Back in 2005 I did an experiment comparing scans made by my Epson 900 scanner and my pictures taken with my Nikon 990 Digital Camera. At that time the Nikon camera had both better quality and was much quicker.
Fast forward to 2021 and I'm looking at some old View-Master slides and I have my newer Epson v600 scanner out anyway so I decide to get some quick scans instead of using my camera. While that was going to be the end of it, I noticed the scanner images were very, very good. I guess it's time to revisit this.
Equipment and Setup
For the camera I'll be using my Canon Digital Rebel EOS T1i (2009) and for the scanner I'll be using my Epson V600 Photo (2009). Yes, I know I'm doing this in 2021, but the Epson v600 is still being sold and I will be using my new Canon EF-S 60mm f/2.8 Macro Lens.
Digital Rebel EOS T1i (4752 x 3168 resolution)
Epson V600 (150 - 12,800 DPI)
Test Images
First up are our test images. These are slides taken by my father circa 1977 in Disney. The slides are your standard 35mm x 24mm Slides with no markings on the slides indicating any brand. The slides are still in excellent condition showing no warping, blemishes, or color issues.
Best Scanning Resolution
With the new scanner I have DPI options from 300-9800 so the first thing to do is find out the maximum useable scanning resolution. Time for some scans at 1200, 2400, and 4800 DPI.
Methodology. Use Photoshop to "upscale" all photos to the same resolution as the 4800 DPI Scan which is about 6300x4000. Then take crops of the same area and compare checking for fine detail.
When I first took a look at this I thought that the scans were maxing out at 2400 DPI but now I'm not entirely sure. Maybe in a few areas you can see a slight difference in quality between the 2400 and 4200 scans but it's incredible subtle unless you're at max zoom. It's not noticeable when viewing the scan full size on a monitor set to 1920x1080 even if you zoom in 200%. I see no difference between the 4800 DPI and 9800 DPI scans at any zoom.
At 2400 DPI I ended up with a resolution of about 3200x2000 and I'm going to call this the best "working" resolution.
Car Wheel Well: There is a difference between the 2400 and 4800 but you can't see this without being zoomed all the way in.
 |
 |
 |
| 1200 DPI | 2400 DPI | 4800 DPI |
Hotel Close-up: The difference between the 2400 and 4800 is incredibly small and again not noticeable without maximum zoom.
 |
 |
| 2400 DPI | 4800 DPI |
For a comparison, I also scanned some of the View-Master slides which at 11.75mm wide by 10.5mm tall are much smaller. I choose the newest reels to get the best possible quality. Interesting even with these being "professional" slides, there is less quality to the slides and they maxed out 1200 DPI with no difference between the 1200 and 2400 DPI scans.
 |
 |
| 1200 DPI | 2400 DPI |
Camera Setup
Unfortunately I don't have a camera slide adapter for my new camera and lens, so I had to get creative. I tried a couple of options starting with simply taping the negatives to a window, but I ended up with the best results when using my new LED studio lights and a soft box. The setup below shows this when I was imaging the View-Master slides but the imaging of the 35mm negatives was similar. Way to may pieces and parts and way to finicky to use this setup for any kind of volume scanning.
There does seem to be one Camera slide adaptor available at the moment which is the Nikon ES-1 52mm Slide Copy Adapter sold for about $60. Many people have said they got very good results with this but all mentioned having to jump through a couple of hoops finding the exact adaptors and extension tubes. No one has done a side by side quality test with most of them saying "the results are good but more importantly I have a lot of slides and this is much faster than scanning".
The closest focusing distance for my macro lens 0.65 feet (7.8 inches). I'm assuming the reason I got such good results from my old Nikon 990 was that it had a minimum focus distance of 2 cm (0.8 in) which is pretty incredible.
Scanner vs Camera
So, what kind of results did I get?
Hands down, the scanner wins. Better details, color, and contrast. It's possible I could do better with a Slide Copying Adapter but I don't think it would be significant. What it would be is quicker. Personally, if I were doing the project now I would purchase the camera slide adapter and test again only because of speed. If it were "close enough" I might go that route for speed alone. Scanning 9 carrousels of slides would take a lot of time and I could always go back and scan any particularly good photos.
 |
 |
| Rebel EOS T1i Camera | Epson 900 Scanner |
 |
 |
| Rebel EOS T1i Camera | Epson 900 Scanner |
 |
 |
| Rebel EOS T1i Camera | Epson 900 Scanner |
View-Master Images
 |
 |
| Rebel EOS T1i Camera | Epson 900 Scanner |
In Comparison with 2005
And how does this compare to the best of 2005?
Again the major issue seems to be contrast. The fine detail is comparable, but the scans really win with the shadows and shadow detail.
 |
|
| Nikon 990 Camera (2005) | Epson 900 Scanner (2021) |
9-6-2021