Nikon Super Coolscan 4000 ED Scanner - First Impressions

Pat asked about the digital ICE on the scanner, so here's a quick sample of what it does. I chose
one of my oldest slides, from the very first roll of slide film I ever shot, back in January 1986.
There's a lot of fine fungus filaments on the slide (primarily the edges), quite a few other marks
(dust that has attached itself to the surface) and a scratch across the slide.

This is a small section of the raw scan at 100%, with no ICE.
http://www.wildwales.fsnet.co.uk/misc/100%25ice0.jpg

This is the same section with ICE set to Normal. If you look closely you can see that there are
still very fine marks in the sky. http://www.wildwales.fsnet.co.uk/misc/100%25ice1.jpg

This is the same section with ICE set to Fine. All traces of the marks have now gone but the image
has been softened slightly. How obvious this softening would be in a print I don't yet know. The
resolution is so high that it may be barely noticeable.
http://www.wildwales.fsnet.co.uk/misc/100%25ice2.jpg

Only the worst slides and negs would need the Fine ICE treatment, and even then, any left over
marks are only obvious in the sky, so for best results two scans could be made, with the sky from
the Fine ICE scan being overlaid onto a Normal ICE scan using layer masking, thus not affecting the
ground detail.

This is a reduced version of the full raw scan without ICE;
http://www.wildwales.fsnet.co.uk/mis...y_original.jpg

And this is the final version of the ICED scan after restoration (colour balance adjusted manually
in Photoshop); http://www.wildwales.fsnet.co.uk/mis...y_restored.jpg

Paul
--
The October Project 2003
http://www.wildwales.fsnet.co.uk/october/october.html
http://www.wilderness-wales.co.uk
http://www.photosig.com/go/users/userphotos?id=118749

"Paul Saunders" <pvs1@wildwales.fsnet.co.uk> wrote in message
news:bn642i$ip0$1@newsg2.svr.pol.co.uk...
>
> This is a small section of the raw scan at 100%, with no ICE.
> http://www.wildwales.fsnet.co.uk/misc/100%25ice0.jpg
>

Hi Paul,

Cannot get to the first three links you specified I'm afraid.

>
> This is a reduced version of the full raw scan without ICE;
> http://www.wildwales.fsnet.co.uk/mis...y_original.jpg
>
> And this is the final version of the ICED scan after restoration (colour balance adjusted manually
> in Photoshop); http://www.wildwales.fsnet.co.uk/mis...y_restored.jpg
>

Yes, looks like ICE is quite effective :-)

Btw, did you make any other adjustments apart from colour balance? Just curious.

Regards, Shaun.

On Wed, 22 Oct 2003 15:26:19 +0100, "Paul Saunders"
<pvs1@wildwales.fsnet.co.uk> wrote:

>Pat asked about the digital ICE on the scanner, so here's a quick sample
<snip>
>
>Paul

Very impressive, Paul. By the way, you accidentally inserted 100%25 in the links for the first three
files, which is why the links don't work.

I have got a lot of slides in that sort of condition, so I am rather envious!

Pat

Pat Bennett From Cheshire in the UK www.cheshirewildlife.co.uk

Here are some more samples, this time of the GEM grain reduction. I chose a very grainy ISO 400
slide to test it on. I used GEM on the maximum setting of 4. All shots shots are scanned with ICE.
No other alterations.

Various before and after shots;

http://www.wildwales.fsnet.co.uk/misc/grain1.jpg http://www.wildwales.fsnet.co.uk/misc/grain1gem.jpg

http://www.wildwales.fsnet.co.uk/misc/grain2.jpg http://www.wildwales.fsnet.co.uk/misc/grain2gem.jpg

http://www.wildwales.fsnet.co.uk/misc/grain3.jpg http://www.wildwales.fsnet.co.uk/misc/grain3gem.jpg

Here's the full pic (at reduced size), this is a raw scan (ICE only), no corrections other than a
little sharpening with Focus Magic.

Sgwd Einion Gam in it's full glory - a far cry from the pathetic little trickles that are currently
masquerading as waterfalls in these parts. http://www.wildwales.fsnet.co.uk/misc/einiongam.jpg

Paul
--
The October Project 2003
http://www.wildwales.fsnet.co.uk/october/october.html
http://www.wilderness-wales.co.uk
http://www.photosig.com/go/users/userphotos?id=118749

Just a few more samples.

For Ste, here's a small section of a newly developed slide, to show what kind of marks are typical
on a brand new photograph; http://www.wildwales.fsnet.co.uk/misc/newslide.jpg

And here's the same section scanned with ICE; http://www.wildwales.fsnet.co.uk/misc/newslideice.jpg

At first there only appear to be a handful of tiny marks, but remember that this is only 2% of the
area of the entire slide, so you can multiply this by 50. That adds up to hundreds of such marks,
which can take some time to remove by hand.

But if you compare these two images very closely (flipping between them with appropriate software,
preferably enlarged) you'll see that there are actually hundreds of very tiny marks in the first
image that have been removed in the second. Many of these may be the infamous "pepper grain" that
Fuji has become notorious for (slight irregularities in the surface of the film), far too small and
awkward to get rid of by hand. These may not be noticeable in small prints, but might be in large
enlargements.

And for those interested in noise, here's a scan of a small section of dark shadow.
http://www.wildwales.fsnet.co.uk/misc/shadownoise.jpg

There is actually some very faint landscape detail in this scan, so what you are seeing is not all
noise. There are options to scan an image multiple times, which would probably cut down noise even
more, not that there's much there to begin with.

All images are full size raw scans (no enlargement or reduction) with no processing of any kind.

Paul
--
The October Project 2003
http://www.wildwales.fsnet.co.uk/october/october.html
http://www.wilderness-wales.co.uk
http://www.photosig.com/go/users/userphotos?id=118749

Code Developer wrote:

>> http://www.wildwales.fsnet.co.uk/misc/100%25ice0.jpg

> Cannot get to the first three links you specified I'm afraid.

Oops, silly me, I probably shouldn't have put the % in the titles.

Try these; http://www.wildwales.fsnet.co.uk/misc/ice0.jpg
http://www.wildwales.fsnet.co.uk/misc/ice1.jpg http://www.wildwales.fsnet.co.uk/misc/ice2.jpg

> Yes, looks like ICE is quite effective :-)

Wait till you see it close up.

> Btw, did you make any other adjustments apart from colour balance? Just curious.

A slight contrast boost I think, plus a bit of sharpening.

Paul
--
The October Project 2003
http://www.wildwales.fsnet.co.uk/october/october.html
http://www.wilderness-wales.co.uk
http://www.photosig.com/go/users/userphotos?id=118749

Pat Bennett wrote:

> Very impressive, Paul. By the way, you accidentally inserted 100%25 in the links for the first
> three files, which is why the links don't work.

Yeah, I shouldn't have put the % in, the 25 seems to have appeared as if by magic.

> I have got a lot of slides in that sort of condition, so I am rather envious!

Good isn't it? I'd actually stopped bothering to scan slides a while back because getting rid of all
the dust and scratches in Photoshop was such a pain.

What do you think of the difference between ICE Normal and ICE Fine?

Paul
--
The October Project 2003
http://www.wildwales.fsnet.co.uk/october/october.html
http://www.wilderness-wales.co.uk
http://www.photosig.com/go/users/userphotos?id=118749

"Paul Saunders" <pvs1@wildwales.fsnet.co.uk> wrote in message
news:bn75of$fh9$1@newsg4.svr.pol.co.uk...
| Here are some more samples, this time of the GEM grain reduction. I chose a very grainy ISO 400
| slide to test it on. I used GEM on the maximum setting of 4. All shots shots are scanned with ICE.
| No other alterations.
|
| Various before and after shots;
|
| http://www.wildwales.fsnet.co.uk/misc/grain1.jpg
| http://www.wildwales.fsnet.co.uk/misc/grain1gem.jpg
|
| http://www.wildwales.fsnet.co.uk/misc/grain2.jpg
| http://www.wildwales.fsnet.co.uk/misc/grain2gem.jpg
|
| http://www.wildwales.fsnet.co.uk/misc/grain3.jpg
| http://www.wildwales.fsnet.co.uk/misc/grain3gem.jpg

Well I'm impressed with all of those before and after shots, you've got yourself a beauty there!

| Here's the full pic (at reduced size), this is a raw scan (ICE only), no corrections other than a
| little sharpening with Focus Magic.
|
| Sgwd Einion Gam in it's full glory - a far cry from the pathetic little trickles that are
| currently masquerading as waterfalls in these parts.
| http://www.wildwales.fsnet.co.uk/misc/einiongam.jpg

Yes, certainly in all its glory, you can say that again! I thought the other picture from your TR
report was fine, but compared to this, I can see why you call it pathetic!

Just an observation, and I know the image above is an older shot, but your digital shots seem to be
of a much higher quality than the film image - by quality, I mean that they generally look 'better,'
ie, shaper, better colours, etc. Is there a reason for this (or is it just my own point of view?),
or is it because you just haven't done much photoshop work on this film one yet?

| Paul

Ste

"Paul Saunders" <pvs1@wildwales.fsnet.co.uk> wrote in message
news:bn8o1d$hh5$1@newsg4.svr.pol.co.uk...
| Just a few more samples.
|
| For Ste, here's a small section of a newly developed slide, to show what kind of marks are typical
| on a brand new photograph; http://www.wildwales.fsnet.co.uk/misc/newslide.jpg

Thanks for posting this, and that's in a *lot* better condition than the other slide!

| And here's the same section scanned with ICE;
| http://www.wildwales.fsnet.co.uk/misc/newslideice.jpg
|
| At first there only appear to be a handful of tiny marks, but remember that this is only 2% of the
| area of the entire slide, so you can multiply this by 50. That adds up to hundreds of such marks,
| which can take some time to remove by hand.
|
| But if you compare these two images very closely (flipping between them with appropriate software,
| preferably enlarged) you'll see that there are actually hundreds of very tiny marks in the first
| image that have been removed in the second. Many of these may be the infamous "pepper grain" that
| Fuji has become notorious for (slight irregularities in the surface of the film), far too small
| and awkward to get rid of by hand. These may not be noticeable in small prints, but might be in
| large enlargements.

Yes, even though there's not that many marks on the sample (compared to the poor example you
previously posted), I can see various 'dots' when I look closely, but they are all removed with
'ICE.' I opened both the images and switched between the windows several times, and it looks perfect
after the ICE has done its magic! This scanner is certainly a great tool for you, and if it can do
this to all your thousands of old images, then it's certainly better value than a 300D!

| And for those interested in noise, here's a scan of a small section of dark shadow.
| http://www.wildwales.fsnet.co.uk/misc/shadownoise.jpg
|
| There is actually some very faint landscape detail in this scan, so what you are seeing is not all
| noise. There are options to scan an image multiple times, which would probably cut down noise even
| more, not that there's much there to begin with.

On my monitor, this just looks like a plain black image - so either you've posted the wrong image,
or it proves that there's no noise in the shadows! )

| All images are full size raw scans (no enlargement or reduction) with no processing of any kind.
|
| Paul

Ste

"Paul Saunders" <pvs1@wildwales.fsnet.co.uk> wrote in message
news:bn6gcg$l4g$1@news8.svr.pol.co.uk...
| Code Developer wrote:
|
| >> http://www.wildwales.fsnet.co.uk/misc/100%25ice0.jpg
|
| > Cannot get to the first three links you specified I'm afraid.
|
| Oops, silly me, I probably shouldn't have put the % in the titles.
|
| Try these; http://www.wildwales.fsnet.co.uk/misc/ice0.jpg
| http://www.wildwales.fsnet.co.uk/misc/ice1.jpg http://www.wildwales.fsnet.co.uk/misc/ice2.jpg
|
| > Yes, looks like ICE is quite effective :-)
|
| Wait till you see it close up.
|
| > Btw, did you make any other adjustments apart from colour balance? Just curious.
|
| A slight contrast boost I think, plus a bit of sharpening.
|
| Paul

That's some improvement! Also, I'm suprised at how many scratches and marks are on a transparency
(my first look at a scan of one though, as I'm 100% digital).

Can the software be used on normal images? It looks like it reduces noise, so might be an
alternative for unsharp mask? (mainly noticeable on the sky)

Ste

On Wed, 22 Oct 2003 19:53:10 +0100, "Paul Saunders"
<pvs1@wildwales.fsnet.co.uk> wrote:

>Pat Bennett wrote:
>
>> Very impressive, Paul. By the way, you accidentally inserted 100%25 in the links for the first
>> three files, which is why the links don't work.
>
>Yeah, I shouldn't have put the % in, the 25 seems to have appeared as if by magic.
>
>> I have got a lot of slides in that sort of condition, so I am rather envious!
>
>Good isn't it? I'd actually stopped bothering to scan slides a while back because getting rid of
>all the dust and scratches in Photoshop was such a pain.
>
>What do you think of the difference between ICE Normal and ICE Fine?
>
>Paul

I've just looked at them using Breezebrowser in slideshow mode, which gives a very good comparison.

Fine gets rid of the last remains of the dust in the sky, but it also removes the fine texture in
the sky, so I think that it is removing noise as well as just the IR channel for removing surface
artefacts. The result is that there is the sort of image softening which you get with Neatimage at
certain settings.

I think I would stick with ICE Normal, and use Neatimage for any further improvement, but would need
to do some more tests to be sure.

You've got some work to do, I should think, before it is full steam ahead on your
transparency backlog!

Pat

Pat Bennett From Cheshire in the UK www.cheshirewildlife.co.uk

"Paul Saunders" <pvs1@wildwales.fsnet.co.uk> wrote in message
news:bn6gcg$l4g$1@news8.svr.pol.co.uk...

> > Yes, looks like ICE is quite effective :-)
>
> Wait till you see it close up.
>

That's pretty impressive! Looks like it copes very well with dirt and scratches.

> > Btw, did you make any other adjustments apart from colour balance? Just curious.
>
> A slight contrast boost I think, plus a bit of sharpening.
>

That's the bit i was wondering about - I thought I detected some unsharp masking. Wouldn't have
noticed if I hadn't compared the two side by side though.

Regards, Shaun.

ste © wrote:

> Well I'm impressed with all of those before and after shots, you've got yourself a beauty there!

Doesn't seem to soften the image too much does it? I'll need to do some print tests to be sure.

> Yes, certainly in all its glory, you can say that again! I thought the other picture from your TR
> report was fine, but compared to this, I can see why you call it pathetic!

I'm beginning to wonder if it will ever rain again, but living in Wales I suppose I shouldn't feel
too pessimistic about that... ;-)

> Just an observation, and I know the image above is an older shot, but your digital shots seem to
> be of a much higher quality than the film image - by quality, I mean that they generally look
> 'better,' ie, shaper, better colours, etc. Is there a reason for this (or is it just my own point
> of view?), or is it because you just haven't done much photoshop work on this film one yet?

It's a combination of reasons, I'll take them one by one, in no particular order;

1. No Photoshop work. You're right, I haven't done any work on this in Photoshop, so that in itself
would improve the image somewhat, in particular the colour balance. Many of my digital shots
don't look that great before Photoshop either.

2. Narrow dynamic range. Digital cameras have a much wider dynamic range than film, which is what I
like most about them. It gives me far greater flexibility when post processing an image, in
particular to tone down the highlights and bring detail out of the shadows, resulting in a more
pleasant balance of light levels within the image.

Slide film is much more contrasty so the highlights are more likely to burn out and the shadows more
likely to go black. It's much more difficult to post process a contrasty image well, so although my
current allegiance still lies with film because of the greater resolution, rest assured that I'll be
defecting to digital as soon as the resolution improves to high enough levels. Digital will be much
better than film once it matures properly.

3. High contrast. That particular scene had a very high contrast (although others that day had far
more). The more water there is coming over the fall, the brighter it tends to be, so as you
reduce the exposure to compensate, that makes the surrounding area much darker by comparison. As
you can see, some of the water is overexposed and some of the shadows are underexposed, so the
contrast range in the scene was too great for the film to handle. Film can usually handle
waterfalls okay, but not when there's a lot of white stuff in the picture.

I usually use a spotmeter when photographing waterfalls, and in very dry conditions I've observed a
contrast range of as little as 3 stops. Normal falls are usually around 5 stops, but in very wet
conditions like these it can reach 7 stops or more. I think some of the falls that day were as much
as 9 stops. In the Sgwd Einion Gam picture I think the contrast was around 6 to 7 stops. Slide film
usually has a dynamic range of about 5 stops, my G3 has around 10 stops.

4. Crap film. I tried using Fujichrome 400 on that day, primarily to try to freeze the water with
fast shutter speeds, but I discovered that the film was crap (for my purposes anyway). Aside
from the horrendous grain, the colours weren't very good at all, there seemed to be a bluish
cast on the film, and it had none of the rich greens that I used to get with Fujichrome 50. So
although I can now alter the colour balance in Photoshop, the shots I took on this particular
film never looked very good compared to my other waterfall shots. (I did experiment with a
variety of film types in my early days, before settling on my favourite - Fujichrome 50, later
superceded by Velvia).

5. One more thing, *SPRAY!* Although a still camera has a habit of making action scenes like this
look pretty placid, you have to bear in mind that there were thousands of gallons of water
literally thundering over that fall. Have you ever *heard* a big waterfall in full spate? It's
really very loud, and more importantly, it throws off a tremendous amount of spray. You can get
drenched from 50 yards away, even if it's not raining. This creates two photographic problems;

5a. A haze in the air. Although you can't actually see the spray in the photograph, it manifests
itself as a watery haze. You said that it doesn't look as sharp, that's because of the haze. The
background of the shot looks slightly out of focus, but it isn't. As you can see from the motion
blur on the water, this was taken on a tripod with a small aperture. It's the spray that ruined the
sharpness.

5b. A watery film on the lens. Because of the spray, the lens can get soaked in a second or two.
Because of this it's necessary to try to protect the camera as much as possible, repeatedly wiping
the water of the front of the lens, while composing the shot. Before taking the shot I clean the
front of the lens as best I can, then hold something in front of it to protect it from spray while I
get behind the camera. Then I take hold of the cable release, reveal the lens and take the shot
instantly, before it gets soaked again. A further problem is that after taking a lot of such shots,
all my cleaning cloths get soaked so I've got nothing dry left to clean the lens with, so it ends up
with a permanent watery film on it even after cleaning, giving a "soft-focus" effect to every shot.

Some falls are more difficult to photograph than others, depending on the distance from the fall
and the direction of the spray coming off the fall. Close up full frontal shots are very difficult,
so a side view is usually preferable. This one wasn't so bad, being a reasonable distance from the
main fall.

The bottom line is that full spate waterfalls are very difficult to photograph, probably one of the
most difficult landscape subjects of all, in my opinion. Even if you have a brolly pod to protect
the camera from the torrential rain above, that won't protect it from the spray.

It will be very interesting to see how my digital copes with such conditions when they
finally arrive.

Paul
--
The October Project 2003
http://www.wildwales.fsnet.co.uk/october/october.html
http://www.wilderness-wales.co.uk
http://www.photosig.com/go/users/userphotos?id=118749

ste © wrote:

> That's some improvement! Also, I'm suprised at how many scratches and marks are on a transparency
> (my first look at a scan of one though, as I'm 100% digital).

Well that's not typical, it's a particularly bad example, which was why I chose it. Without ICE it
would be so much hassle to get rid of all those marks it would hardly be worth bothering.

Having said that, even brand new slides have a surprising number of tiny little marks on them, which
can still take some time to remove by hand. Trying to clean the dust off beforehand only seems to
make the problem worse, and many of the tiny marks can't be blown off, they're fixed on the film,
perhaps tiny particles of dust that fall on the film during development?

I'll scan a brand new slide to show you.

As for the older slides, I have many in that sort of condition, so this scanner is a life-saver.
It would be a criminal waste if I couldn't recover them, there's a lot of good shots amongst those
old slides.

> Can the software be used on normal images?

No, it's not actually software, it's hardware.

"It works by shining an infrared light through the film's emulsion, using the resulting scan
information to create a "defect channel" showing where the dust and scratches are located. The
infrared light passes right through the layers of most color print or slide film, but is blocked by
dust or scratches."

That's how it can tell the difference between marks "on the film" and marks "in the picture". I've
got a great technique for getting rid of marks in Photoshop, but it can get rid of real detail in
the photo too if you aren't careful. I was once removing all the light marks from a picture until I
suddenly realised that I was removing all the sheep from the hillside! :-)

> It looks like it reduces noise,

The ICE Fine setting looks to me as though it's using a slight blur filter, but I can't be sure of
that. The scanner does have a grain reduction feature though - GEM. I haven't experimented much with
this yet, but it seems to work quite well, although that softens the image too. Using a small amount
might be practical.

If you want to get rid of noise, download the Neat Image Demo NOW!
http://www.neatimage.com/download.html

Although it's called a demo, it's actually a fully working program with no timeout. It does have
some limitations compared to the versions that you pay for, basically it only works with jpegs and
doesn't work as a Photoshop plugin. Apart from that it's fully functional. Give it a try.

Be warned though, that the default settings are a bit on the strong side, and can soften image
detail quite a lot. There are loads of settings you can tweak though, so you can get very good
results if you spend a little time playing with it and figuring out how to get the best out of it. I
haven't mastered it yet though, so you can explain it to me when you've got it sussed! :-)

Paul
--
The October Project 2003
http://www.wildwales.fsnet.co.uk/october/october.html
http://www.wilderness-wales.co.uk
http://www.photosig.com/go/users/userphotos?id=118749

Code Developer wrote:

>> A slight contrast boost I think, plus a bit of sharpening.
>
> That's the bit i was wondering about - I thought I detected some unsharp masking.

It wasn't Unsharp Mask, it was Focus Magic.

Paul
--
The October Project 2003
http://www.wildwales.fsnet.co.uk/october/october.html
http://www.wilderness-wales.co.uk
http://www.photosig.com/go/users/userphotos?id=118749