Epson Perfection 3200 Focus Issues

[aoresteen]

I recently bought a used Epson Perfection 3200 scanner for the princely sum of $11 from our local Goodwill. It came with the 35mm slide holder but nothing else. It is working fine with VueScan via USB no issues there. This my first flatbed film scanner. I bought the flatbed to do 120 &127, 617, & 70mm scans - B&W, colour neg & colour slides.

I scanned a couple of 35mm slides with the Epson 3200 but they were not very sharp. I suspected that I had a focus issue. I have an old 35mm Minolta Dimage Scan Speed SCSI scanner that is a lot sharper.

I did a flatbed scan of my stainless steel guitar string gauge (yeh I also work on guitars ) as it is flat and won't flop around like film does. The scan is extremely sharp:

I scanned it at 3200 DPI to a TFF file & converted it to jpg & reduced for posting here. No complaints with the flatbed mode scan.

I then did a scan in transparency mode B&W with the string action ruler lying on the scanner platen glass and got a blurry mess. The edge marks are not sharp (I know steel is not transparent but there is enough light at the edges to scan). This proves that the Epson 3200 has two focus points - one for flatbed mode and one for transparency mode.

So decided to determine exactly how high up off the platen is the focus point for transparency mode. I did some research and found that the Epson film holders are supposed to be 1mm above the platen (correct or not?). So I set up some tests. I decided to use precision feeler gauges to put the steel string action gauge at 1mm above the glass and then scan it. In theory it should be as sharp as the flatbed scan.

I used a new set of gauges that were very oily so I separated them and washed them using Dawn dish soap to keep oil off the platen.

I used the 0.80mm plus the 0.20mm gauges to place the string gauge on the platen at 1.00mm

Here's what I got:

The image is soft. So the focus is off in transparency mode.

I then placed one edge of the sting gauge on the glass and put the other end at 1.50mm above the glass so it is angled up and past 1mm in height. I then scanned it to figure out if the focus point was above or below the 1.00mm point. It was well past 1.00mm.

I won't bore you with all the empirical scans I did to figure out the focus point but here is the one at 1.33mm above
the glass:

I think this is about as good as it will get. The transparency scan focus point is set at 1.33mm.

Great. So how do I fix it? I suspect that there is some adjustment that the factory did to set the lens focus but I am not eager to open up the scanner and start turning screws blindly. Epson must have had special tools to do this in a production mode. I have come up with three options:

1. Shim up the negative holders by 0.33mm and use ScanTech SuperGlass™ ANTI NEWTON RING (ANR) Precision glass sheets to hold the 120 negatives flat. Doesn't solve my need to scan 70mm 127, and 617 negatives as I can't find negative holders for these sizes. And I need to find a 120 film holder to start with.

2. Use two sheets of ANR glass to hold the negative between them and shim the bottom glass to be 1.33mm high and just place in on the platen glass.. I don't know if I can get thin ANR glass that is 1mm thick or less. But if possible I could easily scan anything to include 617 negs, 120, 70mm, 127 (46mm) and 4x5 film. I use a glass negative holder for enlarging so cleaning the 4 sides of the glass holder doesn't bother me.

3. Use a single piece of ANR glass to hold the negative flat on the platen and software control to put the 3200 in flatbed focus mode but have the transparency light on. I don't know if this is possible. But I wonder why this isn't an option with flatbed scanners in the first place.

What is the most practical path forward? I won't bother using this scanner if I can't get sharp scans from it.


Thanks!

 

[shutterfinger]

While your testing procedure is sound I question the use of a non transparent test object.
Document scan: light from the cold cathode florescent light source is reflected off a mirror onto the document, in the test case a stainless steel gauge, which reflects back down to another set of mirrors then into the lens to the sensor.
Transparency Scan: Light from the cold cathode florescent lamp in the transparency adapter is shown through the transparency material onto the set of mirrors that reflect the light into the lens and then the sensor.
The fact that the light must reflect off the stainless steel then back onto the set of mirrors in the image path. This may be causing an erroneous focus height.
Use a loupe and find a critically sharp negative, mount it in the film holder and scan it in transparency mode.

Black poster board or similar can be used to make film guides for scanning. Cut out the image area and secure the edge of the negative/positive to the frame with painters tape as it leaves no residue.
Once you know the correct height then lexan, plexiglass, or acrylic the correct thickness can be used with a single layer of black paper for the mask.
OR
http://www.betterscanning.com/scanning/models/3200.html

 

[aoresteen]

Thanks. I did make a film scan first of a sharp (I use a Rodenstock 4x loupe to look at my transparencies on a light table BTW) 35mm slide using the Epson slide holder that I have and it was very soft compared to slide scans from my Minolta 35mm scanner. That's what sent me down this path. The light coming through a transparency will have no effect on where the scanner lens is focused. The lens is set for one distance only for each of the scanning modes. Scatter may cause reflections issues but not focus. If you can't get the film in the focus plane then you will never have a sharp scan.

I did consider using white plexiglass as the target but I didn't have any that had detailed images on it like the string gauge has. Maybe one with a USAF test chart on it would work better.

 

[shutterfinger]

If your measurement is correct then shimming the slide holder so that the slide is at the distance you measured then it should be sharp.
I've had several scanners apart while the lens is fixed as is the last mirror that reflects the light into the lens the number and position of the others changes the optical path. I have not tested a reflective target in transparency mode to see if it makes a difference in focus.

 

[shutterfinger]

OK, it does not make a significant difference.

Epson V500, EpsonScan, 16 bit grayscale, 6400dpi scan resolution (scanner optical limit) saved as tiff then converted to jpeg in PS.
Left to right: target scanned as document, scanned as transparency positive with target on bed glass, target laying on top of film holder, target inserted as a piece of film in the holder. the target will not fit properly in the film/transparency holder.
Target is a 2 inch X 2 inch standard printed on Kirkland glossy photo paper on a Epson 2200 photo printer at 2800 dpi.

The bed glass on the 3200 is attached to the case with double sided tape. I have seen it turn loose and allow a side to drop slightly. To open that scanner start a scan then pull the power plug once the scan block has moved partway into the scan. Remove the lid by lifting straight up at the hinge. Remove the screw at the bottom of each hinge socket. Working from the rear lift the case off the base then unhook the snaps at the front and unplug the switch plugs. Assembly is the reverse. The scan block will be movable by hand as needed and will lock if pushed back to the home position. Assembly is more difficult with the scan block at the home position. The scan block will return to the home position on the next power on.

 

[jim10219]

I'd just shim the film holders with tape or something simple. No reason to over complicate it. Messing with the focusing height would be too much risk. Besides, I don't even know how you'd go about doing that. If you need to, make your own film holders out of card stock or something. You can spray the card stock with matte black paint or paint it with a thin layer of wood glue if you're worried about paper fibers getting loose.

Scanning is an art unto itself. Each scanner is different and will require a different solution. So even if you went out and bought a brand new scanner, you'll likely have to deal with this same issue, as no scanner that I have ever used actually focused at the height of the included film holders. They all require some adjustment. Some you have to shim. Some, you have to sand down. My current scanner, an Epson 4990, will only focus directly on the glass, so I have to wet mount my negatives.

 

[Jim Jones]

. . . 2. Use two sheets of ANR glass to hold the negative between them and shim the bottom glass to be 1.33mm high and just place in on the platen glass.. I don't know if I can get thin ANR glass that is 1mm thick or less. But if possible I could easily scan anything to include 617 negs, 120, 70mm, 127 (46mm) and 4x5 film. I use a glass negative holder for enlarging so cleaning the 4 sides of the glass holder doesn't bother me. . . .

Placing any solid transparent material between the target and the platen glass will shift the focus, so you'll have to either calculate the focus shift or retest the system for correct focus.

 

[Mainecoonmaniac]

Does your film holder have blank area near the sensor? Sometimes if that is blocked, the scanner can't calibrate for film. I made that mistake many times. Before the scanner starts scanning your batch of film in the holder, it has to go into a calibration cycle first. Blocking the area throws the scanner off.

 

[Alan Edward Klein]

Does your film holder have blank area near the sensor? Sometimes if that is blocked, the scanner can't calibrate for film. I made that mistake many times. Before the scanner starts scanning your batch of film in the holder, it has to go into a calibration cycle first. Blocking the area throws the scanner off.

How does it do that? Does it do it on the Epson V600?

 

[shutterfinger]

I always get an error message or the scanner stops and the power light blinks orange. You should be unable to scan if the calibration area is blocked, you get strange lines if there is dust in the calibration area.

 

[aoresteen]

Does your film holder have blank area near the sensor? Sometimes if that is blocked, the scanner can't calibrate for film. I made that mistake many times. Before the scanner starts scanning your batch of film in the holder, it has to go into a calibration cycle first. Blocking the area throws the scanner off.

Yes, I keep that area clear and the holder I have has the cut out (as do some other Epson carriers that I've bought).

I contacted VueScan and Ed promptly replied that I should try using MANUAL focus with a setting or 1 or -1. So I did some more testing.

While your testing procedure is sound I question the use of a non transparent test object........l

I think shutterfinger is on to something here. I found a transparent plastic ruler and used that as the target. Using Ed's advice to use transparency mode with manual focus I did three scans of the ruler using different settings.

I placed the ruller on the paten glass and scanned it with manual focus value set at 1:

It's sharp. Then I set the manual focus value to 0 and did another scan.

It's sharp as well. Humm Let's try -1:

It looks like nothing changed using different values for manual focus. So I put the feeler gauges under the ruler and raised the ruler up 1mm.

Here's the scan at manual focus value 0 with the ruler up 1mm:

It is sharp! As were -1 & 1 scans. At this point I decided to go back to autofocus and test the scans. Here's the ruler on the glass:

(Continued on next post).

 

[aoresteen]

It is sharp as well.

Now I'm confused so I raise the ruler up 1mm and scan again:

Still sharp. What is going on here? Why isn't the image blurring? I decided to lay the ruler left to right and place it on the sides of the scanner. That raised it up up 2.68mm:

Still sharp.

I am now totally confused as to how the Epson 3200 is focusing the scanner head. The35mm slide I scanned is curved, not flat as it lies in the paper mount. The ruler is flat.

Is it possible that this scanner has real auto focus and not two pre-set focus points? Thanks for your help.

 

[Les Sarile]

Thanks. I did make a film scan first of a sharp (I use a Rodenstock 4x loupe to look at my transparencies on a light table BTW) 35mm slide using the Epson slide holder that I have and it was very soft compared to slide scans from my Minolta 35mm scanner. That's what sent me down this path.

I use a 40X microscope/loupe to examine my film for critical focus. What I can see with this the Coolscan can resolve at 4000dpi. I only use this to show clients that their shots are out of focus and that the Coolscan's AF is reliable which of course doesn't make them any happier.

Anyway, what you can see through this can be resolved by the Coolscan's 4000dpi and the difference between that and the Epsons are as you have found - not close.

I printed out a 4 X 4 arrangement of res charts on tabloid paper to test my lenses. I then used optimal camera settings (tripod, MLU, sufficient lighting for high enough shutter speed across all apertures) and various film types. This one I used Fuji Velvia (RVP) and scanned using my Coolscan, Epson V500 and Epson V7XX.
Fuji RVP50-03_18 Coolscan 5000 full res -> http://www.fototime.com/02BB797801DCA89/orig.jpg
Fuji RVP50-03_18 Epson V500 -> http://www.fototime.com/33269E445D10043/orig.jpg
Fuji RVP50-03_18 Epson V700 -> http://www.fototime.com/11F59FA46FF9497/orig.jpg

Of course with test targets taken under ideal conditions with high resolving film pixel peeped greatly magnifies the difference in resolving ability between these scanners.

Here's a grabshot on Fuji 100 color negative and the difference isn't so huge unless you pixel peep. You could probably apply enough sharpening on the Epson results to make it appear closer to the Coolscan.

Fuji 100 full res -> http://www.fototime.com/535A4899F477014/orig.jpg

 

[shutterfinger]

https://files.support.epson.com/pdf/pr32ph/pr32phpg.pdf does not show any auto focus on this model. Its either the reflectivity of the stainless steel ruler or Vuescan or the combination.

 

[Les Sarile]

https://files.support.epson.com/pdf/pr32ph/pr32phpg.pdf does not show any auto focus on this model. Its either the reflectivity of the stainless steel ruler or Vuescan or the combination.

I don't believe he is suggesting it has auto focus but two fixed focus points.
I believe the Microtek M1(?) is the only flatbed I am aware of with autofocus. My very limited testing of it didn't yield any improvements over the Epson V700.

 

[aoresteen]

The Office Depot ruler is made of clear plastic. not steel. How is the Epson 3200 scanning them at different heights yet they are all sharp? It doesn't make sense to me.

Is VueScan adjusting the focus? Does the Epson scanner have manual focus vs two fixed focus points?

 

[jim10219]

On my Epson 4990, VueScan gives you focus adjustment options, but they don't do anything. And that makes sense, because on that scanner, you can't adjust the focus. However, there is a fairly wide area where it will scan in focus. It probably has an area of about 2mm or so where it will scan in focus. The depth of field isn't large, but it isn't razor thin either. I'm thinking this is what you are experiencing.

 

[GRHazelton]

betterscanning.com has a film holder for the Epson 3200 which allows height adjustment via screws, as opposed to the two positions with the stock holder.. They also sell ANR glass cut to fit the holder. I have their film holder for 120 for my Epson V700 and, after some fiddly adjustment, the scans are now sharp corner to corner which they weren't before. Check their website.

 

[aoresteen]

On my Epson 4990, VueScan gives you focus adjustment options, but they don't do anything. And that makes sense, because on that scanner, you can't adjust the focus. However, there is a fairly wide area where it will scan in focus. It probably has an area of about 2mm or so where it will scan in focus. The depth of field isn't large, but it isn't razor thin either. I'm thinking this is what you are experiencing.

Thanks. I think you are right but I did have one scan that was 2.6mm above the glass and it was sharp as well.

 

[aoresteen]

betterscanning.com has a film holder for the Epson 3200 which allows height adjustment via screws, as opposed to the two positions with the stock holder.. They also sell ANR glass cut to fit the holder. I have their film holder for 120 for my Epson V700 and, after some fiddly adjustment, the scans are now sharp corner to corner which they weren't before. Check their website.

Thanks. I have bookmarked that site. Right now I've decided to order a 4.5" x 9.5" piece of ANR glass and use it to hold the negative flat against the platen glass and see how that works. The max film scan area of the 3200 is 4x9 inches. I will make a black cardboard mask that will place the 4.5 x 9.5 inch ANR glass in the proper scan area of the 3200. I also plan on using thin black paper masks to lie on the top of glass to mask the film area. This should solve any curvature issues.

The only issue that is left are mounted slides. I do not want to remove my slides from the mount. Too much work re-mounting them. Any new 35mm E-6 film I shoot will be returned un-cut and un-mounted.

I am also looking at wet scanning with the 3200. Has anyone tried it?

 

[shutterfinger]

Since you're finding focus is good over a distance of a few millimeters above the glass with Vuscan in manual focus mode then laying the mounted slides on the glass should be fine. I would make a frame to keep them aligned with the scanner's transparency area.

 

[aoresteen]

Since you're finding focus is good over a distance of a few millimeters above the glass with Vuscan in manual focus mode then laying the mounted slides on the glass should be fine. I would make a frame to keep them aligned with the scanner's transparency area.

Not quite. The Epson 3200 slide holder doesn't really hold the slide. The slides actuality lie directly on the platen glass. To use the "holder" you put it on the scanner first, then place the slides in the "holder". If you lift up the holder with slides in it, the slides stay on the glass. I consider it an "arranger", not a "holder". The scans I have tried with slides are not sharp all over - my slides are curved. I have some glass mounted slides - somewhere. When I find them (I have over 100 Carousel 140 & 80 size trays to go through) I'll try scanning them.

 

[Alan Edward Klein]

Not quite. The Epson 3200 slide holder doesn't really hold the slide. The slides actuality lie directly on the platen glass. To use the "holder" you put it on the scanner first, then place the slides in the "holder". If you lift up the holder with slides in it, the slides stay on the glass. I consider it an "arranger", not a "holder". The scans I have tried with slides are not sharp all over - my slides are curved. I have some glass mounted slides - somewhere. When I find them (I have over 100 Carousel 140 & 80 size trays to go through) I'll try scanning them.

After scanning, the slides are a pain in the neck to lift off the glass platen.

 

[shutterfinger]

Not quite. The Epson 3200 slide holder doesn't really hold the slide. The slides actuality lie directly on the platen glass.

Same as the V500 and likely the V700 as well as the in betweens. A 35mm dedicated scanner will produce better results than a flatbed.

 

[GRHazelton]

Same as the V500 and likely the V700 as well as the in betweens. A 35mm dedicated scanner will produce better results than a flatbed.

The V700 slide holder holds the slides with little clips, one can load the holder away from the the scanner, quite convenient. It has three height positions, which may or may not achieve proper focus. Thus it has the same problems as the stock roll film holder or 35mm holder.