Buying a 3200dpi flatbed scanner

[ljupco_ristovski]

Hi there.

I'm thinking of buying a 3200dpi flatbed scanner with 3,2Dmax.

Most of the time I will scan 35mm B&W negatives (mostly HP5+ and APX100).

By my calculations, I could print up to 16"x12" without resampling, but that is

in a perfect world.

I need some real life opinions based on experience.

Thank you and have a nice day.

Lj:)

[Ed_Ingold]

Regardless of the manufacturer's specifications, flatbed scanners underperform with regard to resolution. Expect to get about 1200-1500 ppi resolution. Anything more than consists of resampling during the scan. If you scan at 2400 ppi, you will get about all the scanner can do. With this, you will get a good 4x6 inch print from 35mm, or an 8x10 with considerable sharpening and resampling (if needed for at least 300 ppi at the print size). You can do nearly as well with a $5 Picture CD from CVS.

[wogears]

I think Edward's overstating. He's dead right about the res of the scanners, but you can get a very good 8x10 from a flatbed scanned 35mm neg. There are countless examples on the 'Net to prove this. Beyond 8 times mag, however, it does lose badly to a dedicated film unit, even a cheap one.

[sattler123]

If all you are scanning is 35mm format then get a dedicated scanner - not a flatbed. It is difficult to keep 35mm film (and even 120 format)flat enough to get a decent scan on a flatbed. Look into the Minolta or Nikon scanners - they are much better for your purpose.

[thomas_sullivan]

get a dedicated film scanner for 35mm format. for 120 (medium format) and up, a good 3200dpi or higher flatbed will do a pretty decent job. So, if you are doing 35mm and Medium Format, the best option (meaning price to quality) is to use a dedicated film scanner for 35mm and the flatbed for the Medium Format.

 

The difference is more than just dpi and Dmax.

 

A flatbed, with a transparancy adapter, transmits a diffused light source through the film, then a thick glass of the flatbed and then the sensor. This results in a slightly blurred image with loss in the fine details. The more you enlarge this slightly blurred image, the more noticeable it is. So, a small negative to a large eventual print will look blurred, even if ever so slightly.

 

A dedicated scanner uses a linear light source, thru the film directly to an imaging lens (no glass between them)to the sensor. There is no dispersing of the light AFTER it passes thru the film. Thus, the resulting image is as crisp and sharp as the original piece of film is.

[Ed_Ingold]

8x10 inch enlargements from 35mm would tend to look significantly softer from a flatbed than from a film scanner. Perhaps it is not too soft to be useable, or even "good", but that's a matter of judgment.

 

In my experience, using an Epson 2450, a good scan from a 6x6 image is not quite as sharp as a good 35mm scan using a Nikon 4000 (or 8000). It hardly seems worth the effort and expense of using medium format unless you get results two or three times as good. It is true, however, that you can always get a better scan from medium format in the future, while 35mm is what it is.

[graybrick]

<i>"1200-1500 PPI..."<BR><BR></i>

 

Bollocks. Utter ignorant shite. Don't listen to a word that guy has to say, because he's obviously speaking frome heresay and conjecture, and has no real experience using a good flatbed scanner. Had he actually <i>tried</i> a few and done some comparison, he wouldn't be trying to compete in this DPI pissing contest with numbers which are obviously a load of made-up trash. Probably 'read it' or 'heard it' somewhere, and now trying to spread it on. Maybe he's trying to justify the chunk of cheese he dropped on a dedicated film scanner? Well, for both of you, I'm going to say it one last time. It's a load of hooey. Hogwash. If you don't beleive me, get out there and actually put your hands on some scanners and compare it for yourself, Look at nikon's super-pricey dedicated film scanner and look at Epson's bargain bin 4800DPI scanners with <i>the same film</i> and judge for yourself if the extra $1800 is worth the insignificant boost in resolution. I think you'll tend to agree with me after you've actually <i>done it for yourself</i>.<br><br>

 

I have an Epson 4490 flatbed with film trays that has an optical resolution of 4800DPI. Granted, as has been stated, it is difficult to get the film perfectly flat in the trays, and this will cause some deviation from the published resolution. The dedicated film scanners handle the film automatically, which makes it a whole lot easier to deal with and, in the end, gives overall more consistent results for an inexperienced user. However, the late Epson scanners have a <i>physical sensor resolution</I> of 4800DPI, and the newest ones (haven't checked in a while) may (or will)be even higher than that. That's sensor density. <i>NOT</i> interpolation. I've made 20x30 b/w prints from 35mm Ilford HP5 carefully scanned at 4800DPI in 16-bit greyscale on the Epson 4490 and I would beg anyone who thinks you can't get excellence from a flatbed scanning tiny film to have some impartial observers compare my full-sized prints to their most beloved drum-scanned prints in a blind side-by-side, any time, any day of the week (print quality, not artistic merit, of course). Bottom line, the prints I've enlarged from this scanner look every bit as good as darkroom prints (some maybe better thanks to digital post-processing).

<br><BR>

Be sure that the resolution of your scanner is optical resolution, as often companies will advertise 'maximum' resolution which is an interpolated figure. Many scanners are also capable of scanning more lines down an image than across, which muddles things a bit too (mine can do 9600LPI in one direction, but the sensor is limited physically to 4800DPI). Anyway, don't beleive everything you hear from the 'film scanner' devotees. Yes, they work well, and they do have some advantages in film handling, speed, and ease of use. However, for half the price or often even much much less, you <i>can</i> get an <i>excellent</i> flatbed scanner that will give you very satisfying and useful results. I know. I have one.<BR><BR>

 

Good luck.<BR><BR>

 

GM

[robert_martin5]

I have owned both a 3200 dpi flatbed scanner and now a Nikon 9000 (4000dpi). There is not contest between the flatbed and Nikon - Nikon 9000 wins.

[jtk]

Everybody's right, more or less :-)

 

If you're not determined to get the ultimate best results you may be happy with that flatbed, maybe approaching 11X17 with certain images, assuming you use a focusing film carrier. Wet mounting alone will not help focus.

 

The "calculations" that are taught are virtually always irrelevant, including especially "300dpi", which has nothing to do with the dominant printers (ie. Epson, which defaults to 720dpi and excels at 1440dpi).

 

"Resampling" is not a bad thing, as anyone knows who's viewed a lot of very big inkjet prints from 35mm.

[van_camper]

"I've made 20x30 b/w prints from 35mm Ilford HP5 carefully scanned at 4800DPI in 16-bit greyscale on the Epson 4490 and I would beg anyone who thinks you can't get excellence from a flatbed scanning tiny film to have some impartial observers compare my full-sized prints to their most beloved drum-scanned prints in a blind side-by-side, any time, any day of the week (print quality, not artistic merit, of course). Bottom line, the prints I've enlarged from this scanner look every bit as good as darkroom prints (some maybe better thanks to digital post-processing)." GRAYSON

 

Gray, your lack of experience shows. Epsons are good to 4-5x, and others are more strict with 3x. I own one too, and many of us own high end flatbeds (worth $15-50k), as well as drum scanners. You need to go to galleries and see what a quality print looks like. Also $2k for a Nikon 9000 sounds better to me then $5-10k for a dslr, and you get to keep your old camera and lenses....definately not super expensive like you think compared to the other choices availabe. Check out some other opinions at this site. I suggest you have drum scan done by West Coast Imaging or Nancy Scans and compare to it to your work. That'll change your mind real fast. It is ovbious you have not compared the two, just read bad information. For anyone buying, do the same, ask to try out the scanner using your film while in store, then compare it to a drum scan. You decide, your standards are what really counts. If it is good enough, buy it! Obivously it was good enough for Gray.

[graybrick]

I assure you I've done my homework, and while there is certainly a quality difference, with the proper use and some experience the difference between the flatbed and the dedicated scanner is minimal, if even discernable. I've seen gallery prints. I've seen the dedicated film scanner output. I've seen the output from a $12k flatbed (which was superior, I may add, to many of the drum scans I've seen at full size). You're going to have to print some massive images to really pick out the quality difference between the $200 Epson flatbed and the $2000 Nikon. This assumes, of course, that you have the patience to learn to use your equipment, as the flatbed is slower and a bit more work to get 'just right.' If you're making 'gallery prints' or planning to sell a couple grand worth of scanned prints in the next few months, maybe it's worth the extra money to you, but for the non-professional spending the extra money just doesn't, IMO, make a bit of sense. The perceived quality difference just isn't there. As for the $5-10K for a DSLR, well, there's really no comparisen there either... and you can get a very good DSLR for under $2k that will likely support your lenses if your film setup is decent... however, we're comparing scanners, not cameras, and the truth of the matter is unless you have a special need for ultra-huge ultra high quality prints, the extra $1800 for the dedicated Nikon film scanner is just not going to buy you much, if any difference in final quality. (pissing contest over, Gray crawling back in his happy little hole)

[jtk]

Grayson, if your Epson scans please you you may be thrilled at the improvement with a focusing carrier. Personally, I like to be able to see the finest grain or dye cloud, and deal with it using Vuescan or CS2 if I want...if your flatbed gives you reasonably sharp grain, like a superb enlarger would, congratulations, you're home free. If it doesn't,know that 5400 and V and 5000 et al do achieve sharp grain, and Epson printers will print it if you want.

[van_camper]

As for your pissing contest, your facing the wind, keep learning, eventually you'll realize there is a difference! How old are you anyways? Your not going to convince the rest of us that a 30" prints from a 35mm negative scanned with the flatbed will be barely discernible from one that was drum scanned.

 

"However, the late Epson scanners have a physical sensor resolution of 4800DPI, and the newest ones (haven't checked in a while) may (or will)be even higher than that. That's sensor density. NOT interpolation" GRAYSON

 

Everyone will tell you the optical resolution (what you call physical sensor resolution) is in the range of 1800-2400ppi for the Epson. Scanning the 35mm negative at say 2400ppi, that gives you around 3000ppi lengthwise. If you want to print at 300ppi, that gives you a 10 inch print. If you want to print at 100ppi, that leaves you with a 30" print. Sorry, 100ppi is very low resolution, not what I call a quality image. The Epsons don't have enough horsepower, not even the V750. What do you expect for a couple hundred bucks?

 

Goood luck, and check the weather forecast, it might be windy.

[jtk]

Note: you can't print at 300"ppi" or 100"ppi", but you can print at 300dpi (D, not P) if for some reason you want to get less detail than modern inkjet printers can deliver at their defalt 720dpi.

 

And, incidentally, interpolation doesn't lose visual data, it essentially connects the dots...something like using a developer containing sodium sulfite.

[van_camper]

Your right John, I screwed up. What my mind is thinking and what I should have wrote makes are two different things. I meant 300ppi output resolution @ 1440ppi printer resolution. I was in my mind thinking in pixels to the print, got a bit lazy.

[graybrick]

Van-

 

Who is everyone, anyway? I mean, those that will tell you that a 4800DPI sensor only gives you 1800DPI? I'd love to see that. I'd also like to see your tests that prove it. Yes, there is a small difference in quality, but I have seen it and for $1800 you're gonna need a pretty serious job to make up the cost. The images I'm scanning are at 4800DPI. The actual resolving power, due to glass and other factors, is somewhat less accurate than that, but they're very very close in quality at full resolution to the $2000 Nikon dedicated scanner. I'm not speculating. I'm telling you what I have seen firsthand. At 8x10 there's essentially no differnce. At 12x18 there's essentially no difference. At 20x30, there's a slight difference, but you have to look pretty hard to see it and the average viewer won't ever notice. So it all comes down to your personal needs. No doubt you'll get a slightly better result from using a much more expensive product, but again (if you had trouble understanding me the first time) for the average user, with some patience and care, the Epson flatbed scans are excellent. From your reaction, you'd think I was pissing on you (my legs are dry). So, feel free to send on the tests that 'everyone' has performed refuting Epson's calim of a 4800DPI sensor... I'll be waiting.

[van_camper]

John Kelly, your right. My mind was thinking one thing, the hands another. I meant output resolution of 300ppi @ 1440 printer resolution.

 

"Interpolation doesn't lose visual data, it essentially connects the dots". No it doesn't lose data, but it also does not add real new data either to improve sharpness.

[van_camper]

"Who is everyone, anyway? I mean, those that will tell you that a 4800DPI sensor only gives you 1800DPI? I'd love to see that. I'd also like to see your tests that prove it."

 

Do us all a favor, read up at photonet by doing a search under Epson 4990. Also do the same at

 

http://www.largeformatphotography.info/forum/

 

There are a lot of serious dudes there that find it reasonable for 4x5(some will argue even 16x20 is pushing it), and most are happy with 8x10 film, but no one will agree with you that 35mm gives high quality up to a 30 inch print using a Epson 4990. I've got the scanner, while it does a really nice 8x10 print with 6x7 negs, I have about 5-6 times more film area then you do. You want a real flatbed, get a Screen Cezanne, Fuji Lanovia, Creo IQ2smart/smart3/Eversmart series, then your talking 5000ppi and more and then I will believe you. Epson has always been well known not to give more then about 1/2 of their real optical rating with each generation they made (being generous here). There nice machines, but not for big prints like your talking about, and especially not with 35mm. The rest of us will have to pay 40-60 times more for one of those real flatbed scanners (new), unless you go for used. Even then expect to pay at least 10-15 times what your Epson is worth. Of course the rest of us must be throwing our money away... we all like to burn our money after working hard for it.

 

Ask Ted Harris what his opinion is at the largeformat forum. He tested the Epson 4990 in the View Camera magazine. If I remember, he considered 4x5 film good to 11x14 using the Epson 4990 under critical evaluation (that is about 3x).

[graybrick]

Ugh. OK, just a couple points here. First, this is a ridiculous argument, and you're apparently having a hard time understanding my english. The choice of a scanner is based on need and economy. For your purposes, and I presume you work in a printing type industry, a drum scanner might be the obvious choice. I am an amateur photographer. While I'd love to have an ICG, or even a measley Creo, but I'm not in the commercial scan-to-print business, and if you take a look at the original post here you'll surely pick up that the OP isn't either. As for the 'high quality' 30 inch print, well, it seems subjective but I'm looking at one right now. As for the OPs question, 12x18 prints are easily within the range of my cheap Epson scanner from 35mm. You might get a slightly better-to-your-eye nose in the print result from your drum scanner, or your 'real' flatbed, but again, for sixty(?) times the price or more you better, or you are throwing your hard earned cash away.

 

As for your assertion about DPI of the scanners, your link provided nil for actual data. I've heard a lot that flatbeds don't perform to their specs, but I've never actually seen anyone put that assertion to a properly designed scientific test. I'm willing to bet that you haven't either and you're still talking out your ass, quoting 'everybody' as if heresay must be fact. Do us a favor and show us where you tested the 4800DPI output at 1800DPI. Or, for that matter, where anyone else did. In any case, it's irrelevant to this particular topic. In the end, my scanned prints look as good as my darkroom prints, and I've had many comments from others who have seen them to the same effect. Maybe we're all just not as discerning as you, but hey... I'm no snob. You shoot Leica? 'Everybody' will tell you that 35mm doesn't enlarge well past 11x14 anyway, so the 20x30 is moot, wouldn't you say?

[van_camper]

"First, this is a ridiculous argument, and you're apparently having a hard time understanding my english. The choice of a scanner is based on need and economy"

 

What is ridiculous is your comment below, and that you're willing to match your Epson flatbed against a drum scanned print from 35mm and enlarged to 30 inches (a 24x enlargement). The only one talking out Uranus is you. Do a search at photonet or the other link I gave you, then you will realize how far fetched your statement really is.

 

"have some impartial observers compare my full-sized prints to their most beloved drum-scanned prints in a blind side-by-side, any time, any day of the week (print quality, not artistic merit)"

 

The only right thing you've said is the scanner purchase is based on need and economy. Of course we buy what suits our needs (low or high quality) and price range. However, don't expect it to perform like a drum scanner. As for throwing hard earned cash away, many of us are willing to spend heavily towards artistic expression (many amateurs have a full Hassy system) or hobbies. Call it self-actualization. We do it with cars, boats, camera equipment, jewelry, houses, etc. It is not for you to say how they will spend their money. Enough said.

[graybrick]

1. You haven't seen the prints.

 

2. You still have yet to offer any real substantiation to your assertion that you can't get more than 2400DPI from a flatbed with good technique.

 

Enough said.

[van_camper]

"1. You haven't seen the prints.

 

Yes I have, you forget I own one.

2. You still have yet to offer any real substantiation to your assertion that you can't get more than 2400DPI from a flatbed with good technique. "

Try scrolling up in this thread! See Edward Ingolds comment "Expect to get about 1200-1500 ppi resolution?.and if you scan at 2400ppi, your at limit. Also see comment by Les Berkly (at 8x flatbeds lose badly). Also see Robert Martin?.no contest between flatbed and Nikon 9000 (and drums do better yet). Your stubborn attitude is your weakness, or you would not make outrageous comments that you would match anytime your flatbed against a drum scanned print from 35mm film at 24x enlargement (30" print). LOL.

[van_camper]

....and how will you compete with a drum scanner using your Epson flatbed if it does not even come close on Dmax. See below.

 

"He mentioned a DMax for the 4990 of "maybe 3.5" and that is unlikely as we have tested the DMax of that machine at it is ~ 2.25 which is in the same ballpark as that of all the consumer scanners v. a ballpark of ~3.8 - 4+ for most of the high end scanners." Ted Harris

 

http://www.largeformatphotography.info/forum/showthread.php?t=20869&page=4

[graybrick]

OK, you win. Thanks for quoting portions of this thread and giving me a link to someone's subjective assessment as objective, scientifically derived 'facts' in evidence of your assertions. 'Everybody' says so so it <i>must</i> be true. Again, since your understanding is obviously obtunded and your attention very selective (putting this on a new line so you might read it this time...)<BR>Yes, there <i>is</i> a bit of difference in the results from high-end and low-end scanners (genius) and there better be at 60x (or even 10x) the price.<br>However, in anything but massive enlargements, the average viewer (not you, smarty pants) will never notice the difference, as they won't be likely to scrutinize the print as you might. Even then, with good technique, on some prints the difference probably won't be obvious. On an 11x14, 12x18, or similar sized print what I get is excellent. You might not beleive this, but hey, that's ok. Given that most films in 35mm format are limited to <16MP of usable detail anyway, the film itself becomes the limiting factor. Now, good b/w films give much better numbers, and also much better scans for me. My point here has been that linear increments in scan quality come with exponential increases in price, and the low end of the scale is more than adequate for everyday use unless you have a special need.<br>I don't claim any right to tell anyone how to spend or waste their money. Suggesting that anyone <i>needs</i> a $12k scanner, however, to get good results is absurd. A bit like saying you need a Ferrari and a Rolex to get to work on time (and well)... it's just stupendously ignorant. Next you'll tell me that I <i>need </i> a Hasselblad to get 'quality' pictures. Or that my digital will never approach the 'quality' of film. I'm not pissing here, buddy, it's you who's got your shoes all wet. I understand that you feel the need to defend your reasons for unloading masses of cash for the 'real' thing, though, and if it makes you feel good, then by all means...<br>I'm done here. I hope you are too. You've proved your point- your scanner <i>is</i> bigger than mine :o)

[david_henderson]

I didn't like to intervene in the squabble though as a person close to dropping nearly 2000gbp plus accessories on a medium format film scanner I was following the debate with a little interest, as well as reading older similar threads.

 

Whilst it has never crossed my mind that I should use my current Epson 3200 for serious prints, I have two small anecdotes here that folks might find interesting.

 

First I've just started supplying a new (to me) library, focussing on my b&w portfolios. This library want the photographer to supply 40mb+ rgb scans. Mostly they've selected material for which I have prints and that wasn't an issue given that the dmax requirement of my fibre prints isn't all that high. However they did select some material for which I only had negs, which I scanned at 1200 dpi and uprezzesd by a factor of about 100%. I explained to the library what I'd done, indicating that if they were remotely unhappy then I'd go down a different route. Well, the images are on their website now and so far the silence is deafening.

 

The second concerns the fact that sometimes I use little b&w inkjets as part of the brief to the lab that makes my fibre prints. The colour of these inkjets is frankly useless but its the tonality and relativities and crop I'm after. I've had several circumstances where a very experienced printer has been unable to get as much detail onto paper from an underexposed part of a neg as I have managed using scans from the 3200 and a standard gloss inkjet paper. Given that the 3200 is known to be weak on dmax, this kind of surprised me (though not as much as it surprises my printer).

 

No real conclusions to be drawn here. They are anecdotes. But I guess I might also be interested to see some data on how these scanners actually perform.