Question about Digital ICE on Nikon 5000 ED

I am new to this forum, and there seems to be some very knowledgeable folks on here. I have been scanning slides with the Nikon 5000ED scanner for quite a while now. However, I have noticed that when ICE is turned on objects in the image all have a halo effect around them. Also, stencils on aircraft are virtually unreadable. Here is an example. both are 100% crops of a 4000 PPI scan. I love the way ICE eliminates scratches and dust, but I am a bit alarmed by how much it alters the resolution of the image. Any advice on correcting this is most welcome. By the way, I am using the Nikon Scan 4.0 software.
Thanks,
Ray

With ICE on:



With ICE off:

Ray Wilhite said:

I am new to this forum, and there seems to be some very knowledgeable folks on here. I have been scanning slides with the Nikon 5000ED scanner for quite a while now. However, I have noticed that when ICE is turned on objects in the image all have a halo effect around them. Also, stencils on aircraft are virtually unreadable. Here is an example. both are 100% crops of a 4000 PPI scan. I love the way ICE eliminates scratches and dust, but I am a bit alarmed by how much it alters the resolution of the image. Any advice on correcting this is most welcome. By the way, I am using the Nikon Scan 4.0 software.
Thanks,
Ray

Click to expand...

Hi Ray,

What kind of film is that you are scanning, is it Kodakchrome? Because if it is, digital ice won't be able to handle it properly. The layers of Kodakchrome react differently to IR scanning, which is necessary for digital ice or FARE from Canon.

Cheers,

Cem_Usakligil said:

Hi Ray,

What kind of film is that you are scanning, is it Kodakchrome? Because if it is, digital ice won't be able to handle it properly. The layers of Kodakchrome react differently to IR scanning, which is necessary for digital ice or FARE from Canon.

Cheers,

Click to expand...

Actually this looks like some results I had with scanning Kodachrome and using ICE, but not setting the film type correctly. If you scan Kodachrome then you need to use the Kodachrome setting - other positive films just scan as positive (in Nikonscan).

Using the correct settings I generally get good results with ICE on Kodachrome as well as other colour films.

Mike

Mike Shimwell said:

Using the correct settings I generally get good results with ICE on Kodachrome as well as other colour films.

Click to expand...

Kodachrome processing can result in a significant enough level of silver residue in the film which will prohibit the Digital ICE (IR) light from penetrating unhindered. There shouldn't be much residual silver in the film, but poor quality processing can result in just that. So processing quality plays a role in the scanability of Kodachrome.

Besides silver-based Black and White film (so not the chromogenic B&W kind) and some Kodachrome, all other (color) films should be pretty transparent to IR, which allows the IR/visible light ratio to be a measure of contamination/scratching and be used as a sort of mask for the image in-painting process.

Cheers,
Bart

Thanks for the info!

Bart, Mike, and Cem,
Thanks for the quick response. DOH (slapping head) I can't believe I missed the Kodachrome setting! I am scanning a whole archival collection of images for the US Army Aviation Museum. There are numerous types of slides, but many Kodachrome. I see now that I need to scan those separately. However, as I think the below images will show, ICE distorts the image some no matter what setting you use. Thanks again for the assist. I think I now have the info to get the most out of ICE and Kodachrome. Images are 100% crops of Kodachrome slide scanned at 4000 PPI.
Ray

Original scan with ICE and using general "positive" media setting:



Scan using Kodachrome setting with ICE:



Scan with Kodachrome setting and no ICE:

Ray Wilhite said:

However, as I think the below images will show, ICE distorts the image some no matter what setting you use.

Click to expand...

Hi Ray,

Thanks for the feedback, Kodachrome it is. I suppose that the 'Kodachrome setting' just reduces the amount of ICE correction, but it can't escape the physics of silver being opaque for IR light.

What remains is to remove as much dust as you can before scanning. When you use compressed air cans, make sure they have clean air and don't spray propellant when you move the can. With a lot of scanning those cans become a costly approach. I therefore use a paintbrush compressor to supply filtered air upto a selectable pressure, and the spraybrush allows to adjust the flow as much as I need. For me this has proven to be the best solution for removal of dust instead of moving dust around. Of course one needs to work in a clean environment to start with.

One of the main issues with airborne dust particles + film is static electricity, which could be addressed by regulating the relative humidity and by the use of an ionizer (either a simple home ionizer unit, or even a modified piezoelectric igniter used for gas). I have tried carbon fiber brushes, but they still scratch the film surface a little over time. A clean(!) camel's hair brush seems to be a safe tool for carefully wiping off light dust particles. I suppose that the synthetic brushes used for sensor cleaning also work fine, although I haven't tried that on film myself.

For stuck debris and fingerprints, there is little else than a good commercial film cleaner.

Cheers,
Bart

Ray, another tip is to only use ICE on normal as the 'fine' mode can get a bit messy.

I'll try to scan a slide tobight for comaprison purposes.

I clean with a very soft makeup brush...

Mike

Thanks for the additional advice guys. I have a camel hair brush I am using to clean slides. Besides all the other techniques you guys have mentioned, Kodak recommends 98% Isopropyl alcohol as a cleaner:
http://www.kodak.com/global/en/consumer/products/techInfo/cis145/cis145.shtml
I gave that a try with pretty good results. I used the camel hair brush to clean away any cotton fibers on the slide. I was pleasantly surprised at how well it worked in fact. I'll see if I can get a before and after scan so you guys can see what you think.

Thanks again for all the help.
Ray

Hi Ray

Here are a couple of ICE tests I ran on some old Kodachromes - the bigger (second) one may not show at full res so if you want the full file pm your email (2.2Mb).

Cheers

Mike

ICE

Mike,
I see what you mean. thanks for the demo. One would assume that "fine" would yield better results than "normal" so it's very eye opening to see the dramatic difference between the two settings. I'm thinking now I'll use "normal" ICE for Kodachromes where detail is not critical and just clean the dickens out of those where fine details are important and scan them without ICE. I have thousands of slides to scan so I have to find a compromise to cleaning every dang one! You guys have been incredibly helpful. I have learned more here in a few days than the last two years i have been using the scanner! I just wish I had found this forum sooner. Thankfully, most of the previous 10,000 slides I scanned were not Kodachrome.
Ray

Ray Wilhite said:

Mike,
I see what you mean. thanks for the demo. One would assume that "fine" would yield better results than "normal" so it's very eye opening to see the dramatic difference between the two settings. I'm thinking now I'll use "normal" ICE for Kodachromes where detail is not critical and just clean the dickens out of those where fine details are important and scan them without ICE. I have thousands of slides to scan so I have to find a compromise to cleaning every dang one!

Click to expand...

Hi Ray,

As a compromise, you could scan the more important Kodachromes twice, once without ICE, and once with ICE. When you layer the non-ICE one on top of the Ice one in Photoshop, then you can mask out the visible debris and reveal the mushy version of that region. You should check for alignment between the 2 scans for the best results.

When you use VueScan as a scanner driver then it becomes simpler because you can scan a master Raw including ICE layer once, an process it twice (once without and once with ICE activated). That will save you from physically scanning twice with the problem of alignment between the scans.

Cheers,
Bart

Thanks

Bart,
That's another good idea. I'll have to keep that in mind as an option as I move forward with this project. As always, professional advice is much appreciated.
Thanks,
Ray

Ray

Glad that we were able to help a bit. Please stay around and share some pictures.

Mike

Mike,
Most of my photos are of Army helicopters. I'm happy to share if anyone is interested. Here is a nice shot I scanned of an AH-1G at Middle Wallop in 1975. I also have some fixed wing stuff if there are any aviation buffs on this forum. Thanks again for the help.
Ray

Ray

Middle Wallop, UK? Sounds like Somerset or Dorset? By chance, I visited Yeovil on the day the first EH-101 was handed over and was treated to a discplay that was put on for all the scambled eggs visiting for the ceremony.

Mike

Mike,
Middle Wollap, UK is correct. I should also point out that the above photo is from the Wayne Mutza collection. You have any helo photo requests?
Ray