Minolta DS 4 vs. Plustek 7500i SE results

[Hellashot]

I finally got my Minolta DS 4 from Ebay yesterday and have comparative scans between it and the Plustek 7500i SE that I sent back.

The first photo is the Minolta the 2nd one is the Plustek scanned with trial version of VueScan. The Plustek came with SilverFast but it had no way to adjust scan "exposure" or brightness. Just choose a profile and that was it. Scans with SF of Kodachrome were dark.

VueScan may have had a crude tone curve adjustment capablity and from the result I got, it seems like it did some sort of "auto" exposure assuming there were bright highlights when there really weren't. Is the Minolta scan too green? Maybe. But saturation was at zero.

Does there seem to be a difference so far in Dmax between these scanners? Hard to tell. I have another example I'll post tomorrow that surprised me where I anticipated a bigger difference. But the Minolta software is definately superior for user controls.

[kevinkar]

Hmmmm, seems to be taking a lot of time to approve the images attached to your post. I know we're all busy but how hard can it be?:confused:

Hoping to see your examples soon!

[Hellashot]

I can see the two images, but maybe that's because I uploaded them.

Here is the daisy. First is the Minolta, 2nd is Plustek via VueScan, 3rd is Plustek via SilverFast software. This is the dark results SilverFast gave because there was no brightness or exposure adjustment for slides. All you could do is pick the type of slide (kodak, kodachrome, and another option I forget).

The VueScan color looks better because there was a color adjustment slider in VueScan and I obviously didn't set the color god enough with the Minolta. It was taken in the shade early in the morning. About a 1.5 second exposure.

[Hellashot]

Here are 100% crops of the Minolta and Plustek-VueScan daisy. One note is the VueScan scan was done as 8bit JPG and the Minolta scan was 16bit TIFF. Is the difference seen in how ugly the VueScan shadows are only 8bit vs. 16bit?

And the VueScan scan was at 3600dpi I believe and the Minolta is 3200. The VueScan scan was saved as about 13MB so compression was very low.

1st is Minolta 100% crop, 2nd is PlusTek-VueScan 100% crop

[Hellashot]

And lastly here are two scans from the Minolta. All of these slides were taken this year and are of course on Kodachrome :D

[lxdude]

Is the Minolta scan too green? Maybe.


Can't say for sure without seeing the original, but the Plustek seems more colorful.
I'm sure it would sell better. I can see dollar signs all over it. :D

[Hellashot]

Can't say for sure without seeing the original, but the Plustek seems more colorful.
I'm sure it would sell better. I can see dollar signs all over it. :D

I think the Plustek/VueScan image is dull in color and pushed too far creating highlights in the spruce that shouldn't be there, i.e. auto-exposure.

[lxdude]

I think the Plustek/VueScan image is dull in color and pushed too far creating highlights in the spruce that shouldn't be there, i.e. auto-exposure.

So the much more subdued brownish out of focus highlights in the background of the Minolta scan are more accurate, correct?
The highlights as seen in the Plustek scan certainly make for a less pleasing image. If they don't look like that in the original I could for sure see that as a problem.

[kevinkar]

You are running into what I do and that's lots of different output between which I can not decide which one is more "correct".

Colors are tough. Is the scanner calibrated? Is the software being used set up properly. Is your monitor calibrated? What kind of lighting is in your room. How good are your eyes? Is the slide faded? Etc.

Regarding the crops, it looks like the Plustek is generating more noise and a bluish cast that the Minolta does not. Minolta's whites are brighter. Detail between the two is about the same. No clear winner in detail. Because of the reduced noise and whiter whites, I'd take the Minolta here.

Regarding the spruce images, the Plustek looks more sharpened than the Minolta but the colors are oversaturated and the image is too bright. More and better detail is in the Minolta image. Again, I'd take the Minolta.

Regarding the daisy images, the Plustek/Vuescan images again have more noise and are over saturated as well as displaying a bluish cast. The center of the daisy looks a bit more cheerful but that's got to be due to the color saturation. The Plustek looks over sharpened again. While the color in the Minolta is a bit subdued, the image is overall better than the Plustek so, again, I'd take the Minolta.

The SilverFast one is just wrong so there's a software configuration you are missing. Probably a calibration issue.

The flowers and headwaters images are not bad. I like them.

I think you need to work the two issues separately. Hardware versus software. You have two scanners, X and Y, and three pieces of software, A, B, and C, which result in the following scenarios: XA, XB, XC, YA, YB, YC. It will take some time to figure out the base configurations against which to compare. Time consuming. Even using default values in each software will produce sufficiently different results so you can't really compare them.

It's very frustrating.

However, to me, the Minolta is still winning here!

Kevin

[Hellashot]

I don't think there is a way to calibrate a Minolta scanner. You can load "presets" for a scan but it didn't come with any. With slides, you're scanning the actual colors, unlike negatives film, so there really isn't anything to calibrate. You're scanning the colors the slide film has.

Plus negative film has seel "strong" or "true" or "imbedded" colors compared to slide film. Color correction is often needed with negative films compared to scanning slide film.

[kevinkar]

Hmmm, maybe the Minolta does not have a calibration routine nor a way to use any particular calibration profile files. That's too bad. Then again, that could be part of why it was so inexpensive in the first place. Decent quality scans without any of the other costly high end features like calibration. Fair enough.

However, I will have to disagree with your slides versus negative calibration comments. At first glance you would think that the scanner simply picks up the colors it sees from a slide and needs no calibration. Unfortunately, the scanner doesn't necessarily do that because it actually may see the colors differently than they actually are and produce output that is technically wrong even though the human eye may think they look good.

You probably already know this but, by calibrating, you scan a target with what should be known color values and compare the output to a target reference file containing the actual color values for those colors in the target. A difference is calculated that tells the software how different the scanner's output is from the known values and uses this information to generate a calibration profile which is then applied to all subsequently scanned slides.

So if your target patch at E15 is defined as 150,45,220, and the scanner produces an image with 155,48,216 (+5,+3,-4), a correction needs to be calculated and applied so that the output is corrected to 150,45,220. Not sure of the real math involved but let's be simple and say a difference of -5,-3,+4 needs to be applied in order for your output to match the input based only on the difference between the actual target values and what the scanner sees.

Since all slides are different, you can't necessarily use the same calibration for every slide scanned but calibration generally gives you an even starting point based on a known input which helps but does not solve all issues.

Negatives are actually tougher to calibrate because they also have the orange media containing the embedded image to contend with and that's never a known orange value that can be accounted for in the scanner processing. You can usually take a scan of one image on a roll of film and apply the corrections to the rest of the images on the roll but, next roll you put in will likely be different.

SilverFast does this with their NegaFix algorithm that uses a profile for each type of negative they include since, in general, all Kodacolor Gold 200 is roughly the same. All Fuji 400 is the same, etc., so they can apply roughly the same correction to the same type of film.

Even LaserSoft says they can't accurately *calibrate* a scanner to a negative film but they can calibrate to slides all day long....

I'm no expert. This is just the way I understand it.

Kevin