C750/Silverfast question

[omaha]

Edit: Title should be "V750", not "C750".

Ok, so I decided to try doing my own scanning. I've been getting low-res scans with my processing, and sending out for a drum scan when I wanted to print large.

My V750 just arrived today.

Got it hooked up to my Windows 8 machine. Works fine, by which I mean the Epson software can see it and run it and I've gotten some scans. In fact, my very first scan is right here:



Not too bad for a first effort. I'm pretty happy with that.

Here's the problem. The package also included the Silverfast software, but its giving me trouble. I went through the install, and it completed without error, but when I try to run it, it can't see the scanner. Obviously the scanner is hooked up to the computer: I used it to make that scan. Somehow, though, Silverfast is getting confused.

I guess I have two questions:

First, any tips on getting Silverfast to see the scanner?

Second, and more importantly, what is the purpose of Silverfast? I'm proficient in Photoshop, and while that image needed a bit of massage, it wasn't too bad. Am I correct in my assumption that what Silverfast will do for me is get me a better scan (ie, one requiring fewer adjustments) from the start? Other than saving me Photoshop time, is there anything that Silverfast does that I can't do with the Epson software directly?

TIA!

 

[omaha]

Ok, step 1 is completed. I had to DL new drivers and a new version of Silverfast.

Running as it should.

Lots to play around with here. I'm starting to think that question #2 was dumb...

 

[Doyle Thomas]

Ok, step 1 is completed. I had to DL new drivers and a new version of Silverfast.

Running as it should.

Lots to play around with here. I'm starting to think that question #2 was dumb...

LOL SilverFast vs Epson Scan = no contest

 

[omaha]

That's what I'm trying to understand.

If I'm going to end up in PS anyway to do final adjustments, what is Silverfast doing for me?

 

[artobest]

Once you get your head around it, Silverfast gives much finer control over scanning parameters and, somehow (don't ask me how), simply a better result. I wouldn't use it for things I would ordinarily do in Photoshop however, such as inverting negatives (don't like Negafix) and subtle curves adjustments. I simply scan everything as 16-bit positive and set black and white points for each channel manually, then perhaps a global curve adjustment to get the image histogram nicely in the middle, if need be. Sharpening and grain reduction are kept firmly off.

BTW, half the battle with Silverfast is setting it up. It's worth profiling the scanner with the supplied targets, and also worth searching for information about setting up the options, which can be confusing at first (and always!). Silverfast's own documentation doesn't really help, IMO.

 

[Bruce]

Silverfast does have some very helpful videos that explain operation of the software. Once you have Silverfast installed just look for the videos when you are using the different features.

 

[omaha]

Interesting. So you scan negatives as negatives and then invert in PS? I'll have to try that.

I'm still not sure I understand the differences in the options here. When I look through the Silverfast parameters, it looks like Photoshop to me. It seems like its all about color adjustments and such. If that's all there is to it, then I'm not sure what the payoff is in taking the time to learn a new interface when PS already does all that.

OTOH, if Silverfast somehow has the ability to coax "better bits" off the scanner, then that makes sense. But if the stream of bits coming in the USB port are the same (Epson v Silverfast), I'm not seeing any compelling reason to go with Silverfast.

Of course, I don't know the first thing about scanning (as if I haven't already made that much clear) and am eager to learn more.

Maybe that's a good way to ask this: Is the scanner hardware set up to send the same bits up the wire no matter what, or do the parameters you set in Silverfast actually change that?

 

[jeffgla]

I believe there are still different versions of SF. I have been using their software for a number of years and continue to use SF Studio (8 I think it is 8 I'm not home where it is on my home computer). I use the Mac version and find it to work very well. There should be informational material around. The version I have can be set to lead you through a scan. While I don't use that feature it could be a good starting point to become familiarized with the program. I use it as a stand alone and make adjustments with their program before making the scan and then open in PhotoShop which I used as a "darkroom" for finishing touches. I save both the scans and the PS files. Once you get used to it you will like it.

HOME

 

[artobest]

Just to be clear, I scan negatives as postives then invert, but I'm sure you knew that ...

My experience is that yes, Silverfast does extract 'better bits' from the scanner. As I said, I don't know why, but I seem to get better dynamic range, richer colour etc. Maybe it's the colour management in Silverfast - you really want to use that. For transparencies, I strongly recommend making your own profiles, but not with Silverfast - I get weird gamut holes in my Silverfast profiles. I use the (fairly basic) Monaco (X-rite) software that came with my scanner; however, I don't know if this is still supplied.

Having said all this, it's worth noting that many people are perfectly satisfied with Epson Scan ...

 

[omaha]

Suppose I have an over-exposed negative that (in darkroom terms) needs a long exposure time to burn through the highlights. Does Silverfast have the ability to tell the scanner to use a turn up the light to get more detail? If I adjust the "exposure" slider in Silverfast, does that translate into actual different (physical) behavior in the scanner, or is it just massaging the bits once they arrive in the computer?

 

[Alan Klein]

I use Epson scan with my V600. I don't have Silverfast. But I believe you are right. Notwithstanding all the arguments you'll hear, I am convinced that all the scanning programs are doing is post processing the bits received from the scan. It doesn't change the scan itself so how could it do more? You can scan flat and do that in a post processing program of your choice. The only caveat is that with some scanners, Silverfast may be able to change the speed or brightness of the lighting so that the "bits" come out different. However, I haven't seen any comparisons that really show any difference that matters in the end.

 

[Alan Klein]

One other disadvantage if you are messaging the bits using a scanner program during the scan. You get processed results that cannot be reversed. The only way to change the results is to re-scan, a laborious process. If you scan flat, you can re-do any post processing without having to scan again.

 

[Richard Boutwell]

Suppose I have an over-exposed negative that (in darkroom terms) needs a long exposure time to burn through the highlights. Does Silverfast have the ability to tell the scanner to use a turn up the light to get more detail? If I adjust the "exposure" slider in Silverfast, does that translate into actual different (physical) behavior in the scanner, or is it just massaging the bits once they arrive in the computer?

It is hard to say how recoverable the highlights will be without knowing how over exposed it is—there is a upper limit to the density the scanner is able to resolve data from—i'm not sure what that actaully is, but I doubt that it is the 4.0 as epson claims.

Silverfast and Vuescan both allow you to do an "hdr" scan that scans twice—one for the shadows/midtones, and one for the midtones/highlights and then combines them internally and the saves as a tiff. I couldn't find any concrete information on how this is done, but I gather that both scans are just done at the full native input from ccd and then adjusted two different ways to extract the "best" possible tones from each part of the scale, and then blended afterward into a "full tonal range" scan.

The way I have dealt with this issue in the past is to create a curves adjustment layer set the multiply blend mode, which does a great job of extracting detail from hugely over exposed scans (if done in 16-bit). If that kills or clips your shadow details you can adjust the input/output values of the lower end of the curve—like setting the black point to 50%, and then adjusting the rest of the curve to suit.

 

[omaha]

Here's my first attempt at getting a workable image off the scanner. First the initial scan (original file is a 24 bit TIFF scanned @ 3200 DPI):



Here it is after twiddling it in PS:



I printed it out at 17" x 22", and it looks reasonably good. I think there is some more clarity to be had, particularly in the faces. I'm going to try wet-mounting this negative and see what that yields.

Any advice on the scanner settings to use?

 

[Richard Boutwell]

Any advice on the scanner settings to use?

These look pretty good. I would only say to scan at 48-bit so you are able to edit with smoother gradations and without risking posterization of the 24-bit (8-bits per channel) file.

Are you using the negafix option or scanning as a positive and inverting/color correcting in PS?

 

[Richard Boutwell]

Are you using the negafix option or scanning as a positive and inverting/color correcting in PS?

I looked at the scan again and realized it is probably a slide rather than a negative. If it is a slide, I would just use the IT8 target and reference file to create an input profile, and then convert to your working space (Adobe 1998 or Prophoto RGB) once you get it into photoshop and before you start your corrections.

I tend to do an input sharpening step in photoshop before editing flatbed scans. I create a duplicate of the back ground, and run an unsharp mask filter with an amount of 10-12 and a radius of the same (10-12) threshold at 0. If it looks a little too sharp decrease the opacity to 60%-80%, flatten, and then continue with the rest of the color and contrast corrections.

 

[omaha]

Thank you for your feedback, Richard. I've read your comments on other threads, and am delighted to have the benefit of your expertise!

Actually, this image is off of a 120 format Portra 160 negative. The initial scan was done in "negative" mode using the basic Epson software. The resulting TIFF was loaded in to PS From there.

My next test on this is to try wet mounting the negative. Downside is I don't want to spend $160+ just to do the test. Any ideas on some fluid I could use (maybe baby oil???) that would be good enough for a basic test?

 

[Richard Boutwell]

My next test on this is to try wet mounting the negative. Downside is I don't want to spend $160+ just to do the test. Any ideas on some fluid I could use (maybe baby oil???) that would be good enough for a basic test?

I would still recommend using the kit from aztek. It is really the easiest thing to use with the 700/750. And if you are going to be making a lot of scans it is probably a worthwhile investment.

Don't use baby oil—it has fragrances and who knows what else added to it. You could use food grade mineral oil, which is really just a mineral oil laxative that you can get at the drug store. However, it is a PAIN to clean off platten glass (and films). The only reason I used that was Prazio stopped supplying the heavy oil for mounting on drums—even then it was a pain to clean the drum and the negatives—so you would still need to get the film cleaner if you use mineral oil—I have used drum cleaner to clean films in a pinch, but it can leave a residue. I've since moved on to the Aztek drum mounting fluid (same stuff to get for wet mounting on flatbeds)—it hasn't ruined a drum yet...(knockknock). I don't have any experience using lumina, or lighter fluid...

In any case, you will need the mylar an tape, Kami fluid disolves pretty much everything but the tape aztek sells. You just don't want fluid or oil seeping into the scanner.

 

[Alan Klein]

Omaha: I thin the PS edit is overdone, almost HDR looking. My question about the scan posted image is what settings did you use during the scan? I would like to see the scan "flat" without any adjustments at all except to reduce the resolution for the post here on this site.

 

[omaha]

I don't remember the exact settings...they are written down at the shop...but I used the Epson software and set it (IIRC) to 3200 dpi after selecting the scan area. Then scanned to a 24 bit tiff.

My digital files do tend to have too much "pop", but I find that gives me the look I want when I print on the Z2100 at the office. They don't look as garish on paper as they do on screen.

 

[Alan Klein]

The original scan appears to have some auto exposure settings set. If so, you're getting edits during the scan as well. Without knowing what those scan settings are, it's difficult to understand what's going on. Have you tried to shut of all auto adjustments and scan flat?

 

[omaha]

Quick update as I continue to play around with this thing...

I ran some tests tonight on the film holder. First, with the little "elevation tabs" in place. Then without them. Then I tried laying a negative flat on the glass.

The "flat on the glass" scan was dramatically clearer than the others. No comparison.

All things considered, I'm not feeling really good about this scanner right now. I've got a few rolls of film that will be ready tomorrow. We'll see how they come out, but at the end of the day, the target of my scanning is very large prints (24x30ish), and I'm thinking I'm going to have to stick with sending them out for drum scans. I just don't think this flatbed is going to get me there.

 

[Richard Boutwell]

he target of my scanning is very large prints (24x30ish), and I'm thinking I'm going to have to stick with sending them out for drum scans. I just don't think this flatbed is going to get me there.

What size are the originals again?

That is fairly big for 35mm (and 120)—even for a drum scan—golf ball sized grain and all. 4x5 and 8x10, no problem.

I have seen people who make 24x30 (and larger) prints from 8x10s with this scanner, but they are soft. The largest print I would consider with this scanner is around 16x20, maybe even 20x24. I knows others feel differently, but what is the point of making a big print if it is getting exponentially worse as the size increases. What I have told other photographers that I have worked with about these very issues is: Your time and money is just as valuable when you are making your own scans and prints as it is when having other people do the work for you. If you are going to spend resources making the work it deserves to be made as good as possible. If that means the print size is limited to 16x20 then it is better to have a good 16x20 print than a bad 24x30.

 

[Richard Boutwell]

What size are the originals again?

I'm sorry. I had to search bad a few pages to find it but you said it was a 120 portra.

 

[Doug Fisher]

>>The "flat on the glass" scan was dramatically clearer than the others. No comparison.<<

99 out of a 100 times when there is a huge difference as you describe it is due to the fact the software settings were wrong. If you had the film in a film holder and got the bad results then you most likely did not set the software to "film with film holder" and instead had it set to "film with film area" guide.

Even the Aztek demonstration page made this same error (unless they have corrected it since I last visited).

Doug