I looked at Velvia 50 under a microscope today

<p>Here we got out first 35mm slide scanner back in 1989. Our first 2 pro flatbed cost 3 grand each in the early 1990's. I have Nikon and canon 2700 and 4000 dpi film scanners for 35mm; plus two Nikon 9000 MF units that are 4000 dpi.</p>

<p> I have tested lenses with USAF charts for over 40 years.</p>

<p>If I look that swarm of the publics "sharp" 35mm stuff to scan; about zero is in the 100 line pair per mm league; it is more like say 50 to 60.</p>

<p>Few if anybody records 100 like pairs per mm on film in pictorial useage; unless they are a total bull shitter.</p>

<p>One does have some rare lens tests that did record over 100 line pairs per mm on film; if one looks back and cherry picks data; and uses goosed test conditions; ie 1:1000 test targets; Panatomic-X; granite blocks; timed lights; *AND* one brackets by 1/3 stops and cherry picks the best data.</p>

<p>A film scanner is going to low pass details; depending on the real useable dpi of the device. Use a dump flatbed and one how has 24 to 32 line pairs per mm. Use a 4000 dpi film scanner and a zillion dpi reference etched glass negative; and one gets sub 80 line pair per mm results; ie normally 70 to 75 numbers.</p>

<p>This whole subject is old.</p>

<p>For Will Frost,<br>

<a href="http://www.edmundoptics.com/onlinecatalog/displayproduct.cfm?productID=1331">Test chart with differrent contrast levels</a></p>

<p>Interesting discussion. Sorry I'm late to the party, but after reading through the posts, these are my impressions:</p>

<p>1. Mauro's combined image of the microscope scan and 8000dpi scan of velvia does suggest roughly 160lpmm (*not* 160lp/mm, but l/mm) is conceivable for Velvia 50.<br>

2. Shooting test charts gives an advantage to digital images, where the processing involved in getting from the bayer to the actual image favours black and white images with sharp lines.<br>

3. I haven't seen a conclusive test so far which shows all of the detail film has to offer. Ideally, I'd like to see a Heidelberg Tango scan of various 35mm film shot of an organic target with infinite detail (i.e. a landscape) versus, say, a D3X or similar camera, versus medium format (as the reference).<br>

I would love to do this comparison, but I don't have access to any of those bits of equipment!! Any takers?</p>