Epson V550/V600 MF scanning - worth trying aftermarket film holders to increase scan sharpness?

After a few years spent shooting digital, I'm going back to film - I just seem to enjoy the simplicity of the shooting process and a certain ineffable quality of the final product.

 

For 35mm, I seem to have found a satisfactory workflow: I have my c41 negatives processed by a lab and do my own scanning with my old Minolta Scan Dual III film scanner + Vuescan + Colourperfect plugin. The results are, to me, nothing short of amazing and I have no complaints about the procedure.

 

I am just starting to explore medium format and this brings me to my question. Ideally I'd like to come up with a medium format workflow that resembles the one I use for 35mm. This is, however, not feasible at the moment, as the expense to buy a dedicated medium format film scanner is too high (this includes dodgy used Nikon 8000s/9000s on ebay).

 

I resorted to purchasing an Epson V550, which I believe is an updated V600 with minor differences. I set it up and tested it yesterday, with a few 6x4.5 Ektar 100 negatives shot on a Fuji GA645i.

 

On the positive side, my results *far* exceed the poor quality low-res scan that came from the lab together with the developed negatives. It is shocking how poor the jpeg scan from the "professional" lab were - and I thought the culprit was the camera or film or both!

 

On a negative note, I think I'm still not completely satisfied with the V550's output. Using the standard Epson MF film holder, the scanner clearly does not resolve the grain. I can tell that because, while on my 35mm scans from the Minolta Scan Dual even the particles of dust on the negatives are sharp and in focus, on my V550 MF output the dust is blurry and soft.

 

I have tried placing the negative directly on the scanner glass and rescanning in Vuescan: I see no difference in sharpness w.r.t using the Epson holder to scan.

 

I have read that the V550/V600 models have a fixed optics with relatively high depth of field, compared to the better models (V700/V800) which have tighter DOF but allow for higher focus plane fine tuning. Can people confirm this? If this is true, then I'm suspecting that the (poor) sharpness I see is all I'm ever going to get with this scanner?

 

Else, have people used after-market film holders (betterscanning? Lomography DigitaLIZA?) and seen sharpness improvements with this scanner on MF material?

I have an Epson V500 as well as a Nikon Coolscan 8000. The coolscan does a better job but honestly it's not the night and day difference I would have expected. The V500 is pretty decent actually, but it all depends on what you want to produce in the end. Do you have grain reduction turned on? If so, turn it off. You might also want to try the Epson scanning software rather than Vuescan and compare the results. I'm not knocking Vuescan but EpsonScan might be able to get more out of that scanner.

 

I can't say I've tried different holders on the Epson but the money might be better spent on a nicer scanner. Oh, and if this is a used scanner, it might be worth opening it up and giving everything a good (but gentle) cleaning. I got the Coolscan for almost nothing because dust on the mirror had degraded it to the point that it wouldn't pass the startup tests. Once that was cleaned, it started working beautifully again.

 

In the end, a flat bed scanner isn't going to be able to compete with a dedicated film scanner no matter what you do, especially if you're concerned about the grain being sharp. But again, depending on what the intended output is, that may not matter, and even less so when you're dealing with the larger negatives of medium format film.

Edited October 3, 2017 by tomspielman

I have an older scanner, an Epson Expression 1600, which ships the focus 2 mm for use with film holders. For flat copies, the focal plane is at the glass surface. Other Epson scanners in this class (single lens, deep body) would actually focus on the subject, regardless of height.

 

It's hard to say how newer, less expensive scanners will operate. An easy way to check is to prop up a sharply engraved engineering scale (e.g., Starrett) so that one end rests on the glass and the other elevated with a pencil (or drill rod). This establishes what a toolmaker would call a "sine bar" gauge. See which part of the scale is in sharpest focus when scanned. From the distances and diameter of the support, you can easily calculate the distance. With a slope of 10:1, a 0.5 mm scale registers in 0.05 mm increments.

 

Inexpensive scanners I use have the sharpest focus at the glass surface. That doesn't mean all do, particularly scanners with come with film holders and backlighting.

I tried the Betterscanning holder with my Epson V600 and didn't notice any difference in sharpness. I've settled on using the factory mounted Epson holder and using the Epson scan software with no sharpening during the scan; everything is done in post. Grain doesn't resolve in any case. Of course I usually shoot Velvia 50 or Tmax 100 in Medium Format so there's little grain to begin with. Check my Flickr for scans. I do have 35mm Ektachrome and Kodachromes there as well some other film. Good luck.

Alan Klein’s albums | Flickr

Thank you for your replies so far. Tom, the scanner is brand new and no, I did not have grain reduction activated in Vuescan. I tend to keep Vuescan's processing to the bare essentials, obtain a raw file and invert/post-process in Photoshop. You're right in that I shouldn't be too bothered by details such as visibility of grain, and clearly I have not chosen to shoot film over digital to achieve ultimate detail. Still I was wondering if somehow my process can be optimized or if I can relax and this is the best I'm going to get.

 

Scanning a precision ruler lifted from one side with a pencil - good suggestion and I'll definitely try it. I have so far tried the following

1. scan MF negative directly on glass - no difference in sharpness of fine details

2. scann MF negative in Epson holder, shimmed with 1 level of credit cards or stacks of two levels of credit cards - no difference in sharpness

 

This again would lead me to conclude that the lens in this V550 has been designed to offer rather broad depth of field to allow for decent sharpness from both raised and non raised material. I guess I'll need to do the ruler test to prove this definitively though.

 

Having said the above, I have to add that maybe I've been too negative with my initial reports. I have since kept scanning my negatives, and have found the the final results can be quite satisfactory. I have also tried the native Epson scanning program, finding the final results inferior in terms of colour and tonality than when using Vuescan. So I have returned to Vuescan for all my scanning. Also, it appears that mild usage of Photoshop's unsharp mask seems to help.

 

Here are some scans. These are from Ektar 100 negatives, scanner in Vuescan at 3200dpi 48bit, exported to PS and resized to 1400pixel (longer dimension) + mild unsharp mask. I will post 100% details of the original when I have some time.

 

Edited October 4, 2017 by pablo_escobar

I have a V700, not sure about the sharpness issue yet. I am gonna order my betterfilmholders soon. The issue I have with the Epson holders is that the film is difficult to put on so can take a couple of attempts, often the the top or the bottom side is not held in place, also the film sags in the center quite a bit. I used to have a Coolscan 4000, bought used off a pro but broke within a year ... You can see the grain with the Coolscan yup. To see the grain with the Epson I turn sharpening to 80 amount with Adobe Lightroom.

 

First - is your actual negatives or slides sharp?

Edited October 6, 2017 by RaymondC

There are always mixed reviews on the aftermarket film carriers--but the better ones do provide an improvement over the factory supplied plastic dreck...

 

Not to throw the conversation off the track of the Epson's, but I do have to mention that a goodly number of us are leaving scanners behind when digitizing MF & LF negatives. The advantages of using a DSLR and good macro lens offer improvements to the overall dynamic range, acutance, and incorporation of the 'real grain' of a quality negative.

 

Have you considered this option?

Raymond, your experience with the Epson film holders mirrors mine exactly. As for your question - I can't say with absolute certainty whether the fault is with my negative. I should be able to find out more next week, when I take delivery of a 6x6 camera. For the photos above I used an AF camera (Fuji GA645) and couldn't say if the lens back/front-focuses.

 

Papatango - I haven't considered the option because at the moment I do not own a digital SLR anymore. All I'm left with is a Nikon Coolpix A (compact with DX sensor, pretty good actually) that I use on holiday. So the option is off the table.

I don't know so much about different scanners, but for cheaper digital cameras it is usual for the optics to be not quite sharp enough, to avoid aliasing.

 

I suspect that needs to be done for scanners, too.

 

All scanners I know give the pixel resolution, not the resolution of the (assumed perfect focus)

optical system, that is, MTF curves for the optics.

 

Then there is the possibility that the focus is wrong.

I've been using an Epson 2450 since the early 2000's for scanning medium and large format negs, mostly black and white. I have found even this earlier version of Epson scanners has been adequate for me needs; making prints up to 16x20 or 16x16. After experimenting with different film holders, I now just use the one provided with the scanner, and Vuescan. At the max of "2400" ppi I get an actual 1326 ppi scanning 2 1/4 negs at 2400 setting, and 2200 ppi scanning 4x5 negs at 2400 setting! Anyway, the results are sharp enough for me. Judging by the sharpness of the dust spots on the neg, I am getting pretty good sharpness and actually don't sharpen much at all even on a full size print. Here's an old one from the 70's shot with a Yashicamat.

oops, I miscalculated: my ppi for 2 1/4 negs is the same as 4x5: 2200

So I have performed some non-scientific testing of the focus on my Epson V550. I have purchased a sheet of "museum grade" (anti-reflective, I believe) glass. I wanted to verify if I could extract higher sharpness from the scanner by placing the negatives directly on the scanner glass, with my sheet of glass placed on top - as opposed to placing the negatives in the Epson film holder. I used two pictures, attached below. For each picture, I attach side-by-side comparisons of 100% crops taken from the two, contrasting the two conditions described above.

 

Film: Ektar 100, 120 format

Camera: Fuji GA645i

Scanner: Epson v550

NO sharpness mask or similar enhancements were applied on the 100% crops below (which are from 48bit tif files). The 2 summary images are jpegs (45% sharpness mask+levels applied on these two)

 

Do people note a difference between the two conditions in the 100% crops?

 

Image 1:

 

Image 1 - crop 1:

 

Image 1 - crop 2:

 

Image 2:

 

Image 2 - crop 1:

 

Image 2 - crop 2:

Edited October 13, 2017 by pablo_escobar

I will report back with the betterholders and update this thread :) My main issue is that it holds the film flatter and easier to put the film on. I have had better success with 6x6 negatives and they look sharp enough for me onscreen and when printed A3 I am happy with them. 6x7 is what I have issues with.

Thanks Raymond. Would anyone else like to comment?

 

I'd be interested if anyone can tell apart the picture scanned on the epson holder vs the one held flat on the scanner with a piece of museum-grade glass.

No one looks at crops of 100% , unless you are a pixel peeper :)

 

Images are not meant to be viewed this way. Just stand back and enjoy.

Thank you for commenting.

 

Fundamentally, I agree with you. However there are (rather expensive) aftermarket gimmicks that are being sold as being able to extract the last drop of resolution when scanning transparencies using flatbed scanners. I think it's useful for many people to discuss flatbed scanning baseline resolution to understand if these aftermarket solutions are a waste of money or not.

Thank you for commenting.

 

Fundamentally, I agree with you. However there are (rather expensive) aftermarket gimmicks that are being sold as being able to extract the last drop of resolution when scanning transparencies using flatbed scanners. I think it's useful for many people to discuss flatbed scanning baseline resolution to understand if these aftermarket solutions are a waste of money or not.

Thanks for taking my post in the spirit it was intended :)

I suspect the best way is to buy a drum scanner...

I don't usually jump into these threads but just let people research for themselves, and this isn’t trying to start a flame war, but since you stated "aftermarket gimmicks" I am going to make a few comments...

 

 

There are more advantages to my V550 holders in addition to a potential increase in sharpness. Ease of use and the ability to keep the film flat (either by just using the insert or resorting to the mounting method for really stiff or limp film). Flatness is important for a consistent scan across the entire plane of the image. I am very upfront about what my holders potentially can and can't do. This page link has been at the top of my website all the way back to when I started selling my variable height holder versions: Custom film holders for Agfa, Microtek and Epson film scanners. . A buyer has to decide for themselves if the potential improvement shown on the page is enough to justify the investment. Every increase in sharpness achieved means better potential results during post processing because you have a better base scan from which to start. People have to have reasonable expectations though, and as I state, you must realize the inherent limitations with a reasonably priced CCD-based Epson/Canon/Microtek scanner. With that said, you can see many stunning images all over the internet created using flatbeds.

 

 

Two variables that are out of my control are the characteristics of the scanner as well as user skill/attention to detail/willingness to follow a valid workflow.

 

 

Some scanners come out of the factory is perfect operating condition and thus won't show an increase in sharpness when varying film holder height. Others will show a significant improvement with a variable height holder. Some scanners come with optics so poorly made/installed that even with adjusting multiple millimeters you won't see a difference. That points to a bad scanner (that my holder can’t help!). I can't tell a buyer exactly what they can expect but that is why I promote the other benefits besides potential sharpness increase.

 

 

A person who buys a top of the line Canon camera and lens, on average, isn’t going to produce images better than they would with an entry level Canon Rebel IF they don’t take time to learn and perfect using the advantages of the better camera (and have the aptitude to do it). The sample principle applies to my holders. They are a tool you have to learn to use. It is not difficult but I can’t tell you how many people don’t even read the instructions that are sent with every purchase, much less do a valid attempt at the calibration workflow.

 

 

A final note I would like to make is about testing results posted on the internet. This is a general statement and NOT directed at Pablo or his test. Please realize that testing with the Epson holder is highly likely to be unreliable. Due to their holder design, most films are either arching or sagging to some extent the moment you snap that finicky frame shut. A piece of film might be sagging for one test scan. If a tester removes and inserts the same film again, it might be arching and thus produce a noticeably different scan. This is why I am very hesitant when I read people on the internet in multiple forums (again, this is not directed at Pablo) recommend people should be scanning off the glass in general. Based on a lot of feedback from users of my holders around the world, I am confident that the majority of Epson single lens, single strip flatbed scanners perform best in the 1.5-2 mm range. Yes, of the millions of scanners Epson has sold, there are a number of scanners out there that do have an optimal film suspension height less than 1 mm (and why I offer a .5 mm option) but I am convinced those are a relatively small minority. I am quite sure there are a number of people making claims who have not conducted well thought-out and consistently implemented tests.

I had a CS4000 before, bought off a pro and I have an Epson V700 with the Epson holder so far. I find the Nikon better but at the end of the day printed they are both ok. These are with 35mm film. Having said that a dSLR will be sharper than even a CS4000. Film is Kodak E100G.

 

Thank you for your replies so far. Tom, the scanner is brand new and no, I did not have grain reduction activated in Vuescan. I tend to keep Vuescan's processing to the bare essentials, obtain a raw file and invert/post-process in Photoshop. You're right in that I shouldn't be too bothered by details such as visibility of grain, and clearly I have not chosen to shoot film over digital to achieve ultimate detail. Still I was wondering if somehow my process can be optimized or if I can relax and this is the best I'm going to get.

 

Scanning a precision ruler lifted from one side with a pencil - good suggestion and I'll definitely try it. I have so far tried the following

1. scan MF negative directly on glass - no difference in sharpness of fine details

2. scann MF negative in Epson holder, shimmed with 1 level of credit cards or stacks of two levels of credit cards - no difference in sharpness

 

This again would lead me to conclude that the lens in this V550 has been designed to offer rather broad depth of field to allow for decent sharpness from both raised and non raised material. I guess I'll need to do the ruler test to prove this definitively though.

 

Having said the above, I have to add that maybe I've been too negative with my initial reports. I have since kept scanning my negatives, and have found the the final results can be quite satisfactory. I have also tried the native Epson scanning program, finding the final results inferior in terms of colour and tonality than when using Vuescan. So I have returned to Vuescan for all my scanning. Also, it appears that mild usage of Photoshop's unsharp mask seems to help.

 

Here are some scans. These are from Ektar 100 negatives, scanner in Vuescan at 3200dpi 48bit, exported to PS and resized to 1400pixel (longer dimension) + mild unsharp mask. I will post 100% details of the original when I have some time.

 

 

These look great.

 

I have the V500 and your samples meet or exceed what I see from my scanner. I did abandon the finicky original film holder. I like using an unlikely replacement -- the Lomography DigitaLIZA holders. They really hold the film flat, clamped using magnets. I find them easier to load than the Epson holders too. I have picked up a used BetterScanning ANR holder but I haven't tried it yet.

So I have performed some non-scientific testing of the focus on my Epson V550. I have purchased a sheet of "museum grade" (anti-reflective, I believe) glass. I wanted to verify if I could extract higher sharpness from the scanner by placing the negatives directly on the scanner glass, with my sheet of glass placed on top - as opposed to placing the negatives in the Epson film holder. I used two pictures, attached below. For each picture, I attach side-by-side comparisons of 100% crops taken from the two, contrasting the two conditions described above.

 

Film: Ektar 100, 120 format

Camera: Fuji GA645i

Scanner: Epson v550

NO sharpness mask or similar enhancements were applied on the 100% crops below (which are from 48bit tif files). The 2 summary images are jpegs (45% sharpness mask+levels applied on these two)

 

Do people note a difference between the two conditions in the 100% crops?

 

 

OK -- my preference is for the left-hand image (of the side by side 100% crops.) It looks like a scanned image that will sharpen up well with Unsharp Mask, revealing an amazing amount of detail. (One of the minor miracles of scanning to me. When I look at the raw scans they always seem hopelessly soft and dreary, but a little local sharpening, a touch of global sharpening, and a bit of a curve and VOILA! Nice pictures!)

I don't usually jump into these threads but just let people research for themselves, and this isn’t trying to start a flame war, but since you stated "aftermarket gimmicks" I am going to make a few comments...

 

I don't know anything about your film holders, but I generally agree and suggest a well-thought-out holder is worth the money, to achieve good quality scans in a reliable and repeatable manner.

 

The biggest issue solved by a good holder is flatness. If the film curls at all, you are throwing away sharpness at best; at worst it affects contrast, colour and exposure in localized areas of the film, which is diabolically hard to clean up in post.

 

Finally, I would avoid putting film right on the glass plate of the scanner. You will get Newton rings, maybe not all the time, but when you do it will drive you crazy, and make you use a different technique (holder.)

Just sharing my 2c. The holder is better made than the OEM Epson, it does hold your film flat. I will post an image from a best and worst case scenario. Ie - when the film is already quite flat and when the film is quite curved.

 

For me this improves your flatbed scanner but it isn't going to transform your flatbed scanner into a dedicated scanner. If you are really looking for sharper scans, have a look at using a dSLR with a macro lens and a light table and photograph them 1 by 1. Maybe not all photo's, maybe just the handful. But you would still need to hold your film flat somehow ... when you use the camera to scan it. I read that the Epson V700 effective resolution is around 2300-2400 for a 6x7 neg that's a 38MP file.

My curly films once cut and sleeved has become less curly. Anyway I have two comparison fwiw ...

 

Outline. Left is the Hassie and the right is the RB67.

 

First the 6x6 comparison. Left is the Epson OEM holder with the plastic feet adjusted to the plus mark and the right is the Better Scanning holder calibrated with my own eyes. Using LR both are +1 EV but no other adjustments made, no sharpening added nor with the scanning software.

 

Edited October 21, 2017 by RaymondC