It's a little different. Unlike high end cables, where you have to squint your ears to tell a difference, the D800 and Tango images are obviously different.Quote:
Originally Posted by williaty
Printable View
It's a little different. Unlike high end cables, where you have to squint your ears to tell a difference, the D800 and Tango images are obviously different.Quote:
Originally Posted by williaty
Has anyone tried the Epson Expression 12000XL Photo scanner?
I'm actually pretty interested in it now after hearing it can autofocus or manual focus on negatives. I think it would be really great for 8x10 neg scans vs the v700/800 series
I close the lens 3 stops. That would normally read f/8. But in the macro range it's f/13. still 3 stops from wide open.Quote:
Originally Posted by Greg
Yes, four overlapping quadrants. With the sensor plane a little farther than 15.75 inches, Photoshop's stitching algorithm can't put the quadrants together, although they do overlap. At 15.75 inches, they overlap just enough and the resulting scan is about 2600-2800 ppi.Quote:
Originally Posted by Two23
For 35 mm, the shot is 1:1, 5000 ppi, and the whole rig is much closer to the subject.
I suppose the LED panel could be put on the wall, but this setup is much easier than doing that. Why fight gravity any more than necessary?
I completely agree. I was going to suggest this.Quote:
Originally Posted by williaty
I was also going to suggest shooting a roll of Velvia, same subject, same exposure for the whole roll. Send the samples out to whoever volunteers as "standard observers." Then we could do the scans and everyone who has samples can evaluate them in terms of comparison with the "original" examples they have. Right now I don't have a film camera, but that's solvable.
Pali, are you up for some more comparisons? I could scan some difficult, but otherwise "real world" Kodachromes that have given my colleagues and me some challenges over the years and send them on to you to scan.
Rich
Not easy to do! ;-)Quote:
Originally Posted by faberryman
There are some pretty detailed threads in the diy section on dslr scanners.
Rich,
If that is an LED light box, then quite possibly you will be missing some color. I've been working with different digital camera capture scanning methods myself, but the color of the LED and other light sources can cause problems. I am using the Sekonic C-700 spectrum analyizer to measure the light source.
CRI is one thing but a light source can still report relatively high CRI and have some colors attenuated - often in the Reds. Quite honestly I am skeptical that your D800e with that setup is giving you more accurate colors unless you have somehow made an input profile for that light source for your camera.
Certainly the Tango scan looked much better to me in the images shared.
Quote:
Originally Posted by Rich14
12000XL it is very interesting for ULF until 31 cm x 43.7 cm., or to scan A3 prints.Quote:
Originally Posted by axs810
IMHO 12000XL has a comparable or lower performance than V700-850 for 8x10", depending on how you scan.
The 12000XL it only has 2400 hardware DPI Horizontal, while the V750 has 4800 (for 8x10) delivering around 2000 with resolution USAF 1951 targets. In the V750 the sensor outresolves the internal lens system. If the V750 scans until 5.9" wide then it uses the HiRes lens and has a pixel density of 6400 instead 4800, resolving some true 2300 to 2800 dpi. IMHO the 12000XL won't reach these figures.
Scanner Focus has some importance for 35mm scans, but way less for larger formats. But for 135 use a dedicated roll film scanner to get most in case of challenging jobs.
Quote:
Originally Posted by EH21
This has been (IMHO) an interesting thread, because it makes think how things work.
Beyond stray light, IMHO the better Tango result is a lot from dynamic range, PMT (Photo multiplier RGB sensors) from drum has more DR than the D800, so the D800 should need Multiexposure to match.
And the end the scanner reduces spectral color information to only 3 rgb values, this is after film has reduced scene spectral information to 3 dyes. As we are talking about Spectral to tristimulus information CRI may have a reduced impact that can be solved by color calibration. I'm guessing, more than CRI a significative criterion may be that RGB illumination peaks should match the transmission peaks of film dyes (each film can have different peaks) .
I'm not completelly sure, but I guess that if illumination and (particular) color film have the spectral peaks in the same points then color calibration will work more accuratelly.