file size of 33mb from super coolscan 5000, of 35mm neg (Nikon teck info) what about epson flatbed 750pro . said another way how do the scans compare? thanks
The dedicated Nikon scanner is likely to have a greater D-max, meaning a greater dynamic range of tones that can be differentiated. The Nikon can also do multiple passes to reduce noise/increase detail. If you are dealing with 35mm-6x6cm only, then look into a dedicated film scanner, which of course, can also do black-and-white or color negatives.
aside from D-max (nikon @ over 4 , epson @under 3.8) …so same neg 35mm. do a scam with each machine without interpolation. import to photoshop , do the dance with curves layers contrast etc. output to print on epson 3800, full frame, 16"x 20" , look at the results ? and pick the better ? said another way nikon without fumbleing around said "35mm B&W neg scanner will produce a 33mb file at 16 bit ". so turning to epson I ask the same question. The response was " well that depends on what resolution you scan at ". made me feel like we were going in circles. then epson said "you could buy the 750pro do a scan then you would know the file size" i then desided it was time to regroup……..take a coffee then do some more digging…
The Nikon Coolscan 5000 will give you 4000 ppi optical, which on a 35mm neg is hovering around 60 mb 8 bit RGB and 120 mb 16 bit RGB. The Nikon is a much better scanner than all but the most expensive flatbeds – for scanning 35mm negs and transparencies, including the Epson.
Don’t believe Nikon’s claims of a 4.0 d-max. It’s much closer to 3.0, but that shouldn’t be a problem with negs, only transparencies.
Nikon Coolscan scans will tend to be sharper (even at the same resolution setting) than Epsons, have less flare and more contrast, but it’s more difficult to scan the entire neg including the rebate edge.
The file size you need will depend of course, on your output requirements and whether or not you are attempting to make archival high quality scans of the negs in question. If that’s the case, you may indeed want to scan at the maximum resolution.
1. Neither scanner can approach a D-max of 3.8, let alone 4.0. These are CCD systems. The highest useable D-max obtainable with either is 2.8-3.0. The methods scanner manufacturers use to calculate D-max for these kind of machines is blatantly dishonest and nothing but shameful marketing hype.
2. A 33mb monochrome 16 bit file indicates an original scan of a little under 4000 dpi. Either scanner can output such files.
Using a drum scanner, no additional detail can be resolved from BW 35 mm film beyond about 2800-3200 dpi without using special materials (microfiche) not suitable for general-purpose photography. Color film and slide material maxes out lower than that.
Neither of these scanners, using their so-called maximum optical resolution of 4000-4800 dpi can match a drum scanner operating at the same resolution. They will typically achieve about half as good figures.
Meaning they will just barely resolve most of the detail inherent in the 35 mm format. But that image will not match a drum scanner in sharpness (sharpness is NOT the same as resolution), contrast or smoothness. It will be somewhat soft, flatter and lacking in shadow detail.
Carefully matched in Photoshop and printed to equal size, it will be hard to identify the scan file as being either from the Nikon or the Epson.
Both these scanners perform about as well as CCD scanners can.
Getting a bit off topic on the drum scanning stuff here. A very sharp T-Max100 neg actually can show improvement in detail when going from 4000 to 8000 ppi – at least on my Howtek 8000, but because the scanner has a 3.17 micron and then jumps to a 6.35 micron aperture for the 8000 and 4000 resolutions respectively, it’s hard to say exactly where the limits are, but they are there. I’ve even seen very slight improvements in detail in 35mm Tri-X going to 8000 ppi, at the expense of higher apparent grain. Velvia 50 is another where you can see improvements. For many images it’s just an intellectual discussion as most don’t hold the amount of detail the film is capable of. There is still some outdated Tech Pan floating around eBay and that is the finest grained sharpest film I have ever seen.
I agree with everything else you wrote regarding the scanners.
The real pros seem to think the flatbed scan would approximate the Nikon 5000ED. I scan with an 8000ED and have always felt that 35mm was far better scanned on the Nikon than on the flatbeds …but I do not own the Epson 750 so I must defer to folks who have actually scanned with both devices. Not just read the specs.
I see the same on my Howtek. In fact, the same thing has been shown on D4000s when set to scan at 8000dpi. (!)
The increase in the 4000s is of course a surprise, but is significant and must have something to do with the internal processing of the raw scan data. I’m not sure either D4000 or D8000 machines are deriving any more actual resolution, maybe just delivering everything they should have at the 4000 setting (?). When I had a Tango, it resolved no more detail than the D4000.
Somewhere I have a chart for drum scanners of the relationship between scanning dpi and "real" resolution as measured in line pairs per inch. All I can remember right now is that actual resolution doesn’t double from 2000 to 4000 dpi. I think it’s about a 1.5x increase. And the real resolution increase of 8000 dpi vs 4000 is less than that.
CCD scanners have significantly less resolution at given "dpi" scan frequencies.
"1. Neither scanner can approach a D-max of 3.8, let alone 4.0. These are CCD systems. The highest useable D-max obtainable with either is 2.8-3.0. The methods scanner manufacturers use to calculate D-max for these kind of machines is blatantly dishonest and nothing but shameful marketing hype."
That speaks to a MAJOR pet peeve of mine, with any kind of equipment.
Standardize the parameter testing. Have the tests performed and vetted by an independent lab. Publish the data. Let the consumer make a properly informed decision.
On the 4000 and 4500, when you asked for 8000, you were scanning with the 6.35 micron aperture, but slowing down the drum lead screw speed. This effectively give slightly more resolution in on direction. On the 8000, you can scan with the 3.17 micron spot, and you really can see the difference if and only if you have film and image detail that can show it. The Tango has a 10 micron minimum aperture with give it a pure optical resolution of 2540, but with the same drum mechanics as the Howtek, effectively gets somewhat more in one direction. All scanners have some optical loss and never really reach their stated resolution but at least the drums are closer. The numbers I’ve seen from tests on the Howtek were something like 7300 actual ppi when scanning at 8000 and about 3600 when scanning at 4000.
Phos,
Agreed. Unfortunately there seems to be no standard out there for what is acceptable noise, and that’s where the scanner makers have always fudged their specs. People who test these ought to scan a Kodak density wedge and note where you stop seeing separation in the steps. For drum scanners you have to double the wedge up, as it only goes to a 3.4 density, and drum scanners in good condition can hit around 3.9.
Where I first really started noticing the deficiency back in the ’70s (and which is still rampant today) is in music gear catalogs.
For example: You see a page (or several pages, or several dozen pages-the current Sweetwater catalog devotes over 30 pages to them!) full of microphones, along with their short sales descriptions.
Why is it that a catalog retailer would describe the frequency response specs for one microphone, but leave those specs out for the description of a competing microphone just below it on the page?
That **** makes no bloody sense to me whatsoever.
Same with Dmax and other specs on scanners, same with comparing specs on ANY type of equipment. How the heck ccan a retailer expect a consumer to make a smart buying decision, considering those omissions?
Sure, we can go to each individual manufacturer’s website, and gather the info for offline comparison. But the smartest retailer will have already done that work for us. And there don’t seem to be many that think that way, for ANY type of gear, audio, video, computer stuff, power tools…ANYTHING.
Where I first really started noticing the deficiency back in the ’70s
As a brief sidebar…
Try the late ’50s and ’60s when (as a kid) I was building my first stereo gear. Equipment like microphones, speakers, amps, preamps, tuners, turntables, cartridges, etc. often fudged their specs, whether frequency response, output, s/n ratios…whatever…set their own "standards" and willfully omitted or misstated specs to their advantage. Of course, a lot of folks didn’t understand there was a difference between continuous, RMS, peak and unqualified output, so this all played beautifully into their hands.
Some reasonable standards, fortunately, were eventually established for a lot of sound gear, but there were/are exceptions, such as (I think) PA, portable radio and auto sound equipment.
Today, there is still a lot of fun with other kinds of specs: the absurd point ‘n’ shoot camera megapixel horse race; or the idea that a $699 1080p TV is necessarily going to provide the same viewing or feature experience as a same-size one selling for twice the price; or (if you watch the commercials) that an all-wheel drive SUV is going to save your behind while speeding along on slippery pavement, or on gravel or snow; or that regular consumption of large burgers and super-sized fries is perfectly fine for your health.
To much is implied or omitted; too little is asterisked and referenced to an acceptable reference standard.
Where I first really started noticing the deficiency back in the ’70s
As a brief sidebar…
Try the late ’50s and ’60s when (as a kid) I was building my first stereo gear. Equipment like microphones, speakers, amps, preamps, tuners, turntables, cartridges, etc. often fudged their specs, whether frequency response, output, s/n ratios…whatever…set their own "standards" and willfully omitted or misstated specs to their advantage. Of course, a lot of folks didn’t understand there was a difference between continuous, RMS, peak and unqualified output, and other audio details, so this all played beautifully into their hands.
Some reasonable standards, fortunately, were eventually established for a lot of sound gear, but there were/are exceptions, such as (I think) PA, portable radio and auto sound equipment.
Today, there is still a lot of fun with other kinds of specs: the absurd point ‘n’ shoot camera megapixel horse race; or the idea that a $699 1080p TV is necessarily going to provide the same viewing or feature experience as a same-size one selling for twice the price; or (if you watch the commercials) that an all-wheel drive SUV is going to save your behind while speeding along on slippery pavement, or on gravel or snow; or that regular consumption of large burgers and super-sized fries is perfectly fine for your health.
To much is implied or omitted; too little is asterisked and referenced to an acceptable reference standard.
I believe you, Neil, and I can certainly imagine it was much worse then. As a naive kid just starting to understand—and then, REALLY grok—the tech stuff I was interested in during the early ’70s, I just couldn’t understand why these people I looked to for the true dope on specs would fudge, omit or deliberately misstate that info I was looking for.
VHS vs Beta. To those who knew, Beta had the superior recording technology and hardware. VHS marketing, with the emphasis on its longer maximum recording time (despite its miserable picture quality) won out to shoppers who didn’t know (or care) about the format differences.
Right up there with the kind of marketing that’s pushing 12+ megapixel point-and-shoot cameras at amateurs who will rarely make a print larger than 4"x6".
VHS vs Beta. To those who knew, Beta had the superior recording technology and hardware. VHS marketing, with the emphasis on its longer maximum recording time (despite its miserable picture quality) won out with shoppers who didn’t know (or care) about the format differences.
Right up there with the kind of marketing that’s pushing 12+ megapixel point-and-shoot cameras at amateurs who will rarely make a print larger than 4"x6".
Right — I had forgotten about the price "advantage" of VHS! We used Beta machines until they died and were pretty much forced to switch to VHS. We finally got our revenge: DVRs are clearly superior to VHS for time-shift recording. And movies are best on Blu-ray.
That said, and in a feeble attempt to get us back on track, I think we can say that a dedicated slide scanner is gonna be superior for 35mm slides than most other scanner types short of drum scanning.
Lemme ask this of folks already participating in this thread:
Am I the only one who remembers reading—during the nascent age of home scanners—someone who wrote in a prominent magazine article that the effective resolution of film grain was the equivalent of around 30,000 ppi? I clearly remember reading that, and then wondering about what it would take for manufacturers to ever reach an equivalent level of quality.
"Am I the only one who remembers reading—during the nascent age of home scanners—someone who wrote in a prominent magazine article that the effective resolution of film grain was the equivalent of around 30,000 ppi? I clearly remember reading that, and then wondering about what it would take for manufacturers to ever reach an equivalent level of quality. "
The grains size of different film will obviously differ somewhat, but nothing is even close the number you’re quoting here. TMax100 is about the finest grained film available now and it clocks in around 3 microns in grain size. That just about matches up with the 3.17 micron aperture that the Howtek and ICG drum scanners use to achieve 8000 ppi.
Color neg films are coarser, with most of them running between 13 and 19 microns, or a range of something like 1333 to 2000 ppi.
For those who have drum scanners and can manipulate the aperture independently, a quick series of detail scans will show you where the most optimum resolution for a particular film is, but virtually everything I’ve scanned has fallen in the range I’ve mentioned.
I’ve never actually heard of anything with a higher optical resolution than 8000. It gets mighty hard to pump enough light through a 3 micron hole in a disk behind the scanner lens and if you were to try an even smaller aperture, there might not be any visible gains.
Whatever point the author of that article was trying to make, he spoke rather saliently about the 30,000 ppi value. That much I remember.
What context he was putting it into…that’s the part I don’t remember.
Now, given what the article Ann linked to said (Good one, Ann. Just my cuppa!) and what Peter said, that makes scientific sense to me.
That "30,000" bird flew through the window that day perhaps 20 years ago…and may have just as well bashed into the TV, for all I know. I remember it made me shake my head in awe and confusion, since I knew practically nothing about computers and digital imagery then.
I would imagine, that just like a camera lens’ iris diaphragm, if the aperture is very small, you start getting less image sharpness due to light refraction effects at the edge of the aperture.
30,000 spi scanning sounds like putting racing tires on a garbage truck.
The aperture on a drum scanner is little different than that on a camera lens.
There may be slight variations of this, but on the Howtek, there is a lens focused precisely on the film emulsion via a patented autofocus sytem. The light source come through the center of the drum via a fiber optics bundle and provides a focussed light directly into the lens through the film. Behind the lens in a box called the "optical chamber", there is a metal disc with a series of precisely made holes, or apertures. The model of the scanner determines how many apertures and what size they are. The disc is rotated to align in the optical path with the aperture that corresponds to the resolution called for in the software, but that can be overrided if desired.
That aperture feeds the focussed light through three dichroic separation filters directly into three analog photomultiplier tubes, which then convert the light into electrical energy.
I’m obviously not an engineer, but I’m guessing that there is a reason that three microns has been the smallest aperture available on drum scanners. One eight-thousandth of an inch is pretty damn small.
Learn how to rescue details, remove flyaways, add volume, and enhance the definition of hair in any photo. We break down every tool and technique in Photoshop to get picture-perfect hair, every time.
Related Discussion Topics
Nice and short text about related topics in discussion sections
