Getting tired with scanning. Epson 4870 failure.

[Clement Apffel]

Hi everyone.

I come to you to get some help with scanning.
I own an Epson 4870. I’m scanning my 4x5” with it.
So far, every single scan has been a disappointment.
Not sharp, no accuracy in color rendering. Just junk scans.

As softwares, I tried EPSON scan. to me it’s a pain to use. Not intuitive at all, colour management is just never accurate and settings are so awkward that I never manage to get the right colour before scan.
I tried Silverfast. Even worst.
I am currently trying VueScan. Firstly I thought it might be better… but it’s just the same awkwardness in colour rendering.

Hoping it will improve focusing on my 4870, I bought the betterscanning kit.
I did the calibration. 16 scans of a crop at 2400 dpi.
I did it twice and ended with the same range of height : from 1.6 to 2.2mm no discernable difference in sharpness. So I set the height at 1.8mm.

Did 2 comparisons between OEM Epson holder and the recently calibrated betterscanning holder…
One of the test didn’t showed an ounce of difference between the 2 scans.
And the other one was slightly in favour of the betterscanning kit. But by so few that it could be my imagination tricking me to get me away from thinking it was a bad purchase.
I’m still waiting samples of mounting fluid to test fluid mounting.

And I can’t wait, cause dry mounting… is just a pain. No matter what anti-static cleaning fluids or anti-static brush or anti-static cloth I use, I always end with a scan riddled with dust. And by that I mean an immense amount of dusts. On 2400 dpi scans, I think I could need an entire hour to remove all of them on PS.

Not mentioning that cleaning fluids always leave whitish residues on both scanner glass and ANR glass. They are not necessarily causing a fall in terms of quality, but I guess they can increase flare.

But I don’t think wet mounting will be the magic solution to dust. For the only reason that the polyester sheets I already got attracted all the dust available in the universe as soon as their box was opened. And I don’t see all that dust getting off the sheet in my life time. Unless putting them in the dish-washing machine…

So as you see, I can not understand how some of you manage to get acceptable scans out of that hardware.
I can not understand how some of you can print 18x24” with that hardware.

My purpose with this thread is not to complain and to blame the hardware, either 4870 or betterscanning kit.
My purpose is to get advice from those stating in this section of the forum that they improved their scan quality with Doug Fisher’s kit or from those printing good 18x24”.
Because I’m ready to hear that I’m doing it all wrong. But the more I read stuff explaining how to proceed to get better IQ, the more frustrated I get about it because nothing seems to improve my IQ.

I could end this first message with some simple questions because I might do some basic errors with those manipulations:


1) How did you calibrated your betterscanning kit ? By that I mean the precise method you followed. Because “stopping when the next scan is less sharp than the previous” as Doug Fisher explains in his tutorial does not sound like a real-world solution to me.
2) How do you manage dust when dry mounting?
3) How do you manage dust with polyester sheets from local art store?
4) What scan software are you using on your Epson and do you manage to easily get accurate colour rendering before scanning?
5) I never saw anywhere in any software the lens choice that is supposed to be given. Between film on holder and film on glass. Where is that setting in common softwares?

Finally, I would like to add that I’m of course well aware of the fact that EPSON low-end flatbeds won’t ever match high-end scanners IQ.
But as I keep reading people’s opinion about their V700 not being that bad with XYZ solution that they have, I got the idea that I could improve my scans.

It never happened so far. To me my scan remains just good to print 8x10” out from 4x5”. Great.
Please help me soothing my frustration.

[Charles Carstensen]

Clement, you have so many issues in this post that it is way hard to attempt a reply that would be any serious help. Silverfast is top software. Your scanner should be able to produce a sharp 18x24 print. DPI is the print output - PPI is the scan resolution (or files size). Can you simplify your request to one issue at a time?

[Bob McCarthy]

I can reinforce the last comment. The Epson will do a very credible 4X.

You might start by profiling the scanner with a IT-8 slide.

Sharpening skills post scan are also a requirement. More-so than the fancy holders.

bob

[Joanna Carter]

As softwares, I tried EPSON scan. to me it’s a pain to use. Not intuitive at all, colour management is just never accurate and settings are so awkward that I never manage to get the right colour before scan.
Are you scanning trannies or negs? If trannies, then you will have problems trying to use the colour settings in the Epson software. The recommended workflow is to calibrate the scanner using a test target, like those from Wolf Faust, and switch off all adjustments in the scanning software. When you have the image in Photoshop, you then apply the profile that you created from the test target.


Hoping it will improve focusing on my 4870, I bought the betterscanning kit.
I did the calibration. 16 scans of a crop at 2400 dpi.
I did it twice and ended with the same range of height : from 1.6 to 2.2mm no discernable difference in sharpness. So I set the height at 1.8mm.
I used to use a 4870 and, with the original holder, I got scans sharp enough to do a print of 40" x 32" for an exhibition.

Now, I have a V700 and a BetterScanning holder and found that the scanning height was significantly more than expected. Did you go higher than 2.2mm?


I’m still waiting samples of mounting fluid to test fluid mounting.
I found that dry mounting was more than adequate.


No matter what anti-static cleaning fluids or anti-static brush or anti-static cloth I use, I always end with a scan riddled with dust. And by that I mean an immense amount of dusts. On 2400 dpi scans, I think I could need an entire hour to remove all of them on PS.
That sounds about normal for a 4x5 sheet; don't forget some of the dust could be from when you exposed the film.


1) How did you calibrated your betterscanning kit ? By that I mean the precise method you followed. Because “stopping when the next scan is less sharp than the previous” as Doug Fisher explains in his tutorial does not sound like a real-world solution to me.
I did a "coarse" adjustment of 0.5mm per step until I found the sharpest using that method, then I did a series of finer adjustments up and down from that point. I think my final height was around 3.2mm.


2) How do you manage dust when dry mounting?
I start by using microfibre cloth to clean all theglasses, then (and some may scream at me for suggesting it) I use my fingers as a "squeegee" to clear the film of any visible dust before taping it to the ANR glass.
You might like to consider 1. an ioniser and 2. a gentle humidifier like a dish of water on a heater in the room to help lay the dust.


4) What scan software are you using on your Epson and do you manage to easily get accurate colour rendering before scanning?
As I mentioned, I use Monaco EZColor software and Wolf Faust targets to profile the scanner. Then I use the Epson software with all adjustments turned off. Finally, the profile is applied to the image in Photoshop.


5) I never saw anywhere in any software the lens choice that is supposed to be given. Between film on holder and film on glass. Where is that setting in common softwares?
The lens choice is made in the Epson software by choosing the film holder type.


Finally, I would like to add that I’m of course well aware of the fact that EPSON low-end flatbeds won’t ever match high-end scanners IQ.
I was extremely happy with the results from my 4870, even without the BetterScanning holder, but I did ensure it was profiled.

[James Beck]

As a drum scan provider, may I offer my services for your consideration?

Respectfully,
Jim

www.jamesbeckdigital.com

[Bill McMannis]

Clement: I just want to chime in that I have used an Epson 4870 sinc '04. Like a previous post, I recommend the MonacoEZ color for profiling and use SilverFast AI for scanning. I have very satistfactory results creating 11x14 print either from Tri-X negatives and chromes. I prefer the Tri-X as I have had a lot of trouble with losing detail in shadows with the chromes. Overall a cost effective system that I plan to use for a few more years.

[Keith S. Walklet]

Unfortunately, there is no way to ascertain what your personal expectations are through a post on a forum, but there is lots of good advice here and rather than trying to convince you that something is acceptable, here is what I do to get the most out of the Epson.

1. Use a small halogen lamp to illuminate the glass bed of your scanner at a sharp angle. This will highlight all the particles adhering to the bed before you even start. Make sure to check the lid glass as well. Canned air will get rid of most of the dust. A microfiber cloth, PEC pad or anti-static brush can be used to remove the rest.

2. After using the canned air to dislodge loose dust from my film, I place it on a PecPad and clean the top surface of the transparency with another PECPAD folded in the shape of a teabag and moistened with PEC-12 film cleaning fluid, making sure to draw the pad completely off the film edge so the dust goes with it.

3. I don't use film holders, instead wet mount my transparencies. Fluid, plex, fluid, film, optical mylar (AZTEK) and a layer of diffuse Mylar (art supply). The thickness of plex I use for my scans was determined after testing much like you did with your holders.

I, for sure, don't use the EPSON holders, having already determined the advantages of wet mounting. The film doesn't lay flat and I used to get Newton rings with the EPSON holders.

4. I scan in 48 bit color, NO sharpening. I use the EPSONscan software, but intervene by setting the black point to "0" and I do test scans to accurately determine the true white point, which I give five or so additional units of brightness for good measure. I set the gamma to 2.0.

5. I scan at the highest claimed optical resolution of the scanner (4800spi for the 4870). This means a 1gb raw scan for a 4x5, so your computer would need to be able to deal with large files initially. The scan itself will appear high-key, washed out, but there will be information in the shadows.

6. I use the Layer Transfer Method to work on my files, which means I downsize the resulting raw scan to 8x10 at 360dpi and save it as a discreet file that will be my working file. The black point will need to be properly set in the file as part of the editing process. But, by bringing the shadow values down, rather than trying to open them up, there is no noise in the shadows.

7. I use non-destructive digital editing methods to refine the file.

8. I save the working file in its unflattened state.

9. I upsize the unflattened image to its output size and open the original scan, which I resize to the exact dimensions as the working file, copy and paste it into the unflattened working file. Throw away the background layer and flatten the file with its high res information.

10. I sharpen carefully, sometimes with selective sharpening, sometimes sharpen only edges, sometimes unsharp mask.

Every step of the way, I do my utmost to preserve the data in its highest quality, and the results go well beyond what some seem to feel is the limit for this machine. I do think the newer scanners do a better job than the 4870, as does the new version of Epsonscan software that came with my V750.

One issue that seems to come up again and again in these discussions is how much information is necessary to make a quality print. There does seem to be a strong feeling in the B&W community that grain is a desirable thing to have in a print. In my own work, which is almost entirely color transparency, I'd like the scanner to resolve the grain, but I don't want it to show up in my prints, especially continuous tone areas of sky and water.

But, the bottom line is, if your machine doesn't meet your expectations after all that work, then you'll need to consider having your work scanned with a drum scanner.

Good luck.

[Clement Apffel]

Thanks for the answers.
Nothing new apart from the dish of water on heater trick. I'll try that.
And Keith's way of working scan files. Very interesting.

And of course the IT8 calibration would help. But I think this is far enough $$$ spent for a low IQ scan. I’ll get IT8 solution when I have the opportunity to get an IQsmart3.
Let's forget about colour rendering and scanner softwares in this thread. too many issues indeed.

For the betterscanning kit, I did 16 scans at 16 heights. From 0mm to 3mm. with 0.2mm steps.
I would really like to know how you manage to achieve precise 0.5mm step… cause even a quarter of turn isn’t easy to achieve with precision… 0.5 would be 5/8 of a turn. Come on…

Those 16 scans were ranked by 3 operators excluding me without them knowing which is which. The 3 classifications put the 0 to 1.4 and the 2.4 to 3 under the range from 1.6 to 2.2.
And within this range, all 3 classifications were inconsistent, meaning the operators couldn’t accurately say which were sharper than another. So I picked 1.8mm as an average value. (I didn’t pick 1.9 because as I said I do not think ¼ of a turn is easy to achieve with precision on all 8 nylon screws so I won’t try 1/8)

I’ll redo the test including values from 0 to 4mm just for the peace of mind.

Concerning the dust, I use a halogen lamp at sharp edge to spot every single dust on the glass and film. With canned air and anti-static brush. And when I could swear there is not a single grain of dust on anything, I perform the scan and end with tons of dust everywhere.
And they end white on the scan meaning they are scan dusts and not dusts from the shot.
The only trick I didn't try yet is the PecPad / Pec-12 clean off of the film.
I'll try that as well. nothing to lose.

Anyway…

And yes, that first post was very confusing, too many issues indeed. But after spending so much time reading about all the issues I had everywhere on the web... I wanted to lay them down. Not a good idea, I agree.

thanks again for the answers, I'll perform some other tests using your tips.

PS : just a side note : in my epson software, there is no selection of holder type anywhere. nothing about lens selection. and I explores every single menus and options. same in VueScan. I guess we are not talking about the same versions. I could use some newer soft as well I guess.

[Keith S. Walklet]

Clement, I've attached two jpegs illustrating the EpsonScan interface from the 4870 and the V750. The 4870 software is version 2.3, the V-750 is version 3.2.4.

These settings appear under the "Professional" mode, a choice made in the upper right corner.

The "film holder selection" often cited is part of the newer software, and controls which of the two lenses the V700 and 750 use to make the scan. It is found under "Document Type." There are three choices -- "reflective" "film (with film holder)" and "film (with film area guide).

The 4870 only has one lens, and only gives the option of scanning "film" or "reflective."

I also must apologize. I stated the highest claimed optical resolution of the 4870 was 4800spi, but it has a limited memory buffer, so the highest resolution a 4x5 can be scanned in one pass is 3200spi. You can use the 4800spi setting if you don't mind scanning the transparency in two pieces, later joined in Photoshop. Each half will be about 1gb!

Another thought about your dust problem. You don't mention if you purchased the scanner new or not. If it was used, you might also look to see if there is dust inside the scanner. Some people have discussed taking the platen off and cleaning it, which might also introduce dust.

[neil poulsen]

It sounds like you're scanning transparencies. Is this correct?

[Clement Apffel]

I scan color negs, chromes and BW.

I tried fluid mounting yesterday. And it was an improvement concerning dust.
And I was reassured when I saw that fluid mounting isn’t that hard. My second try was already quite good and accurate.

In this concern, the betterscanning kit is a plus. Flat neg, plus the ability of fluid mount.

That said, I still have that whitish residues issue. On both glass and negs or chromes!
On glass I’ll use some common house degreasing drug, like Doug Fisher advised me to do. But on negatives? I used the technique suggested by Keith S. Walklet with 2 pecpads and some film cleaner.


After using the canned air to dislodge loose dust from my film, I place it on a PecPad and clean the top surface of the transparency with another PECPAD folded in the shape of a teabag and moistened with PEC-12 film cleaning fluid, making sure to draw the pad completely off the film edge so the dust goes with it.

I was quite successful in removing dust with that technique, thanks for sharing. But it leaves traces on the film: whitish lines. I do the wipe 2 times for the upper part of the film and then for the lower part. And most of the time, the first pass doesn’t leave any trace, but the second does.
With the “teabag” technique do you wipe the entire film frame once at a time?
Note that the traces does not appeared in the scans.

Oh and how do you manage not to get a splitting headache from using all those nocive mounting fluids? I only performed 3 scans and wow…

[Keith S. Walklet]

I'm glad the previous post helped. From my experience, both the Kami fluid and the Pec-12 initially leave a high-key residue on the film and platen that eventually disappears.

I sympathize with regard to the fumes. Not my favorite part of the process!

Marvin Silver turned me onto the diffuse mylar. He doesn't like the fumes either, so doesn't wet mount his work. Instead, he uses a film holder from an Imacon, and places the diffuse mylar over that. I noted a marginal improvement from that approach with my wet mounting, so I wet mount.

As for the workflow, it was Rich Sieling of West Coast Imaging that first told me about the Layer Transfer Method, which is how it is possible to work quickly with larger files. I know I've described the process at length in an earlier thread.

I've attached a couple more jpegs to illustrate the PecPad approach. I owe gratitude to Bill Atkinson for showing me this method when he trained me to use the TANGO drum scanner.

Fold the PecPad lengthwise in half, and then half again so that it is 1/4 its original width. Grasp the ends and make a triangle shape similar to the familiar Lipton flow-thru teabag. The bottom surface of the PecPad is moistened with Pec-12 cleaning fluid prior to use. When one side of the film has been cleaned, the pad can be refolded to expose a clean surface. In all, there are 8 potential clean surfaces to use before the pad's disposal.

While applying enough pressure to flatten the "teabag" shape so that its bottom surface is fully in contact with the film, draw the moistened PecPad across and OFF the edge of the film to ensure that all the dust ends up off the film.

[Clement Apffel]

Thank you very much for that.
It seems to improve the dust issue and the residue on film.
clever technique indeed.