Who needs a 5400dpi scanner?

[robert goldstein]

Minolta has announced their newest 35mm film scanner, the Dimage

5400, which has a resolution of 5400dpi. Since 4000dpi scanners can

already pick up film grain, what advantage is there to even higher

resolutions?

[scott_eaton]

The higher DPI tends to make lower rez scans sharper because the optics are inherently better in the scanner to achieve this resolution.

 

Othwerwise, about the only need I see for 5400dpi is to downscale to 4000 or 2000 dpi.

[beauh44]

Good question, Robert!! I don't have an answer. I use a Canoscan 4000US to scan my 35mm and certain films drive me nuts because I'm staring at grain and have to "back off". I have to wonder if it's marketing. Best wishes . . .

[beau 1664876222]

That's like saying, what's the point of buying an expensive lens when I can see individual hairs on the head of my subject using a cheap zoom? There's seeing it and there's seeing it in all its texture and tonality. I personally love grain, and I want a scanner to capture the look and texture of grain accurately. What I like about high-res drum scans is how pleasingly natural the grain looks. Current 4000 dpi scanners tend to render the grain as little 2-dimensional blotches. My hope for this new Minolta is that it will be a reasonably-priced way to make film look like FILM.

[ellis_vener_photography]

As I recall certain Imacon scanners already on the market can produce 8000ppi

(scanners don't produce dots, the produce pixels) scans. So clearly there is a need for

large scans. The other interesting features of this scanner are the price (about $9000

and that unlike the Canon or Nikon 4000ppi scanners, the Minolta Scan Elite 5400

produces

true 16 bit per channel images. A 16bit per channel (or 48bit) depth scan of a 1" x

1.5" image (a 35mm negative) at 5400ppi makes for a huge file size: 250Mb by my

calculations. The other difference is the maximum size, 13 x 20 @ 300dpi prints

(4000ppi scans); 18x27 @ 300ppi prints (5400ppi scans).

[beau 1664876222]

Ellis, I think you meant $900 for the new Minolta.

[jbq]

People who shoot Velvia (or other high-definition films) want this kind of dpi (and they'll probably want more).

[dalius_gilvydis]

Where is this thing? It's supposed to be on the shelves in June. I called Minolta and they said it was released on May 12th.

[edward_c._nemergut]

FWIW, I find that 35 mm scans I made with my Minolta Dimage Scan Multi Pro (at

4800 dpi) do have more detail (not a lot, but a noticeable amount) than scan made

with my Nikon LS4000 (4000 dpi)...Given that fact, I would anticipate 5400 dpi scans

to be incrementally better...

[ellis_vener_photography]

Yes, I meant $900.00 not $9,000.00.

[sergey_oboguev]

There is life beyond 4000 DPI, as noted above.

 

However, according to the specs on the site, this scanner has two other important features: high dynamic range (higher than 4000ED, whether this is true or not is another matter, to be seen) and light diffusion control.

 

The latter, as well as higher resolution, matters to fight grain (and grain aliasing) effects.

[charles_miller]

As a former owner and user of a Nikon 4000 ED scanner, I agree that there is life above 4000dpi. And a diffuse light source and a better dynamic range could be equally important improvements too.

 

It is well documented that the Nikon LED light source pops dirt, scratches, and grain more than scanners with diffuse light sources. What this means in practice is that you must subsequently use stronger settings of ICE and GEM to get a clean and smooth image. And that in turn softens the image, so you get less than 4000 dpi as a final result.

 

What the Minolta 5400 seems to promise is not only 5400 dpi, but a diffuse light source which would allow milder settings of ICE and GEM to accomplish the job. So the Minolta 5400 should actually perform closer to 5400dpi than the Nikon 4000 does 4000dpi, after the visually equivalent amount of image processing is accomplished. That is the theory, anyway. It is hard to argue that a diffuse light source would provide a higher dynamic range image since there is no clear trend that way in existing scanners. But it would be another very nice bonus if the Minolta 5400 somehow delivers on that promise too. I suspect that SilverFast software will be an important accessory for the best color purity, just as it is with many prior scanners. But why not hold off judgement on the Minolta 5400 until some trustworthy photo.net citizen like Ellis or Carl can report "the straight skinny" and reveal "the gouge" on line? Know what I mean?

[marshall]

Given some tests I've seen about the actual resolving power of film (wish I could find it again to cite the source), it will be interesting to see if the 5400 dpi of the Minolta starts to surpass the detail in the film. I've seen tests showing that, even with good equipment shot at good apertures, many slide films sorta peter out for resolving power around 5000. Now, I don't think that means that the Minolta will get everything possible out of the film (affordable PMT's, anyone?), but maybe we're nearing the point where the only reason to go to drum scans of normally-exposed 35mm is for dynamic range?

[carl smith]

Well forgive me if I disagree with the optimism here, but most film won't go anywhere above 4000, and in fact 4000ppi is often above what film contains. We had a discussion about this just a few days ago in another thread. 4000ppi does extract more detail other than just grain compared to a 2800 or 2900 ppi scanner. However for many slides and color negs as well as B&W material you've already blown through peak resolution.

 

The Minolta will resolve grain finer but won't resolve more actually useful image detail in most situations. Similar situation occurs with the flextights where bigger doesn't really mean better image detail. Drum scanners are no different.

 

What interests me about the Minolta is the speed, true 16 bit capability and hopefully the better Drange and Dmax. The diffusion capabilities are also excellent benefits for the unit. Resolution, IMO, is not the most important selling point. However for many people it will be the most effective.

[stephen_ratzlaff]

the Minolta 5400 at 5400ppi to provide scans with better tonality than older scanners' scans at 4000ppi.

 

Again, this is one of the reasons to use large format films as well, the higher pixel density of the 5400ppi scanner should, in theory, provide the same advantage.

 

Also, as the mfg.'s introduce newer, better models, you can probably count on the speed to improve. I.e., the 5400 scanner scanning at 4000ppi will most likely be faster due to improvements in design and scanning algorithms, etc.

[scott_eaton]

This is one of those threads that tries to prove that 35mm slide film has more resolution than larger formats. We also have the usual drastically exagerated claims of the resolution of slide film.

 

If film were this good, including Velvia, we woulnd't be shooting larger formats to avoid having to scan above 2000dpi in the first place.

 

I'll still bet my 6x7 scans off my beater Epson 1640 look much better than your weasley 35mm scans off a 5400dpi scanner. If you're really that obsessed with getting better results, put your money into bigger film vs sharper scanners.

[daniel_staver1]

Actually, your 6x7 scans on a flatbed probably doesn't look much better than a 35mm 5400dpi scan on a real filmscanner. I have an Epson 3200 and a Canon Canoscan FS4000US, and a 6x7 on the Epson just about equals a 35mm scanned on the Canon in terms of sharpness and detail. Maybe it's a little bit better, but it's definitely hard to tell.

 

I do agree that the scanning process being equal, having a larger film format is much preferable to increasing DPI.

[carl smith]

who knows, scott has a mutant flatbed

[oskar_ojala]

I got to agree with Carl, 5400 ppi is useful only for a fraction of chromes, the other features are more significant with the new scanner. Hopefully speed would be close to the speed of an Epson 3200 (2.5 min for a 6x6 @3200 ppi). But the scanner will probably be pretty sharp at lower resolutions too... However, ir remains to be seen how close the actual resolution will be to the advertised.

[jonathan_bundick]

"This is one of those threads that tries to prove that 35mm slide film has more resolution than larger formats."

 

No one made this claim/comparison anywhere in the thread. You are the first to attempt to distract/deviate the thread to larger format comparisons.

 

Lets get back on track. It would be doubtful that most of us, even under the most optimal conditions, can put more than 50lpp/mm on film. The question comes down to what is the "actual" resolution of this or any other scanner. The manufactures stated resolution us usually far from what it actually can see. If we assume 50 - 60% of stated resolution as a rough estimate, then, the scanner in question can see between 50-60 lppmm, or more than any of us can hope to put on film. Will you get better results from it as opposed to a 4000dpi scanner? How strong or weak is the rest of your photographic chain?

[carl smith]

Jonathan, the numbers quoted by some do make that claim and that's Scott's point.

[ethan hansen]

Getting away from the "how many grains fit on the head of a pixel" debate, I find it curious that the Minolta makes 16 bit output. Photoshop's implementation of high bit editing discards the topmost bit, leaving only 15 bits to work with. If you edit and save a file from the scanner, half your colors disappear. This sounds as though Minolta has a really cool A to D converter they couldn't wait to show off.

[jonathan_bundick]

"Jonathan, the numbers quoted by some do make that claim and that's Scott's point."

 

No Carl, this is his, and apparently your, conclusion. I have yet to see a scientifically valid study presented by anyone on Photo.net that shows film and camera can�t resolve beyond XX lpp/mm under conditions X,Y, or Z. Opinions based on personal experience, well meaning for the most part, are what one reads here. (Including my own.)

 

The difference between theoretical resolutions and what most of us actually produce is a point of contention between many. I tend to agree that we don't put much more than 30 lpp/mm (if that much in many situations) on film, and as a result wouldn't benefit much from a 5400dpi scanner, but as it is not an accepted fact, live and let live. This argument wou't be solved by anyone on this thread.

[marshall]

Well, we're clearly getting into the resolution range where improvements in scanning are somehow inexorably tied to film resolution. You can't read detail that isn't there. I, for one, by mentioning that 5400dpi is beyond the limits of most film wasn't trying to a) suggest that most 35mm shots offer anywhere near that much, b) trash the possibly-quite-good new Minolta scanner, or c) claim that I'm an expert on any of it. I just wanted to suggest that, based on what I've seen and read, the extra resolution here may have surpassed even the BEST 35mm can offer.

 

Scott's made the point in other threads that all of us who go out and spend our efforts trying to extract the last bit of detail from 35mm should be shooting other formats, and I'm not surprised to see it repeated here, slightly off-topic though it may be. That's not really practical for many shooters, but I don't want to get the thread bogged down in it.

 

Onward.

[charles_miller]

But what if 5400 dpi, a diffuse light source, and Minolta's "Grain Dissolver" software all work together for results better than GEM? Or faster than GEM? Or both? And what if there is still money left for SilverFast software and a calibration target. WOW dude!