Scanner accuracy

[couldabin]

How accurately does a flatbed scanner (I'm using an Epson 4990) record distances? For example, how certain can I be that if I scan a 4x5 negative at 4800 dpi, the points that 4800 pixels apart in the electronic file are truly 1.0000 inches apart on the physical negative? I would think that the use of stepper motors and gears would ensure a very high degree of accuracy. On the other hand, output devices pretty much have to be accurate but I'm not sure the same could be said for input devices. Does anyone know if the DPI can be used as a tool for accurately measuring distances on negatives?

Thanks.

 

[Doyle Thomas]

Try scanning a ruler, i know that some scanners are more accurate than others and can be different for reflective ~vs~ transmission. use one of those clear office rulers.

 

[couldabin]

Try scanning a ruler, i know that some scanners are more accurate than others and can be different for reflective ~vs~ transmission. use one of those clear office rulers.

I think I have just the ticket -- a glass slide, marked with 0.01mm lines, that is used to calibrate a microscope. Will see how it scans ...

 

[RalphLambrecht]

Try scanning a ruler, i know that some scanners are more accurate than others and can be different for reflective ~vs~ transmission. use one of those clear office rulers.

very interesting question and test but,I have little faith in the accuracy of plastic office rulers.I'll try a metal machinist ruler instead

 

[Doyle Thomas]

very interesting question and test but,I have little faith in the accuracy of plastic office rulers.I'll try a metal machinist ruler instead

lol idk we were talking rocket science and a metal rule wont scan transparent

 

[Alan Klein]

You need an X-ray scanner for metal rules.

 

[RalphLambrecht]

lol idk we were talking rocket science and a metal rule wont scan transparent

of course it will scan just fine with a reflection scanner or mabe when layed edge to edge witha transparent ruler; I doubtYou'll get a ruler into a negative scanner or get it to focus anyway, but feel free to try if you insist

 

[couldabin]

I think I have just the ticket -- a glass slide, marked with 0.01mm lines, that is used to calibrate a microscope. Will see how it scans ...

I scanned it at 12800 dpi and it came up as a solid glob. But it was a 1mm wide glob, and surprisingly close -- 501 pixels vs 504 theoretical. Surprising, because Epson evidently isn't prepared to make any promises about the scanner's accuracy. I asked what the dimensional accuracy of the 4990 is, and this is what came back from Epson today:

"Since the Perfection 4990 is a general purpose "photo" scanner, this type of specifications are not available."

 

[Arctic amateur]

A metal ruler might scratch your scanner glass.