[interested_observer]

Here are two articles. One comparing a digital camera & macro lens against a drum scanner vs an epson v700. The second article is a how to with the digital camera & macro lens. My thoughts are that a K5 with the DA 100 macro lens should work just as well, using an increased number of shots. Also, Microsoft ICE apparently was not considered for stitching (I somewhat wonder why) - as it should perform just as well as the others.

• Best film scanner: Canon 5D Mark II vs Drum scanner vs Epson V700 | Addicted2light
• http://www.addicted2light.com/2012/11/29/how-to-scan-films-using-a-digital-camera/

[Medex]

Here are two articles. One comparing a digital camera & macro lens against a drum scanner vs an epson v700. The second article is a how to with the digital camera & macro lens. My thoughts are that a K5 with the DA 100 macro lens should work just as well, using an increased number of shots. Also, Microsoft ICE apparently was not considered for stitching (I somewhat wonder why) - as it should perform just as well as the others.

• Best film scanner: Canon 5D Mark II vs Drum scanner vs Epson V700 | Addicted2light
• How to scan films using a digital camera | Addicted2light

I think that digital camera without anti-aliasing filter would be better choice, e.g. Pentax K-1 or Pentax K-5IIs or Pentax K-3II or similar. Even more better with pixel shift resolution. Ir really works when you need higher resolution, even when you use kit lens. In my opinion the quality of digitized film is better with DSLR than flatbed scanner, e.g. Canon Canoscan 9000 mkII

[Jonathan Mac]

I have used my K200D & K3 with the DA 35mm macro to "scan" B&W 135 negatives. I think 35mm is a better focal length as the DoF is greater and should be better for getting a possibly curved negative in focus, plus it provides a more comfortable working distance between camera and negative.

This method produces much sharper results than a flatbed scanner but produces very flat, lifeless tones that need a lot of work in PP to get looking good. In the end I went back to my Epson V500.

[pathdoc]

I have used my K200D & K3 with the DA 35mm macro to "scan" B&W 135 negatives. I think 35mm is a better focal length as the DoF is greater and should be better for getting a possibly curved negative in focus, plus it provides a more comfortable working distance between camera and negative.

This method produces much sharper results than a flatbed scanner but produces very flat, lifeless tones that need a lot of work in PP to get looking good. In the end I went back to my Epson V500.

I've used the same trick with an Asahi Pentax slide-copier holder, a few 49mm filters to give some standoff distance and get the whole frame in*, and then the 35 macro. I use my AF540FGZ-II flash for illumination, dialled down manually to an appropriate power and triggered with a PC sync cable. I agree on the tones; right now I'm happy enough to be able to process film in the downstairs bathroom and digitise the images myself.



* = Even with a 35mm it's touch-and-go to squeeze the whole frame in on a K-5. If I'm right about this, I anticipate better results when I move to the K-1 and the field of view widens. I know I will have to crop at that point, but I have to do that anyway and my Raw Therapee presets seem to incorporate it into the process pretty nicely.

[pathdoc]

What it boils down to is this: Flatbed scanners always come in last place. Dedicated slide scanners do not give as much resolution but have useful features like DigitalICE and built in color toning that produce a more pleasing image with less work. A DSLR/macro setup can give you the highest resolution, but DR is lower, you will have to tweak the curves manually, and you are on your own with dust and scratches. Wet drum scanning doesn't quite match the resolution of DSLR, but provides vastly better dynamic range and inherently corrects for things like dust and scratches.

An interesting and useful summation of the pros and cons of each; many thanks.

[rgknief60]

Also, Microsoft ICE apparently was not considered for stitching (I somewhat wonder why)

If memory serves, I rejected ICE years ago when I read its licensing--that it cannot be used for any image you hope to sell. I've never sold anything, but did not want to redo my work if I ever wanted to do so. I'm not connected with either article. I uninstalled ICE once I saw the terms.

[LesDMess]

If memory serves, I rejected ICE years ago when I read its licensing--that it cannot be used for any image you hope to sell.

I can't conceive of a way they can determine a print used MS ICE but I wonder if they embed/encrypt something in the file?

[Murfy]

thank you for this, I did test a dslr set up and had some issues but I will give it a try again. I read somewhere, maybe here, that someone took a slide viewer that uses the trays and stuck their slides in, and then stuck the camera up to the lens in the slide tray. This gives me reason to try again

[rgknief60]

I can't conceive of a way they can determine a print used MS ICE but I wonder if they embed/encrypt something in the file?

Forensics have come far enough that they can take a file and tie it to a specific sensor. Sensors are like fingerprints. There is a way to encode virus/malware information in a JPG on a website that usable by a script/program. Shouldn't be too hard for MS to embed info they can track into files, or that they can scan by taking a picture of a picture. They are going to have to pay for Windows 10 somehow!

All that aside, I do try to respect the wishes of the software developer. Makes life easier for all concerned.

[LesDMess]

Forensics have come far enough that they can take a file and tie it to a specific sensor. Sensors are like fingerprints. There is a way to encode virus/malware information in a JPG on a website that usable by a script/program. Shouldn't be too hard for MS to embed info they can track into files, or that they can scan by taking a picture of a picture. They are going to have to pay for Windows 10 somehow!

All that aside, I do try to respect the wishes of the software developer. Makes life easier for all concerned.

I have a constantly changing sensor so I doubt any two frames - even from the same roll, would exhibit the same fingerprint pattern . . .

I am not aware that something can be activated from within a JPEG file but it might be interesting enough to examine some files for signs of any embedding other then EXIF data.

[rgknief60]

I have a constantly changing sensor so I doubt any two frames - even from the same roll, would exhibit the same fingerprint pattern . . .

Some days my sensor seems to have a mind of it's own as well. One day I have this picture taking thing down, then that sensor changes everything on me...

You might find this article interesting. Also corresponded with a university researcher who did some of the studies on this. She told me that for us it is important because if someone steals our copyrighted image, even if our copy is destroyed, as long as we can prove ownership of the camera that produced the image, it can be proved who owns the image.

A security newsletter I get reported on a researcher's proof of concept that it was possible to encode data into a picture (not metadata or append or prefix the virus to the file) and use a second file to harvest the info and act on it. So I have to assume it is somewhere out in the wild as well. I think is was in a newsletter that knowbe4.com published, but can't find it right off. The owner used to own Vipre antivirus, and now does security training for companies, in order to keep employees from giving away the keys to the kingdom.

[jcdoss]

One comparing a digital camera & macro lens against a drum scanner vs an epson v700.

My reaction when I saw the v700 vs DSLR scan in the first article...

Youtu.be

Youtu.be

[FoxbatK]

There've been several discussions on this forum regarding this subject before, and I've scanned slides and negatives using the DSLR technique before myself.

What it boils down to is this: Flatbed scanners always come in last place. Dedicated slide scanners do not give as much resolution but have useful features like DigitalICE and built in color toning that produce a more pleasing image with less work. A DSLR/macro setup can give you the highest resolution, but DR is lower, you will have to tweak the curves manually, and you are on your own with dust and scratches. Wet drum scanning doesn't quite match the resolution of DSLR, but provides vastly better dynamic range and inherently corrects for things like dust and scratches.

Thanks for the summary!

[skierd]

Wet drum scanning doesn't quite match the resolution of DSLR, but provides vastly better dynamic range and inherently corrects for things like dust and scratches.

I can only imagine the file sizes, but one of the methods I've been contemplating once I move in to DSLR scanning is to use pixel shift and multiple exposure stacking to pull the most out of the film. Example, take 3 pixel shift shots at -2.0EV, 0, and +2.0 EV, then combine the three.

[Jonathan Mac]

I can only imagine the file sizes, but one of the methods I've been contemplating once I move in to DSLR scanning is to use pixel shift and multiple exposure stacking to pull the most out of the film. Example, take 3 pixel shift shots at -2.0EV, 0, and +2.0 EV, then combine the three.

Pixel-shift + HDR would yield massive files but potentially tremendous resolution and dynamic range. I have thought of using HDR with my K3 but would need to get a different light source from the flash I use at present. A constant source such as a lamp or video LED panel would work, I imagine.