Scanning: Silverfast question - histogram and dynamic range et al

Started 8 months ago | Discussions thread
SilverFast-LSI
SilverFast-LSI New Member • Posts: 6
Re: Scanning: Silverfast question - histogram and dynamic range et al
2

Hi,

cdeyoung wrote:

Hi,

Quick question on Silverfast best practices (and no, I'm not looking for "use Vuescan" :-), been there...).

If I have a relatively low dynamic range negative, so the histogram is mostly clustered in the middle, I can move the histogram end points to set the black point and white point before I do the scan, and that looks like what I would expect in the preview (and pretty much in the final scan).

My question is, does this actually make any change to the data captured by the scanner?

Nowadays SilverFast will always acquire the maximum dynamic range (raw scan) available from the hardware during the capture phase. Basically it will always get the raw unprocessed data from the scanner and either apply the settings before exporting the final scan, or save the raw scan as an HDR-file. In the very early days of SilverFast there were some scanners, mostly production level, to which we had to upload a gradation curve when scannning so the settings in the software did make a difference to the data captured by the scanner.

Or is it effectively doing the same scan it would do regardless of settings, and then applying post-processing to the result, which I could just as well do in Photoshop afterwards with probably more flexibility? Am I actually throwing away data by doing this adjustment as part of the scanning process?

Not in the capturing phase.

And similarly, what other adjustments (exposure, for example) are beneficial to do before running the scan, if any?

MultiExposure is beneficial which captures more dynamic range in 2 passes from the hardware. One pass for the highlights and one for the shadows with the additional benefit of reducing noise which also increases the dynamic range captured - not to be confused with Multisampling, which just performs multiple passes at the same "exposure".

It occurred to me that the data captured by the scanner might be basically hardware-limited, and a flat un-adjusted scan might capture the most data - which would be best, as long as I'm willing to take the time to post-process elsewhere later.

That's correct, with the exception of MultiExposure. The flat un-adjusted "raw" scan also serves as the ideal archiving format if during scanning you are not yet sure for which output you actually want to process further to avoid having to rescan the original.

What's the best practice for best quality? If it matters, I'm using Silverfast AI Studio 9 with an Epson V850 Pro.

In general profiling your scanner with an advanced IT8-part 2 chart, applying MultiExposure during the capture/scanning phase, save as raw HDR(i) Tiff or DNG and process in SilverFast HDR-Studio in reference to your intended output.

Thanks!

-Chris

My pleasure! I hope my replies are sufficient to answer your questions. Don't hesitate to send follow-ups or drop a request. We're here to listen to user and customer demands for future updates as well as answer any questions. Safe scanning, y'all

- Jan

Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow