Why do most scanners scan larger negs at lower resolution?
For example, the Hasselblad X5 scans 35mm negs at (true, non-interpolated) resolution of 8000dpi, but 4x5 negs at 2040dpi
http://www.hasselbladusa.com/product...xtight-x5.aspx
The Epson v750 uses its high quality lens to pull 6400dpi from negs up to 4x5, and then switches to its dedicated worse-resolution lens when it scans 8x10 negs to give you 4800dpi.
What's the point?
Re: Why do most scanners scan larger negs at lower resolution?
If you think about how a scanner works, you realize that a lot of precision is required. Precision is expensive. To precisely scan a large area is commensurately expensive.
The Hasselblad scanner is probably only giving 2040 dpi, and the manufacturer is honest about it. Epson claims 6400 dpi, but in fact the scanner delivers a much lower number. There have been many discussions on this forum and elsewhere about the actual resolution of the Epson.
Re: Why do most scanners scan larger negs at lower resolution?
Point taken, but an X5 is 30K. That's pretty expensive!
And why should the v750 deliberately switch to a shitty lens for 8x10?? The CCD and transport system are the same - why switch the lens? If I'd known that I never would've bought it.
Re: Why do most scanners scan larger negs at lower resolution?
And isn't the main factor simply time?I'm happy for the X5 to take half an hour to scan my 4x5 at 8000dpi - how is it more expensive?
Re: Why do most scanners scan larger negs at lower resolution?
There are several factors: optics is one, and accurate "stepping" or stitching is another. Even with a superb lens, the scanner needs to be able to move along the target and assemble a large image out of small samples. Drum scanners solve the problem by spinning the film and moving the head in only 1 direction. Flatbed scanners need precision in 2 dimensions. And then there's the issue of keeping the target in focus. And of course there's the whole computer-side issue of scanning: try writing the software to interpret and save the electronic signals as data files. It has to be done well (and quickly) or the images will be mediocre.
Compared to Hasselblad, Epson is a large company with vast technical resources.
If you had to make your own scanner and sell only a handful of units, how would you price them to recover your investment ?
Re: Why do most scanners scan larger negs at lower resolution?
I see. Thanks for the explanantion. I still say 30K should buy a lot of stepping :)
Re: Why do most scanners scan larger negs at lower resolution?
Quote:
Originally Posted by
Uri A
Point taken, but an X5 is 30K. That's pretty expensive!
And why should the v750 deliberately switch to a shitty lens for 8x10?? The CCD and transport system are the same - why switch the lens?
Simple, because the CCD is not wide enough. As a matter of fact, I do not know whether the V750 switches the lens only - I never had one, on all flatbed scanners I've owned that had similar restrictions they switched the entire lens/CCD unit. The Imacon does resize with its lens (presumably by zooming), but the reason is essentially the same - it has a 8000px wide scan line, and has to optically fit whatever it scans to these constraints.
Re: Why do most scanners scan larger negs at lower resolution?
Thanks Sevo!
Anyway, it seems to defeat the purpose of shooting (or at least scanning) 8x10.
Re: Why do most scanners scan larger negs at lower resolution?
[QUOTE=Sevo;1061183]
Quote:
Originally Posted by
Uri A
Simple, because the CCD is not wide enough. As a matter of fact, I do not know whether the V750 switches the lens only - I never had one, on all flatbed scanners I've owned that had similar restrictions they switched the entire lens/CCD unit. The Imacon does resize with its lens (presumably by zooming), but the reason is essentially the same - it has a 8000px wide scan line, and has to optically fit whatever it scans to these constraints.
The V700/V750 definitely has two lenses and the scanner switches based on which scanning area you choose in software. If you select film holder the scanner uses a shorter focal length lens and the maximum area that is covered by the lens is about 6.5 X 8.5". If you select film area guide the scanner switches to a second lens that is positioned farther from the CCD to cover a wider area, which allows one to scan a full 8X10" negative. The lenses are probably not very different in optical quality, but the closer you place a lens to the CCD the higher resolution you can get. Imacon/Hasselblads and the old Leafscans do the same thing, position the lens closer to the CCD to cover small format media, farther away to cover 4X5. For example, with the Leafscan 45 it was possible to scan a 35mm negative or slide at 5400 dpi, medium format was only 2400 dpi, and 4X5 was 1200 dpi. The Imacon, which was probably based to some extent on the design iof the Leafscan, uses a similar method of zooming. It is rather like the concept of using an enlarger, the closer the lens and film is to the baseboard, the higher the resolution on paper.
The Eversmart/IQSmart scanner use a high quality lens placed very close to the CCD and scan in rows, like mowing your lawn, then the files are stitched in proprietary software. This is called XY scanning. Cezanne scanners will also do XY scanning, but there is no stitching in the software so you have to do that yourself after the scans are made.
Sandy
Re: Why do most scanners scan larger negs at lower resolution?
Quote:
Originally Posted by
Uri A
Thanks Sevo!
Anyway, it seems to defeat the purpose of shooting (or at least scanning) 8x10.
V700 scans 8x10 at 2400 dpi making quite reasonable results except with colour film. The problem is that 8x10 can only be scanned with the film on the glass and using the film area guide. With black and white films you can scan the film emulsion down and flip the results, this seems to successfully eliminate Newton Rings. Colour film doesn't do this, the results have Newton Rings no matter which way down you lay the film...
I'd love to know of any scanner that can scan 8x10 without making Newton Rings...
RR