Finding focus - Epson V550 + Silverfast 9

[janew]

Total noob here. Bought a barely-used Epson V550. Developed and scanned my first role of film...

I'm trying to figure out where the "sweet spot" is for focus.

Got frustrated so I carefully scratched a pattern into the emulsion on a blank bit of the negative strip, using a very sharp mat knife.

I tried...
1. using the negative holder (and note the negative is curled, so it remained that way when scanned)
2. putting the negative directly onto the scanner bed, with a clean piece of plain glass on top of it
3. putting the negative _between_ two pieces of clean glass, and putting that on the scanner bed.

Scanned all three in Silverfast 9, at 600dpi with the sharpness correction set at zero.

All three of the images look acceptably sharp to me. But what bothers me is that they all look the same. Even at very high magnification. The ones that were smooshed under or between glass maybe look a tiny bit better. Why would 2 and 3 look exactly the same, when 3 is farther away from the scanner bed (by the thickness of one piece of glass)? Don't scanners have a "sweet spot," an optimum focusing distance?

Also, scanning the actual monochrome negatives at varying settings (monochrome and color), I never had Newton rings when using the (totally glossy) pieces of glass in 2 and 3. I thought I was supposed to...??

 

[George Collier]

There probably is a distance of optimum sharpness, and there might also be some depth of field. And your lack of newton rings is likely to do with low humidity in the air (your location indicates this is probable).
I would tape the neg, base side to the glass, which should use the curl to maintain contact. Use the blue painter's tape, which leaves no residue, being sure to pull slightly on the second piece to ensure that the film is pulled flat.
Set the software to indicate that you are using a film holder. Shim the glass up from the scanner glass, emulsion down, with spacers of some kind and keep records. You should see it increase in sharpness, then decrease as you find and then pass over the sweet spot. I'm going to do a post of this soon, in case anyone else cares to try it, with more detail.

 

[janew]

[Y]our lack of newton rings is likely to do with low humidity in the air (your location indicates this is probable).

Oh I didn't know that! And yes you're correct, it is _very_ dry here.

Thanks for the rest of your advice. It's very helpful!

I'm going to do a post of this soon, in case anyone else cares to try it, with more detail.

I'm looking forward to it!

 

[Alan Edward Klein]

The scanners are factory-focused for the film holder's height. Of course, due to lens aperture, just like in a camera, there's a depth of field range where the picture will appear to be in focus.

Try to use flat negatives in the holder. I found that was best in my Epson V600.

Note that I scan at 2400 and use the scanner's furnished program Epsonscan. You should be able to get more resolution than 600. Feel free to ask other questions. Good luck.
Alan

 

[Adrian Bacon]

The scanners are factory-focused for the film holder's height. Of course, due to lens aperture, just like in a camera, there's a depth of field range where the picture will appear to be in focus.

Try to use flat negatives in the holder. I found that was best in my Epson V600.

Note that I scan at 2400 and use the scanner's furnished program Epsonscan. You should be able to get more resolution than 600. Feel free to ask other questions. Good luck.
Alan

I’ll second here. Start with the supplied film holders. I’d only deviate from that when I needed to.

You won’t notice much difference at 600 dpi as you’re just not capturing enough spatial resolution to resolve the differences in focus, so the first step after using the supplied holders is to Jack the scanning resolution up to the native resolution that the scanners sensor has.

 

[janew]

I tried the same test again, but better. I made a fine scratch with a sharp mat knife, lengthwise on a 4-inch piece of thin transparent plastic.

I put it on the scanner, with one end propped up about 1/4" and the other resting on the glass. Scanned at 6400dpi, with the sharpening correction turned off.

Looking at the scan of the scratch at high maginification, I can't see any change in focus along its length (and therefore, at varying distance from the scanner).

To make it easier to handle the negatives, and to de-curl them, I think I'm just going to put them between two pieces of glass. I tried this and compared it to using the film holder. Couldn't see a difference. And IMO it's a lot less fiddly than the film holders. At least for the V550, they're really flimsy and hard to use.

Thanks to everybody for the replies, and good advice.

 

[shutterfinger]

I don't think a scratch on a piece of plastic or film is a good scanner focus target.
Download one of the lens test targets at http://www.normankoren.com/Tutorials/MTF5.html , print it on your printer at its highest resolution on heavy plain paper or use multiple copies of the 1951 USAF target http://www.takinami.com/yoshihiko/photo/lens_test/USAF.pdf
Mount the targets on a 20x30 piece of foam core, photograph the targets with your camera on a tripod, film plane parallel to the target, on ISO100 or slower film.
Inspect the processed film for sharpness with a loupe. Scan the film at different heights from the bed glass.

 

[janew]

I don't think a scratch on a piece of plastic or film is a good scanner focus target...

Yes I think I get this. And, for now, for me, I think I'm good to go. I got the scanner cheap, wanting to get started. I realize this might not be the best option.

And having scanned a few of my photos and looked at the resulting images quite carefully, I think the limiting factor is my skill with the camera. So that's job one, pretty much

 

[Les Sarile]

The scanners are factory-focused for the film holder's height.

I assume you mean that it was intended to be that way by design because I'm very certain there is no factory checking going on.

And having scanned a few of my photos and looked at the resulting images quite carefully, I think the limiting factor is my skill with the camera. So that's job one, pretty much

If you want to verify that the film itself has the detail captured on it and you don't think the scanner is getting the detail out of it, you might want to get something like a 40X microscope/loupe and a light box.

Untitled by Les DMess, on Flickr

 

[janew]

If you want to verify that the film itself has the detail captured on it and you don't think the scanner is getting the detail out of it, you might want to get something like a 40X microscope/loupe and a light box.

Thank you!

 

[Alan Edward Klein]

I assume you mean that it was intended to be that way by design because I'm very certain there is no factory checking going on.
...

Yes. I have no way of knowing if they actually check and adjust the nominal focus point. At least, it must be designed that way and the lens is inserted accordingly. THat's one advantage of using the film holders on the Epson V850. They're adjustable;e and you can check to see which height is the most accurate. I was surprised to find that there was a difference that is clearly notiecable.

 

[grat]

I assume you mean that it was intended to be that way by design because I'm very certain there is no factory checking going on.

Oh, do you work for Epson?

I suspect part of the problem, especially for used scanners, is no one remembers to put the locks back on before they move the scanner.

 

[Les Sarile]

Oh, do you work for Epson?

No just general manufacturing knowledge working in the industry given the cost of labor vs cost of product.
In fact, I'm surprised to see a premium product like Silverfast running on this model.

 

[Alan Edward Klein]

The Silverfast version is the low quality "freebie" version. They're trying to induce you to upgrade to their expensive and"unfree" version. It's like a lot of software included with hardware you buy.

 

[Les Sarile]

The Silverfast version is the low quality "freebie" version. They're trying to induce you to upgrade to their expensive and"unfree" version. It's like a lot of software included with hardware you buy.

Well at least it doesn't emblazon your scans with an intrusive watermark!

 

[Alan Edward Klein]

Well at least it doesn't emblazon your scans with an intrusive watermark!

I use Epsonscan. It doesn't emblazon my scans with watermarks. Which one does that?

 

[MattKing]

I use Epsonscan. It doesn't emblazon my scans with watermarks. Which one does that?

The free trial version of Vuescan.

 

[grat]

The Silverfast version is the low quality "freebie" version. They're trying to induce you to upgrade to their expensive and"unfree" version. It's like a lot of software included with hardware you buy.

1) The "free" version is a full version minus a couple of features that most users won't really care about
2) The "upgrade" is heavily discounted because you own an Epson

 

[Alan Edward Klein]

1) The "free" version is a full version minus a couple of features that most users won't really care about
2) The "upgrade" is heavily discounted because you own an Epson

I believe the free version won't properly scan color negative film. You need the upgrade to properly reverse the colors. I don't use color-negative film because of these problems and use Epsonscan for scanning BW and chromes.

 

[grat]

I believe the free version won't properly scan color negative film. You need the upgrade to properly reverse the colors. I don't use color-negative film because of these problems and use Epsonscan for scanning BW and chromes.

No, negafix works just fine (at least with SilverFast SE, which is what came with my V800). You're limited to 24 bit output, and you can't edit the negafix curves. Multi-exposure isn't available either.

The upgrade to SE+ was $99 when I provided them with my SE serial number.

 

[Alan Edward Klein]

No, negafix works just fine (at least with SilverFast SE, which is what came with my V800). You're limited to 24 bit output, and you can't edit the negafix curves. Multi-exposure isn't available either.

The upgrade to SE+ was $99 when I provided them with my SE serial number.

That's my point. You have to upgrade Silverfast to do it right.

 

[grat]

48 bit is not required. Nor is multi-exposure. 24 bit means you're "only" getting 16.7 million possible colors.

 

[Alan Edward Klein]

48 bit is not required. Nor is multi-exposure. 24 bit means you're "only" getting 16.7 million possible colors.

What about Negafix? Without that, you might as well use Epsonscan.

 

[grat]

What about Negafix? Without that, you might as well use Epsonscan.

No, negafix works just fine (at least with SilverFast SE, which is what came with my V800).

 

[Adrian Bacon]

48 bit is not required. Nor is multi-exposure. 24 bit means you're "only" getting 16.7 million possible colors.

I can't say I agree with that, though, to be fair, I'm not familiar with Silverfast or Negafix, however, having more than 8 bits per color channel isn't about the number of unique colors, it's more about how many discrete tone values you'll have available to you per stop of image density, and in that regard, more than 8 bits per color channel on the input does result in a lot less posterization and banding on the output. 8 bits per color channel on the output is fine as long as you're not going to do a lot of color manipulation, but it falls apart pretty quickly if you have 8 bits on the input and start converting from negative to positive with that. You really want to have at least 10 bits per (see cineon), as the minimum for input with more bits per just giving more flexibility for color manipulations without showing banding or posterization artifacts as a result.