Fake out Epson scanner to scan without negative carrier?

[David Reynolds]

Look on the back of your negative carrier. There should be a patch of white material (a calibration patch) that the scanner tries to reference.

You could mock up a template with black construction paper with a bit of white tape in the same position, but depending on how the tape reads, it may give you wonky results.

Using a holder the scanner reads the patch and scans with reference to it ( the plane the film lies on above the glass)

With the older Epson flatbeds, all you have to do is keep the top inch of the scanner glass clear. That's what I do with the 4990 and have no problems.

Using a 'film area guide' keeps the top inch of the scanner glass clear. It allows one to scan any size film flat on the glass

The Epson software has a setting to scan the negative laying flat on the screen. So the scanner either focuses on the height of the holder or directly on the glass.

Some people on this forum disagrees with this statement. I agree wholeheartedly and have been scanning both with holders and flat on the glass for many years

 

[brbo]

Maybe the 4990 scanner doesn't but the 4990Photo does. If you want proof I will put together both glass scan and holder scan and you won't be able to tell the difference

Sure! Every now and then somebody comes along with claims that Epson flatbeds prior to V700 use two different lenses or two different focus point. But, till now nobody produced a proof of that... My attempt also failed as you can seen above.

 

[Don_ih]

Well, the 1640xl had a setting in software to adjust focus - and that scanner was from 2001. The proof of that is freely available on Epson's site here.

I, however, only find reference to the 4990 being fixed focus.

I don't have either of those. I have v550, a v600, and one that scans up to 4x5 (I don't know the number).

 

[David Reynolds]

Take a look. It may be focus or it may be depth of field.

 

[beemermark]

I have a Epson V700. I infrequently scan negatives larger than 4x5 (the V700 comes with a 4x5 holder and an 8x10 mask that lays flat on the screen). I'll also scan 120 negatives that are too curly and a PIA to fit in the holder. I lay the negatives flat on the glass and use another please of glass to hold them flat. There is a big difference between the settings. Why would Epson have two different settings for a holder and one for no holder if they thought DOF would work. For people with an Epson scanner check it out your self. If you don't have an Epson scanner than no comments are necessary.

Another trick is too flatten flowers on the glass and scan with the two different settings.

 

[koraks]

I have been using an Epson 4990 photo for over 20 years

Nice, same here - it's a great piece of kit! Still use it very often.

Maybe the 4990 scanner doesn't but the 4990Photo does.

The 4990 Pro retailed with IT8 calibration targets and some additional software; I think the SilverFast version included in the bundle was one of those items. The hardware is/was exactly the same.

The 4990 has no hardware provision for focus adjustments. There's no extra lens, no focus servo etc. It's a fixed-focus device. Open it up and check for yourself. Yes, I did that when cleaning it, out of curiosity, many years ago.

Also:

Same 35mm B&W negative scanned with film holder (left) and straight on the glass (right). scanned at 3200dpi and downsampled to 1600dpi. 100% crop from comparison image.
The only difference in processing was that the left one was scanned with the Film Holder setting selected and the right with the Film Area Guide. Contrast & curve settings were left alone between scans. Contrast was adjusted in GIMP after scanning with an identical adjustment for both images.

There's a bigger version here, but WordPress has downscaled it a bit. Still, the lack of acutance in the fine detail in the treetops is very apparent in the glass-level scan, even in this dramatically down-rezzed version.

Really, this scanner doesn't adjust focus - because it physically can't do this.

 

[brbo]

Take a look. It may be focus or it may be depth of field.

Or it might be that you are scanning at 300dpi where humans can't possible tell the difference? Grab a resolution target or a negative with lots of fine detail and scan at 4800dpi to see the difference.

I have a Epson V700.

We are talking about 4990. V700 has two lenses. 4990 has only one and it's position is fixed.

Well, the 1640xl had a setting in software to adjust focus - and that scanner was from 2001. The proof of that is freely available on Epson's site here.

Yes, thank you for correcting me. I should've stated "lesser than V700 Epson scanners".

 

[David Reynolds]

Or it might be that you are scanning at 300dpi where humans can't possible tell the difference? Grab a resolution target or a negative with lots of fine detail and scan at 4800dpi to see the difference.



We are talking about 4990. V700 has two lenses. 4990 has only one and it's position is fixed.



Yes, thank you for correcting me. I should've stated "lesser than V700 Epson scanners".

You are correct I am scanning at a low dpi (300 to 800) as that is all one needs to produce fine images for either here (internet) or produce prints up tp 13"x17" (most printers). I have worked with some of the finest art book publishers in the world and thats all one needs to produce coffee table book illustrations. When I make prints up to 40" x 60" I would have the scans produced on a drum scanner. You would be naive to think you could produce flatbed scans for large format prints. Why people here are scanning at 4800dpi I have no idea other than that is overkill. If one is producing images for human consumption (Internet images and /or smaller format prints) why produce an image beyond a human's grasp.

 

[koraks]

I am scanning at a low dpi (300 to 800) as that is all one needs to produce fine images for either here (internet) or produce prints up tp 13"x17" (most printers).

That may work fine for 4x5 and 8x10 negatives, but for 35mm 800dpi scans don't cut it for even smallish prints. Scanning those at e.g. 1600dpi using the film holder will produce adequate scans for moderately sized prints. Straight on the glass, it's going to remain a severely compromised situation. That's why the difference matters on the 4990.

You would be naive to think

That's not a very kind way of formulating a response when talking to people who seem to be perfectly aware of real world resolution requirements, limits of hardware and technical possibilities.

 

[David Reynolds]

Nice, same here - it's a great piece of kit! Still use it very often.



The 4990 Pro retailed with IT8 calibration targets and some additional software; I think the SilverFast version included in the bundle was one of those items. The hardware is/was exactly the same.

The 4990 has no hardware provision for focus adjustments. There's no extra lens, no focus servo etc. It's a fixed-focus device. Open it up and check for yourself. Yes, I did that when cleaning it, out of curiosity, many years ago.

Also:

Same 35mm B&W negative scanned with film holder (left) and straight on the glass (right). scanned at 3200dpi and downsampled to 1600dpi. 100% crop from comparison image.
The only difference in processing was that the left one was scanned with the Film Holder setting selected and the right with the Film Area Guide. Contrast & curve settings were left alone between scans. Contrast was adjusted in GIMP after scanning with an identical adjustment for both images.

There's a bigger version here, but WordPress has downscaled it a bit. Still, the lack of acutance in the fine detail in the treetops is very apparent in the glass-level scan, even in this dramatically down-rezzed version.

Really, this scanner doesn't adjust focus - because it physically can't do this.

Again you are correct. I was speaking from low res scans . Thanks for putting me straight on the actual physical operations of the 4990 scanner. Good to know.

 

[koraks]

I appreciate the response, David. I'd be happy to see you continue sharing your experience in publishing here for the benefit of all of us.

Also, my apologies for the somewhat sour comment above. I agree that the resolution of these scanners is inherently limited.

 

[David Reynolds]

That may work fine for 4x5 and 8x10 negatives, but for 35mm 800dpi scans don't cut it for even smallish prints. Scanning those at e.g. 1600dpi using the film holder will produce adequate scans for moderately sized prints. Straight on the glass, it's going to remain a severely compromised situation. That's why the difference matters on the 4990.


That's not a very kind way of formulating a response when talking to people who seem to be perfectly aware of real world resolution requirements, limits of hardware and technical possibilities.

Again You are correct . I work in 4x5 for most reproduction work and speak from that experience. Using Silver Fast and a holder has been more than adequate . I have very little experience with 35mm.
I used the word naive not in a derogatory way but to address most of the people reading this forum who make internet images or 8x10 prints not people who obviously have greater technical knowledge than I have.

 

[koraks]

I have very little experience with 35mm.

Sensibly so; having briefly looked at your work and given your emphasis on reproduction, I think the limited real estate just doesn't warrant much flexibility for your purposes.

I used the word naive not in a derogatory way

Gotcha; sorry for taking it that way!

 

[brbo]

Why people here are scanning at 4800dpi I have no idea other than that is overkill. If one is producing images for human consumption (Internet images and /or smaller format prints) why produce an image beyond a human's grasp.

A lot of consumer flatbeds (all?) benefit from scanning at a higher nominal resolution (to reach the highest effective resolution possible). So, if you want absolutely the best possible results, you scan at higher nominal resolution and downsample later, you scan at optimum hight...

It can make a slight (but noticeable) difference when you need to scan 35mm and you only have a consumer flatbed scanner available to you. If you only need 300-800dpi from your scanner it's something you will never notice, though.

 

[David Reynolds]

A lot of consumer flatbeds (all?) benefit from scanning at a higher nominal resolution (to reach the highest effective resolution possible). So, if you want absolutely the best possible results, you scan at higher nominal resolution and downsample later, you scan at optimum hight...

It can make a slight (but noticeable) difference when you need to scan 35mm and you only have a consumer flatbed scanner available to you. If you only need 300-800dpi from your scanner it's something you will never notice, though.

It makes a lot of sense to scan to achieve the best resolution possible, then downsize when applicable. I agree and see your point. In my case the file is going on to a printing process (usually Offset Litho for commercial publication) so the end viewer only sees the finished product. Therefore it only had to be within the human eye's capabilities. As well as; from a business standpoint the less time spent scanning (meaning smaller dpi) the better the profit margin. For my own photographic and Art work I always went from a negative or transparency and printed from them directly on photographic paper (the resolution on a silver negative or color transparency could never be improved by a scan). The only time I used a scan (Drum type) for my artwork was when the prints were so large (40" x 60") that a digital file necessary by the printing lab.