Scanning objects - any tips?

[Marjolein Katsma]

Don ([email protected]) wrote in

A-ha! The software is "doing you a favor"! That's when I shriek in
horror and ask: How do I turn this "fantastic" feature off? ;o)

Right - and I only recently found the workaround (made invalid at first
with the "leave the cover off" trick)

Admittedly, it _is_ clever and in most cases I don't need to turn it off
either. But I've never used this scanner for the types of things I'm
using it for now ... suddenly causing a need to turn it off.

But when it straightens the image it does it in a flash - as opposed to
PSP which takes quite a while resampling and produces far sharper
results. I don't know what algorithms are used, but just the *time*
taken by PSP suggests it must be better - and results show it is.

The sad thing is the "programmer" (and I use the term very loosely
here) is probably very proud of this "feature" and how "smart" he was
to think of this. :-/

Really, the only thing left out of this feature was the option to turn
it off - and I can't blame the programmer (whatever her name) for that:
if it wasn't in the spec it didn't make it into the product. If you want
to blame anyone (I don't) blame the product manager.

I'm realistic: this scanner wasn't really made for the sorts of purposes
I'm squeezing it into operation for now, and I never bought it for those
purposes either. If I'd known what I'd be doing now when I bought the
scanner I'd probably have bought a higher-end one.

 

[Marjolein Katsma]

Don ([email protected]) wrote in

Cloth is fun to scan all by itself! ;o)

Oh, yes - cloth and various types of paper, too. At one time I did a lot
of that, colorizing the results and making seamless tiles of them to
produce nice desktop backgrounds.

No, what are they? Is it a snack (like roasted corn or chickpeas) or
is it a staple food? As a vegetarian, I'm always curios about these
things. ;o)

It's a snack - though I'm not sure if vegetable oil is used (I can't
read the ingredients in Chinese!) Google for "wasabi peas" (including
quotes) and you'll find sample images and descriptions. But most
descriptions say they're "hot" while the ones I've bought (several
times) never were.

Although I know most people using newsgroups hate web-based forums (I'm
one of those though I tolerate a few) I'm discussing different aspects
of my little project of the bag of peas here (with a few images
attached):
http://www.desktoppublishingforum.com/bb/showthread.php?t=1578
We're having lots of fun with the Chinese text on the package, too ;-) -
quite off-topic for this NG.

One good way to get rid of casts is to change the color space. Lab is
usually recommended for removing a cast.

Hmm - 'lab' is a PS term, I know. I don't think PSP 7 has an equivalent
but PSP X might (installing that requires a major reorganization so I'm
holding back on that - basic techniques can be practiced with version 7
as well).

Also, it appears you need a complex curve with multiple points.
Normally a cast is removed by setting the gray point, but if the cast
is non-linear (like my infamous Kodachromes) then you need several
points because the cast is different in different areas.

I end up doing different curves for different colors - but I haven't
quite worked out the best approach. I keep working on it though because
I think it *should* be possible to get the bright green of the peas in
the scanned image but I don't know what role the plastic bag is playing
in this (if any).

if the setting causes more pixels without a rescan then it does
sound like interpolation. In that case, as I mentioned before, I would
use native resolution and interpolate later in an image editor.

I tend to agree - it seems PSP is better at upscaling as well (or at
least as good). I now have arrived at more less the following process:
Scanner:
- scan with the bag angled, (mostly) cover off, preventing straightening
with the white paper strip and setting selection border
- set at 300dpi and 100% size
- save as TIFF
PSP:
- FIRST enlarge to 200%
- then rotate carefully to straighten, using a guide line to check the
result (the extra pixels of the enlargement help to get a sharper result
after rotating)
- clear up any dust specks etc. (clone brush and some copy-and paste
with lassoed areas)
- save back over the original TIFF: that's the "master" then
Then I do the "extra" stuff like retouching the areas with "bloom"
(background) - a LOT of work! - and whatever color correction is needed
(adjustment layer) and save the result as .PSP (PSP's native format
which keeps the layers etc.). Maybe do some slight sharpening before
saving. I'll never actually *use* the image in this form but it gives me
my basic material.

I'm then ready to reduce in size or do any cutouts needed or maybe
combine front-and-back into one image (tricky).

But I'm still learning and don't have a single set of front and back
images done with this process. I think I'm getting a lot closer to what
I can squeeze out of the scanner + PSP though.

I do have two front-and-back images of smaller objects, but with what
I've learned by now I'll probably redo them... but they're nice with a
shadow on the white cover.

 

[Don]

It's a snack - though I'm not sure if vegetable oil is used (I can't
read the ingredients in Chinese!) Google for "wasabi peas" (including
quotes) and you'll find sample images and descriptions. But most
descriptions say they're "hot" while the ones I've bought (several
times) never were.

Oh, I like hot food! I also find that most food described as "hot"
never really is. I guess I must just have higher tolerance.

Although I know most people using newsgroups hate web-based forums (I'm
one of those though I tolerate a few) I'm discussing different aspects
of my little project of the bag of peas here (with a few images
attached):
http://www.desktoppublishingforum.com/bb/showthread.php?t=1578
We're having lots of fun with the Chinese text on the package, too ;-) -
quite off-topic for this NG.

I don't like web forums because it's not possible to read them offline
(right now a requirement for me because I only have POTS access).

Hmm - 'lab' is a PS term, I know. I don't think PSP 7 has an equivalent
but PSP X might (installing that requires a major reorganization so I'm
holding back on that - basic techniques can be practiced with version 7
as well).

Lab is actually a color space like RGB or CMYK. PS supports it but I'm
not sure about PSP. I only have a very old PSP 4 installed here.

I end up doing different curves for different colors - but I haven't
quite worked out the best approach. I keep working on it though because
I think it *should* be possible to get the bright green of the peas in
the scanned image but I don't know what role the plastic bag is playing
in this (if any).

You can set all three channels simultaneously when you use the gray
point. What you need to do is find an area which should be neutral
(i.e. where all three channels should have the same value). Because of
noise scanned images don't have such smooth areas which is why a color
sampler is usually set to a 3x3 pixel or 5x5 pixel sampling radius. I
go one better and select an area and then read out the average values.
That way the sample is much bigger than the little 5x5 square. I then
use that to set the neutral point.

I tend to agree - it seems PSP is better at upscaling as well (or at
least as good). I now have arrived at more less the following process:
Scanner:
- scan with the bag angled, (mostly) cover off, preventing straightening
with the white paper strip and setting selection border
- set at 300dpi and 100% size
- save as TIFF
PSP:
- FIRST enlarge to 200%
- then rotate carefully to straighten, using a guide line to check the
result (the extra pixels of the enlargement help to get a sharper result
after rotating)
- clear up any dust specks etc. (clone brush and some copy-and paste
with lassoed areas)
- save back over the original TIFF: that's the "master" then
Then I do the "extra" stuff like retouching the areas with "bloom"
(background) - a LOT of work! - and whatever color correction is needed
(adjustment layer) and save the result as .PSP (PSP's native format
which keeps the layers etc.). Maybe do some slight sharpening before
saving. I'll never actually *use* the image in this form but it gives me
my basic material.

I'm then ready to reduce in size or do any cutouts needed or maybe
combine front-and-back into one image (tricky).

Given your context that sounds about the best workflow. You end up
with both the archived and processed versions.

But I'm still learning and don't have a single set of front and back
images done with this process. I think I'm getting a lot closer to what
I can squeeze out of the scanner + PSP though.

I do have two front-and-back images of smaller objects, but with what
I've learned by now I'll probably redo them... but they're nice with a
shadow on the white cover.

Yes, re-doing scans happens all the time. When I look at the first few
images I edited 3-4 years ago I can't believe I though that was good!

Don.

 

[Marjolein Katsma]

humanoid ([email protected]) wrote in

I know a woman, a jewelry maker, who scans all of her work. She
drapes a dark opaque cloth over the jewelry instead of using the
scanner lid. It works very well. Make sure it is a finely made cloth,
like a silk scarf, so it will drape over the object and not just cover
it.

I tested this approach with one piece of jewelry (a Tibetan necklace)
that I brought from my trip and I wasn't happy with the results.

Instead, I got a piece of cardboard large enough to amply cover the
scanner bed and wrapped and fixed a (not so fine) white scarf around it
so I got a flat "white cloth" surface and used that instead of the
scanner cover. The necklace has a few very large beads, and these keep
the "cloth cover" away from the scanner bed - with the result that it's
not entirely white since apparently the strength reflected light on the
CCD diminishes quite fast with distance from the scanner bed.

What mostly surprised me was the depth of field my little scanner has
(probably by design so you can do usable scans of books that don't open
flat!): I expected the cloth to be quite blurred but while not entirely
sharp you can still see the individual threads.

I can post a result if anyone's interested.

I tried the same "cloth background" with other objects but wasn't happy.
I'm beginning to think each (type of) object requires its own approach,
and even different type of jewelry require their own approach. What
works for your friend doesn't necessarily work for me if what I'm
scanning is something entirely different...

Still, your post prodded me to do a lot of experiments and I've gotten a
lot more creative in trying different (including "no") backgrounds for
different objects! And I'm trying more different object than I started
out thinking of "scannable" at all - I'm experimenting with a small
brass statue of Ganesh now. ;-)

 

[Marjolein Katsma]

Marjolein Katsma ([email protected]) wrote in

As long as I don't have a film scanner (and supporting new computer)
to get my photographs into digital form, I'm amusing myself with
making scans with my little HP 3300c flatbed scanner of the objects I
brought from my last trip to serve as illustrations in my travel blog.
Well, mostly it's amusing, but it's also instructive and causing quite
a few grumbles.

Some of these things are easy: things like (hotel) business cards and
admission tickets for various places are just flat and easy enough to
scan. I'm saving those for last. It's the not-so-flat objects that are
more interesting and sometimes really challenging.

Getting ever more creative in trying out techniques (thanks to all the
hints in this thread!), I found two new ones I'd like to share with you
all:

1. Transparent objects.
These are quite a challenge. I have one plastic water bottle that may
just be too large - whatever I try, it just looks darkish gray instead
of "clear". The cause: I found the amount of light picked up by the CCD
decreases sharply with distance (much more than DOF, it seems).
Flatter transparent objects work well though but to make the material
really appear transparent you need some light falling _through_ - how to
accomplish that when the scanner light is just light falling _onto_ the
object? I had a little cellophane bag with a snack that I wanted to scan
(after my failed experiments with the water bottle)... taking a break in
the kitchen did it: I used aluminium (aluminum for those in the USA)
foil as a background. Nice highlights in the transparent material. In
PSP I selected the background which had rather stark white and gray
shades, and applied a Gaussian blur with a radius of 20-30 to it: that
way the little bag stands out nicely.

2. "Very 3D" objects.
I found my little scanner has a lot larger DOF than I expected: material
away from teh glass place isn't quite sharp but still has a lot more
detail than I thought it would. But since the light intensity picked up
decreases fast, with "really" 3D objects some details may be lost in the
shadows. As an experiment I put a little brass Ganesh statue onto the
scanner bed (leaving the cover off): what's close to the glass plate
comes out nicely but what's a little further off is "sharp" enough but
not light enough, and some parts disappear completely.
So: I needed some "studio lighting" to light up the darkest parts. I
experimented a bit with an (oblique) shiny metal surface to hopefully
throw some scanner light back but the result was disappointing. The next
thing I tried was successful though: take a sheet of paper (I used A4
format), cut it in half lengthwise and stick the parts together to get a
loop half the height of the width of the paper that's wide enough to fit
around the object on the glass place and leave some distance around it.
Now there is a light surface *near* the glass (and the source of light!)
that reflects a bit of light back on the object. Not perfect, but a lot
better...

I can post results if anyone's interested.

 

[DaveG]

Sometime on Mon, 05 Dec 2005 00:52:28 +0000, Marjolein Katsma scribbled:

2. "Very 3D" objects.
I found my little scanner has a lot larger DOF than I expected: material
away from teh glass place isn't quite sharp but still has a lot more
detail than I thought it would. But since the light intensity picked up
decreases fast, with "really" 3D objects some details may be lost in the
shadows. As an experiment I put a little brass Ganesh statue onto the
scanner bed (leaving the cover off): what's close to the glass plate
comes out nicely but what's a little further off is "sharp" enough but
not light enough, and some parts disappear completely.
So: I needed some "studio lighting" to light up the darkest parts. I
experimented a bit with an (oblique) shiny metal surface to hopefully
throw some scanner light back but the result was disappointing. The next
thing I tried was successful though: take a sheet of paper (I used A4
format), cut it in half lengthwise and stick the parts together to get a
loop half the height of the width of the paper that's wide enough to fit
around the object on the glass place and leave some distance around it.
Now there is a light surface *near* the glass (and the source of light!)
that reflects a bit of light back on the object. Not perfect, but a lot
better...

Although I've not tried anything like what you are currently attempting, I
did try slides on a nasty cheap flatbed, the type that uses three
different coloured LEDs instead of a white light and found that using high
brightness inkjet paper as a backing gave better (but still useless in
this case) results.

I wonder if a gloss finish "photo" inkjet paper would be good for your
tests? Or even a silver/mirror finish thin card?

 

[Marjolein Katsma]

DaveG ([email protected]) wrote in

Although I've not tried anything like what you are currently
attempting, I did try slides on a nasty cheap flatbed, the type that
uses three different coloured LEDs instead of a white light and found
that using high brightness inkjet paper as a backing gave better (but
still useless in this case) results.

I wonder if a gloss finish "photo" inkjet paper would be good for your
tests? Or even a silver/mirror finish thin card?

Hmm, I think I'll have a look what my local arts materials store has in
the way of cardboard or thick paper with various finishes... good idea!

Beauty is in the eye of the beerholder

(I'll steal that!) A nice Tuborg here

 

[DaveG]

Sometime on Mon, 05 Dec 2005 17:04:20 +0000, Marjolein Katsma scribbled:

DaveG ([email protected]) wrote in

Hmm, I think I'll have a look what my local arts materials store has in
the way of cardboard or thick paper with various finishes... good idea!

You;re welcome.

(I'll steal that!) A nice Tuborg here

That would mean you are receiving stolen goods LOL