Comparison of Howtek 8000 Resolve to Epson V850

[Sasquatchian]

If they used a 12.7 micron aperture - and DPL rounds those numbers up - to 13 - you'd really only be getting 2000 ppi, which would help explain why your Howtek scans don't look any better than the Epson. Then, there's also what aperture you shot the original film at and whether you have diffraction limiting what the film can record.

[Torontoamateur]

Where is this picture taken? Where were you ? I am going to the Grand Canyon and would like to see this vista.

[Ken Lee]

Good question indeed. And I would guess when I start printing above 16 x 20. I plan to make same rather large panos and 32x40 images. Plus, for my astrophotography and astronomy research, the drum scanner will give me better images suited for the science aspect of things.

For 4x5 film, a 16x20 is roughly a 4-to-5x enlargement. This tallies well with the notion that that actual usable resolution is around 1500 dpi, since a 300 dpi print at 5x requires 1500 dpi.

That's what forum member Nathan Potter determined when he provided this graph and pointed out that 50% contrast is a typical cutoff point when evaluating lens performance. The Epsons reach 50% at 1500 dpi.

So while we may be able to specify a scanning resolution as high as 6000 dpi - and while we may be able to just barely distinguish tint bars on a USAF target at 2400 dpi - the scanner starts to go south well before. For greater enlargements, we get superior results (as you have shown) with a superior scanner.

Thank you !

[Steven Ruttenberg]

Everything aside, I know from my own work, that when I scanned at 2400 dpi even 3200 dpi on the Epson, the scans were considerably softer than at 6000 dpi. I choose to scan at 6000 to 6400 then bin to 3200dpi and when I print I size to 300. Some have even said to just let the printer make the necessary conversions from whatever your resolution is.

Based on what I see from the Howtek scans of my images compared to my Epson scans of same images, I am going with the Howtek. Just what it will be. Thought I would put up real images and not the typical colored beads, and fruit baskets, but real images by a real photographer, good or bad.

Anyways, if someone wants to post up their drum scans compared to their flatbed scans of actual images, I encourage that. I have made available the actual scans of each for download and the differences are obvious, not just in color, but in detail, etc. While charts and graphs are useful, I base my final product whether science related or photographic on the final state/outcome. Sometimes we can't see the forrest for the trees.

[Steven Ruttenberg]

Where is this picture taken? Where were you ? I am going to the Grand Canyon and would like to see this vista.

This is taken from the North Rim (closed for the winter). This is east of Cape Royal on the south side of the road.

[Steven Ruttenberg]

I will give the Epson one advantage, software that uses it can bin the images with a 2x2 which smooths out things, like grain, but at a small expense of sharpness.

[Sasquatchian]

Steve - I just downloaded two of the high res images from your dropbox and have a few comments.

One, the 100% crops you posted were not anywhere near 100%. They were, in actuality, right around 19%. Second, why on earth would they make untagged RGB gamma 1.0 scans? No reason ever, well, almost ever to do that. These are some of the worst, filthiest drum scans I've ever seen. And both from the taping around all four edges AND from the moiré patterns in both the color and black and white GC shots, indicates that these were dry mounted with no scanning fluid or overlay. Fluid mounting gives cleaner, sharper scans with measurable improvement in the d-min. It also looks like they did bother to clean them at all. And these folks who sold you the scanner were trying to give you an idea of what to expect? The Provia scan at 100 percent does look pretty soft overall even after you sharpen it. It feels like the film itself is not optimal in terms of what it can produce. Likely a case of where hi-res reproduction reveals deficits in technique. And finally, saving with zip compression dropped the color image down to approximately 1.3 gb, a bit better for both up and downloading. I would be embarrassed to deliver anything like that especially if I were trying to sell the scanner. Sorta speechless really.

[Steven Ruttenberg]

Here is a final comparison.

[Sasquatchian]

At this point Steve, it's sorta pointless to do comparisons when the scans from the Howtek are nowhere near the best they can be. As has been said many many times - the scanner operator matters when operating a drum scanner and therer is a huge learning curve to get really good at it. Maybe I have to go buy an Epson and make my own comparisons - like I have time to do that - but I just might.

[interneg]

The Epsons reach 50% at 1500 dpi.

50% at about 650ppi is more like it & a max resolution somewhere a little above 1200ppi but well under 2400ppi matches more realistically with what I've had to deal with off various Epsons (including the A3 models with autofocus). 1500ppi at 50% might not however be wide of the mark for an average LF setup (if lens, film flatness, camera solidity, transparency film performance are factored in).