the scanners?
>
> *Very* unlikely.
>
> Bart
There is another SE5400 user who has exactly the same problem as me.
He has posted over on photo.net:
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=006I3Q
Someone has posted a response on there saying their Scan Elite II has
always had this problem and has provided a link to the review on
imaging-resource.com:
http://www.imaging-resource.com/SCAN/DSEII/DSEIIPICS.HTM
and has also provided a link to a test image from this review which
shows lines that look VERY similar to the lines I'm getting:
http://www.imaging-resource.com/SCAN/DSEII/DSETRAA602PS.HTM
The poster seems to think that this is a characteristic of Minolta
film scanners. If you can turn the gamma up in a scan of a slide with
blacks such as this and you do NOT see any lines, then maybe it's just
that Minolta's quality control isn't so good.
Anyway though, I will be receiving a brand new unit on Monday. If this
one shows the same lines as the previous two examples I've had, then I
will be getting a refund. Maybe I should have stuck with the FS4000. I
used to get minor lines (a little bit similar to this) in shadows with
the FS4000, but no where near as bad, and doing a 'Long Exposure' pass
in Vuescan eliminated them.
How well does the 'Long Expsosure' pass work with the SE5400, and do
you think it would help in this case?
Thanks,
Ricky.
Ricky
> fault at Minolta or one of their suppliers.
> The P/S in the 5400 is fairly robust and would likely protect against
> most surges, other than a lightning strike on the mains outside your house.
> The fact that the stripes are in the same place suggests something is
> wrong with the scanning guts (the sensor array, dirt there, or electronics)
>
> Send the critter back.
I've just done a search of this group and there are LOADS of messages
posted about this EXACT SAME problem with Minolta scanners. I refuse
to believe that if I want to own a Minolta film scanner that I have to
accept these lines as 'normal'. These lines render the scans
absolutely USELESS. If it was a cheap low-end flatbed I'd find it
easier to swallow but as it is I've paid a fair amount of money for it
and it is unacceptable! As I posted earlier, I am expecting a new unit
on Monday. If it has the same problem then I will conclude that this
is a characteristic/defect in Minolta film scanners and that they are
all CR*P, and get my money back.
Here are links to a couple of other posts regarding this same problem
with other Minolta film scanners (copy and paste into your browser):
An example of a scan someone made with a Scan Dual III:
http://hem.passagen.se/alis/vuescan.jpg
This is from this post:
And also this post about the same problem with a Scan Elite II:
Thanks all,
Ricky.
CSM1
From the images that I looked at from your posts. The lines look like
processing scratches from the negative development. Or maybe scratches on
the film from the automated print processor, made when the negative was
pulled through the exposure gate, which was dirty.
I have scratched negatives myself in my own darkroom, just pulling the
negative through the enlarger. (Years ago!)
--
CSM1
http://www.carlmcmillan.com
--
Kennedy McEwen
>posted about this EXACT SAME problem with Minolta scanners. I refuse
>to believe that if I want to own a Minolta film scanner that I have to
>accept these lines as 'normal'.
Unfortunately, Ricky, if the problem is as common as your research
indicates then I don't think you have any choice - if you STILL want to
own a Minolta film scanner.
>These lines render the scans
>absolutely USELESS.
I wholeheartedly agree, but your only recourse would appear to complain
to Minolta themselves. :-(
Reading some of the posts from others with similar problems, I do not
hold out much hope of success. Magically all scanners returned to
Minolta seem to work perfectly for them. I can only suggest that their
QA inspectors are either completely incompetent or are working to an
extremely poor specification, if it even includes test and acceptance
criteria for scan lines at all.
This is a sorry state of affairs, because this does appear to be a
scanner which has so many other things going for it that it would be
almost criminal for Minolta to let poor QA ruin its reputation.
Mike Hide
at it logically many people are having the same problem with a variety of
scanners .
One other point the "lines" are always in the vertical direction of the
exposure, regardless of whether the shots are portrait or landscape, whereas
the scanner scans in one direction only and is unaware of the type of image
being scanned [landscape or portrait].
Considering the above, the past processes cannot be remedied . The question
now is can these flaws however produced be corrected by one or all of the
imaging softwares without severely compromising image quality ?
--
mike hide
"CSM1" <nom...@nomail.com> wrote in message
news:mDimb.934$uA....@newssvr22.news.prodigy.com...
Mike Hide
???
--
mike hide
"Ricky" <ricky...@hays.com> wrote in message
news:1b3b5b53.03102...@posting.google.com...
Mike Hide
--
mike hide
"Kennedy McEwen" <r...@nospam.demon.co.uk> wrote in message
news:l3yowtDYZbm$Ew...@kennedym.demon.co.uk...
William D. Tallman
> Until then though, I was wondering if anyone out there has suffered
> from the same problem, or if anyone has any suggestions as to what
> could be causing these lines, as I can't take much more of this, and
> my wife is getting fed up with me hunched over my desk grimacing and
> cursing!
>
> Many thanks,
>
> Ricky.
Hi Ricky,
I've got one of those beasts too. So far, I've not seen this effect at any
working magnification. If I crank up to 300% I can see a very small amount
of this, but only at the point where the pixels are well defined.
I don't buy the film scratch explanation. The lines are far too regular and
color specific. Nor do I buy the dust explanation, as the colors would be
random, I would think, dust not being color specific. Or is that true?
You say the lines vary in width from one to three pixels. Is the width
color specific; I noticed that the magenta lines are of approximately the
same width.
From what you say, may we assume that the color of the lines is the same for
both negs and slides? If so, then we might imagine that whatever the
mechanism of malfunction is, it must somehow be color specific.
I guess the first question I would ask is whether the lines remain the same
from strip to strip as well as from frame to frame. Also, do the position
of the lines change when the scanner is cycled? Dunno exactly what the
answer to either question might mean, but they seem worth asking.
Perhaps someone reading this thread who knows the mechanics of scanner
function, especially the physics of charge coupled devices, could shed some
light on this? The point here is that if one is able to provide a
plausible mechanism to present to Minolta, that might be worth more than
just a complaint.
Yeah, I'm concerned. Hasn't happened to me. Yet....
Bill Tallman
Ed Hamrick
> I don't buy the film scratch explanation. The lines are far too regular
and
> color specific. Nor do I buy the dust explanation, as the colors would be
> random, I would think, dust not being color specific. Or is that true?
It's obviously not film scratches - the lines are too straight and
they extend the whole length of the scan in the scan direction.
It's clearly a CCD calibration problem.
> Perhaps someone reading this thread who knows the mechanics of scanner
> function, especially the physics of charge coupled devices, could shed
some
> light on this?
All the scanners that seem to have these problems are the
Minolta scanners made by Avision (Scan Dual II, Scan Dual II, Scan
Elite II, Scan Elite 5400). They all use similar black/white point
calibration.
There are two possibilities - problems compensating for the black
point of the CCD and problems compensating for the white point.
I could figure out the problem if people would be more systematic
about describing the problem. It's confusing, because sometimes
people find these lines in negative scans, and sometimes in positive
scans, sometimes in the dark parts of scans and sometimes in the
light parts.
It would be useful if people who have this problem could be more
precise about when the problem occurs - is it in the dark parts
of negatives (i.e. the bright parts of the final scans), or the
bright parts of negatives (i.e. the dark parts of the final scans)?
Is it in the dark parts of slide scans (i.e. the dark parts of
the slide) or the bright parts of slide scans (i.e. the bright
parts of the final scans)?
Regards,
Ed Hamrick
Bart van der Wolf
"Kennedy McEwen" <r...@nospam.demon.co.uk> wrote in message
news:l3yowtDYZbm$Ew...@kennedym.demon.co.uk...
> Reading some of the posts from others with similar problems, I do not
> hold out much hope of success. Magically all scanners returned to
> Minolta seem to work perfectly for them.
I've read about similar experiences with Nikon service. Not what it used to
be.
SNIP
> This is a sorry state of affairs, because this does appear to be a
> scanner which has so many other things going for it that it would be
> almost criminal for Minolta to let poor QA ruin its reputation.
I agree with that. I have little trouble with my SE5400, other than VueScan
related imperfections. Minolta has a need to let this scanner succeed, they
want to gain back some reputation in digital imaging. Since they have so
much lose, it must be possible to persuade them to do a decent job on
servicing.
Bart
Bart van der Wolf
"Mike Hide" <mike...@comcast.net> wrote in message
news:Xtmmb.9352$mZ5.52359@attbi_s54...
SNIP
> One other point the "lines" are always in the vertical direction of
> the exposure, regardless of whether the shots are portrait or
> landscape, whereas the scanner scans in one direction only and
> is unaware of the type of image being scanned [landscape or portrait].
Correct. This points to an inability to calibrate.
> Considering the above, the past processes cannot be remedied.
> The question now is can these flaws however produced be
> corrected by one or all of the imaging softwares without severely
> compromising image quality ?
After tonal corrections, that would be practically impossible. On a RAW
image it is doable, but not something one would try without adequate
software. It is too time consuming to do it routinely.
Bart
Ricky
> they extend the whole length of the scan in the scan direction.
> It's clearly a CCD calibration problem.
I'm sure these are not scratches on negatives/slides as I've scanned
some of these before on my old FS4000 with absolutely no trace of
lines.
> All the scanners that seem to have these problems are the
> Minolta scanners made by Avision (Scan Dual II, Scan Dual II, Scan
> Elite II, Scan Elite 5400). They all use similar black/white point
> calibration.
So this IS a common problem Minolta scanners. What I'd like to know is
when I get my *third* SE5400 on Monday what are the chances that it
will display the same problems as the previous two. There are
obviously people that own this and other Minolta scanners that do not
have any of these problems. Can I expect to go through several of
these scanners before getting one that works properly, or have I just
been unlucky? If I knew that eventually I would get a model that did
not have this problem then I would stick with it, because as a
previous poster has pointed out, this scanner has many other things
going for it.
> I could figure out the problem if people would be more systematic
> about describing the problem. It's confusing, because sometimes
> people find these lines in negative scans, and sometimes in positive
> scans, sometimes in the dark parts of scans and sometimes in the
> light parts.
These lines appear on both negative and slide scans. The only scans
without lines so far have been very low contrast images shot on Fuji
1600ISO negative film.
> It would be useful if people who have this problem could be more
> precise about when the problem occurs - is it in the dark parts
> of negatives (i.e. the bright parts of the final scans), or the
> bright parts of negatives (i.e. the dark parts of the final scans)?
> Is it in the dark parts of slide scans (i.e. the dark parts of
> the slide) or the bright parts of slide scans (i.e. the bright
> parts of the final scans)?
The lines always seem to be in the mid-dark to dark areas of the
negative (the bright areas of final scans, eg skies and white tops
etc) and the mid-dark to dark areas of slide scans. In the example I
posted of a scan from Provia 100F I had to turn the gamma way up to to
make the lines visible (it is a fairly dark image), so probably
wouldn't be noticeable (but it's still unacceptable), but negatives
are a completely different story. I scanned a negative of my Father
standing on a sunny beach, and there were lines completely throughout
the scan. It looked terrible! Recalibrating does not improve the
situation, neither does multi-sampling (in fact it makes the lines
worse).
If this is a calibration issue, are there any great differences
between the way DiMAGE scan utility calibrates and the way Vuescan
calibrates? The lines do seem to be more noticeable when scanning with
Vuescan. Funny thing is, the lines can never be seen in the preview
image, either with Vuescan or with DiMAGE scan utility. Whether this
is due to the low resolution of the preview image I don't know.
Many thanks for everyone's contribution, I know there will be many
potential SE5400 owners out there very concerned about this problem.
Ricky.
Kennedy McEwen
<wtal...@olypen.com> writes
>
>I've got one of those beasts too. So far, I've not seen this effect at any
>working magnification. If I crank up to 300% I can see a very small amount
>of this, but only at the point where the pixels are well defined.
>
>I don't buy the film scratch explanation. The lines are far too regular and
>color specific. Nor do I buy the dust explanation, as the colors would be
>random, I would think, dust not being color specific. Or is that true?
>
of the lines are green, some brown and some dark blue, however each line
remains a fixed colour throughout the scan. That is a strong indicator
that the problem is dust or defects on the scanner's sensor. The sensor
in this scanner has different lines for detecting red, green and blue
colours and dust, dirt or a defect on one pixel of any individual line
would give rise to a consistently colour biased line on the image. The
actual colour would depend on the amount of signal each of the remaining
pixels get from the image - and also on the exact output of the
defective device, but the bias would remain consistent.
This is in contrast to the Nikon approach which has a single row of CCD
cells and the sensitivity to colour is achieved by changing the
illumination source through red, green and blue LEDs. Dust or defects
on such a device would result in the same defect appearing in all colour
channels.
>
>I guess the first question I would ask is whether the lines remain the same
>from strip to strip as well as from frame to frame. Also, do the position
>of the lines change when the scanner is cycled? Dunno exactly what the
>answer to either question might mean, but they seem worth asking.
>
>Perhaps someone reading this thread who knows the mechanics of scanner
>function, especially the physics of charge coupled devices, could shed some
>light on this? The point here is that if one is able to provide a
>plausible mechanism to present to Minolta, that might be worth more than
>just a complaint.
>
Having worked with CCDs in scanning systems, albeit not film or flatbed
but high performance military surveillance systems, almost since their
invention I do understand the issues with CCD physics and their
deficiencies. The problem with the issue is that without additional
diagnostics it is impossible to tie the cause down to more than
dust/defect at the sensor itself - and there are several potential CCD
defects which could give rise to this effect, including channel traps,
output non-linearity, excess noise and even electro-static damage during
assembly.
As Ed Hamrick points out on his post, the problem is rooted in the CCD
calibration, the process of correcting for the individual response and
dark level of all the elements in the CCD and converting their unique
outputs to a common response for the array with equal outputs for black
and white and every colour in between. However, unless Minolta have
deviated significantly from the common practice in their film scanners,
CCD calibration is usually performed on aerial images. ie. not viewing
a focussed surface which can accumulate dirt and dust. (This is
completely different from flatbed scanners which often calibrate on near
focus surfaces and can suffer lines induced by dirt on the references.)
Assuming that Minolta have not been stupid enough to break this golden
rule, then the black and white references themselves should not result
in this type of fine line structure, which leaves dirt and defects at
the sensor level.
During the calibration procedure the output of each cell in the CCD must
be very accurately determined since any error will be applied to all
subsequent outputs during the scan process and appear as lines long the
image. Since you can see line structure in the image much more readily
than random noise, this means that the calibration data must be
determined to much better accuracy than the noise floor of the CCD
itself and this is generally achieved by averaging many calibration
samples together. This usually improves the SNR of the calibration data
by the square root of the number of samples taken. If, for any reason,
the SNR of any CCD cell is poorer than the other cells in the device
then additional averaging is necessary to ensure that the quality of the
calibration data is maintained. Usually the number of samples averaged
in the calibration process is determined by the range of SNR that is
present on all of the CCD cells. Obviously, the smaller the cells the
greater the variability in response for a given silicon fab technology,
simply due to geometric variation from cell to cell. Hence smaller
cells generally require more averaging to prevent line structure in the
final image.
That is fine if the CCD behaves within its expected performance range,
However, it is possible for CCD cells to be degraded beyond the normal
expected range. This can happen if the cell is defective, becomes
defective as a result of temperature rises within the unit or if only a
small spec of dust reduces the amount of light falling on the cell. For
example, if dirt reduces the light level reaching the cell to one
quarter of what it should be, then the SNR will reduce by the same
amount and this will require 16x as many samples bring the SNR of the
calibration data back to normal.
If you follow this through, you can see that the number of samples
required to achieve the necessary quality of calibration rapidly
increases as the signal reduces or the noise in the CCD cell increases
above the average of the remaining cells. If insufficient samples are
obtained then the random error on the calibration will result in a
randomly dark or bright line from that cell throughout the entire scan.
To make matters worse, there are mechanisms in the CCD which give rise
to a noise with a 1/f to the power n frequency spectrum. These are
generally associated with defects resulting in charge traps in the bulk
of the silicon. The consequence of these are that after a certain
amount of samples, no further SNR improvement is obtained at all by
increasing the number of samples. In short, if the response of a
particular CCD cell is significantly deficient then you can rapidly
reach the point where it is impossible to achieve suitably good
calibration data to prevent it from producing a very visible line on the
scan.
The problem with all of this is that as the number of cells in the CCD
increases, to provide higher resolution scanners, the probability of a
rogue cell or two being present in the device increases dramatically. In
a 2-dimensional CCD, such as those used in digital cameras this problem
is overcome simply by concealing the output of these rogue cells with
the average of their nearest neighbours. This is not possible with a
scanner, since such concealment would be much more visible so
effectively every cell in the CCD has to perform almost perfectly. There
are ways around this issue, involving additional lines so that the
output of any defective cell can be replaced with the output from the
spare line, but these are clearly more expensive devices and I am not
aware of any being used in the current generation of commercial
scanners.
All of the above is consistent with the type of defects that Ricky has
posted in his scanned image, which is why I am fairly confident that
this is a CCD dirt/defect issue and not a film scratch, PSU noise or
processing defect.
Kennedy McEwen
<bvd...@no.spam> writes
>
>"Kennedy McEwen" <r...@nospam.demon.co.uk> wrote in message
>news:l3yowtDYZbm$Ew...@kennedym.demon.co.uk...
>SNIP
>> Reading some of the posts from others with similar problems, I do not
>> hold out much hope of success. Magically all scanners returned to
>> Minolta seem to work perfectly for them.
>
>I've read about similar experiences with Nikon service. Not what it used to
>be.
>
spell on their service engineers, Bart. Its all too common a problem
these days.
It would be interesting to get the Minolta view of the acceptability of
Ricky's scans - if they don't recognise the defect from his results then
it is a clear indication of the level of success he is likely to have
with other examples of the scanner.
Kennedy McEwen
>negative (the bright areas of final scans, eg skies and white tops
>etc) and the mid-dark to dark areas of slide scans. In the example I
>posted of a scan from Provia 100F I had to turn the gamma way up to to
>make the lines visible (it is a fairly dark image), so probably
>wouldn't be noticeable (but it's still unacceptable),
That is a contradiction, Ricky - if it isn't noticeable then it is
acceptable, but I can see your point that something even barely
noticeable is still unacceptable.
>but negatives
>are a completely different story. I scanned a negative of my Father
>standing on a sunny beach, and there were lines completely throughout
>the scan. It looked terrible! Recalibrating does not improve the
>situation, neither does multi-sampling (in fact it makes the lines
>worse).
>
That is consistent. Multisampling just reduces the random noise on the
image, making systematic noise such as this line structure much more
perceptible - and objectionable.
Some questions resulting from your observation above:
Although recalibration does not make the situation any better, does it
change the position of the lines or do they remain in exactly the same
place from scan to scan irrespective of the recalibration?
I am unfamiliar with the Minolta 5400 driver software, but does it have
an option to calibrate automatically prior to every scan? If so, do you
have that enabled? Also, if calibration is a manually initiated process
either by default or by the settings you have selected, how long after
the recalibration do you make the scan?
It may be that the internal temperature of the scanner is rising so that
the dark current at the time of the calibration is not the same as the
dark current at the time of the scan - resulting in erroneous
calibration data which is more significant on a few CCD cells.
>Funny thing is, the lines can never be seen in the preview
>image, either with Vuescan or with DiMAGE scan utility. Whether this
>is due to the low resolution of the preview image I don't know.
>
Almost certainly.
Marko B.
the day it was announced i noticed the lines problem twice.
Once by accident and once cause i was looking for them.
Both times the lines were not visible in 'normal' viewing situtation. When i
bumped up the middle tones to 4 in PS i noticed them.
The lines disapeared the first time i found them by accident and i tried
duplicate them with second scan - no luck.
Mine apear random.
Yesterday i scaned a few night slides and not one had those nasty lines.
I'm using driver from Minolta web page - 1.1.1?
I scanned a lot of slides and negatives with this thing and i never had a
problem. Used some fashion scans in fashion magazine without any problem.
Did you tried the scanner on different PC or MAC?
Good luck fixing your problem.
Marko
Bart van der Wolf
"Ed Hamrick" <use...@hamrick.com> wrote in message
news:bndhi3$frj$1...@ngspool-d02.news.aol.com...
SNIP
> I could figure out the problem if people would be more systematic
> about describing the problem.
The only times I have seen it myself, before 7.6.23/24, was in the low well
count areas (shadows in slides, highlight areas in negatives, IAW high film
density areas).
> It's confusing, because sometimes
> people find these lines in negative scans, and sometimes in positive
> scans, sometimes in the dark parts of scans and sometimes in the
> light parts.
Perhaps describing it in terms of film density/transparency areas would be
less ambiguous. The areas of occurence can even be described in density if
people press and hold the CTRL/Command key while hovering the mouse pointer
over those areas.
Bart
Ed Hamrick
> The lines always seem to be in the mid-dark to dark areas of the
> negative (the bright areas of final scans, eg skies and white tops
> etc) and the mid-dark to dark areas of slide scans.
Thanks for the info - I'll add this to my list of things
to look into.
> Funny thing is, the lines can never be seen in the preview
> image, either with Vuescan or with DiMAGE scan utility.
Hmm, I'll look into this. This might provide a clue.
One thing for you to check is whether the lines appear
in the Scan (not the Preview) when you have "Input|Lock exposure"
set and "Input|RGB exposure" set to 1.0.
If they don't, then I might need to add some logic that
changes the black point at each CCD element as a function
of the RGB exposure time. Right now, the black point is
assumed to be independent of RGB exposure time.
Regards,
Ed Hamrick
Bart van der Wolf
SNIP
> The lines always seem to be in the mid-dark to dark areas of the
> negative (the bright areas of final scans, eg skies and white tops
> etc) and the mid-dark to dark areas of slide scans.
This is more prominent with VueScan than with the Minolta Scan Utility in my
case (but nowhere near problematic with my SE-5400). Only the most dense
parts of film show a trace of irregularity, but that's partly due to a lower
S/N ratio.
The fact that VueScan (7.6.64) shows these patterns better is, IMHO,
probably due to an incorrect Black/White point slope determination at
calibration time, which causes a too low D-max and flat contrast at high
densities. If people start curves adjusting those areas, the low S/N shows
up more clearly than necessary. I've sent a more detailed report to Ed
Hamrick, so I'll await his findings.
The Minolta DiMAGE Scan Utility does not cause these issues with me, but it
is problematic (it clips) with negative scans. The more challenging slides
are not clipped. Go figure.
SNIP
> If this is a calibration issue, are there any great differences
> between the way DiMAGE scan utility calibrates and the way Vuescan
> calibrates? The lines do seem to be more noticeable when scanning with
> Vuescan.
Yes, see above.
> Funny thing is, the lines can never be seen in the preview
> image, either with Vuescan or with DiMAGE scan utility. Whether this
> is due to the low resolution of the preview image I don't know.
Probably a (lack of) resolution issue.
Bart
Alan Browne
Ricky wrote:
>>If it is electronic or sensor related it is most likely a manufacturing
>>fault at Minolta or one of their suppliers.
>>The P/S in the 5400 is fairly robust and would likely protect against
>>most surges, other than a lightning strike on the mains outside your house.
>>The fact that the stripes are in the same place suggests something is
>>wrong with the scanning guts (the sensor array, dirt there, or electronics)
>>
>>Send the critter back.
>
>
>
> I've just done a search of this group and there are LOADS of messages
> posted about this EXACT SAME problem with Minolta scanners. I refuse
> to believe that if I want to own a Minolta film scanner that I have to
> accept these lines as 'normal'. These lines render the scans
Can you read?
Where do I say it's 'normal'?
I said above SEND THE CRITTER BACK.
Alan.
Alan Browne
Kennedy McEwen wrote:
> In article <1b3b5b53.03102...@posting.google.com>, Ricky
> <ricky...@hays.com> writes
>
>>
>> I've just done a search of this group and there are LOADS of messages
>> posted about this EXACT SAME problem with Minolta scanners. I refuse
>> to believe that if I want to own a Minolta film scanner that I have to
>> accept these lines as 'normal'.
>
>
> Unfortunately, Ricky, if the problem is as common as your research
> indicates then I don't think you have any choice - if you STILL want to
> own a Minolta film scanner.
Horseshit. It is not 'normal', it is a defect. Nobody should accept
this performance. I have a 5400 and it works flawlessly, as do most
5400 owners ... the silent majority.
>
> I wholeheartedly agree, but your only recourse would appear to complain
> to Minolta themselves. :-(
>
> Reading some of the posts from others with similar problems, I do not
> hold out much hope of success. Magically all scanners returned to
What is your agenda? You seem to be on a crap-on-Minolta rant.
> Minolta seem to work perfectly for them. I can only suggest that their
Talk to Tony Spadaro about Nikon's charging him $300+ to tell him "no
fault found" after 3 months.
> QA inspectors are either completely incompetent or are working to an
> extremely poor specification, if it even includes test and acceptance
> criteria for scan lines at all.
>
> This is a sorry state of affairs, because this does appear to be a
> scanner which has so many other things going for it that it would be
> almost criminal for Minolta to let poor QA ruin its reputation.
Fact is, theses are complex devices, and fact is, some will pass tests
at the factory and then manifest problems after delivery.
Alan.
William D. Tallman
< snip exactly the sort of post that makes this venue so valuable: real
information>
Kudos, sir!
It would seem, then, that Minolta has a tiger by the tail here, and may get
bitten in the process! Ed says that Avision scanners are most prone to
this type of thing, which suggests that it is in fact the CCD array itself
that is the problem. I think that if dust were the problem, we'd see it
much more often; dust gets in everywhere, and if a scanner design is such
that dust would cause this, we'd all experience it at some point or other.
I would think that this is a QC problem, not with Minolta, but with Avision
(or the manufacturer of the array, if not Avision). Here's why: We're
just now hearing about this problem. If it were inherent in the scanner
design, it would have been noted from the beginning, but that's apparently
not the case. The DSE5400 has arguably been a runaway success for Minolta
for obvious reasons, and they may be taxing the ability of the array
supplier(s) to ship dependable product.
You note that there are other causes for this phenomenon, and it seems to me
that any one of them could find its way past a harried QC section. So I
reject the dust explanation as it does not satisfy the evidence, and
suggest that it is in fact Minolta's problem, although probably not of
their causing.
Hope Ed gets a handle on some of this, although I suspect that the law of
diminishing returns will keep him from applying a complete solution. Ergo,
it seems that Ricky must needs persevere in his insistence on getting an
acceptable machine from Minolta.
Wonder if anyone at Minolta is monitoring this newsgroup! If not, they
should.
Hmmm.. glad I got a good machine <offers a prayer of thanks to the patron
saint of film scanners for his good luck....>, lol!!!
Bill Tallman
Kennedy McEwen
<"Alan Browne"@videotron.canospam> writes
>
>
>Kennedy McEwen wrote:
>
>> In article <1b3b5b53.03102...@posting.google.com>, Ricky
>><ricky...@hays.com> writes
>>
>>>
>>> I've just done a search of this group and there are LOADS of messages
>>> posted about this EXACT SAME problem with Minolta scanners. I refuse
>>> to believe that if I want to own a Minolta film scanner that I have to
>>> accept these lines as 'normal'.
>> Unfortunately, Ricky, if the problem is as common as your research
>>indicates then I don't think you have any choice - if you STILL want
>>to own a Minolta film scanner.
>
>Horseshit. It is not 'normal', it is a defect.
Never really excelled in the logic class, did you Alan?
Read what I wrote: "IF... then if... then...". That's a doubly
conditional statement!
>Nobody should accept this performance.
Nobody is suggesting he should - a conditional statement is NOT an
unconditional instruction!
> I have a 5400 and it works flawlessly, as do most 5400 owners ... the
>silent majority.
>
Whoop-de-do!
A totally unhelpful "I'm all right so stuff the rest of you" attitude!
Sufficient people have reported this problem publicly that there is no
longer any question that it is a common issue. As Ricky has already
stated, he HAS returned samples of this scanner already - TWICE!! Whilst
nobody should accept this level of performance, only a rabid
Minoltaphile, such as yourself, would consider that continuously
returning goods to Minolta is acceptable either! Hence the final "IF"
in the paragraph you quoted above!
>
>What is your agenda? You seem to be on a crap-on-Minolta rant.
>
Stop letting emotions get in the way of fact and logic! The fact is
that the scanner has problems and they are not being recognised by
Minolta - if that continues then the scanner's deserved reputation will
be damaged. That is not an anti-Minolta rant, but a statement of fact!
Simply sweeping the problem under the carpet by continuously returning
the scanner until you get one that works is just making the situation
worse. Inevitably more faulty scanners remain in circulation and the
odds of a new buyer getting a faulty one increases. The only solution
is for Minolta to recognise and deal with the problem. but if everyone
behaves like a rabid Minoltaphile then they never will.
>
>Talk to Tony Spadaro about Nikon's charging him $300+ to tell him "no
>fault found" after 3 months.
>
Some people would like to charge TS a lot more than $300 over 3 months
just for breathing rights, or at least having to listen to his rants, so
I wouldn't be too quick to criticize Nikon for that! Had you been
around in these parts for any reasonable time you would find plenty of
criticism of Nikon's scanners and software too! However Nikon does not
appear in the subject line of this thread and it is therefore of little
value criticising their deficiencies within it!
>
>Fact is, theses are complex devices, and fact is, some will pass tests
>at the factory and then manifest problems after delivery.
>
Again, you let emotion get in the way of logic. Anyone who has been
involved in setting up a production facility will tell you that the
lowest cost tests are those which detect and correct failures at the
earliest possible stage of manufacture - before further value is
invested in the manufacture of the product. Faults detected at the
final test before shipping are the most expensive faults in any
production line, because the entire value of the unit is at stake -
faults detected AFTER shipping cost even more!
The fact is that scanners are NOT particularly complex devices - they
have been engineered to be very simple to manufacture in high volume at
low cost and to retail with a significant margin. That margin is
severely dented by unnecessary field repair and support costs due to
faulty product being shipped. This applies just as much to Minolta as
it does to every manufacturer and was recognised as long ago as Henry
Ford's first production line for the Model T, which was a damn sight
more complex than a film scanner in real terms - assembly from the
bought in parts. Some faults inevitably get out the door - the whole
point of modern production methods is to MINIMISE the number, and hence
the cost, of those that do!
Ricky, for example, has already returned two scanners without Minolta
making a single cent of profit, in fact they are currently making a
considerable loss servicing this *potential* sale alone! Even if his
third sample of this scanner proves to be perfect and remains so
throughout its warranty period, Minolta will probably never make any
profit at all on the deal now. If the third sample proves the equal of
the first two and is also rejected then Minolta need to sell a lot of
good scanners just to recover the loss! Perhaps you should calculate
how many scanners Minolta need to successfully sell for their deliveries
to qualify as a "SIX SIGMA" supplier as a consequence of the two faulty
samples that are the subject of this thread - and that ignores the
others that have been identified with the same problem!
Kennedy McEwen
<wtal...@olypen.com> writes
>I think that if dust were the problem, we'd see it
>much more often; dust gets in everywhere, and if a scanner design is such
>that dust would cause this, we'd all experience it at some point or other.
>
Remember that this is dust would have to be on the focal plane of the
CCD itself - inside what should be a completely sealed semiconductor
package. Unless the dust is in there at manufacture then you would
never see it at all.
>
>You note that there are other causes for this phenomenon, and it seems to me
>that any one of them could find its way past a harried QC section. So I
>reject the dust explanation as it does not satisfy the evidence,
In light of the above, I suggest it does.
> and
>suggest that it is in fact Minolta's problem, although probably not of
>their causing.
>
That would be consistent with the evidence and my suggestion of
dust/defects on the sensor.
>Ergo,
>it seems that Ricky must needs persevere in his insistence on getting an
>acceptable machine from Minolta.
>
If it was me, they would be on their last chance - three counts and out!
;-)
Olaf Ulrich
Alan Browne
William D. Tallman wrote:
> Kennedy McEwen wrote:
>
> < snip exactly the sort of post that makes this venue so valuable: real
> information>
>
> Kudos, sir!
>
> It would seem, then, that Minolta has a tiger by the tail here, and may get
> bitten in the process! Ed says that Avision scanners are most prone to
> this type of thing, which suggests that it is in fact the CCD array itself
> that is the problem. I think that if dust were the problem, we'd see it
> much more often; dust gets in everywhere, and if a scanner design is such
> that dust would cause this, we'd all experience it at some point or other.
Dust getting into the scanner and landing on a lens, as has been pointed
out, would likely go unnoticed in 99.99% of scans. Dust sealed into the
linear array against or very close to the sensor during a manufacturing
step, would affect that one scanner on every scan.
>
> I would think that this is a QC problem, not with Minolta, but with Avision
> (or the manufacturer of the array, if not Avision). Here's why: We're
> just now hearing about this problem. If it were inherent in the scanner
> design, it would have been noted from the beginning, but that's apparently
> not the case. The DSE5400 has arguably been a runaway success for Minolta
> for obvious reasons, and they may be taxing the ability of the array
> supplier(s) to ship dependable product.
Yes.
>
> You note that there are other causes for this phenomenon, and it seems to me
> that any one of them could find its way past a harried QC section. So I
> reject the dust explanation as it does not satisfy the evidence, and
> suggest that it is in fact Minolta's problem, although probably not of
> their causing.
If they contract for the design and manufacture of a product that they
sell, it is of course Minolta's problem.
>
> Hope Ed gets a handle on some of this, although I suspect that the law of
> diminishing returns will keep him from applying a complete solution. Ergo,
> it seems that Ricky must needs persevere in his insistence on getting an
> acceptable machine from Minolta.
I don't believe Ed should waste his time on it.
I believe Ricky will get a good machine, he's just been a tad unlucky.
MOST people we hear about have received good scanners.
I have been a Minolta customer for 10+ years and posess 22 'value'
pieces of Minolta photography equipment (n/i odds and ends), all bought
new except for two lenses. I have NEVER had a defective Minolta product
'out of the box' (or used) and I've had one piece of Minolta equipment
fail a couple months after the warranty expired. Minolta (Toronto)
repaired that free of charge in any case. (5400HS flash).
I don't baby my equipment. I'm very confident in Minolta quality.
>
> Wonder if anyone at Minolta is monitoring this newsgroup! If not, they
> should.
I have had mixed success with direct e-mails to Minolta Toronto. Often
blatently bad replies (don't understand, don't know, don't care),
sometimes bang on good advice.
Steve Stanford
I had the 5400 and scanned many hundreds of transparancies and negatives
without problem. Then the ambient temperature went to over 30C. At this
point ALL scans where unusable - dozens of scan lines in the direction of
scan. The original 5400 was replaced and the problem has not recurred.
Hence, I wonder whether the posters are in a high temperature environment
or not.
This conculsion may be totally in error but it is my observaton.
Alan Browne
Well, their temperature is certainly rising...!
What you say does go to some theories relating to electronics deficiencies.
William D. Tallman
> In article <vplk96m...@corp.supernews.com>, William D. Tallman
> <wtal...@olypen.com> writes
>
>>I think that if dust were the problem, we'd see it
>>much more often; dust gets in everywhere, and if a scanner design is such
>>that dust would cause this, we'd all experience it at some point or other.
>>
> Remember that this is dust would have to be on the focal plane of the
> CCD itself - inside what should be a completely sealed semiconductor
> package. Unless the dust is in there at manufacture then you would
> never see it at all.
ARGHHH!!!! Massive wetware failure!!!
Of course! This is a problem I've foreseen for a while now: too many
assembly setups are using laminar flow in place of clean-room technology.
Obviously the former is much much less expensive to operate and maintain.
It simply doesn't do the job when the "chips are down".. hehheh....
This was part of the change taking place during my last contract (HPSJ), and
the engineers had this quiet little pool on how long it would be before the
failure rate started to climb. PMs ignored the pool and DMs didn't know.
I didn't stay long enough to see it completed and part of the production,
so have no idea what happened. But given HP's record for proactive QC, I
bet the whole thing went offshore where clean rooms are a lot cheaper....
So I think I may be correct about the mechanism here. You've shown that the
dirt problem is most likely a manufacturing problem. Perhaps there is a
single lot of scanners out there with arrays from a single bad batch that
got through QC.
And you're right: scanners are rather simple devices, especially compared
to some of the equipment it appears we've both seen and used.
Another post in this thread suggested, or cited the suggestion, that the
arrays may have a small thermal load tolerance... wonder if they have an
"elephant" running in the QC section. [An 'elephant' is a test system that
provides a calibrated temperature environment, usually consisting of a
rather long hose that drops down to cover the DUCT].
If you can confirm that this may be a factor, us 5400-o-piles may be well
advised to watch that part of our operating environment. I've no problem,
as it rarely gets anywhere close to the suggested temperature (30C?), but
others might. Comments?
Bill Tallman
Ed Hamrick
> Remember that this is dust would have to be on the focal plane of the
> CCD itself - inside what should be a completely sealed semiconductor
> package. Unless the dust is in there at manufacture then you would
> never see it at all.
It's unlikely to be dust - dust spots wouldn't be a single pixel wide.
It's probably something to do with the way the CCD responds
under different exposure time conditions, and probably it has
to do with the black point of the CCD.
Regards,
Ed Hamrick
Ricky
> Although recalibration does not make the situation any better, does it change the position of the lines or do they remain in exactly the same place from scan to scan irrespective of the recalibration?
The lines do appear to be in the same position each time irrespective
of re-calibration. If there is a shift then it must only be a couple
of pixels at most.
The lines in this scanner (and the previous scanner) get worse and
worse with every scan as the scanner gets warmer.
From Ed H
> One thing for you to check is whether the lines appear in the Scan (not the Preview) when you have "Input|Lock exposure" set and "Input|RGB exposure" set to 1.0.
> If they don't, then I might need to add some logic that changes the black point at each CCD element as a function of the RGB exposure time. Right now, the black point is assumed to be independent of RGB exposure time.
Here’s what I did (not sure if it was right or not). 1. Preview
2. Check ‘Lock Exposure’ 3. Changed RGB exposure to 1.0
(left IR exposure) 4. Preview 5. Scan. I’m presuming this would
be the same as turning off ‘Auto-exposure for slides’ in
DiMAGE scan utility. The lines were still there. I un-checked
‘Lock Exposure’ and then did another preview and scan and
the lines were in exactly the same places.
One thing I did notice however, is that the lines are not only much
more prominent when scanning with Vuescan, the lines are visible in a
much wider tonal range. The lines are even visible in mid-tone and
highlight areas. The same slide scanned with DiMAGE scan utility the
lines are visible ONLY in the dark areas.
From Bill T:
> Ergo, it seems that Ricky must needs persevere in his insistence on getting an acceptable machine from Minolta.
I am disappointed that I have received two defective scanners (I
should have expected it though – I am the worlds unluckiest
man!). I will persevere with this though as I believe that I will
(eventually) get a properly working machine, and there are obviously
people out there very satisfied with this scanner.
I would point out that the ambient temperature goes no-where near 30C,
maybe 22C at most.
Again, my thanks go out to everyone who has posted a response. I am
expecting the new unit tomorrow. I’ll post again and let
everyone know how it goes (fingers crossed eh?)
Ricky.
Kennedy McEwen
<s...@stanford-uk.com> writes
>Looking at all the posts leads me to ponder tho following.
>
>I had the 5400 and scanned many hundreds of transparancies and negatives
>without problem. Then the ambient temperature went to over 30C. At this
>point ALL scans where unusable - dozens of scan lines in the direction of
>scan. The original 5400 was replaced and the problem has not recurred.
Are you saying that the scanner continued to produce scan lines even
after the ambient temperature fell below 30C again or that you replaced
it before the temperature returned to a normal level?
Kennedy McEwen
>
>> Although recalibration does not make the situation any better, does
>>it change the position of the lines or do they remain in exactly the
>>same place from scan to scan irrespective of the recalibration?
>
>The lines do appear to be in the same position each time irrespective
>of re-calibration. If there is a shift then it must only be a couple
>of pixels at most.
>
>The lines in this scanner (and the previous scanner) get worse and
>worse with every scan as the scanner gets warmer.
>
black level uniformity degrades as a consequence. However, calibration
should correct for this - that is what it is there for.
We are back to either a defective calibration or defective pixels.
Steve Stanford
$Ew...@kennedym.demon.co.uk:
> In article <106712453...@lotis.uk.clara.net>, Steve Stanford
> <s...@stanford-uk.com> writes
>>Looking at all the posts leads me to ponder tho following.
>>
>>I had the 5400 and scanned many hundreds of transparancies and negatives
>>without problem. Then the ambient temperature went to over 30C. At this
>>point ALL scans where unusable - dozens of scan lines in the direction of
>>scan. The original 5400 was replaced and the problem has not recurred.
>
> Are you saying that the scanner continued to produce scan lines even
> after the ambient temperature fell below 30C again or that you replaced
> it before the temperature returned to a normal level?
The scanner always had severe scan lines following the initial occurrence
during a period of high temperature. Of course the replacement has not had
to endure anything like that temperature since.
The scan lines actually started partway through a scan - that is the scan
was perfect and then the scan lines began to be evident throughout the rest
of the scan and every scan following.
Alan Browne
Kennedy McEwen wrote:
Ricky wrote:
*** I refuse to believe that if I want to own a Minolta film scanner that
*** I have to accept these lines as 'normal'.
>>> Unfortunately, Ricky, if the problem is as common as your research
>>> indicates then I don't think you have any choice - if you STILL want
>>> to own a Minolta film scanner.
>>
>>
>> Horseshit. It is not 'normal', it is a defect.
>
>
> Never really excelled in the logic class, did you Alan?
>
> Read what I wrote: "IF... then if... then...". That's a doubly
> conditional statement!
You put up a conditional statement predicated on a problem being
'normal'. No logic in that.
>
>> Nobody should accept this performance.
>
>
> Nobody is suggesting he should - a conditional statement is NOT an
> unconditional instruction!
See above.
>
>> I have a 5400 and it works flawlessly, as do most 5400 owners ... the
>> silent majority.
>>
> Whoop-de-do!
>
> A totally unhelpful "I'm all right so stuff the rest of you" attitude!
No. And again, my advice to Ricky was to send the unit back. Just
pointing out that you have no statistics here, so you have no idea what
the real number of sold units is, no idea how many defective units have
been identified, or the nature of all defects. This scanner is
extremely popular and selling pretty hot. Not all scanner users (by a
great long long long shot) participate in NGs. So you don't know squat
either way. Or if you do, please put put up the numbers.
>
> Sufficient people have reported this problem publicly that there is no
> longer any question that it is a common issue. As Ricky has already
remember that only those with problems report problems. See above.
Unless you can post numbers you don't know if it is common, rare or in
between.
> stated, he HAS returned samples of this scanner already - TWICE!! Whilst
Indeed, and that is unacceptable and leave it at that. Personnaly I
would be every bit as upset as Ricky is; probably more. But your
"lather it up" attitude is suspicious.
TWICE! could be related to many things including a bad "batch" (if it is
produced in batches, which is a disappering trend) or just plain bad luck.
> nobody should accept this level of performance, only a rabid
> Minoltaphile, such as yourself, would consider that continuously
> returning goods to Minolta is acceptable either! Hence the final "IF"
> in the paragraph you quoted above!
Where the hell did I say it was acceptable? My advice to ricky was to
send it back. In my case, after two units, I would have gotten my
money back, period and gotten the Nikon.
>
>>
>> What is your agenda? You seem to be on a crap-on-Minolta rant.
>>
> Stop letting emotions get in the way of fact and logic! The fact is
It's not emotion, it is your blatent attack on Minolta that appears very
strange. Using terminology like "rabid Minoltaphile" above certainly
doesn't help.
> that the scanner has problems and they are not being recognised by
> Minolta - if that continues then the scanner's deserved reputation will
Do you have factual knowledge that it is being ignored?
> be damaged. That is not an anti-Minolta rant, but a statement of fact!
> Simply sweeping the problem under the carpet by continuously returning
> the scanner until you get one that works is just making the situation
> worse. Inevitably more faulty scanners remain in circulation and the
I would hope that with any returns Minolta are getting on this or any
product that they will be investigating the source of the problem and
rectifying it. I have the confidence that Minolta will do so. Feedback
is not instant when a customer returns a defective product. We hope it
is fast, but it is never instant.
> odds of a new buyer getting a faulty one increases. The only solution
> is for Minolta to recognise and deal with the problem. but if everyone
> behaves like a rabid Minoltaphile then they never will.
Again, my advice was to return the scanner.
You are misconstruing my statements which are simply to say that Minolta
is a responsible supplier and I suspect no more happy about this problem
than Ricky.
>
>>
>> Talk to Tony Spadaro about Nikon's charging him $300+ to tell him "no
>> fault found" after 3 months.
>>
> Some people would like to charge TS a lot more than $300 over 3 months
> just for breathing rights, or at least having to listen to his rants, so
> I wouldn't be too quick to criticize Nikon for that! Had you been
> around in these parts for any reasonable time you would find plenty of
> criticism of Nikon's scanners and software too! However Nikon does not
> appear in the subject line of this thread and it is therefore of little
> value criticising their deficiencies within it!
I put it there as a fair example that perfection does not exist in this
industry, even with the supposed top-dog. That you use what I said to
turn around and attack someone else says a lot about your character.
>
>>
>> Fact is, theses are complex devices, and fact is, some will pass tests
>> at the factory and then manifest problems after delivery.
>>
> Again, you let emotion get in the way of logic. Anyone who has been
> involved in setting up a production facility will tell you that the
> lowest cost tests are those which detect and correct failures at the
> earliest possible stage of manufacture - before further value is
> invested in the manufacture of the product. Faults detected at the
> final test before shipping are the most expensive faults in any
> production line, because the entire value of the unit is at stake -
> faults detected AFTER shipping cost even more!
I will tell you that the first principle is that manufacturing that
depends on testing (at any level) to root out quality problems is a
design-for-manufacturing disaster. I agree that earlier found defects
prevent costly rework or rectification; but the *real* way to do it is
to design the product and processes to be defect free.
The best known (to me) example is that most Sony televisions are not
even turned on at the factory. The first power-on test is by the
customer. (I don't know if sub-assy's are tested with power on).
>
> The fact is that scanners are NOT particularly complex devices - they
Compared to what? I don't know the parts count but being:
-optical
-elctromechanical (in at least 2 axis)
-electronics
is *is* more complex than, say, a radio with no moving parts and
nothing requiring precise alignment or ability to self-align.
> have been engineered to be very simple to manufacture in high volume at
> low cost and to retail with a significant margin. That margin is
> severely dented by unnecessary field repair and support costs due to
> faulty product being shipped. This applies just as much to Minolta as
> it does to every manufacturer and was recognised as long ago as Henry
> Ford's first production line for the Model T, which was a damn sight
> more complex than a film scanner in real terms - assembly from the
> bought in parts. Some faults inevitably get out the door - the whole
> point of modern production methods is to MINIMISE the number, and hence
> the cost, of those that do!
Henry's process had its greatest impact in manufacturing cost, not quality.
Been there, done that. Seems you have too. And so you should recognize
that new products when introduced typically DO have a higher return rate
than a product that has been on the floor for many months. Where I
worked a few years ago, despite our best efforts with
cross-company/customer/supplier teams on several products, all had
initial in-service fail rates higher than once the line was 'matured'.
>
> Ricky, for example, has already returned two scanners without Minolta
> making a single cent of profit, in fact they are currently making a
> considerable loss servicing this *potential* sale alone! Even if his
> third sample of this scanner proves to be perfect and remains so
> throughout its warranty period, Minolta will probably never make any
> profit at all on the deal now. If the third sample proves the equal of
On Ricky's 'deal' probably not. However for every Ricky there are 50,
100 or 500 who have had 0 problems. (I don't know the number, but please
publish it if you have it).
> the first two and is also rejected then Minolta need to sell a lot of
> good scanners just to recover the loss! Perhaps you should calculate
> how many scanners Minolta need to successfully sell for their deliveries
> to qualify as a "SIX SIGMA" supplier as a consequence of the two faulty
> samples that are the subject of this thread - and that ignores the
> others that have been identified with the same problem!
Let's take your last sentence: You don't know how many are sold. You
don't know how many have defects. NGs are not databases.
Given the moderate complexity of the product I'd guess each unit has on
the order of 1000 defect opportunities ... most of these opportunities
being at lower tier supplier level. In that case 1 unit in 300 being
defective would meet six-sigma. But, like you, I have no real knowledge
of the defect opportunity number for the product. In any case, this is
false analysis as the types of defects that are being reported seem to
be in the optical area. That would make the number of defect
opportunities in that area much lower and much more costly to Minolta
(and the supplier(s)). All this before considering the unit cost and
price when delivered.
BTW: Six-Sigma is a process, not a qualification. A company can declare
that a product, a product line, a production facility, etc. has reached
six-sigma defect rates; but there is no certifying 'agency' other than
the statistics the company gathers on its own behalf. (Compare to say
ISO 9000 where a registrar audits your processes and asserts your
qualification). You could, of course, get an outside agency to endorse
or certify your numbers through document inspection, indeed upper
management or the board of D's might insist on it.
Alan.
Kennedy McEwen
<"Alan Browne"@videotron.canospam> writes
>
>
>Kennedy McEwen wrote:
>
>Ricky wrote:
>*** I refuse to believe that if I want to own a Minolta film scanner that
>*** I have to accept these lines as 'normal'.
>
>>>> Unfortunately, Ricky, if the problem is as common as your
>>>>research indicates then I don't think you have any choice - if you
>>>>STILL want to own a Minolta film scanner.
>>>
>>>
>>> Horseshit. It is not 'normal', it is a defect.
>> Never really excelled in the logic class, did you Alan?
>> Read what I wrote: "IF... then if... then...". That's a doubly
>>conditional statement!
>
>You put up a conditional statement predicated on a problem being
>'normal'. No logic in that.
>
again, READ what was written: "IF...THEN...". NOTHING predicates either
qualifier in the statement made:
IF Ricky still wants to own such a unit
AND
IF the problem is as common as it appears
THEN I don't think he has any choice but to accept the problem.
The IF statements are absolutely conditional under all interpretations
EXCEPT that the reader is a logic deficient rabid Minoltaphile!
>
>Unless you can post numbers you don't know if it is common, rare or in
>between.
>
Common does not mean that it is a majority or even a large proportion of
owners who suffer from the problem. It means that the problem exists in
sufficient numbers to be recognised as a specific fault type - as
opposed to a random series of unconnected faults from several different
underlying causes. Further, the fact that few scanner owners use news
access does not necessarily skew the statistics in either direction, it
merely samples them. A similar argument can be made that the number of
satisfied Minolta 5400 buyers has been enhanced by its positive recent
discussions on newsgroups, since many have never actually owned a film
scanner in their life and are more inclined to accept such deficiencies
as normal because the "experts" on the groups say it is the best thing
since sliced bread!
>
>> nobody should accept this level of performance, only a rabid
>>Minoltaphile, such as yourself, would consider that continuously
>>returning goods to Minolta is acceptable either! Hence the final "IF"
>>in the paragraph you quoted above!
>
>Where the hell did I say it was acceptable?
By continually promoting the same process loop without any breakout
clause - faulty scanner, send it back, faulty scanner, send it back...
He's already been round the loop three times (as of today) yet you still
offer no alternative but to accept the process.
> My advice to ricky was to send it back. In my case, after two
>units, I would have gotten my money back, period and gotten the Nikon.
>
So you are suggesting less tolerance of Minolta than I was! I said
three counts and out - you are suggesting that even two would switch you
to the competition. Just what is YOUR agenda?
>
>It's not emotion, it is your blatent attack on Minolta that appears
>very strange.
Again, I suggest you read my post again. There has been NO blatant
attack, as you put it. In fact, the final paragraph of my original post
that you took so much offence to makes an implicit assumption that the
features offered by this MINOLTA product are significant and states that
IMO they should not be permitted to be undermined by poor QA.
What is strange is your overprotection of one particular manufacturer,
one that you just so happen to have made significant investment with.
Given your own statements of product ownership, you do not appear to
have any capacity for INDEPENDENT comment on this matter whatsoever.
>
>
>> that the scanner has problems and they are not being recognised by
>>Minolta - if that continues then the scanner's deserved reputation will
>
>Do you have factual knowledge that it is being ignored?
>
From the quotes of those who have returned the scanner to Minolta - no
fault, but it is still there when it returns. Did you read any of the
reports that were referenced?
>
>I would hope that with any returns Minolta are getting on this or any
>product that they will be investigating the source of the problem and
>rectifying it. I have the confidence that Minolta will do so. Feedback
>is not instant when a customer returns a defective product.
But when feedback is negative then it indicates nothing is being done!
>
>I will tell you that the first principle is that manufacturing that
>depends on testing (at any level) to root out quality problems is a
>design-for-manufacturing disaster. I agree that earlier found defects
>prevent costly rework or rectification; but the *real* way to do it is
>to design the product and processes to be defect free.
>The best known (to me) example is that most Sony televisions are not
>even turned on at the factory.
Rubbish! Having visited Sony factories they have a conveyor belt of TVs
passing in front of, mainly young women, each checking the picture
quality. Incidentally, one of the standard tools in their final test is
actually a rubber mallet which is used to tap the side of the unit if
the picture has significant line structure! The trinitron tube design
is prone to the wires in the aperture grille sticking during manufacture
and a sharp tap with the mallet reduces the number of customers who
receive badly lined TVs - even though the installation instructions warn
of this potential problem and the solution!
>The first power-on test is by the customer. (I don't know if
>sub-assy's are tested with power on).
>
Of course they are tested with power on - that is in keeping with the
intent to weed out defects while they still have minimum value added.
>> The fact is that scanners are NOT particularly complex devices -
>>they
>
>Compared to what? I don't know the parts count but being:
> -optical
> -elctromechanical (in at least 2 axis)
> -electronics
> is *is* more complex than, say, a radio with no moving parts and
>nothing requiring precise alignment or ability to self-align.
>
Few radios are in the same price bracket as these scanners and those
which are have considerably more complexity in their electronic design
and manufacture! Compared to, say a 35mm SLR of similar cost, they are
exceedingly simple devices.
>> have been engineered to be very simple to manufacture in high volume
>>at low cost and to retail with a significant margin. That margin is
>>severely dented by unnecessary field repair and support costs due to
>>faulty product being shipped. This applies just as much to Minolta as
>>it does to every manufacturer and was recognised as long ago as Henry
>>Ford's first production line for the Model T, which was a damn sight
>>more complex than a film scanner in real terms - assembly from the
>>bought in parts. Some faults inevitably get out the door - the whole
>>point of modern production methods is to MINIMISE the number, and
>>hence the cost, of those that do!
>
>Henry's process had its greatest impact in manufacturing cost, not quality.
>
Quality IS cost! Quality is consistent performance within
specification. Failure to meet specification after delivery results in
additional cost. You sound like a novice who cannot distinguish between
quality and prestige or cachet!
>
>Let's take your last sentence: You don't know how many are sold. You
>don't know how many have defects.
Neither do you - you ASSUME a silent majority!
> NGs are not databases.
>
No, but they are a sample of the database.
>Given the moderate complexity of the product I'd guess each unit has on
>the order of 1000 defect opportunities ...
Err, only a few lines back you were extolling the virtues of design for
zero defects... what a shame Minolta didn't rely on you to set up their
manufacture facilities - they would reduce the defects considerably -
but probably never brought the scanner to market!
>Compare to say ISO 9000 where a registrar audits your processes and
>asserts your qualification
ISO9000 audits your process but does NOTHING to quantify the process or
the reliability of the final product. The process can virtually random
assembly with almost 100% return rate, provided it is documented as such
and that such returns are expected. In short, you can have a process
which assembles garbage and, provided that the process is documented and
enforced you can still achieve ISO9000 certification. Like many of the
so called quality initiatives it is yet another example waiting for
someone to point out its nakedness.
Kennedy McEwen
>Kennedy McEwen <r...@nospam.demon.co.uk> wrote in news:O6H218ILW8m
>$Ew...@kennedym.demon.co.uk:
>
>> In article <106712453...@lotis.uk.clara.net>, Steve Stanford
>> <s...@stanford-uk.com> writes
>>>Looking at all the posts leads me to ponder tho following.
>>>
>>>I had the 5400 and scanned many hundreds of transparancies and negatives
>>>without problem. Then the ambient temperature went to over 30C. At this
>>>point ALL scans where unusable - dozens of scan lines in the direction of
>>>scan. The original 5400 was replaced and the problem has not recurred.
>>
>> Are you saying that the scanner continued to produce scan lines even
>> after the ambient temperature fell below 30C again or that you replaced
>> it before the temperature returned to a normal level?
>
>The scanner always had severe scan lines following the initial occurrence
>during a period of high temperature. Of course the replacement has not had
>to endure anything like that temperature since.
>
>The scan lines actually started partway through a scan - that is the scan
>was perfect and then the scan lines began to be evident throughout the rest
>of the scan and every scan following.
That certainly sounds like something a lot more serious than temperature
dependent dark current. Its almost like some element of the scanner is
actually being irrecoverably damaged by temperature. There are a couple
of possibilities, but I find it surprising that this would occur on,
what is designed to be essentially, a domestic product.
Alan Browne
Kennedy McEwen wrote:
> Demonstrating once again that you cannot follow any form of logic! Once
> again, READ what was written: "IF...THEN...". NOTHING predicates either
> qualifier in the statement made:
> IF Ricky still wants to own such a unit
> AND
> IF the problem is as common as it appears
> THEN I don't think he has any choice but to accept the problem.
All you are demonstrating is a propensity to argue. The context was
what the context was. Wiggle all you like and snip out text all you
like: you will be correct to your self only.
>
> The IF statements are absolutely conditional under all interpretations
> EXCEPT that the reader is a logic deficient rabid Minoltaphile!
Who's emotional now?
>
>>
>> Unless you can post numbers you don't know if it is common, rare or in
>> between.
>>
> Common does not mean that it is a majority or even a large proportion of
[common > rare] regardless of the proportion assigned. W/O data, you
just can't say.
> owners who suffer from the problem. It means that the problem exists in
> sufficient numbers to be recognised as a specific fault type - as
> opposed to a random series of unconnected faults from several different
> underlying causes. Further, the fact that few scanner owners use news
> access does not necessarily skew the statistics in either direction, it
> merely samples them. A similar argument can be made that the number of
Again, without real data, you don't know sqaut.
> satisfied Minolta 5400 buyers has been enhanced by its positive recent
> discussions on newsgroups, since many have never actually owned a film
> scanner in their life and are more inclined to accept such deficiencies
> as normal because the "experts" on the groups say it is the best thing
> since sliced bread!
There are certainly other things besides scanners that are the best
things sliced bread.
Those 'experts' who have the scanner and have seen no problems, indeed
great performance should say so.
>
>>
>>> nobody should accept this level of performance, only a rabid
>>> Minoltaphile, such as yourself, would consider that continuously
>>> returning goods to Minolta is acceptable either! Hence the final
>>> "IF" in the paragraph you quoted above!
>>
>>
>> Where the hell did I say it was acceptable?
>
>
> By continually promoting the same process loop without any breakout
> clause - faulty scanner, send it back, faulty scanner, send it back...
> He's already been round the loop three times (as of today) yet you still
> offer no alternative but to accept the process.
The word "acceptable" was in the context of the performance of the
particular scanner.
He has been around the loop twice. I haven't seen any posting of his
with results with a third scanner. I would also suggest that if he has
similar problems with a third scanner, then perhaps the problem is not
really in the scanners. I also hope he has tracked serial numbers ...
not to sound suspicious, but it does happen (regardless of manuf.) that
you get back what you returned by accident or otherwise.
>
>> My advice to ricky was to send it back. In my case, after two
>> units, I would have gotten my money back, period and gotten the Nikon.
>>
> So you are suggesting less tolerance of Minolta than I was! I said
> three counts and out - you are suggesting that even two would switch you
> to the competition. Just what is YOUR agenda?
I was just aluding to -my- nature which can be as impatient as anyone
else in certain circumstances. When you plunk down a few days pay on
something, you expect it to work. I do, ' expect you do, Ricky
certainly does.
>
>>
>> It's not emotion, it is your blatent attack on Minolta that appears
>> very strange.
>
>
> Again, I suggest you read my post again. There has been NO blatant
> attack, as you put it. In fact, the final paragraph of my original post
> that you took so much offence to makes an implicit assumption that the
> features offered by this MINOLTA product are significant and states that
> IMO they should not be permitted to be undermined by poor QA.
>
> What is strange is your overprotection of one particular manufacturer,
> one that you just so happen to have made significant investment with.
> Given your own statements of product ownership, you do not appear to
> have any capacity for INDEPENDENT comment on this matter whatsoever.
>
>>
>>
>>> that the scanner has problems and they are not being recognised by
>>> Minolta - if that continues then the scanner's deserved reputation will
>>
>>
>> Do you have factual knowledge that it is being ignored?
>>
> From the quotes of those who have returned the scanner to Minolta - no
> fault, but it is still there when it returns. Did you read any of the
> reports that were referenced?
No, as a matter of fact. When Ricky posted his image, I and others
suggested that that image be returned with the unit along with a precise
explanation of how he got his results. I also understood Ricky's 2nd
unit to be exchanged, not the same one returned NFF.
>
>>
>> I would hope that with any returns Minolta are getting on this or any
>> product that they will be investigating the source of the problem and
>> rectifying it. I have the confidence that Minolta will do so.
>> Feedback is not instant when a customer returns a defective product.
>
>
> But when feedback is negative then it indicates nothing is being done!
Again, was the unit replaced, or returned?
>
>>
>> I will tell you that the first principle is that manufacturing that
>> depends on testing (at any level) to root out quality problems is a
>> design-for-manufacturing disaster. I agree that earlier found defects
>> prevent costly rework or rectification; but the *real* way to do it is
>> to design the product and processes to be defect free.
>> The best known (to me) example is that most Sony televisions are not
>> even turned on at the factory.
>
>
> Rubbish! Having visited Sony factories they have a conveyor belt of TVs
Which factory? The president of a company I worked for, who was our
Six-Sigma "instigator" told the story. He had seen this. I have no
reason to believe he lied about it.
> passing in front of, mainly young women, each checking the picture
> quality. Incidentally, one of the standard tools in their final test is
> actually a rubber mallet which is used to tap the side of the unit if
> the picture has significant line structure! The trinitron tube design
> is prone to the wires in the aperture grille sticking during manufacture
> and a sharp tap with the mallet reduces the number of customers who
> receive badly lined TVs - even though the installation instructions warn
> of this potential problem and the solution!
>
>> The first power-on test is by the customer. (I don't know if
>> sub-assy's are tested with power on).
>>
> Of course they are tested with power on - that is in keeping with the
> intent to weed out defects while they still have minimum value added.
>
>>> The fact is that scanners are NOT particularly complex devices - they
>>
>>
>> Compared to what? I don't know the parts count but being:
>> -optical
>> -elctromechanical (in at least 2 axis)
>> -electronics
>> is *is* more complex than, say, a radio with no moving parts and
>> nothing requiring precise alignment or ability to self-align.
>>
> Few radios are in the same price bracket as these scanners and those
> which are have considerably more complexity in their electronic design
> and manufacture! Compared to, say a 35mm SLR of similar cost, they are
> exceedingly simple devices.
That depends on the radio. I meant that as an example of a solid state,
non mechanical device (except a button or dial) v. a scanner with motors
and sensors in it. The scanner also has a microprocessor and the
optical processing channel(s) which increase its complexity betond that
of most commercial radios.
>
>>> have been engineered to be very simple to manufacture in high volume
>>> at low cost and to retail with a significant margin. That margin is
>>> severely dented by unnecessary field repair and support costs due to
>>> faulty product being shipped. This applies just as much to Minolta
>>> as it does to every manufacturer and was recognised as long ago as
>>> Henry Ford's first production line for the Model T, which was a damn
>>> sight more complex than a film scanner in real terms - assembly from
>>> the bought in parts. Some faults inevitably get out the door - the
>>> whole point of modern production methods is to MINIMISE the number,
>>> and hence the cost, of those that do!
>>
>>
>> Henry's process had its greatest impact in manufacturing cost, not
>> quality.
>>
> Quality IS cost! Quality is consistent performance within
> specification. Failure to meet specification after delivery results in
> additional cost. You sound like a novice who cannot distinguish between
> quality and prestige or cachet!
>
>>
>> Let's take your last sentence: You don't know how many are sold. You
>> don't know how many have defects.
>
>
> Neither do you - you ASSUME a silent majority!
With regards to NG's, absolutely.
>
>> NGs are not databases.
>>
> No, but they are a sample of the database.
Skewed and small.
>
>> Given the moderate complexity of the product I'd guess each unit has
>> on the order of 1000 defect opportunities ...
>
>
> Err, only a few lines back you were extolling the virtues of design for
> zero defects... what a shame Minolta didn't rely on you to set up their
> manufacture facilities - they would reduce the defects considerably -
> but probably never brought the scanner to market!
A 'defect opportunity' is just that. A part or manuf-process that might
fail or be out of tolerance. 1000 defect-opportunities for a product is
very small beer. The lowest possible count for defect opportunities is
the number of parts. Then add all the operations on those parts,
including tests (which can fail too). These are all the defect
opportunities that are part of Six-Sigma.
I made it clear that I (nor you) really know the number. I could have
used 100 or 10,000 defect opportunities. The point is I don't know and
neither do you. Further, I never said Minolta's supplier was Six-Sigma.
I never said the product was designed for zero defects (which is a
goal, and rarely a destination IAC).
Your last sentence is very insightful, if you actually meant it. I have
brought products to market, not alone, but with small to large teams.
Everyone working very hard to make the product perform, meet
requirements and be producible at a reasonable margin. The pressures to
bring a product to the revenue line are huge. Launching production with
only a few successfully built pre-prod or LRIP units always carries the
risk of unknowns in larger volume production. I won't regale you with
tales of the unexpected in production.
>
>> Compare to say ISO 9000 where a registrar audits your processes and
>> asserts your qualification
>
>
> ISO9000 audits your process but does NOTHING to quantify the process or
> the reliability of the final product. The process can virtually random
> assembly with almost 100% return rate, provided it is documented as such
> and that such returns are expected. In short, you can have a process
> which assembles garbage and, provided that the process is documented and
> enforced you can still achieve ISO9000 certification. Like many of the
> so called quality initiatives it is yet another example waiting for
> someone to point out its nakedness.
I mentioned ISO-9000 as an independantly audited qualification. Period.
The point (which you hastilly snipped from your reply) is that Six-Sigma
is NOT a qualification. It is a process. It is bean counter friendly
as it IS measurable within the company if properly implemented.
Aris
I just received my Minolta 5400 tonight.
I had discovered this discussion just after i had ordered the scanner.
Needless to say that i was really anxious to test it for those
"lines"...
So, the answer is... yes, there are some lines...
But
1) they are not as prominent as in the exemples i have seen.
2) there are 2-3 red lines that can be easily found and 2-3 green ones
that are a little harder to find.
3) they occur only in very dark areas. Areas which give RGB values
less than 10 (in the 1-255 range), usually 6-7. And they become
apparent only if you boost the shadows to over 35 (RGB values again).
4) they are no more than 1 pixel wide and they are not continous.
Especially the green ones can only hardly be called lines.
Here you can see a sample of what I found: (the dark part on the right
side is the edge of the slide)
http://users.forthnet.gr/ath/arvid/images/5400line003.jpg
A red line is portrayed here and some hardly discriminable green ones.
You can also see the extreme manipulation needed to bring them
forward.
I have observed that in the very dark areas you can see the tendency
of this scanner to produce little linear patterns which is indicative
of the sensor's movement but are not posing a problem.
For me and for the present those "lines" are not a serious problem.
I just hope they do not deteriorate.
On a general notice, i can say i am impressed with the scanner's
ability to catch detail in very dark and very light areas. (but it is
my first dedicated film scanner)
I am looking forward to showing some scans to a friend of mine who
swears by his Crossfield drum scanner... He can surely judge better
than me.
Aris
Ricky
Yes I can read, thanks Alan. The reply I posted certainly did not make
any reference to your 'advice'. BUT the fact is this does SEEM to be a
common problem with Minolta scanners. As other people's posts about
their problems with Minolta scanners have stated, when they return the
scanner to Minolta, Minolta do not find any problem with it, and say
it is performing 'normally'. I therefore stated that I refuse to
accept these lines as 'normal'.
Calm down Alan.
Ricky.
Ricky
I received my third SE5400 yesterday. So....
Using the DiMAGE scan software, I scanned the same slide (contrasty
scene with bright highlights and fairly dense shadows) three times in
a row (preview, scan, preview, scan, preview, scan). I then brought up
the scans in Photoshop one by one and...all three had lines in, and as
with the previous two scanners each scan was worse than the previous.
After swearing for a few minutes, and taking out my frustrations on
the boy next door who had just kicked his football against my window,
I decided to give Vuescan a try. I wasn't holding out for much
success, given that the lines in the previous scanners were much worse
when using Vuescan. I fired Vuescan up, recalibrated and followed the
same process with the same slide (preview, scan, preview, scan,
preview, scan). I then brought up the scans in Photoshop one by one
and...NONE of them had any trace of these lines whatsoever! Almost
perfect in fact, although the shadows were fairly noisy.
The only thing I had changed within Vuescan that I hadn't done with
the previous two scanners was change 'Exposure Clipping' from 0.1 to
0.01. Could this have anything to do with it?
I will do a few more test scans tonight. Hopefully all will go well.
I am concerned about something though...there seems to be a 'clicking'
noise (kinda like a train rolling down the track, although obviously
quieter, and without the 'Whooo Whooo') when the holder gets part way
into the scanner, and also when the holder is ejected. When I scan,
the clicking noises start intermittently about 20% of the way into the
scan. Although the quality of the final scans do not seem to be
affected by this, would you be worried? Not sure if the clicking
noises would cause vibrations that could affect sharpness.
Also, the scanner seems to pause a few times during scanning (you can
hear it actually slowing to a stop, pausing for a second or two, and
then starting again). I did not notice this happening with the
previous two scanners. Is this normal?
Thanks,
Ricky.
Bart van der Wolf
"Aris" <ari...@in.gr> wrote in message
news:c6e08a2.03102...@posting.google.com...
SNIP
> You can also see the extreme manipulation needed to bring them
> forward.
That's nothing to worry about. If it needs such extreme manipulation to get
marginal visibility, the lines will have identical RGB values to their
neighbors under normal conditions.
> I have observed that in the very dark areas you can see the tendency
> of this scanner to produce little linear patterns which is indicative
> of the sensor's movement but are not posing a problem.
That's why you would normally maximize exposure, while avoiding highlight
clipping. This maximizes the Signal to Noise ratio, and it minimizes the
visibility of those slight calibration errors.
SNIP
> I am looking forward to showing some scans to a friend of mine who
> swears by his Crossfield drum scanner... He can surely judge better
> than me.
If he's objective, then that would be interesting to hear. On the other
hand, if he really swears by his Crossfield, he might search very hard to
find anything that could justify his own preference. In the latter case,
gently remind him of the price difference ;-)
He may find the Blue channel noise to be higher than for Red and Green,
which is only a minor issue given the low color resolution of the human
visual system, and Blue only accounts for some 7% of the luminance.
Happy scanning,
Bart
Bart van der Wolf
> Hi everyone,
>
> I received my third SE5400 yesterday. So....
>
> Using the DiMAGE scan software, I scanned the same slide (contrasty
> scene with bright highlights and fairly dense shadows) three times in
> a row (preview, scan, preview, scan, preview, scan). I then brought up
> the scans in Photoshop one by one and...all three had lines in, and as
> with the previous two scanners each scan was worse than the previous.
If the scans get worse with each scan, then the calibration is shifting.
Have you tried re-initializing (CTRL+Shift+I while in the Scan Utility)
after some warm up time? You may also want to avoid strong ambient lighting
on the scanner while operating it.
> After swearing for a few minutes, and taking out my frustrations on
> the boy next door who had just kicked his football against my window,
> I decided to give Vuescan a try. I wasn't holding out for much
> success, given that the lines in the previous scanners were much worse
> when using Vuescan. I fired Vuescan up, recalibrated and followed the
> same process with the same slide (preview, scan, preview, scan,
> preview, scan). I then brought up the scans in Photoshop one by one
> and...NONE of them had any trace of these lines whatsoever!
Which seems to point in the direction of Calibration again, although it's
strange that previously (with the same VueScan version?) VS was worse.
> Almost perfect in fact, although the shadows were fairly noisy.
>
> The only thing I had changed within Vuescan that I hadn't done with
> the previous two scanners was change 'Exposure Clipping' from 0.1 to
> 0.01. Could this have anything to do with it?
This will only have a marginal effect (depending on the image at hand) on
the exposure time. In your case it would have sightly reduced exposure
times, giving more reason for dense film area noise.
> I will do a few more test scans tonight. Hopefully all will go well.
>
> I am concerned about something though...there seems to be a 'clicking'
> noise (kinda like a train rolling down the track, although obviously
> quieter, and without the 'Whooo Whooo') when the holder gets part way
> into the scanner, and also when the holder is ejected.
The clicking noise may just have been the Grain Dissolver being moved in and
out of the lightpath.
> When I scan, the clicking noises start intermittently about 20%
> of the way into the scan.
I have never heard that during the scan, only in between scans, so maybe
it's something else.
> Although the quality of the final scans do not seem to be
> affected by this, would you be worried? Not sure if the clicking
> noises would cause vibrations that could affect sharpness.
>
> Also, the scanner seems to pause a few times during scanning (you can
> hear it actually slowing to a stop, pausing for a second or two, and
> then starting again). I did not notice this happening with the
> previous two scanners. Is this normal?
That just indicates it can't off-load its data fast enough, and has to wait
for its buffer to empty (are you using USB or FireWire?). Longer exposure
times give more time for the interface to process the data stream, but when
you use ICE, that will most likely claim all processing power from the CPU
it can get and the scanner has to wait.
Bart
Ed Hamrick
> I decided to give Vuescan a try. I wasn't holding out for much
> success, given that the lines in the previous scanners were much worse
> when using Vuescan. I fired Vuescan up, recalibrated and followed the
> same process with the same slide (preview, scan, preview, scan,
> preview, scan). I then brought up the scans in Photoshop one by one
> and...NONE of them had any trace of these lines whatsoever! Almost
> perfect in fact, although the shadows were fairly noisy.
This seems to point to the possibility that the scanner needs
to warm up a bit before calibrating it. You might try the
Minolta software again, run the calibration and see if the problem
goes away.
Also, I did find one thing about the Scan Elite 5400 some time
ago - the signal from the black point of the CCD is really,
really noisy, and I was using the median of this signal to
correct for black point. However, this was too noisy, and caused
lines in the scans.
VueScan 7.6.64 uses the average of all the black points instead
of the black point at each CCD element, and this seems to solve
the problem with the lines. It's a kludge, but it seems to work.
Regards,
Ed Hamrick
Alan Browne
Sorry Ricky, the context in which I read your post was not a context of
Minolta claiming the performance is normal. I was among those who
advised you to return the scanner along with samples of bad scans. And
if it came/comes to it, just abandon Minolta. I would not be really
surprised if Nikon come out with a scanner to 'answer' the 5400 in
coming months ...
Alan.
Alan Browne
Ricky wrote:
> Hi everyone,
>
> I received my third SE5400 yesterday. So....
>
> Using the DiMAGE scan software, I scanned the same slide (contrasty
> scene with bright highlights and fairly dense shadows) three times in
> a row (preview, scan, preview, scan, preview, scan). I then brought up
> the scans in Photoshop one by one and...all three had lines in, and as
> with the previous two scanners each scan was worse than the previous.
Please post examples. I only get lines , extremely faint if I punch the
brightness of the scan and turn down the room light and zoom to 400% on
a 5400 dpi scan of a negative exposed to turn black. These lines are at
a level a hair above the noise in surrounding pixels.
In vuescan they are there but smudged out a little.
Question: Do you know for a fact that each scanner is a different one?
eg are the SERNOS different?
>
> After swearing for a few minutes, and taking out my frustrations on
> the boy next door who had just kicked his football against my window,
Funny how there's always a kid throwing a ball around when we're
frustrated... if a kid's not available, a barking dog is usually
provided..... ;-)
> I decided to give Vuescan a try. I wasn't holding out for much
> success, given that the lines in the previous scanners were much worse
> when using Vuescan. I fired Vuescan up, recalibrated and followed the
> same process with the same slide (preview, scan, preview, scan,
> preview, scan). I then brought up the scans in Photoshop one by one
> and...NONE of them had any trace of these lines whatsoever! Almost
> perfect in fact, although the shadows were fairly noisy
The noise in the shaddows is good news. If there were no noise, then
I'd be suspicious.
>
> The only thing I had changed within Vuescan that I hadn't done with
> the previous two scanners was change 'Exposure Clipping' from 0.1 to
> 0.01. Could this have anything to do with it?
Change it back....to see, but you know that.
>
> I will do a few more test scans tonight. Hopefully all will go well.
>
> I am concerned about something though...there seems to be a 'clicking'
> noise (kinda like a train rolling down the track, although obviously
> quieter, and without the 'Whooo Whooo') when the holder gets part way
> into the scanner, and also when the holder is ejected. When I scan,
> the clicking noises start intermittently about 20% of the way into the
> scan. Although the quality of the final scans do not seem to be
> affected by this, would you be worried? Not sure if the clicking
> noises would cause vibrations that could affect sharpness.
Occasionally, mine will let out a 'Twick', but not regulalrly.
Sometime, say 10 minutes after a scan, and the holder removed, there
will be a similar twick.
>
> Also, the scanner seems to pause a few times during scanning (you can
> hear it actually slowing to a stop, pausing for a second or two, and
> then starting again). I did not notice this happening with the
> previous two scanners. Is this normal?
Likely bandwidth/CPU, and more common with VueScan than the Minolta s/w.
AFAICT, no effect on any scans to date. I searched your messages
and I can't find your 'setup', but if you have very fast CPU ( >1 GHz)
and firewire or USB 2.0 then this should not happen too much.
>
> Thanks,
>
> Ricky.
Alan Browne
Ed Hamrick wrote:
> "Ricky" <ricky...@hays.com> wrote:
>
>
> Also, I did find one thing about the Scan Elite 5400 some time
> ago - the signal from the black point of the CCD is really,
> really noisy, and I was using the median of this signal to
> correct for black point. However, this was too noisy, and caused
> lines in the scans.
Ed, How much is "really noisy" (range?) and would you attribute this to
the 16 bit depth?
>
> VueScan 7.6.64 uses the average of all the black points instead
> of the black point at each CCD element, and this seems to solve
> the problem with the lines. It's a kludge, but it seems to work.
Ah, now, I'll DL .64!
>
> Regards,
> Ed Hamrick
>
>
Alan Browne
Ed Hamrick wrote:
> "Ricky" <ricky...@hays.com> wrote:
>
>
> VueScan 7.6.64 uses the average of all the black points instead
> of the black point at each CCD element, and this seems to solve
> the problem with the lines. It's a kludge, but it seems to work.
I just DL'd .64 and the very faint lines I had before with the DSE5400
are completely gone. (The lines I had were so faint as to be not really
worth bothering over...)
>
> Regards,
> Ed Hamrick
>
>
Andreas Schmidt
> I would not be really
> surprised if Nikon come out with a scanner to 'answer' the 5400 in
> coming months ...
yeah, if you change this to "the coming weeks" ;-) See the link here
(Japanese): http://www.nikon.co.jp/main/jpn/whatsnew/2003/coolscan_03.htm
Andreas
Alan Browne
Andreas Schmidt wrote:
Yeah, Littleboy also mentioned this. But you were nice enough to put in
link.
Reading that it looks to be a 4000 dpi, but at 16 bits/channel.
Probably the quickest upgrade they could make to the existing machine.
On the other hand this will handle med format too and that is
impressive. The pros will really snap this up.
They (like Minolta) also claim Dmax 4.8 which is clearly bullshit.
Cheers,
Alan
Ricky
> Have you tried re-initializing (CTRL+Shift+I while in the Scan Utility)
> after some warm up time? You may also want to avoid strong ambient lighting
> on the scanner while operating it.
I tried this last night: I fired up the scanner and started up the
DiMAGE utility (and it went through it's normal setting up routine). I
waited 5 minutes and did 3 previews in a row to warm the scanner up. I
then did the CTRL+shift+I thing, then when that had finished, did
another preview and then a full res scan. The resulting scan only had
faint traces of lines, so I would agree that some warm up time is
needed.
> It's strange that previously (with the same VueScan version?) VS was worse.
It is the same VS version. After the above test with the DiMAGE
utility, I tried Vuescan. Instead of doing the first few steps (as the
scanner was already warmed up), I just recalibrated and then did a
preview and scan. No traces of any lines anywhere in the image.
> This will only have a marginal effect (depending on the image at hand) on
> the exposure time. In your case it would have sightly reduced exposure
> times, giving more reason for dense film area noise.
To be honest, I don't really know exactly what the Exposure Clipping
setting does. Could you explain?
> That just indicates it can't off-load its data fast enough, and has to wait
> for its buffer to empty (are you using USB or FireWire?). Longer exposure
> times give more time for the interface to process the data stream, but when
> you use ICE, that will most likely claim all processing power from the CPU
> it can get and the scanner has to wait.
Thought I had better ask, just because I hadn't noticed it with the
other two scanners. I'm on firewire.
Thanks,
Ricky.
Ricky
> Ricky wrote:
>
> Please post examples. I only get lines , extremely faint if I punch the
> brightness of the scan and turn down the room light and zoom to 400% on
> a 5400 dpi scan of a negative exposed to turn black. These lines are at
> a level a hair above the noise in surrounding pixels.
I'll post a couple of examples today, one from DiMAGE scan and one
from Vuescan.
> Question: Do you know for a fact that each scanner is a different one?
> eg are the SERNOS different?
They are definately three different scanners. The first scanner I
returned to the supplier for a refund rather than exchange as they had
no others in stock. I then sourced the second scanner from a different
supplier, and rather than return the second scanner to them and then
they send me a new one, the supplier was kind enough to arrange for a
courier to deliver me the third scanner and take the second scanner in
one go.
> Funny how there's always a kid throwing a ball around when we're
> frustrated... if a kid's not available, a barking dog is usually
> provided..... ;-)
I was grateful to have someone to take my frustrations out on,
although a little while later his mother knocked and gave me a piece
of her mind!
BTW I changed the Exposure Clipping setting back to 0.1, didn't make a
difference. Again, there were no lines (that I could see) in the final
scan.
> Occasionally, mine will let out a 'Twick', but not regulalrly.
> Sometime, say 10 minutes after a scan, and the holder removed, there
> will be a similar twick.
Yeah, I also get the 'twick' if I haven't used the scanner for 10
minutes. Any idea why it makes that noise?
I hear the clicking noises when I insert the holder, during a scan
intermittently and also when the holder is ejected.
> Likely bandwidth/CPU, and more common with VueScan than the Minolta s/w.
> AFAICT, no effect on any scans to date. I searched your messages
> and I can't find your 'setup', but if you have very fast CPU ( >1 GHz)
> and firewire or USB 2.0 then this should not happen too much.
I have an Athlon XP2000 with 768mb RAM and I'm on firewire.
Thanks,
Ricky.
Bart van der Wolf
> To be honest, I don't really know exactly what the Exposure Clipping
> setting does. Could you explain?
VueScan tries to maximize the exposure time (because that gives better
quality), but it has to prevent overexposure (results in clipping and other
issues).
VS uses the histogram data of the preview to determine the maximum exposure.
It determines the exposure that produces the "Exposure clipping" percentage
of overexposed pixels, and then reduces that exposure by 5%. This will
almost certainly produce non-clipped but well exposed scans.
It's a clever way of setting the exposure time, but there are a few images
that still need a small correction. Therefore the control was made available
for the user to tweak. The default setting is probably good for 99% of the
images. So changing the percentage will result in (depending on the image
content) small changes in the exposure time.
Bart
Alan Browne
Ricky wrote:
I assume that means 2 GHz? If so, the scan with Minolta or Vuescan
should go through without any pauses at all (at least mine does at 2.4GHz).
Do you have a lot of other processes running at the same time? Some
operations in PS-Elements take a lot of CPU time and -do- slow down
scans when they occur.
Alan
Dierk Haasis
<alan....@videotron.ca> wrote:
>Yeah, Littleboy also mentioned this. But you were nice enough to put in
>link.
Here's one in English:
http://www.nikon-image.com/eng/Lineup/index.html
Actually the press release has been seen by me Monday or Tuesday
somewhere; but this is the source.
--
Dierk
Ricky
Thanks for the explanation.
I'm sure this must have been brought up before but, to quote from the
VS users guide:
"A nominal value of .01 works well for most images, but if you find
that the CCD exposure time is too low for most images, try increasing
this to 0.1 or 1.0."
I notice within VS the default value is 0.1. Should I be setting this
to the 'nominal' value of 0.01, or is the default setting of 0.1 the
best setting for 'most images'?
Thanks,
Ricky.
Alan Browne
Thanks!
Marko B.
> > Sometime, say 10 minutes after a scan, and the holder removed, there
> > will be a similar twick.
>
> Yeah, I also get the 'twick' if I haven't used the scanner for 10
> minutes. Any idea why it makes that noise?
I think it goes to stand by or some other eco crap. Makes sense the thing
should rest a buit if not working.
m.
William D. Tallman
<snip>
So the lines were caused by misconfiguration of the black point, just as Ed
Hamrick suggested? As it appears that Ricky has just discovered this
adjustment, and the other two scanners were also misconfigured, does this
suggest that the dirty CCD array is probably a misdiagnosis?
I always scan at 5400dpi. The scanner chuckles along softly; presumably the
sound is the mechanism advancing the film holder.
Incidentally, the problem I had with the mottled images that resulted from a
fast scan speed is resolved. I think. It turned out I was not returning
to the first tab before I started the scan. Starting the scan in the color
tab may be the culprit here. Further experience will test this.
Also, my setup may be of interest. I am running a 500Mh Celeron on an Asus
CUBX (Coppermine processors, BX chipset). I've got 512Mb of RAM with a 2Gb
swap partition. It's slow, but I can do all sorts of other stuff while the
scan is running with nary a hitch. It's Linux, dontchaknow.... <grin>
Bill Tallman
Bart van der Wolf
> to the 'nominal' value of 0.01, or is the default setting of 0.1 the
> best setting for 'most images'?
A setting of 0.01 results in a slightly shorter exposure time. If this
causes the histogram to leave a gap between the left side and the first bin
data (histogram display set at Logarithmic for better judgement) for
negatives, or the right side for slides, you can try a higher value. If 0.1
clips too much, use a lower value.
It all depends a little on how *your* films have been processed and what
output profile you chose. Start with the default, Ed Hamrick usually chooses
very sane default settings. Also, we are talking about tweaking the last bit
of performance out of a scanner. The differences could be small.
Bart
Ricky
> should go through without any pauses at all (at least mine does at 2.4GHz).
It's 1.67 GHz. I have recently ordered an XP2400 chip, which is 2 GHz,
and also an extra 256Mb RAM, to take me up to 1 Gb (just waiting for
them to be delivered).
> Do you have a lot of other processes running at the same time? Some
> operations in PS-Elements take a lot of CPU time and -do- slow down
> scans when they occur.
I don't do anything with the computer while the scanner is busy. I
prefer to just go away and come back when it's finished.
Thanks,
Ricky.
Ricky
> Hamrick suggested? As it appears that Ricky has just discovered this
> adjustment, and the other two scanners were also misconfigured, does this
> suggest that the dirty CCD array is probably a misdiagnosis?
We'll never know. All I know is the lines got worse scan after scan
with the other two scanners, and recalibration did nothing to improve
it. After warming up this scanner and then calibrating then there are
no lines (well, nothing to cause me too much concern).
> I always scan at 5400dpi. The scanner chuckles along softly; presumably the
> sound is the mechanism advancing the film holder.
Seems like it is, but I wasn't sure if the clicking sound of the
mechanism advancing the holder could cause some kind of vibration
which could affect the sharpness of the scan.
Thanks,
Ricky.
Ricky
> causes the histogram to leave a gap between the left side and the first bin
> data (histogram display set at Logarithmic for better judgement) for
> negatives, or the right side for slides, you can try a higher value. If 0.1
> clips too much, use a lower value.
>
> It all depends a little on how *your* films have been processed and what
> output profile you chose. Start with the default, Ed Hamrick usually chooses
> very sane default settings. Also, we are talking about tweaking the last bit
> of performance out of a scanner. The differences could be small.
>
> Bart
Ok...I think I'm starting to get it!
Thanks Bart,
Ricky.
Alexandru Rosu
> Alan Browne <alan....@videotron.ca> wrote in message news:<vrxnb.44344$Pt3.1...@weber.videotron.net>...
> > Ricky wrote:
> >
> > Please post examples. I only get lines , extremely faint if I punch the
> > brightness of the scan and turn down the room light and zoom to 400% on
> > a 5400 dpi scan of a negative exposed to turn black. These lines are at
> > a level a hair above the noise in surrounding pixels.
>
> I'll post a couple of examples today, one from DiMAGE scan and one
> from Vuescan.
>
I'm currently in the market for a scaner and I've been following this
thread withn great interest.
Could you please post some examples of the 'new' faint lines?
Thank you in advance,
Alex
Alan Browne
William D. Tallman wrote:
> Also, my setup may be of interest. I am running a 500Mh Celeron on an Asus
> CUBX (Coppermine processors, BX chipset). I've got 512Mb of RAM with a 2Gb
> swap partition. It's slow, but I can do all sorts of other stuff while the
> scan is running with nary a hitch. It's Linux, dontchaknow.... <grin>
Meaning? With my Scan Dual (SCSI / Win 98) and now with the 5400
(Firewire / WinXP) I can be doing other things (e-mail, web, photoshop,
etc, etc, whatever) while scanning (with nary a hitch). It's not Linux
dontchknow ...
>
> Bill Tallman
>