tr?id=2006659569564293&ev=PageView&noscript=1
 
Theme: Quality issues
Q:

What are the optimal settings for Line-Art scans?

A:

when the preview is there, select as Scan type = "1bit line Art" in the SilverFast Frame-card > Scan type.

now you minimize the scan frame - in the prescan window - so that it covers just 3 or 4 lines of the text of your original and click the zoom button in SilverFast Ai window. SilverFast will generate a zoomed preview.

Next step is to call the "gradation dialog" in order to adjust the brightness. In the SilverFast Ai window three buttons left of the zoom-button, the brightness controll will be the only live slider.
You adjust the brightness so that in the zoom-preview the characters of the text are clearly seen - they should not merge with neighboring characters nor should they fade.

When the brightness is correctly set, click the zoom-button again to return to the normal pre scan and now enlarge the scan frame again to cover all the text you want to scan. Set the scan resolution to approx 600dpi.

Start the scan.


transp transp transp

Q:

Does it make a difference whether to apply the image optimization directly in SilverFast or later in Photoshop?

A:

Yes, it does. It is better to apply all the image corrections as early as in the SilverFast interface, because SilverFast is using the entire dynamic range of the hardware (scanner) thus applying all corrections on the true data BEFORE this data is "compressed" into 8bit data and transfered to the computer's harddrive or into the host application. SilverFast optimizations will NOT effect the data dynamic and no losses in tonal values occur, as would be the case when optimizing a ready-scanned 8bit file.


transp transp transp

Q:

Every time I try to scan b/w films I get really bad results, almost looking like line-art scans. What is wrong here?

A:

When scanning black and white film material make sure your film scanner's dust and scratch removal function is NOT IN USE. This function can not be used in combination with b/w film material.


transp transp transp

Q:

Bad scan result with black/white films of ASA speed 400, what can I do?

A:

- scan image with SilverFast as raw data file (16bitHDRgreyscale) at full optical resolution.
- load file into host application.
- (if negative: convert after scan)
- apply a tonal correction (levels) to your image.
- convert to 8bit image
- downscale to desired resolution.


transp transp transp

Q:

When I scan with SilverFast into Photo Impact 8 that scan is cut into pieces and set together in the wrong order. Is this a bug?

A:

This happens when the image preview of SilverFast was rotated before the scan.
Workaround: rotating the scan result in Photo Impact 8 various times will put the image together in the correct order.


transp transp transp

Q:

When scanning from photos or even film material (transparency) sometimes the scan result shows a regular pattern all over the image - this can't be a moiré pattern, can it?

A:

This pattern is often due to calculation algorithms in the software (problems with high frequencies in image).
Workaround 1: set SilverFast's interpolation algorithm to "Anti Aliased" (Option > General)
Workaround 2: only use optical scan resolutions. Set the "dpi"-slider in SilverFast to your desired resolution and click into the "dpi"-box at the right end of this slider. Now press the CTRL key on your keyboard and SilverFast displays the nearest optical scan resolution supported by your scanner. If the value in the entry box does not change, you selected a true optical resolution.
If the value changes, however, we recommend to release the CTRL button and set the value in that box to the value you saw when the CTRL button was pressed, thus using the nearest possible scan resolution for your scan.


transp transp transp

Q:

Using Nikon film scanner my negative prescan with NegaFix appears too dark.

A:

This effect may occur with some negatives and is due to the default values of the NegaFix Auto tolerance (NegaFix > Expert dialog).
Workaround: increase the auto tolerance in the NegaFix dialog.


transp transp transp

Q:

Final scan does not correspond to prescan in SilverFast. What is wrong?

A:

With RGB scans, make sure that the values for "Gamma correction" correspond in SilverFast (Options > General dialog) and in Photoshop (color settings dialog).
Also make sure that your RGB colour space profile chosen in SilverFast (Option > CMS dialog > internal) corresponds to the RGB working colour space profile selected in Photoshop.

With CMYK scans, make sure you selected the same CMYK separation table in Photoshop AND in SilverFast.


transp transp transp

Q:

Greyscale scans from my Nikon LS-30 are way too bright when I do a greyscale scan from an original colour negative.

A:

This is due to the nature of the hardware. The Nikon LS30 scanner's maximum output is 8bit (per colour channel). The NegaFix algorithm calculates on 16bit values, and if a device like the LS30 does not deliver this data, SilverFast just expands the delivered values (press the Alt key when in the SilverFast Histogram) you can see there that the values are expanded 8bit values.
There is currently no remedy to this. However, scans from b/w negatives or grey scans from colour positive films are o.k.


transp transp transp

Q:

Using a flatbed scanner I draw multiple frames in my SilverFast preview window. All frames show a color cast. This colour cast is not on the original.

A:

cause and remedy:
This effect may be due to some bad settings in SilverFast. You'll find out by re-setting SilverFast back to factory settings using the trashcan-icon.
in the vertical button-bar on the left-hand side of the pre-scan window. The first icon is an "i", then a question mark, etc.
Go down that row of buttons until you see a trashcan-icon. Click this icon until you get the message "do you want to delete all scan frames and previews" ? Confirm this dialog and the SilverFast prescan window goes blank. Now SilverFast is re-set to factory values.


transp transp transp

Q:

The SilverFast user interface has an entry field for Q-factor. What does the Q-factor actually do?

A:

In SilverFast you have the possiblility to determine the scan resolution by setting the Quality Factor (Q-Factor). This Q-Factor indicates the relation between line screen resolution (lpi) and scan reslolution (ppi resp. dpi).
A Q-Factor of 1.5 means that the scan resolution is 1.5 times larger than the line screen resolution.

How to calculate the maximum scan resolution with no information loss.?
desired output screen x quality factor x scaling factor = scan resolution

Example
scanning a photo that is destined to be printed at a print pattern of 175lpi, and the photo is scanned 1:1 size (not enlarged or downscaled) this is your calculation:

175lpi x 1.5 x scaling factor (1:1 = output scaling 100% = 1) = ? 175lpi x 1.5 x 1 = 262,5dpi

SilverFast runds it off to 262dpi


transp transp transp

Q:

SilverFast 5.2.x had problems with IT8 calibration on SilverFast(MicroIT8) for Artixscan models 4000, 2020 and 1100:
SilverFast's IT8 software versions up to ver. 5.2.0 could not completely eliminate strong green colour cast in scans.

A:

Cause:
the direct data output from the hardware delivers this color cast and the manufacturer software corrects this via a profile. SilverFast versions up to 5.2.0 could not use this profile, thus calculation is based on bad data.

remedy:
LaserSoft recommends to use the Microtek ICC profiles installed with manufacturer software instead of using SilverFast calibration profile. SilverFast can use this ICC target (setting SilverFat CMS to "ColorSync" and select that ICC Microtek profile).


transp transp transp

Q:

Horizontal lines in final scan on Macintosh. Occurs only when SF has a calibration profile selected in SilverFast's CMS dialog.

A:

cause:
when in Macintosh ColorSync calibration profiles for input (scanner) and for output (monitor) are preselected, and at the same time SilverFast is set to work with a profile - either SilverFast's own calibration input profile or a ColorSync profile - the calculation for output levels was allocated wrongly. So with RGB scans the output was considered CMYK (4 levels instead onf 3), and vice versa. This leads to horizontal lines in the final scan.

solution:
this bug is fixed in SF v.5.2.1

demo_06 update_06