Silverfast users?

Bormental · Aug 8, 2020

OpticFilm 120 Pro is out and I am hesitating. For $2,400 including taxes it needs to be a massive improvement over my current DSLR workflow. It comes with SilverFast. I've spent some time on https://www.silverfast.com/ reading the online manual and it promises to be a colossal pile of manure.

My fundamental problem with what I'm seeing is all these "editing" features. WTF they're for? Why do I need to "prescan"? Why does a scanning software need midtone, contrast and saturation settings? I want raw bits from the hardware saved directly to RAW. The only transformation I need from it is color inversion and orange mask removal. Can I just click a single button and get a nice 16-bit TIFF or (even better) RAW file so I can edit it using modern software I already have? Are these adjustments mandatory to get optimal quality?

Basically I am looking to minimize the amount of "Silverlight time" (ideally, just click a button to get 16 RAW files from the tray) but I do not want to sacrifice image quality. Is it possible? I have been googling for a blog post or something that offers a sane, modern scanning workflow with Silverfast without mentioning "special toolbars", "expert mode" and "image optimizations". I have Capture One, Photoshop, Lightroom and Affinity photo for that.

Thanks!

Alan9940 · Aug 9, 2020

You can scan to a raw file, then use NegativeLabPro or ColorPerfect to do the conversion. Personally, since I've used Silverfast software for 20 years I can nearly always obtain better results manipulating the software myself. But, learning to use Silverfast competently is certainly a rather steep learning curve IMO. If you're satisfied with your DSLR scanning, why not stay with that?

Bormental · Aug 9, 2020

@Alan9940 what is your workflow? i.e. what do you do usually in Silverfast to get a RAW file? I am trying to figure out what buttons/toolbars/steps are completely and utterly unnecessary. Thanks!

To answer your question: I absolutely love my DSLR workflow for 35mm B&W. But medium format is another matter: I shoot color in MF. Stitching is annoying, and I only get 6000x6000 image out of it (I know Portra 160 is capable of more). Also, color inversion and orange mask removal for color negatives is very time consuming. I have not figured out how to automate color inversion, for some reason every shot requires separate level adjustment for each channel.

Wallendo · Aug 9, 2020

The version of SilverFast usually bundled with scanners scans at 16-bit (28-bit color) and down converts to 8/24 bits on export. For an additional payment, you can but an upgrade which allows you to output as 16/48 bit TIFF files. It seems that the bundled version is crippleware to some extent with useful features turned off.

With a little effort and practice, you can scan negatives as TIFF files with proper inversion and mask removal, but few other changes. Export as aTIFF file, and do your final editing in whatever program you choose.

There are a lot of things you CAN change in SilverFast, but don't need to.

I personally use infrared dust removal. I export as an 8/24 bit TIFF which I later edit with PhotoShop and Lightroom.

I hated SilverFast when I first used it. The workflow seemed overly complicated and slow and the documentation was nearly useless. Once I figures out that most of the features promoted for this were unnecessary, I was able to deactivate the un-needed modules and now have a fast efficient workflow.

Hopefully a useful hint: Run the Pilot program only once for each film type when starting scanning, After doing this, remove all modules you don't want. After that for each new image, select pressman from the top of the window, adjust the scanning borders, take a quick peek at the histogram, and then click scan.

Alan9940 · Aug 9, 2020

Bormental said:
@Alan9940 what is your workflow? i.e. what do you do usually in Silverfast to get a RAW file? I am trying to figure out what buttons/toolbars/steps are completely and utterly unnecessary. Thanks!

To answer your question: I absolutely love my DSLR workflow for 35mm B&W. But medium format is another matter: I shoot color in MF. Stitching is annoying, and I only get 6000x6000 image out of it (I know Portra 160 is capable of more). Also, color inversion and orange mask removal for color negatives is very time consuming. I have not figured out how to automate color inversion, for some reason every shot requires separate level adjustment for each channel.

Take a look at this page for your scanner. It explains what's needed to create a raw scan:

https://www.colorperfect.com/scanning-slides-and-negatives/creating-linear-scans/

Conversion doesn't have to be time consuming. Take a look at NegativeLabPro. If he has a profile for your camera (probably does because there is a lot of 'em), you're all set. Scan as directed per his instructions, run NLP, tweak a few settings, if desired, and bam..nicely converted image in <10 secs.

PhilBurton · Aug 9, 2020

Wallendo said:
The version of SilverFast usually bundled with scanners scans at 16-bit (28-bit color) and down converts to 8/24 bits on export. For an additional payment, you can but an upgrade which allows you to output as 16/48 bit TIFF files. It seems that the bundled version is crippleware to some extent with useful features turned off.

With a little effort and practice, you can scan negatives as TIFF files with proper inversion and mask removal, but few other changes. Export as aTIFF file, and do your final editing in whatever program you choose.

There are a lot of things you CAN change in SilverFast, but don't need to.

I personally use infrared dust removal. I export as an 8/24 bit TIFF which I later edit with PhotoShop and Lightroom.

I hated SilverFast when I first used it. The workflow seemed overly complicated and slow and the documentation was nearly useless. Once I figures out that most of the features promoted for this were unnecessary, I was able to deactivate the un-needed modules and now have a fast efficient workflow.

Hopefully a useful hint: Run the Pilot program only once for each film type when starting scanning, After doing this, remove all modules you don't want. After that for each new image, select pressman from the top of the window, adjust the scanning borders, take a quick peek at the histogram, and then click scan.

@Wallendo Can you describe your combned Silverfast/Lightroom and Photoshop workflow. What parts of Silverfast are important and which parts should simply be ignored or not installed?

George Collier · Aug 9, 2020

I use SF with a V850 Pro almost exclusively for B&W neg scanning. I, too, was initially put off by the seeming unnecessary complexity of it's workings and controls, but perhaps without quite the level of hostility you might have. I have used German software products for decades though, and so I didn't have quite the reaction.
I agree, though, that much of the automation you won't want, but as someone suggested, you might try some of the time saving options, like the dust and scratch removal, just to convince yourself that they are not for you.
One comment I would make - you are converting an image from an analog (true continuous tone) to a digital (stepped) image. I use the Histogram and Curves control only to get me as close as possible to the final image. It's a quick adjustment from neg to neg, with the other aspects of the frame (saving location, size, resolution, etc) staying the same. I save to TIF, which allows opening in Raw and making adjustments, then PShop to finalize editing, things like adjustment layers, masking, things I can't do in SF or Raw. The reason I take the time in SF to come close regarding dynamic range and value placement is that all editing after that is digital on top of digital (I know, RAW is not destructive) so in the first conversion from CT to Digital, I want to get as close as I can. I could be overvaluing this difference, and if there are others who disagree, I'd like to hear.

Wallendo · Aug 9, 2020

For color, I form on Negafix and iSRD. I turn off sharpening and SRDx.

I have SilverFast automatically open the file in Photoshop and crop as needed and do any scratch removal that slipped past SilverFast.

I then import to Lightroom and tweak color, shadows, highlights, and exposure.

it works well for me.

Alan Edward Klein · Aug 9, 2020

George Collier said:
I use SF with a V850 Pro almost exclusively for B&W neg scanning. I, too, was initially put off by the seeming unnecessary complexity of it's workings and controls, but perhaps without quite the level of hostility you might have. I have used German software products for decades though, and so I didn't have quite the reaction.
I agree, though, that much of the automation you won't want, but as someone suggested, you might try some of the time saving options, like the dust and scratch removal, just to convince yourself that they are not for you.
One comment I would make - you are converting an image from an analog (true continuous tone) to a digital (stepped) image. I use the Histogram and Curves control only to get me as close as possible to the final image. It's a quick adjustment from neg to neg, with the other aspects of the frame (saving location, size, resolution, etc) staying the same. I save to TIF, which allows opening in Raw and making adjustments, then PShop to finalize editing, things like adjustment layers, masking, things I can't do in SF or Raw. The reason I take the time in SF to come close regarding dynamic range and value placement is that all editing after that is digital on top of digital (I know, RAW is not destructive) so in the first conversion from CT to Digital, I want to get as close as I can. I could be overvaluing this difference, and if there are others who disagree, I'd like to hear.

I have a V850 and previously used a V600. I've always used Epsonscan. I limit my scan adjustments to setting the black and white points to just past the histogram on both sides. I do all other adjustments and clean the dust spots in Lightroom. If you do too many adjustments for the scan, then you have to rescan if there's something you want to change. Plus you're basically learning another editing program - SIlverfast on top of Photoshop and/or Lightroom. I just don't see the advantage of that.

Bormental · Aug 9, 2020

George Collier said:
I use SF with a V850 Pro almost exclusively for B&W neg scanning. I, too, was initially put off by the seeming unnecessary complexity of it's workings and controls, but perhaps without quite the level of hostility you might have. I have used German software products for decades though, and so I didn't have quite the reaction.

George, thank you for responding. Can you clarify something for me please? Do image adjustments* actually affect scanning, or they're applied to the same scan data? Obviously, settings like ICE and double-exposure have direct effect on hardware, but others - it's unclear to me. In other words, what can I skip/ignore without affecting the image quality, assuming that I'll be doing my post-processing in something like Photoshop?

* Exposure compensation, curves, histogram, color correction, etc

P.S. I do not have anything against German software I am just allergic to "kitchen sink" software that tries to do several unrelated things within the same user interface. My perfect imaginary scanner shouldn't even need software. It should auto-mount as a new volume similar to a USB stick, with DNG files for each frame on the holder. Scanning would take place when you try to read/copy the files.

shutterfinger · Aug 9, 2020

The most critical correction settings are the White and Black points. These are the Zero and 255 settings on a histogram. The other controls are similar/equivalent to Levels or Curves in PS.
I learned to scan on a Microtek 8700 Pro with Silverfast AI 6. I took a sharp correctly exposed negative and scanned it on default settings at 600 dpi and saved it as my reference. Next I went to manual mode and started making scans at 600dpi changing one editing variable 10% and repeating until I got delectable distortion and compared each setting scan to the reference auto scan. Once I found the limit of that control I did the same with the next. Once through all controls singularly I adjusted them in pairs then in three. This exercise took about 18 hours spread out over a few days. I can pick up any scanner and software and get good to excellent scans at the start. Vuscan is the only scan software that uses its own terminology that is not relatable to CanonScan, EpsonScan, Microtek's software, or Siverfast. Its the utility wash basin IMO, the kitchen sink is an upgrade.

Alan Edward Klein · Aug 9, 2020

Bormental said:
George, thank you for responding. Can you clarify something for me please? Do image adjustments* actually affect scanning, or they're applied to the same scan data? Obviously, settings like ICE and double-exposure have direct effect on hardware, but others - it's unclear to me. In other words, what can I skip/ignore without affecting the image quality, assuming that I'll be doing my post-processing in something like Photoshop?

* Exposure compensation, curves, histogram, color correction, etc

P.S. I do not have anything against German software I am just allergic to "kitchen sink" software that tries to do several unrelated things within the same user interface. My perfect imaginary scanner shouldn't even need software. It should auto-mount as a new volume similar to a USB stick, with DNG files for each frame on the holder. Scanning would take place when you try to read/copy the files.

As far as I can tell, Epsonscan edits the data after the scan. The only hardware effects are things like ICE where the image is scanned twice. It's basically an image editing program which is why I save editing for afterwards. That way I don't have to scan a second time if I screw up the settings or want to change the results.

A scanner's resolution and dMax is limited by its hardware. You can't get blood from a turnip. Some people claim SF can get blood. But I haven't seen evidence of it.

Ko.Fe. · Aug 10, 2020

It took me long time to get how SF works. After I understand how it works it is superior to VS with my Plusteck.
SF has good feature for whataeffers. Hoover over the function and short description will show up.

warden · Aug 10, 2020

Bormental said:
.....and it promises to be a colossal pile of manure.

I never got comfortable with SF due to its user interface. Epson Scan and Vue Scan are both more to my liking and are simpler to use, although neither make a "better" scan. Like you I'm of the opinion that the ideal scanning software should have very few controls. Just get the data to the image editing software, which is where the complexity should be.

PhilBurton · Aug 10, 2020

Ko.Fe. said:
It took me long time to get how SF works. After I understand how it works it is superior to VS with my Plusteck.
SF has good feature for whataeffers. Hoover over the function and short description will show up.

@Ko.Fe. I'm still trying to decide which scanning software to use. In what ways do you find that SF is better than VS?

Alan9940 · Aug 10, 2020

jawarden said:
I never got comfortable with SF due to its user interface.

Funny, I'm exactly the opposite...never really could get comfortable with Vuescan's interface. Maybe, it's simply because I've used SF for 20 years and have had VS for a lot less time (and use.)

Billy Axeman · Aug 10, 2020

For a Plustek user with SilverFast you don't need much discussion here. If you want to try Vuescan how it works just download the free trial, and when it fits your needs the professional edition is very cheap (less than 3% of the cost for the Plustek 120 Pro scanner).

George Collier · Aug 10, 2020

Bormental - you ask a good (and germain) question about the point at which adjustments are applied in the scanning operation. I don't really know, but I do notice that when the file is saved, the progress bar has a couple of steps before the "Saving" step, so it could be afterwards. I have a friend in the graphic arts business who has been involved in product development for decades, kind of a very smart user that the companies listen to. I'll ask him.
Compared to Epson software, though (which I've used exclusively with Epson scanners up until the 850) the controls for curve and end point adjustments are dramatically more discrete, and the preview window (which can be significantly enlarged) is very accurate (at least on my 27" iMac display). My final adjustments in PShop are minimal, and I've never had to rescan due to an error in judgement. I've also done two scans for an image, to capture enhanced separation in different parts of the range (no change to size or crop) and dropped the second one into the file from the first scan, masking off what I don't need. This gives me a control over separate parts of the range that one scan can't possible accomplish without flattening some part of the curve, at least in my view.

138S · Aug 10, 2020

Bormental said:
Also, color inversion and orange mask removal for color negatives is very time consuming. I have not figured out how to automate color inversion, for some reason every shot requires separate level adjustment for each channel.

To me Silverfast has only two interesting features, one is Multi-exposure to recover deep shadows from ultra dense velvia, the othe one is color inversion for color negatives

See here min 21:50 :

Alan Edward Klein · Aug 10, 2020

138S said:
To me Silverfast has only two interesting features, one is Multi-exposure to recover deep shadows from ultra dense velvia, the othe one is color inversion for color negatives
...

How can multi exposure improve the dMax of the scanner?

Bormental · Aug 10, 2020

@Alan Edward Klein Great question! I've assumed that "dMax" means "max range of densities" but if it's tied to a maximum negative density (is it?), I suppose the answer is "lower noise". By increasing the exposure for the 2nd scan, a scanner can lower the signal/noise ratio in the shadows somewhat.

I have never owned a scanner with this feature. I assume it is only relevant to scanning slides

grat · Aug 10, 2020

Alan Edward Klein said:
How can multi exposure improve the dMax of the scanner?

From Silverfast's website:

How does Multi-Exposure increase the Dynamic Range?
SilverFast Multi-Exposure records an original's maximum Dynamic Range by performing a double scan with an increased exposure time of the second scan. This procedure captures the light image area's details in the first pass and the shadow details in the second. Afterwards an algorithm calculates the final scan, which now contains any detail, from each single scan.

George Collier · Aug 10, 2020

Grat - you are speaking of scanning a positive, right? In the case of a negative, the highlights would be the the second scan, correct, as they are dense on the neg? I've seen this also, and that's how I interpret it.

Alan Edward Klein · Aug 10, 2020

If slowing down the scan allows you to capture more detail and raise the dMax, why doesn;t Epson just do it? My thinking is the scanner is optimized for the best recovery the way it is. It doesn't make sense that Epson would underperform it and spec a lower potential dMax. Just slow the speed a little all the time and raise the dMax if slowing the scan improves it. That's like thinking your camera which has let's say an optimum ISO of 100 will collect more data if you overexpose the shot by one stop. You're going to have too many photons and overexposure the shot by 1 stop. You're not going to increase the "speed" of the sensor to an ISO of 50. The sensor is limited to its design just like in a scanner.

Bormental · Aug 10, 2020

Alan, I'm not sure which Epson you're talking about, but their V8xxx models have Dmax 4.0 without having to slow down, so why bother?

Silverfast users?

Privacy & Transparency

Privacy & Transparency