Fluid Mounting Epson V750

[more photography]

Now - I have tested my scanner and it seems to focus at the factory setting. My question is now I want to use the EPSOn FMA systems, is there any advatnges to using the Better scanning holder other the Quality and variable height adjustments.
Will it produce better quality images - due to the ANR glass and undermounting.

[Walter Calahan]

I would test the FMA frame that comes with the scanner first to learn if you like its quality. If not, then test the Better Scanning holder. The Better Scanning holders are made far better, and have the hight adjustment, but if you get perfect results with the Epson holder, why spend the money?

[timbo10ca]

I was looking at the Betterscanning mount as well. With some research I found a way to simulate it using the epson fluid mount system. It may not be as good, but it gives me pretty good results and I would be surprised if the Betterscan system made a noticeable improvement at printing sizes less than 16x20. When I have the money to buy it, maybe I'll test the theory.

I mount the film emulsion side down (toward the scanner base) onto the *bottom* of the fluid mount tray using Prazio anti-newton spray and blue 3M painter's tape (no residue left behind). I have fluid as well, but haven't tried it yet (more messy) because I'm happy with my results and I try to Keep It Simple. I found the sweet focus spot by doing coarse tests changing focus height by 0.5mm using pennies, dimes, nickles and quarters. I then found fine focus with different combinations giving closer to 0.2mm height changes. I didn't see the need to go finer because I was not seeing a noticeable enough difference to get into days of scanning using pieces of paper as my steps. My sweet spot with the film mounted on the bottom of the tray was 1 Canadian quarter and nickle high. I determined this by looking at 100% crops onscreen. I printed them off as well, and found them prints to coincide with my impression of onscreen sharpness (saving more time, paper and ink for future testing if necessary).

A couple of things I've found:
1) don't use multipass for large pieces of film if you are scanning to high resolution. The film moves slightly over time and when the averaging is done, it actually blurs your lines (really noticeable in brick walls)
2) Silverfast is a bloated program but gives better scans than the epson software every time
3) The bloat of Silverfast is still no match for post processing in Photoshop in my hands. Probably because I'm not very good with either one of them. For a beginner I can't see the benefit of learning Silverfast over PS. What I did was use some of the Silverfast tools to get the best result I could. I then used the PS equivalent (levels, curves, lightness/darkness, color balance, etc). PS was better for me.
4) regardless of what technique you use to scan, you need to get a good handle on sharpening in PS. I'm learning that it's not as simple as just applying USM.

I've read that multiexposure gives a much better dynamic range, but it requires a Silverfast upgrade.

Tim

[Keith S. Walklet]

1) don't use multipass for large pieces of film if you are scanning to high resolution. The film moves slightly over time and when the averaging is done, it actually blurs your lines (really noticeable in brick walls)

This would be a deal killer for me. While the Epson does a credible job, to achieve the highest quality, I find the SF multipass essential for minimizing noise in the shadows.

The rest of your approach under-mounted method is similar to one of the list I looked into with my V-750, except I used the KAMI fluid.

I am not certain if the blurring I experienced with that test was sagging, shifting, expansion, or Kami fumes ;-) or what, but it was not an issue with the approach that achieved my best results, top mounted with a full fluid envelope. With that approach, multipass worked well up until 16x, at which point there was a blurring and chromatic fringing. I have more tests to complete to see if that was a one time event or a consistent issue.

[timbo10ca]

This would be a deal killer for me. While the Epson does a credible job, to achieve the highest quality, I find the SF multipass essential for minimizing noise in the shadows.

I thought that's what multi-exposure was for...? I don't know why SF split up these 2 functions for any reason other than to make more money on an upgrade.

[Keith S. Walklet]

So I take it you are using the SE version?

The AI version ships with the V-750. I have to say that in this case, the software bundle, Monaco, SF AI with it IT-8 targets and multipass feature really add value to the upgraded purchase price of the scanner alone.

[aphexafx]

I thought that's what multi-exposure was for...? I don't know why SF split up these 2 functions for any reason other than to make more money on an upgrade.

They are split up because they are different:

Multi-sample) scans x times with the same exposure and averages them to increase the s/n ratio as a whole. This is simple, so I cannot believe SE is without it. SE Plus has it.

Multi-exposure scans x (4?) times and varies the CCD exposure on each pass. Each of these scans is then normalized and combined to produce a cleaner scan, especially where color resolution becomes limited (shadow/dark areas). This is more complex and exists as a newer feature in SE Plus, and Ai, etc.

The version of SilverFast Ai that came with my 750 is pre- multi-exposure, but it looks likes some people are getting upgrades simply by calling...

It seems like people are having variable luck with multi-pass scanning (either type) on the Epson V scanners. I am thinking that Keith's plexi laminate mount is probably highly stable in that regard. Another issue is the accuracy of the individual scanner's hardware, I suppose.

Tim, at what resolution do you begin to see misalignmet from multi-sampling?

[sanking]

Why would the plexi laminate mount have anything to do with the stability of multi-pass scanning? The problem with multi-pass scanning with Epson flatbeds, as I understand it, is that the stepper motors lacks the precision necessary. I don't understand how a system of mounting is going to change that.

Sandy





It seems like people are having variable luck with multi-pass scanning (either type) on the Epson V scanners. I am thinking that Keith's plexi laminate mount is probably highly stable in that regard. Another issue is the accuracy of the individual scanner's hardware, I suppose.

Tim, at what resolution do you begin to see misalignmet from multi-sampling?

[timbo10ca]

Why would the plexi laminate mount have anything to do with the stability of multi-pass scanning? The problem with multi-pass scanning with Epson flatbeds, as I understand it, is that the stepper motors lacks the precision necessary. I don't understand how a system of mounting is going to change that.

Sandy

There you go- makes sense. I thought it was the film popping from the heat from the light.

[timbo10ca]

So I take it you are using the SE version?

The AI version ships with the V-750. I have to say that in this case, the software bundle, Monaco, SF AI with it IT-8 targets and multipass feature really add value to the upgraded purchase price of the scanner alone.

I have the Ai version, but it's 6.5. I think you need 6.6 to get multi exposure.

[timbo10ca]

The version of SilverFast Ai that came with my 750 is pre- multi-exposure, but it looks likes some people are getting upgrades simply by calling...

Tim, at what resolution do you begin to see misalignmet from multi-sampling?

I tried emailing them a couple weeks ago but have received no response so far. I've only done multi pass at 2400 ppi and above, but from what Sandy is saying, should it matter? The weird thing is that it has only happened a couple of times.

Tim

[aphexafx]

Why would the plexi laminate mount have anything to do with the stability of multi-pass scanning? The problem with multi-pass scanning with Epson flatbeds, as I understand it, is that the stepper motors lacks the precision necessary. I don't understand how a system of mounting is going to change that.

Sandy

Agreed, about Epson hardware and alignment and precision. The mounting method would not change that, obviously.

It was mentioned, and I have heard others mention before, that film expansion was an issue for multi-pass scanning. Given this possibility, I assume that the plexi mount would hold the film and insulate it from lamp heat better than other methods. Therefore, I just meant that if film expansion IS a problem, then this method might be more stable. Personally I have never had an issue with film expansion, creep, or what-have-you.

[aphexafx]

I've only done multi pass at 2400 ppi and above, but from what Sandy is saying, should it matter? The weird thing is that it has only happened a couple of times.

Tim

There would be more alignment error for smaller steps at higher resolution...or rather, the smaller samples reveal the alignment error to a larger degree. So, yeah, I think you would see more alignment error the higher the scan resolutions. (if that is what you meant!)

So far mine has been doing fairly well at around 2400/4800 ppi. (fingers crossed)


On that note, both multi-sampling and multi-exposure could be implemented physically in a single pass and this would remove any hardware alignment and/or film alignment errors. I wonder if the scanner firmware prevents the driving software from doing this…???

[more photography]

I have Silverfast AII and upgraded to studio ME, tried it last night on a very dark shadow area, and there is a very marked improvement is shadow detail, tonality and noise reduction. I used the standard 5x4 epson holder, and there was no shift in the imag or halo effects.

Certainely early signs are very encourgaing, combine that with fluid mounting, we should have results

[Philip Harris]

Hi All

One of the most frustrating things about Silverfast (I'm using Ai version that comes with the Epson V750) is that whenever you get a good preview it seems impossible to keep these settings for the next image due to the auto-exposure being permanently active. I also binned SF as a result - it was an impossible workflow for making scans that had consistent colour. However, he control for turning the auto-exp off is hidden away in the Negafix control panel. Click on the expert button, then select expansion, go to the bottom and uncheck the auto box. This makes ALL the difference.

I think Vuscan is an excellent scanning application. If I didn't get SF with the scanner I would be using it right now.

Issues with multi passing - I really like this feature of SF, but the problem here, like many of you have notes below, seems to be the scanner. I've had some very good scans with a extension of tone using x4 pass, but I have found that most of the time there are problems with the images not overlapping properly. I've tried a number of methods, including letting the film warm up in the scanner, letting it run over-night to reduce any vibrations. To no avail. I think the problem is mechanical. Sure, the V750 is a good machine, but I'm not convinced that for high res scans the tolerances of the gearing are good enough to achieve accurate scans using the multi-pass method. I think what happens is that when the scanner head makes each pass it is in a very slightly different position each time.

I would be very happy if someone would prove me wrong!

I've got the fluid mount tray and I will purchase the fluid kit in the next month or so as I want to achieve a high standard and standardised work flow. I'll post feedback on the results as soon possible.

Kind regards

Phil Harris

[Philip Harris]

Timbo made a comment about there being more to sharpening than USM. Very true!

I use a method developed by Bruce Fraser (digi guru). Here it is below. Is also appears in Martin Evening's Photoshop for Beginners, CS3 etc. When you come to the USM settings you might want to increase the radius to 1.0 (or a bit more) since film users scan at image file sizes far exceeding the file sizes of DSLRs. Let me know how you get on.

Bruce Fraser Sharpening method
(see Martin Evening)


1. duplicate layer and double click to open Layer Style

2. change blend mode to 66%

3. Set Blend If to
This layer: 0, 230/250
Underlying layer: 10/20, 255

To split the values press the alt key when selecting

This has the effect of selecting out a range of shadows and highlights that will not be effected by the sharpening. The visual effect over the dup layer/high pass techinique is that the sharpening effect is more subtle and realistic.

4. Apply Unsharp Mask to this layer:
Amount: 320%
Radius: 0.6 pixels
Threshold: 4 levels

5. Go to Edit - Fade:
Opacity: 70%
Mode: Luminosity

6. Change layer blend from Normal to Overlay

7. Filters - Other - High Pass, set Radius of 2 pixels

Done