LaserSoft Imaging

Hello,

My problem:

*Scans of negatives look normal.

*Scans of positives are always too dark.
(I tried different kinds: Fuji Provia 400X, Kodak Elite Chrome 100 & 200 - same result)


*Scanner: Plustek Opticfilm 7600i
(with latest driver: v4.2.0.0)

*Silverfast Ai Studio v6.6.2r4a

*OS: Windows XP SP3

*Monitor: Samsung SyncMaster 910T (19" LCD monitor)

*Automatic IT8 calibration gave me an average delta E value of 1.0.


Software settings I use for both negatives and positives:

*Multi-Exposure: always on.
*Scan type: 48 Bit Color.
*Resolution: 3600 dpi
*Embedded color profile: sRGB (for positives)



I set my monitor contrast to 100 (range: 0 - 100),
and I set monitor brightness to 50 (range: 0 - 100)
(the normal for all my daily computer use).
When scanning negatives I set gamma gradation to the
standard value of 2.2 and everything is fine.
I do not need to adjust brightness or exposure.
Most negatives I scan look nice and natural.


But when I scan positives, the results are ALWAYS too dark. Even when I set the gamma gradation to the
maximum value of 3.0 and I raise my monitor brightness to 100 (maximum value), all my scans of positives are
still too dark. And even with multi-exposure the scans seem to lack dynamic range.

What am I doing wrong?
Is the darkness I see in my scans of positives a limitation of the scanner?
Maybe because of the low density range (3.5), the scanner does not scan positives so well?
Or is the problem something else?

Dear Mr. Fantastico,

Please make sure you have not accidentially set the scantype to 48bit HDR.
That is a raw image format which uses a gamma value of 1.0.

Also did you use the image automatic or set the highlights/shadows manually?
That is necessary to get the correct color-range in the final image.


kind regards,
ArneKetelhohn.

LSI_Ketelhohn wrote:Please make sure you have not accidentially set the scantype to 48bit HDR.
That is a raw image format which uses a gamma value of 1.0.

I made sure. The scan type is normal "48 Bit Color", not HDR, so that is not the problem.

LSI_Ketelhohn wrote:Also did you use the image automatic or set the highlights/shadows manually?
That is necessary to get the correct color-range in the final image.

No, I did not use any of the image enhancing functions in SilverFast.
I just do a raw scan and do all the processing in my photo editor software.

Only by setting gamma gradation in SilverFast to 2.8 and increasing the exposure by +0.5, +1.0 or +1.5
(depending on the photo), do I get acceptable results that are not too dark.

I guess in a way I solved my problem. I have to set the gamma gradation to 2.8 each time I scan
slide film, and then increase the exposure according to the requirements of the photo I am scanning.
Sometimes I need to increase shadows too to get good results.
It is just strange that I have to set the gamma gradation so high to get acceptable results,
when everyone else on the forum seems to get good results with the standard value of 2.2.

I still think the low density range of my scanner (Dmax 3.6) is a big weakness when scanning slides,
because it can maybe not see enough details in dark areas.

Mustavio Fantastico wrote:
No, I did not use any of the image enhancing functions in SilverFast.
I just do a raw scan and do all the processing in my photo editor software.

In that case a dark and or "foggy" image is normal.
Setting the Highlights and shadows is an essential setp in scanning.
If you want correct images you will have to do those adjustments in silverFast.
In Photoshop it is too late and you will loos image information.

This tells SilverFast where your image content starts and stops and which information should be stored in the final scan.
The scanner can see a much wider range of shades than available in the image.
e.g. the brightest part in your image might deliver a value of 220 to your device but in you image you would want that to be 250.
This is because the brightest value to your scanner is the direct light of its lamp.

Uncorrected highlights and shadows also compress the available image information because the color range will always have to be rendered in the smaller target color space.
If yo do not correct the values you will give away precious color space values to unused space which does not contain any color information.

Mustavio Fantastico wrote:Only by setting gamma gradation in SilverFast to 2.8 and increasing the exposure by +0.5, +1.0 or +1.5
(depending on the photo), do I get acceptable results that are not too dark.

That is normally not the best way to compensate the issue.

kind regards,
Arne Ketelhohn

Hi,

Thanks for the advice.

But then the question is: why do I not have this problem of dark scans when scanning negatives?
Why only with positives?

When I scan negatives I leave the gamma gradation at 2.20 and most scans look good.
But when I scan positives even with the gamma gradation on 2.80, my scans are still too dark.

Is there a reason for this?
Maybe positive film just has a higher dynamic range and is more difficult to scan?

Hello,

I will show an example from a scan of a photo made on Fuji Provia 400X slide film.

Both scans were made with Multi-Exposure, 3600dpi and 48-bit color (not HDR)
and originally saved in uncompressed 48-bit TIFF format and then resized
to 1280x868 resolution and compressed as 24-bit JPEG with 100% quality setting.

Scan #1: Gamma Gradation 2.8 and Exposure "+1.50" in SilverFast.
No other changes. No color manipulation. Result: http://i55.tinypic.com/2n9lrw7.jpg

Scan #2: Gamma Gradation 2.2 (standard value). No Exposure change (0.0).
Manually increased Brightness, Contrast and Shadows in SilverFast using button
"Gradation-Adjustment". No more changes in other programs.
Result: http://i54.tinypic.com/5ff0pc.jpg


As you can see, Scan #1 looks very nice, but the settings used in Scan #2
produces an ugly and bright "washed out" look and the shadows are still much
too dark and the colors are weak.
Only problem with Scan #2 (in my opinion) is...if you look on the far left
side of the image, you can see a building in the distant - and a very small part
of that building is washed out and invisible in Scan #2 because of the exposure
compensation and strong highlights. This is a small compromise, because in this
photo (which was taken under a bridge) the important thing is to get good natural
colors in the graffiti and to be able to see the top part of the bridge (which looks
totally dark and black in Scan #1).

Another problem I have is that when I adjust Shadows in a scan, they make dark
parts brighter, but also more purple, and then I have to use color compensation
to decrease the purple color cast, but this only works in an image that does not
have any other purple colors. In this photo I get a purple color cast on the top part of
the bridge but the graffiti also contains a strong purple color, and trying to decrease
the purple color cast in the top, also affects the colors of the graffiti and make them look ugly.

Now I am learning that with scanning there are always compromises.
The result is never 100% perfect or optimal.

I am not sure what to do anymore. I just think my method from Scan #1 produces
better results, even if it is theoretically the incorrect method.

I guess it is wrong to expect my 400 Euro Plustek CCD scanner to produce results
like a 20,000 Euro drum scanner.

Dear Mr Fantastico,

We further analyzed your issue and it is possible this could be caused by an error in the scanner itself.
If there is a problem with a heavily shifted color space (e.g. due to a lamp issue) the scanner might not be able to see the entire range of colors in a positive scan.

It is possible this does not effect negatives very much because the color information there is compressed due to the orange mask.
That is why the entire image information of your negatives still can fit into the shifted color space while the positive with its wider range of colors does not.


kind regards,
Arne Ketelhohn.