Scanning Software Matters

This is a quick comparison of how VueScan and Epson Scan produce high quality images from negative film. I’m not a scanning or film services vendor; I’m just a film user who tries to get the most from his tools, and believes this information may be helpful to others who feel the same.

Introduction

You’ve gone to all the trouble to set the right exposure and composition in your film camera. You’ve been really careful developing the film. Now it comes time to scan the negatives, and it could all fall apart unless you know the ins and outs of your scanner software.

But that’s all automatic, isn’t it??  Isn’t all scanning software the same?

I admit it’s a rhetorical question. Of course they’re not the same, in some of the most important aspects, including the ability to fully manage the scanner and make it do what you need  it to do: Get all the detail possible from the negative without clipping (or dumping) the shadows and without clipping the highlights.

Well, you say, if the exposure is perfect and the development is perfect, won’t any scanning software accomplish the same thing? And my reply is… theoretically. But anytime you deviate (intentionally or unintentionally) from the ideal situation, such as pulling or pushing film, or departing from normal development time or temperature (or developer formulation), or missing the exposure reading by a 1/2 stop or so because of an aging light meter, then all bets are off. So you need a scanning software program that can direct the scanner to ‘record’ all the grains of silver in as many situations as possible. Or, you can continue throwing useful negatives away based on imperfect scans.

If you think about all the images (and/or negatives) you’ve trashed because the shadows were dumped or the highlights clipped once scanned, it might be time to question your scanning software and how you’re using it.

But first, you’ll need to trust that most modern scanners are highly capable machines. My 10 year old Epson V700 (rated at DMax 4.0) can easily record tonal differences between film base + fog (FB+F)(0% Luminance, L) and an adjacent shadow luminance of 0.1% L. It can record differences in the highlights contained WITHIN the brightest specular highlight, recording separate tones on a negative within a specular highlight of 100% L at the center and 99.7% L from a peripheral area of the same specular highlight. The dynamic range of my V700 easily records metered values from 2EV (I’ve got a bit of flare in my Sekonic meter) and >18 EV, and it’s all captured on my negatives and, most importantly, in my scanned images.

Of course, such fine differentiation won’t matter in a print except to create smooth tonal transitions and discernible shadow and highlight textures, which are important to fine art prints of large size. If your scanner (or how you’re using it) can’t differentiate very fine differences (which are actually delicate transitions) in film densities on the negative, no amount of processing and/or printing expertise will overcome that shortfall.

Let’s compare scans from VueScan® and Epson Scan®.

I know lots of photographers who use film. I know none (except myself) who use VueScan (from https://www.hamrick.com/).  Most if not all of them use Epson Scan that came with their scanner.

Before I get into this, let me say that I’m not associated with VueScan nor Epson in any way. What I share here is only from the perspective of a user, okay?

My strong preference is for VueScan because of its much greater capability to direct the V700 (or many other scanners) to read negatives having vastly different levels of density, contrast, color, and grain structure. I’ll cover some of those capabilities below.

I have rarely used Epson Scan. I think it’s grossly simplistic and constraining to users who want to get the most from their negatives. And even though its User Manual doesn’t even cover any of its so-called advanced adjustments for levels or curves, I’ve experimented enough with them to feel okay about sharing my experiences with Epson Scan.

On the contrary, I’ve used ViewScan for years, and have scanned several hundreds of negatives, both B&W and color. Its User Manual is thorough, and it needs to be because of the comprehensive range of settings available for a great number of scanner models. I’ve already covered my basic approach to getting the most from my negatives in this article, so I won’t go into them again. Regardless, my approach is merely a starting point and you’ll want to experiment and develop your own workflow.

For this article, I’ll use a single negative having an extreme range of subject brightness (EV -5 to +6, not including specular highlight reading). I scanned the same negative using both VueScan and Epson Scan, each optimized to record as much of the density range in the negative as possible.

The scene is test scene I use routinely to quality check my workflow. I’m lucky to have a pure white van. When  I open the rear doors on a sunny day, I get a composition having a full scale of brightness ranges, from the deep shadows within the van to the bright highlights (including specular highlights) on the exterior. I exposed TMY rated at 400 by metering the passenger seat up front and placing that EV on -2 (i.e., Zone 3). The brightest reflections from the white paint, not including the linear and specular highlights fell on EV +6 (i.e., Zone XI) using my Sekonic 758. Of course, the specular highlights, if I were to try to measure them specifically, would likely fall around EV +9 or above (or Zone 14).

Below is a picture of the scene and the negative made from it. The luminance range was extreme: from fine shadows in the cargo area of the van to very high and specular highlights on the sunlit areas of white paint. The goal in exposure, development, and scanning is to retain every bit of density range above the film base + fog into useable information for processing the negative, without clipping!

Test Image having wide range of subject brightness and dynamic range
Test Image
Test negative
Test negative (TMY rated at 400 ASA, developed in Gainer’s PC at 100%, 75 deg F, 6 min)

Ok, it’s a difficult negative due to the extreme density ranges, but let’s compare the optimized scans using VueScan and Epson Scan to see if there are significant differences in the image produced.


VueScan Record

Epson Scan Record

 

Shadow Detail Results

Shadow detail (1:1) from VueScan scan. Blue areas indicate clipped shadows created by selectively setting black point to allow direct comparison. Inset shows area of original image.
Shadow detail (1:1) from EpsonScan scan. Blue areas indicate clipped shadows created by selectively setting black point. Inset shows area of original image.

Comparison

Even though the levels and histogram settings for both scanning software was intentional to avoid clipping black and white points, this was much more difficult using Epson Scan due to its user interface (or perhaps my inexperience with Epson Scan). Regardless of the fact that I had to adjust levels in LR6 to make a comparison to the VueScan image, it was clear that Epson Scan compressed shadow detail, even to the point of losing details below Zone IV.  While the histogram for the VueScan image was nicely evenly spread out, the histogram for the EpsonScan image pushed nearly all shadow densities to the left.

The effect of this was particularly obvious in the 1:1 details of the same shadow area on the negative. Compared to the VueScan image, the tones of the passenger seat back and much of dashboard became completely featureless. Note these values weren’t clipped, merely compressed, by Epson Scan.

Both Epson Scan and VueScan treated highlights about the same. VueScan treated the highlights with slightly less aliasing of the grain clumps, creating more pleasing transitions across highlight tones.

Wrapping Up

If I were a photographer tied to Epson Scan as my sole tool to convert negatives into digital images, I probably would have thrown the test image described in this article in the trash after scanning it. Even after adjusting levels and curves to get the most information from the negative, Epson Scan’s inability to record shadow densities gave poor results. Where my light meter differentiated relative exposure values (rEV) between 0 (Zone V) and -3 (Zone 2), Epson Scan dumped Zone V to Zone IV and compressed Zones III, II, and I all into about Zone I (no discernible detail). This is a subjective conclusion more than objective.

Being confronted with the Epson Scan result, a smart photographer would go back to his ISO and film development workflow to try to get better separation of the shadows by shooting at less than film box speed and tweaking development time to compress highlight densities to avoid clipping. Sound familiar?

It’s possible that so many photographers recommend shooting film at 1-2 stops less than box film speed is because of scanning failures more than shooting failures. Without being able to fully control the scanner, via the right software, you’re left with settings and electronic algorithms developed by engineers, not photographers, and they develop them based on rather ideal shooting/developing scenarios that are actually quite rare in practice.

I used this extremely challenging scene to learn more about the differences in scanning between VueScan and Epson Scan scanning software. The negative was perfectly exposed and developed. The scan using VueScan was fairly straightforward and required very little adjustment to my normal scanning workflow. It gave me a image having useable information from Zone 1(L*=4.4) to Zone X (L*=100). The negative itself did not challenge the capacity of the scanner to record this wide range of densities (i.e., no clipping low or high).

This direct comparison between VueScan and Epson Scan indicated a couple things. First, the two software programs do not read or record the same negative equally, even after optimizing the levels pre-scan. Second, built-in settings are locked from the user in Epson Scan. Unlike Epson Scan, VueScan permits directing the scanner to adjust Exposure, RGB Gains, Film base colors, and Image color to tailor the scanning process to any individual negative. This flexibility is a great enabler when scanning any challenging negative.

The most obvious explanation for the difference between EpsonScan and VueScan might be their use of a different gamma correction built into the software, but I think it goes way beyond that. I’m working on a Mac and routinely rely on the native Mac gamma setting of 2.2.  Adjusting the VueScan curve in LR to represent a decreased gamma created a histogram much like the Epson Scan histogram. However, adjusting the Epson Scan curve in LR to mimic the VueScan histogram was not possible. The lost detail in the deepest shadows was still missing, and the shadow contrast too severe. It might be that an adequate correction before scanning in Epson Scan could give better results.

I’ll leave you with this:  Get to know your scanning software; scanning software matters. If your current program doesn’t let you manually adjust settings like Exposure, Levels, Curves, RGB Gains, FB+F color, and Image color, try VueScan or other program that does. Modern scanners are highly capable machines, but if you’re tied to an engineer’s guess as to what you actually need, you may not be getting everything you paid for, or that you need.

 

Do I Need To Drum Scan My Negatives?

Out of the Sack by J Riley Stewart
Out of the Sack

Photographers who scan film will eventually ask themselves whether it’s worth the cost to get a drum scan of their negatives. Let’s discuss how drum scanning compares to flatbed scanning to learn when you might need the extra quality provided by that mystical drum scanner.

I’ve used film as my image recording tool since the early 1980s. I once printed the images in a wet darkroom on sensitized paper, but have since moved to scanning the negatives and using the resulting digital images to modify and make prints using inkjet printers. This is called the hybrid workflow, and most film users use it today, partially or totally.

Resolution and acuity are two very common ways to assess technical quality of photographic images. Resolution refers to how much subject detail is retained in the image or print. Acuity refers to the sharpness of fine edges and lines.

I own and use both the Epson V700 flatbed scanner and the Howtek 4500 drum scanner. I’ve made thousands of scans from each, so I think I’m qualified to help you answer “Do I need a drum scan?”

My needs for scanning will likely differ from yours. I like to make large high resolution prints. I’ve sold hundreds of such prints, and many up to 32×40 inches in size from medium- and large-format negatives. My experience with 35mm negatives isn’t vast, but I’ve scanned perhaps hundreds of them for evaluation or printing.

Large size prints demand the highest level of scanning resolution. Any lapse in quality will utterly destroy the feeling of a print needing a high sense of texture or subject clarity. But how much resolution is enough? Can you get by with a more affordable, small flatbed scanner or do you need to drum scan your negatives?

I’m going to show you a real life test that will illustrate the difference in resolving power of a drum scanner and a modern Epson V700 flatbed scanner. I’m using a 4×5 negative, but you can reasonably extrapolate my findings to any size negative.

The image below “Out of the Sack” is from my “Afterglow” project. The same 4×5 negative was scanned two ways: (1) Epson V700, undermount, wet mount technique, 4000 dpi using VueScan™ and (2) Howtek™ 4500 drum scanner at 4000 ppi and a 6 micron aperture using DPL™ software. Both resulting scans produced negatives in the 320 MP range, which means printing them at 300 dpi makes a print around 60 x 48 inches.

The small white rectangle represents an approximate 2″ wide section of the large 60 x 48″ print. We’re going to look at this section in detail to compare the V700 to the Howtek scan.

Rectangle shows area of interest for comparing V700 flatbed and Howtek 4500 drum scans.
11 inch wide section of the 60×48″ print, showing the delicate fine details and textures in the burlap fringe around the area of interest.

Remember the 2″ wide area of interest is a very small section of the 60×48″ print, but is still large enough to elicit a sense of fine texture and detail in someone viewing the photograph on a wall.. But in a 30×24 ” print, the area of interest would be 1″ wide, and in a 15 x 12 ” print, 0.5″ wide. And at some smaller print size, you’d need a loupe to see those same details and texture.

This is important, because any difference in quality between the two scans will be diminished merely because of the size of the print, regardless of viewing distance. As the print gets smaller, the relative significance of any small section of the print also diminishes. So if you routinely print no larger than say, 16×20 inches, the conclusions I draw from this experiment will have far less importance than I draw from the large reference photograph.

Let’s see how the V700 and Howtek compare in a real-world scenario.

The images below are 1:1 depictions of the small rectangle in the image above (i.e., the area of interest). On the left is the drum scan (Howtek 4500) and on the right is the flatbed (Epson V700). Neither image has been sharpened. Which is “better?”

First, there are differences between the drum scan and the unsharpened flatbed scan. Resolving power of the Epson V700 is very, very close to that of the Howtek 4500. Almost every element (lines, shapes, tones) in the image can be found in both scans. To assess resolution I like to look at lines that create a “V” pattern (or intersectional angles) and at parallel lines that create a line of shadow between them. You can see that almost 95% of such patterns appear in both, in both highlight tones and shadow tones. There are only a small number of very insignificant angles and inter-line shadows that can’t be found in the V700 scan. This indicates that resolving power in the V700 is very close to the Howtek drum scanner.

Second, the main difference between the drum scan and the flatbed scan appears to be due to the higher acuity possible in the drum scanner. All edges are just a bit sharper on the drum scan, giving the appearance of higher resolution, but in fact it is not. Acuity can be best enhanced by sharpening the image (but remember, excessive sharpening can also degrade resolution).

The next image permits side-by-side comparison between the unsharpened drum scan (Howtek 4500) and the V700 scanned image slightly sharpened in Lightroom. Which is “better?”

Indeed, a slight degree of sharpening to the V700 scan improves, but does not equal, the acuity of the drum scan. But it comes really, really close. The settings I used to sharpen the V700 image in Lightroom were Amt 91, Radius 2.0, Detail 33, and Masking 94. This is a small degree of sharpening on a 320 MP image.

What do these comparisons tell us?

My results are not inconsistent with those reported by others. They did the same thing (perhaps not as definitively as I have) showing the differences in image quality between a highly magnified portion of a image when scanned with a flatbed scanner and with a drum scanner. The drum scanner always looks ‘better’ than the flatbed scanner.

I’ve gone one step beyond and showed that the V700 appears to resolve details almost as well as the Howtek 4500 drum scanner, but the V700 suffers a bit in obtaining the same level of acuity. Luckily, minor lapses in acuity can be nearly corrected by careful sharpening.

So, let’s put all this in context again. We began with a scan of a 4×5 ” negative that would produce a 60 x 48 ” print when printed at 300 dpi. 300 dpi is above the 240 dpi below which most healthy human eyes begin to detect a loss of resolution. I showed that the V700 (using my undermount wetmount workflow using Vuescan can resolve nearly all the detail resolved by the Howtek, but that minor sharpening of the V700 image is required to produce nearly all the acuity provided in the unsharpened Howtek scan. This is actually pretty remarkable.

Is the difference between the Howtek and V700 acuity important enough to warrant the extra cost of a drum scan? To answer that, you’d need to ask:

  • Could your eyes detect the difference I’ve shown if you were to closely inspect a 2″ wide section of a 60 x 48 ” print and say “..that’s not as good as it could be.” You’d not have the benefit of having a drum scanned print as a comparator as we’ve done here. So I’d hazard a guess that most would not be able to detect that small difference.
  • Do you routinely produce prints that are over 12x linear enlargements of your negatives, as used in this example? If so, you may need to eake out every bit of resolution and acuity that only a drum scanner can provide. Having said that, I’ve created many exquisite, fine prints to 40 x 32 ” from 120 size negatives, a linear enlargement factor of around 15x.
  • Do you produce high-quality negatives having sharp subjects? If sharpness (i.e., resolution + acuity) isn’t a consideration in your final prints, then it won’t matter during scanning either.
  • Do you know how to get the most out of your flatbed scanner? I’ve written several articles about scanning with the Epson V700 to maximize image quality as well as several other articles about the Craft of Photography. You may enjoy reading them.

So, do you need to drum scan your negatives?

The short answer is “it depends on the size of the negative and the size of the prints to be made from that negative.” For most of us shooting 35mm film and printing to not larger than 16×20 inches, a high quality, 4000 ppi scan from a modern flatbed scanner could make an exceptional print. But you’ll need every bit of scanning power to make a quality print larger than 16 x 20 inches, and that may mean investing in a drum scan. This also assumes a perfectly sharp, full crop negative from which you take the scan, of course. Still, you will be limited in print size: even drum scanning will not likely get much beyond 18×24 inch fine print from perfect 35mm negative. If you want large fine prints, you need to start with large negatives.


Update 6/23/20:

Some of you have asked for a 35mm comparison as I’ve done above for a 4×5 negative.  I don’t have many as I rarely shoot 35mm, but I can offer one example. The first image is the complete 35mm frame showing an inset box, from which I then show a 1:1 view of a small pink bicycle on a roof top, first from a V700 Scan and then from a drum scan. Both were sized to 4000 dpi (~24mp) and under a wet mount protocol. Both images were sharpened optimally (and modestly). 

Full frame of a 35mm negative
Comparison of 35mm drum scan to flatbed scan
The left frame is a 1:1 view of a 24mp image scanned with the Epson V700 at 4000 dpi. The right is from the same negative, but scanned using the Howtek 4500 drum scanner.

You’ll probably agree that there are noticeable differences in both acuity and resolution comparing the flatbed with the drum scans in the 35mm format. Keep in mind this is approximately what you would see in an 18.5 x 13 inch full-frame print (printed at 300 dpi). The differences would be less detectable in smaller prints, to a point where you’d need a magnifying loupe to see them. 

End of Update


Since I have both my own drum scanner and a good flatbed scanner, which do I use for my final scans of my larger format negatives? Well, I tend to be a perfectionist and I know the drum scanner can get me closer to perfection, so I use my drum scanner for final production, especially for my larger prints for exhibition. But I have no qualms about relying on my Epson V700 for all pre-production proofs, book images, or anytime the expected print sizes are less than 40×32″ or so, depending on negative size..

Today’s modern flatbed scanners are really quite good once you’ve mastered their capabilities. The linear CCD sensors used in these printers will likely never compare exactly to the PMT sensors of the drum scanners of yesterday, but who knows? As the PMT machines become more rare, they may become completely obsolete. While we have them, they definitely have their place in creating the highest quality of images in large prints. For small prints, especially from larger negatives, the cost of getting a drum scan is probably not worth it.

All comments and questions welcomed.

J. Riley Stewart, copyright 2020.

Wet Mount Scanning: The Undermount – Overmount Dilemma

Lowcountry Cypress, J. Riley Stewart, High resolution 40×32″ fine art print.

Let’s talk about different ways to wet-mount negatives on a scanner for the best scanning results.

For years I’ve been scanning my 35mm, 120, and 4×5 negatives using a wet mount workflow, using either a drum scanner or an Epson V700. I made the move to wet mount on my V700 following testing that indicated a slight improvement in image resolution, a slight lessening of dust spots, and since I could wet mount sections of 120 size film easier than using the native MF carrier.  Plus, after a bit of practice, I realized that wet mounting was just as easy as anything else.

To wet mount on any scanner, you need a mounting base (i.e., platen), the negative, mounting fluid, and a clear cover sheet. I use Kami wet mount fluid and high grade mylar cover sheets, both from Aztek. The film lies on a rigid platen. Between the film and platen is applied a thin coat of mounting fluid. Over the film lies the cover sheet. Between the film and cover sheet is a thin coat of mounting fluid. This set up provides an optically correct path for the illumination light to pass through the negative to the sensor, with (ideally) no diffraction and reflection as it passes through the platen, mounting fluid, or mylar.

Since any disruption to the light path could affect resulting image quality, carrier system design (i.e., platen, fluid, and cover) must minimize their inherent bad effects on  image quality. Thus, it’s worth talking about undermounting vs overmounting on a flat bed scanner.

By far, the drum scanner does the best job, but that’s because drum scanners use adjustable apertures and photomultiplier tubes instead of CCD sensors to record the data from the negative. Like flatbed scanners, drum scanners use a carrier system of platen (usually made of acrylic), fluid, and mylar to hold the negative during the scan.

How you sequence the  light source->>sensor is fixed when using a drum scanner. The sequence is always source->platen->negative->mylar->sensor. 

When applied to a flatbed scanner, this same sequence is referred to as undermounting, i.e., the negative is illuminated by the light source after the light has passed through 1 pane of glass (i.e., the negative carrier glass or platen) then the negative, the mylar, and finally the scanner’s platen glass before reaching the sensor. 

But typically, most negative carriers adapted for wet mounting employ the overmounting sequence, i.e., source->mylar (or cover glass)->negative->carrier platen> scanner platen-> sensor.

For my own work, I’ve adapted my V700 to permit wet mounting as described here and typically use the undermount method.

Betterscanning(TM) offers a new wet mounting carrier for the Epson V7xx/V8xx scanners and claims that with under mounting  “.. sharpness can be improved compared to other methods.”

I was intrigued about the undermount vs overmount dilemma and its potential difference on resolution of the resulting image.  Theoretically, either overmounting or undermounting wouldn’t seem to affect resolution. The single largest factors on a scanner’s ability to record with precision is the size of the light source (which is fixed on a flatbed scanner) and light scatter (a collective term to account for diffraction and reflection of light) between its source and the sensor.

Whether you overmount or undermount, the potential light scatter is the same (i.e., 2 layers of glass, 2 layers of fluid, the negative, air spaces, and the mylar. Resolution should (theoretically) be the same. But it’s possible that undermounting might be better if the mylar has significantly less post-negative scatter of the data then the carrier platen.

I wondered how much difference this phenomenon would make on my own V700 (wet mount with Kami fluid; neg mounted to plate glass with a  mylar overlay), and film planes equal at focal distance of my scanner. I scanned part of the film mark from a sheet of TXP 320 both ways (VueScan) at 4800 ppi. This is a screen shot from a LR comparison of the two scans; 100% view; no sharpening. On the left is the overmount scan and on the right, the undermount scan. Conclusion: No appreciable difference in resolution. 

Undermounting may have other important benefits; perhaps to reduce chromatic aberrations, but it didn’t affect resolution in my short test. I’m still intrigued, tho, and will explore further.

Other possible reasons to explain a lack of difference between overmounting and undermounting might be that other factors are limiting resolution, regardless of the configuration of the carrier system.  In my case, that could include the quality of my carrier (a piece of window glass), and internal flare.

-Carrier platen: For years I used a plain piece of window glass as my carrier platen. Because of this study, I wanted to rule out its quality being a limiting factor in image resolution. So I bought the V700 platen assembly, which includes the platen glass and holder that screws into the scanner bottom (Epson Part #1432338 housing assy., upper (includes glass), then carefully removed the glass from the assembly and set the distance to the correct focal length of the V700 lens. I re-scanned several negatives using the new glass in an undermount configuration, and then compared the results between old and new carrier platens and between overmount and undermount techniques.

Conclusion:  The benefits of the new optical glass over the old window glass were subjectively insignificant. I could not compare overmounting and undermounting using the optical glass, but that comparison was already made using the window glass, and since the optical glass and window glass were the same, I would not expect any difference between the configurations using optical glass.

-Internal flare:  Some have suggested that internal system flare caused by extraneous light (from the illuminator) around the negative might affect data quality coming from the negative. It makes sense, and Betterscanning(TM)’s new system includes a mask to apply around the negative to reduce internal system flare.  I’ve tried in the past to mask negatives using my own device, but found that it made no difference in image quality.

I recommend you test your own system of wet mounting to see if undermounting, carrier platen quality, or masking makes a difference in your scans.

See more J. Riley Stewart images here.

Unlocking Your Scanner to Get Great Scans from Film

B&W fine art image of the Nez Perce River in Yellowstone NP
“A Moment on the Nez Perce” by J Riley Stewart

Why Scan Film?

This tutorial covers how I manage my film scanning to get every bit of tonal information from a negative and avoid clipped shadows and highlights, which ensures I get great scans from my film images.

Scanning film is a necessary evil if you shoot using film cameras and want to convert the film image into digital format for proofing, sharing to the web, making a digital negative for alternative processes, or for inkjet printing. Scanning film requires a special machine (a transparency scanner) that passes light through the film and to the scanner’s sensor, and records minute differences in density and color at millions of sampling points on the film, and records those differences in a digital file.

For some images on film, we can relay on automatic settings in scanning software to adequately read and convert the film image to an acceptable digital image. But there will always be those film images that fail miserably during scanning; automatic settings and presets just can’t provide a digital image that we can process into a fine image for printing or sharing. For those film images, we need to learn how to force the scanner and scanner software to do the best job possible to get all the useful information from the film. 

Software can make a big difference in user experience. I’ve tried Epson Scan®, SilverFast®, and VueScan®. By far, I’ve found VueScan is the most flexible and effective, and what I say below relates to its use. VueScan, because it includes so many controls for so many scanners, can be frustrating to use sometimes, but if you follow my workflow, and refrain from experimenting too much, you’ll be okay (I think). Try VueScan® here: (I have no association with it). I have no doubt that if all your negatives (color or B&W) are perfectly exposed and developed, you won’t need the advanced tools built into VueScan (however, it will take care of those too) and so the basic scanning software like Epson Scan will work fine. But there are those times when things don’t go as planned, and you’re left with a great composition on a crappy negative. That’s when you need the tools and the skill to get every bit of useful information from the negative before you can make a great image. I’ve compared Epson Scan to VueScan in scanning B&W film here. Take a read if interested.

Take Home Points

  • Scanning film is a craft that requires learning, but folks, it’s not rocket science. You don’t need a degree in scanning to get the job done. 
  • Scanning film is a technical task. There is little usefulness for creativity during scanning.
  • Don’t ask too much of your scanner and scanning software: it’s only good for one thing, and that’s to get every bit of useful information from the film image as is possible.
  • There is no reason a properly exposed and developed negative can’t be scanned without clipping any shadow details or any delicate highlights, even specular highlights. But to scan some film images, you must know how to control the scanner and its software.

Unleashing the power of the film scanner

The image above (“A Moment On the Nez Perce River”) is an example from a scene having extreme dynamic range that can give automatic scanner settings the fits. Automatic scanner settings will often assume a perfectly exposed negative, perfectly developed negative, a ‘common’ film or type of film, a ‘common’ type of scene, and/or it assumes you want sharpening, dust removal, color balance, resolution, and any number of other assumptions. Perfect rarely happens in real life photography.

Scanning film is a technical task. There is little room for creativity, so don’t try to make it that. But, because it’s technical, realize that you must control the machine (i.e., the scanner and its software) to get the most out of it. Luckily, controlling the machine isn’t difficult if you use the right scanning software. Like any consumer machine, its controlling software has built-in defaults and automatic settings for folks who just want something fast. If you’re serious about getting great scans, you need to know to turn off all/most of the built-in presets.

Effective Scanning Workflow (VueScan®)

Initial settings:

  I scan the negative as a negative, both for color and B&W films. I sometimes deviate from this with color negatives I’m having issues with, but that’s not common.  Here’s my initial settings in VueScan using my Epson V700.

  Using these settings, click <Preview> and it will give a quick scan of the negative converted to a positive (like normal). Here’s what I’m scanning. It’s a Kodak Tri-X negative that’s been exposed through a Stouffer’s Step Wedge®. The step wedge has a minimum optical density at step 1 of 0.05 and a max density at step 31 of 3.05 in 1/3 stop increments. So it can mimic a scene having exactly 10 stops of luminance range.  Much of this is altered by negative development, which I won’t discuss here. The step wedge is merely a tool I’m using here to give you a standard ‘negative’ to talk about. Development of the negative can alter the actual densities from those 10 stops of light (i.e., by pushing or pulling exposure and development). But you’ll notice that while the darkest step (Step 31) is nearly black, the lightest step (Step 1) is far from pure white. The Epson V700 scanner has the latitude to record the 10 stops of density on film, and so does yours, probably.

Now the fun starts. We’ll set the scanner to capture every bit of information from a negative having 10 stops of density range (and more).  

This is a histogram VueScan produces using these initial settings. You can change this histogram (and the resulting image) remarkably by adjusting certain settings. I’ll show you.

  The initial histogram produced an image from 0% Luminance (L) to Zone VIII (82% L) above film-base + fog (FB+F). Shadows are on the left; highlights are on the right, just like a normal histogram. I count 27 separate peaks, including all tones that are combined within the largest peak at the far left of the histogram. So, I’ve lost 4-5 stops in the shadow areas of the step wedge negative. If I printed this image as scanned, it would clip the darkest 6 zones or so to black. That’s not good enough for me, and it’s important to know that we can get more information from the shadows than the initial scan provided.

You probably know that film development modifies the overall contrast of densities on the film. The measure of contrast is called the Contrast Index (CI). The more development you give a specific film, the greater its CI will be, normally from 0.4 (low contrast negative) to 0.8 (very contrasty) or so. Vuescan includes  several models (target film profiles), including many that enable matching of actual negative CI to model CI for B&W films. Go into the <COLOR> tab and you’ll see you can change the target film profile to several “model” renderings. The initial histogram above shows a <Generic Color Negative> model even tho I’m scanning a B&W step wedge. 

  You can experiment with varying film models under the <Color> tab. I like the Kodak TMAX 400 model as it allows me to select different contrast index (CI) levels so that I can better capture the shadows and highlights. Below is exactly the same scan, but using a different setting in the <COLOR> tab.

  You notice the left and right side of the range are now completely inside the extremes (levels) shown by the little red-green-blue triangles. You can stop here and be assured you’re getting every bit of data from the negative, with plenty of room to adjust levels and curves in post without clipping.

  If I do anything beyond this, it’s to set RGB Exposure (<Input> tab) to move the film FB+F peaks as far left as possible without clipping any important shadows. I want to eliminate FB+F as it has no useful information. In other words, during scanning I want to “expose to the left,” just the opposite of shooting a digital camera. Remember that when scanning a negative, digital noise (if any) will occur on the far left of the histogram, where negative densities are least. So don’t be too aggressive about this setting; don’t move your deepest important shadows too far to the left. When done right, you should still still the FB+F peak.

Setting RGB Exposure in VueScan is the one of the pre-scan settings  (i.e, machine settings) you can select on the Epson Vxx models. VueScan is the only software I’ve tried that is able to direct the scanner to slow the light emitter bar based on the RGB Exposure setting.  All other adjustments you make in software is performed on the image after scanning it, similar to PS or Lightroom does, which adjusts pixels and can cause artifacts. So there’s value in adjusting the RGB exposure to fit the negative.

The other important machine setting enabled by VueScan is the R-G-B analog gain controls that can become important when scanning color films. Film models will often presume specific R-G-B gain settings, so you need to check to see if those settings are causing your histogram to ‘dump’ tones.

Below is the exact same negative scanned using 20% more exposure so that the film base plus fog (FB+F) falls as near to the left as possible: That first little peak is FB+F.

  To recap, I’ve selected a film model that gets the shadows away from the left edge so I can clearly differentiate their peaks on the histogram, then I’ve increased RGB Exposure to ease them back to the left without clipping any important shadows (FB+F is not important).

  This is where I stop and hit <SCAN>. It’s the best I can do with this negative under any circumstances. Granted, I may need to optimize my exposure and development of TXP to get the deepest shadows away from the film base.

I’ll give another example of an Ilford FP4 negative exposed through the Step Wedge and given my standard development for FP4.

The first histogram results by scanning the film using a <GENERIC FILM> preset. Note how both the shadows (empty areas + Steps 1-3 on left of histogram) and the highlights (Steps 29-31 and blocked areas on right of histogram) are not well separated. In a final image or final print, this “automatic” scan would have clipped shadows and highlights. 

The next histogram is from the exact same negative, but the settings have been explicitly set to optimize the output image file to separate the various step wedge densities. In this scan, the RGB Exposure was set to 1.12 to move everything (all densities) to the left. Then I chose a film model having a contrast index (CI) of 0.70 to best match the actual CI of the negative. You’ll see now that every one of the 31 steps of the step wedge is discerned in the histogram, and can therefore be realized in a post-processed image or print. No important shadows nor any highlights will be clipped in the image.

I hope this was helpful, and I hope it doesn’t sound too complicated; it’s really not. It takes some time to learn the essentials of scanning negatives, but once you’ve invested in that time, you’re good to make great scans forever.

 I wrote a tutorial on how to easily convert your flatbed scanner to do wet mount scanning without buying special holders. I can perfectly wet mount 2 strips of 120 or 2 4×5 negatives using this technique in about 30 seconds. Here’s the tutorial:

https://www.jrileystewart.com/blog/2018/03/02/easily-convert-your-flatbed-scanner-to-a-wet-mount-workflow/

Happy scanning!

Subscribe to my newsletter “Under the Darkcloth” here.

 

Easily verify the focus on your flatbed scanner

Wet Mounting on the Epson V700 V750 V800 V850

This article describes a way to make sure the placement of negatives above your scanner’s platen is within the depth of field of your scanner’s lens for best sharpness.

If your film holders don’t place the film where the scanner focuses its lens, then you’re going to get fuzzy images. And that may be causing you to pull your hair out. Neither are good situations to find yourself in.


Now, let’s talk about checking the focus of your scanner.

Many flatbed scanners have a fixed focal distance set by the manufacturer, and they provide holders that place the negative “precisely” at that distance.  Most of the time, this works well.

But there are times when you might suspect that the fixed focus (or even auto focus) isn’t behaving right. If you are getting consistently blurry scans or scans that you believe should be sharp and aren’t, you may have a scanner that isn’t focusing precisely, or a holder that isn’t placing the negative where the lens is focusing.

In this short article, I’ll show you an easy way to verify the depth of field on your flatbed scanner’s lens.  Then, you can take steps to make sure your negative carrier places negatives precisely at the optimum distance above the scanner platen for maximum sharpness and resolution.

To verify the focusing capability of your flatbed scanner, all you’ll need is a transparent ruler and a second ruler to measure with.

I’ll be using the Epson V700/V800 for this instruction, but the procedures should work for any flatbed scanner.

Concept

Like any lens, your scanner’s lens has a range of acceptable focus, or depth of field (DOF) where it performs best.  You can determine DOF by scanning a transparent ruler that has been placed on an incline, with one end well below the minimum DOF height and the other end placed well above the maximum DOF height. The graduations on the scanned image of the transparent ruler will be sharp within the scanner’s DOF.

Procedure

  1. The first step is to place the transparent ruler on an incline relative to the scanner platen. In the following picture, I placed one end of the ruler on the platen and placed the other end of the ruler to rest on the top of the scanner bed, shimmed by approximately 1/4″.   Shimming the elevated end of the ruler will steepen the incline and help differentiate the sharpness along the entire length of the ruler.  It’s important to check that when you lower the lid on the scanner that you don’t move or warp the ruler. You don’t need to lower the lid all the way; just close it enough to avoid touching the end of the ruler.  Now, open the lid and tape each end of the ruler down to avoid movement during scanning.

    Shows the inclined placement of the transparent ruler on the V700 platen. The upper end is shimmed to increase the steepness of the incline. The white sheet of paper is only for illustration purposes so you can ‘see’ the scanner’s glass platen in this illustration (it’s not used during the actual scan).
  1. Scan the ruler using your normal scanning settings (i.e., high resolution, 2400 ppi) without sharpening, and open the file in your photo viewing software, like Preview, Photoshop, or Lightroom. Do not move the transparent ruler yet; you’ll need it in exactly the same location in Step 4.
  2. On the scanned image, examine the ruler along its entire length, looking for the sharpest delineation of the numbers on the ruler. There should be a range of numbers/graduations that will be sharper than anywhere else along the ruler. If your scanner’s depth of field is very wide, you may not see any difference in sharpness along the ruler. This is good, because it means you shouldn’t be experiencing out of focus situation no matter where the negative is placed above the platen. But you need to examine closely to rule out minor differences in sharpness as you examine along the length of the ruler.  In my example below, maximum sharpness was seen about 23-24 cm on the ruler, call it 23.5 cm. The differences in sharpness between 0 and 4 inches on this image is difficult to show, but careful inspection at 50% in Preview clearly shows the difference.  You should see a clear difference. V700FocusTest003
  3. Now you need to determine the distance above the platen where you found the sharpest graduations on the ruler. Use a different ruler for this step. A micrometer may also be used if you have one, but you don’t need that much precision. I used a simple wooden ruler that  began graduations at the very end of the ruler. So, at the point on the transparent ruler (still resting on the scanner) where I found maximum resolution, I measured the distance above the platen surface and found it to be 1/8″. Your results may vary.  This measurement told me that my scanner’s best performance was 1/8″ above the platen. IMG_2946
  4.  Now compare this distance to the height of the negative plane on your negative holder to see if your holder is placing the negative where the scanner has its maximum resolution. First place your negative holder on a flat hard surface (e.g. a table top) and find the lip on the holder that holds the negative during scanning. Measure the distance from the table top to the top of the lip. This tells you exactly where your holder is placing the negative during scanning. IMG_2978 copy
  5. Interpretation

    If you’re lucky, the height measured in Step 5 is exactly the same as the height measured in Step 4. It means your holder is placing the negative exactly where the scanner lens is focusing.

    If the holder doesn’t place the negative within an acceptable distance for the scanner’s lens, you’ll need to modify your holder or change to a holder that does. The Epson holder comes with adjusting feet; merely changing the setting on the feet may fix the problem.

    For the vast majority of scanners, the distance you find is sharpest will be very close to the distance measured on the negative carrier. But this quick test of focus on your scanner should at least confirm that scanner focusing is or is not optimal, and you can take actions to fix it.

    It’s possible that after doing this test, you find that there is no region of perfect sharpness along the entire length of the transparent ruler. There are two possible reasons: 1. the ruler moved or deflected during the scan when the lid is closed (re-read Step 1 in Procedures above), or the scanner’s lens is faulty and you may need to replace or repair the scanner.

Bonus utility: I used this test to determine where my custom negative carrier had to sit for maximum sharpness. I modified my scanner to accommodate a wet-mount under-mount design, and knowing my scanner’s specific DOF told me how much shimming I had to do for correct placement of the negative in this custom carrier. We’re not constrained to using the manufacturer’s film carriers once we know the scanner’s true DOF.

What does it mean when the negative is placed optimally, but the scans still appear ‘fuzzy?’

First, make sure your shooting technique is perfect.  Barring the case where the scanner is not functioning properly, finding that your images still appear unsharp most likely suggests a fault in technique. BY FAR, this will be the cause of unsharp images (i.e., not the scanner). It’s been the same since the the advent of modern cameras and lenses; it’s not the tools but the technique that causes most situations of unsharp images. You already know this, right?  Do you use a tripod when sharpness is important to the image? Do you practice appropriate aperture/shutter speed relative to subject motion? Technique remains one of the most important factors in acquiring an image on the negative that is sharp or not sharp. It’s always been that way.

Second, make sure to minimize your subjective bias toward those hyper-acute images shot with modern digital cameras. I suspect this bias is a leading cause for people new to film   thinking their scans are ‘blurry.’ This topic is probably worth a separate article; because it took me a couple years after returning to film scanning after a year shooting with digital to realize that comparing the two technologies is an apples and oranges comparison. In short, never compare a 1:1 depiction of a digital image to a 1:1 depiction of a scanned image unless the two images are exactly the same resolution (i.e., 20 MP to 20 MP), and even that is a stretch.  The best evaluation of sharpness for a film scan is by looking at an appropriately sized print, not a file on your computer monitor.  Don’t toss your scanner or give up on shooting film before making a print comparison!

I hope this article helps you determine whether your scanner is functioning properly, and how to determine where the optimum focal point of your scanner’s lens is.  Feel free to ask questions.

 

 

Let me hear your questions or comments below.