Epson vs. Nikon - MF

[Ukko Heikkinen]

At long last I managed to locate a Nikon Coolscan 9000 demo scanner so that I could scan these examples. They are crops of about super A3-sized files, both "photoshopped" a little bit.

Ukko Heikkinen

 

[Ukko Heikkinen]

Sorry again - the names of the pictures are not shown. The left one is Epson, the right one Nikon.

Ukko Heikkinen

 

[Ara Ghajanian]

What size negative are these scans from? The Nikon looks pretty nice.

 

[Ukko Heikkinen]

6 x 6 cm.

Ukko Heikkinen

 

[T_om]

Not to denigrate your tests but I would make the following observations.

Strictly from a scientific inquiry point of view, comparison images that have been PhotoShopped are not really very enlightening.

Also, both scanners have default settings that you did not list, further obfuscating the results.

As most people that work in a digital environment come to know (and I am sure you know), it is a process, a progression of steps, not a single sample instance. You have presented the images as processed, not as the scanning step alone. This brings in so many variables that the results are not very useful.

A true comparison would have all the scanner defaults set as closely as possible to each other (sharpening, levels, curves, density, etc.) and then viewing the results without further processing.

I have no doubt the Nikon will be a stellar performer, I own the previous model, the LS8000.

But to do a "Epson vs. Nikon" test, a bit more precise testing structure would be more helpful to all considering these scanners.

Tom

 

[P C Headland]

Tom

I fully understand what you're saying, but what is important is the final output,. i.e. the end of the process.

If you can achieve a final output that is largely the same, surely it matters less than a comparison of the raw scanner output. If however a lot more steps are involved, or one requires an additional processing component, then it would be important to note that.

 

[Ukko Heikkinen]

Hello Tom

Hello Tom

I agree.

One should have a "perfect " standard scan.

Then one should try to make the test scans as close to this standard as possible.

I scanned the Nikon file in a camera shop using Nikon's "Super fine" mode and Nikon Scan. Black point was at 5 and white point at 250.

I scanned the Epson scan at home on my own scanner, using Vuescan and it's "Sharpen" setting,. This is a well considered and deliberate choice after a few years of trial and error. (I do not let Vuescan sharpen my 35 mm scans.)

Input resolutions: Nikon 4000, Epson 4800.

I "capture sharpened" both files gently, exactly in the same way.

Because the files were scanned on different machines using different software, and not very "scientifically" at that, I tried to equalize their tonality and contrast, again very gently, but not trying to achieve a 100 per cent perfect result.

Since both files were huge, I downsampled both to about super A3-size,, using a good Lanczos algorithm.

So far so good. However, such files are only raw material one has to transform into prints, and the prints are the only things that matter. That requires processing. And tools. And skill.

Naturally I printed both pictures after this "processing" which I did again in an "easy-does-it"-way. One has to inspect them very carefully in order to note the minor differences, and, honestly, I cannot say if one is "betrer" than the other.

One cannot see the differences at a normal viewing distance.

BTW, the camera was Rolleiflex Planar 1:3.5, on a tripod, f 8, released with a cable release, and the film was Ilford XP2.

My conclusion: the assertion that a good flatbed is good enough for MF is true. 35 mm is of course another thing.

We have discussed previously why this is so. Meanwhile I have been doing some additional research. Mayabe I will post some results some other time.

I have the raw scans of both machines. IMHO, posting small jpgs from them would be foolish and of no use to anybody, but if somebody would like to inspect them, I'd be glad to email small crops.

Regards

Ukko Heikkinen

 

[Nikon Bob]

My conclusion: the assertion that a good flatbed is good enough for MF is true. 35 mm is of course another thing.

Ukko Heikkinen[/QUOTE]

I was thinking the same thing myself after having read reviews of flatbed scanners that mentioned this. Looks like I'll take a chance on a MF flatbed scanner for my limited needs in MF. Thanks.

Bob

 

[Ukko Heikkinen]

Good evening, or mornig, Kris

Good evening, or mornig, Kris

Thank you for your private message, If I only coud figure out how to answer a private message, except in this way < g >.

The Epson scan is 29.092 inches x 29.533 inches. Downsampled it measures 12.402 x 12.57 inches. The posted jpg is about 100 per cent crop of a part of this file.

Input resolutions: Nikon 4000, Epson 4800, the highest optical in both cases.

Ukko Heikkinen