How sensitive is the Epson V700 to source height above glass?

[Leonard Evens]

I recently got a betterscanning film holder. I am using it in the dry mounting mode. I tape the emulsion side of the film to the dull side of the holder glass and place the glass in the frame with the film base facing downard towards the scanner glass.

I've been trying to determine the optimal height over the scanner glass. I've scanned 4 x 5 Ilford HP5 negatives with the frame at 0.0, 0.2, 0.4, 0.6, and 0.8 mm above the scanner glass, and so far I can't see any differences in how sharp the image is. The images are not quite as sharp as what I can see looking at the negative on a light table under a loupe, but they are all pretty close to that.

It may just be that images formed on HP5 film with my camera and lenses is not going to show enough detail for me to see any differences in that range. But I think my scans capture enough detail to look quite sharp in prints at normal viewing differences. It may also be that there is something else I don't understand.

It would help me if I had some idea of how the optics of the V700 work. The negative image could be considered a subject which the scanner is taking a picture of with its optical system. So there must be something we can call depth of field, i..e, a range of distances the negative can be in which will produce essentially the same result given the capability of the scanner. It would help me to know what to expect if I knew what that range might be.

The scanner is capable of scanning at 6400 ppi, although I don't know if it actually does that for a 4 x 5 negative. I believe it can capture that many pixels, but I don't believe the scanner can resolve that well, which would amount to 1/2 x 6400/25.4 = 126 lp/mm. In my trials I am scanning at 3200 ppi, which would provide 63 ppi, and I doubt it it can actually do even that. I would be surprised if it can even provide 40 ppi optical resolution.

So I would appreciate anything anyone can tell me about the optical performance of the Epson V700 and how that might affect the tests I am doing. How far will I have to raise the frame above the glass before I see degraded performance with typical large format negative films. Also, can anyone suggest another source to scan which will have enoough fine detail to allow me to see a progression?

Anything else of course would be of value.

[biedron]

Leonard,

check out post 7 in this thread:

http://www.largeformatphotography.info/forum/showthread.php?82286-Epson-v750-optimum-resolution&highlight=scanner%20focus

Bob

[Leonard Evens]

Leonard,

check out post 7 in this thread:

http://www.largeformatphotography.info/forum/showthread.php?82286-Epson-v750-optimum-resolution&highlight=scanner%20focus

Bob

I see now why I don't see any differences. I would have to start well over 1 mm and go upp in step of .4 mm before I would start to dee any differences.

But I also clearly need a source with finer resolution.

[D. Bryant]

I see now why I don't see any differences. I would have to start well over 1 mm and go upp in step of .4 mm before I would start to dee any differences.

But I also clearly need a source with finer resolution.

See if you can source a high contrast focus negative. This will make your height adjustment a little easier.

[Heroique]

...It would help me if I had some idea of how the optics of the V700 work...

Leonard, keep in mind that selecting “Film (with film area guide)” activates the lower resolution lens; selecting “Film (with film holder)” activates the higher resolution lens. Typically, the two lenses will have two different optimal scanning heights. (These options are inside the “Document Source” box, for Epson Scan.)

Note to 4990 users: strangely, you also have these options, but the 4990 has only one fixed lens, so it doesn’t matter which option you pick.

[genotypewriter]

It would help me if I had some idea of how the optics of the V700 work. The negative image could be considered a subject which the scanner is taking a picture of with its optical system. So there must be something we can call depth of field, i..e, a range of distances the negative can be in which will produce essentially the same result given the capability of the scanner. It would help me to know what to expect if I knew what that range might be.


Try the maximum height and the minimum height and one in between. If you do this in a strategic way, you can narrow down on the height that gives the best height fairly easily.

One thing I've noticed with the V700's focus is it comes to a point where you're balancing between low CA and sharp focus. But at that resolution (6400 PPI) the image is not something you'd want to view at 100% anyway.



The scanner is capable of scanning at 6400 ppi, although I don't know if it actually does that for a 4 x 5 negative. I believe it can capture that many pixels, but I don't believe the scanner can resolve that well, which would amount to 1/2 x 6400/25.4 = 126 lp/mm.

Don't be too quick to blame the scanner. At 6400 PPI diffraction starts to set in at f/5. So you need to have a lens that resolves 126 lp/mm at f/5 or faster. Not many 4x5 lenses are up to that:

http://www.flickr.com/photos/genotypewriter/6806479707

And not many people shoot faster than f/5 on LF to begin with...

In all fairness, 6400 PPI is equivalent to a 54.8 megapixel full frame 35mm sensor (the current state of the art/D800 is 36.3) ... even if the whole capture process was digital and five-$-figure DSLR lenses were used, you're not going to be seeing many 1 pixel wide details :) So the V700's pretty awesome in what it does for a few hundred $.

[Leonard Evens]

Leonard, keep in mind that selecting “Film (with film area guide)” activates the lower resolution lens; selecting “Film (with film holder)” activates the higher resolution lens. Typically, the two lenses will have two different optimal scanning heights. (These options are inside the “Document Source” box, for Epson Scan.)

Note to 4990 users: strangely, you also have these options, but the 4990 has only one fixed lens, so it doesn’t matter which option you pick.

I work under Fedora Linux, so I can't use the Epson software. I use Vuescan, which I think is superior anyway. I don't know how Vuescan selects the lens to use. I presume it is based on the scan resolution chosen.

[SergeiR]

I work under Fedora Linux, so I can't use the Epson software. I use Vuescan, which I think is superior anyway. I don't know how Vuescan selects the lens to use. I presume it is based on the scan resolution chosen.

it gives you weird option with v700 of regular negative and 8x10 negative

[Leonard Evens]

I've now done 0.8, 1.6, and 2.4, i.e., one, two, and three full turns from base. (I tried to measure the extensions of the white screws with a dial caliper, and I don't get those measurements, but I will use the figures given by betterscanning.) They get progressively sharper. I am not sure what to do now, since there isn'tt much further I can go with the screws provided by betterscanning. The reference provided above showed the optimal distance to be 3.2 mm. I may be able to get that high by one last full turn of the screws, but the support is likely to be pretty wobbly. Also, I won't be able to go beyond that to determine what my optimum is if it is a bit hgiher.

Any suggestions?

[Marc B.]

More height, simple...Shims!

[Leonard Evens]

I've now done 0.8, 1.6, and 2.4, i.e., one, two, and three full turns from base. (I tried to measure the extensions of the white screws with a dial caliper, and I don't get those measurements, but I will use the figures given by betterscanning.) They get progressively sharper. I am not sure what to do now, since there isn'tt much further I can go with the screws provided by betterscanning. The reference provided above showed the optimal distance to be 3.2 mm. I may be able to get that high by one last full turn of the screws, but the support is likely to be pretty wobbly. Also, I won't be able to go beyond that to determine what my optimum is if it is a bit hgiher.

Any suggestions?

It is not as bad as I thought. I forgot to take the thickness of the support for the glass in the film holder into account. It added more than a mm to the overall height.

So my last measurment was in fact at about 40 mm above the scanner glass. I've now also tried going down to about 36 mm above the scanner glass, and I can't see any difference from 40 mm. I think the optimal height is between 30 and 40 mm, as noted in the above reference. I may do some more measurements, but if I leave it at about 36 mm, I think it will be more than good enough for scanning 4 x5 negatives.

[Marc B.]

Don't forget those pesky decimal points. I'm sure you mean 3.6mm to 4.0mm(?).

[Tony Evans]

Leonard,

"I tape the emulsion side of the film to the dull side of the holder glass and place the glass in the frame with the film base facing downard towards the scanner glass."

For dry scanning, I think it is "tape the film base to the ANR side of the glass and place the glass in the frame with the emulsion side downward towards the scanner glass" ie: Film is under glass with emulsion side facing downwards. Measuring height then does not include the glass thickness.

[Leonard Evens]

Don't forget those pesky decimal points. I'm sure you mean 3.6mm to 4.0mm(?).


Yes indeed!

[Greg Lockrey]

FWIW, the maximum resolution for the lens in those Epson v700 is about 2000 dpi.... any thing more than that is just wasting memory. 6400 is an interpolated number that "looks good" on the box.

[Leonard Evens]

Leonard,

"I tape the emulsion side of the film to the dull side of the holder glass and place the glass in the frame with the film base facing downard towards the scanner glass."

For dry scanning, I think it is "tape the film base to the ANR side of the glass and place the glass in the frame with the emulsion side downward towards the scanner glass" ie: Film is under glass with emulsion side facing downwards. Measuring height then does not include the glass thickness.

Perhaps I didn't explain it well enough. Indeed, the film was taped to the underside of the glass in the holder, so the holder glass thickness didn't add to the height. But, betterscanning suggested that I tape the emulsion side to the ANR side of the holder glass since in that case I would be putting the tape on the film base, which was preferable to putting it on the emulsion side. This does have the disadvantage that the film base lies between the image in the emulsion and the scanner glass, whereas if I did it as you suggest, there would be nothing between the image in the emulsion and the scanner glass. Be that as it may, I don't think the optics is affected significantly by having the light rays pass through the film base. And in addition to betterscanning's argument, if I have the film base pointing downward, I don't reverse the orientation.

[Heroique]

Actually, looks like the illustrations at Betterscanning show base side on glass.

(The film notches above the white glove indicate emulsion side is “up.”)

I’ve done both ways, and my personal results are best the opposite way – emulsion side on glass – but others here w/ reasons of their own prefer base side on glass, like the photo below.

The critical point, of course, is to tape the film sheet on the ANR (etched) side of the glass – and use more than four pieces of tape! I usually use at least 6 or 8 pieces to better ensure flatness.

[Kirk Gittings]

FWIW, the maximum resolution for the lens in those Epson v700 is about 2000 dpi.... any thing more than that is just wasting memory. 6400 is an interpolated number that "looks good" on the box.

[Brian Ellis]

Yes indeed!

That's a relief. I was puzzled as to how a scanner with a negative held almost two inches above the glass would work. : - )

[D. Bryant]

Be that as it may, I don't think the optics is affected significantly by having the light rays pass through the film base. And in addition to betterscanning's argument, if I have the film base pointing downward, I don't reverse the orientation.
Optically this is exactly analogous to printing sheet film in an enlarger with a glass negative carrier.

I've never had any IQ issues with the Betterscanning wet mount sheet film holder. Usually crud on the bottom of the scanner platen is more of a concern. Seems like you are making this more difficult than need be, at least that is what I take away from your posts.

[genotypewriter]

FWIW, the maximum resolution for the lens in those Epson v700 is about 2000 dpi.... any thing more than that is just wasting memory. 6400 is an interpolated number that "looks good" on the box.

It's definitely more than 2000 DPI (PPI). Most people don't get beyond that because the stock holders are bad.

Also the 6400 DPI number is not a software interpolated number. I say this because VueScan allows me to do raw scans at 6400. It doesn't prove much because the scanner hardware itself could be interpolating things are you're suggesting. Also I'm not saying you can get that much out of it but I'm quite certain it's a proper optical resolution. Would like to know if you're saying this based on some experimental evidence because I'm very interested in knowing.

[Brian Ellis]

It's definitely more than 2000 DPI (PPI). Most people don't get beyond that because the stock holders are bad.

Also the 6400 DPI number is not a software interpolated number. I say this because VueScan allows me to do raw scans at 6400. It doesn't prove much because the scanner hardware itself could be interpolating things are you're suggesting. Also I'm not saying you can get that much out of it but I'm quite certain it's a proper optical resolution. Would like to know if you're saying this based on some experimental evidence because I'm very interested in knowing.

In View Camera magazine the late Ted Harris published results of his tests of the V700/750 shortly after they first hit the market. I belive his tests showed that the scanner could resolve about 2100 ppi.

[genotypewriter]

In View Camera magazine ... shortly after they first hit the market.

Then it's very likely that he tested it with the stock holders that came in the box. Not sure how much later people started making custom holders. But they make a difference like night and day. It's the same as shooting a camera out of focus vs. in focus.

In no way I'm saying the 6400 DPI setting will give scans that are usable at that resolution. Even a 6400 DPI digital camera sensor will struggle greatly with the best of lenses, as I noted earlier.

When I make scans they're scanned at the maximum resolution the scanner hardware supports in the "raw" mode, which I suspect is the native resolution. I of course rescale these down to usable sizes and if the shot is up for it, 3200 DPI and seeing 1 pixel wide detail is not very difficult. That is with a plain dry scan. Wet scans will make things even better.

[Leonard Evens]

Actually, looks like the illustrations at Betterscanning show base side on glass.

(The film notches above the white glove indicate emulsion side is “up.”)

I’ve done both ways, and my personal results are best the opposite way – emulsion side on glass – but others here w/ reasons of their own prefer base side on glass, like the photo below.

The critical point, of course, is to tape the film sheet on the ANR (etched) side of the glass – and use more than four pieces of tape! I usually use at least 6 or 8 pieces to better ensure flatness.

Betterscanning's advice about putting the emulsion side up taped to the ANR glass was in an email he sent to me in response to a question I asked.