Noob question... scanner for 4 x 5...

[Ron Marshall]

To settle this question for myself, I had the same neg scanned on my 4990, an Imacon, a pro-flatbed (Creo) and a drumscanner.

There was a significant improvement in resolution of detail comparing the Creo to the 4990. By significant, I mean that in comparison prints at 16x20 a difference was visible, at a normal viewing distance. Let me stress, it was a slight difference, and in many cases not an objectionable difference. But the important thing for me is that there was a visible difference.

For an image that I don't intend to print larger than 16x20 I use the 4990. For anything that will be printed larger I get a drumscan.

That said, I think some images, would print well at 24x30 from a 4990 scan.

Since there is some subjectivity involved, I advise anyone in the market for a scanner to have a couple of test scans made from the same neg, and test prints to judge for themselves.

[Brian K]

It's all subjective in a way. It's what the each photographer or viewer considers "good enough". Some people simply don't have a high degree of critical observation, others do. Some people are satisfied with lesser quality, others are not. If you really want the best quality for your images you have to spend the money, if money is a factor you have to find a scanner that is the best compromise between quality and cost. But understand that it is a compromise. You're not going to be able to compete head to head using $1000-$2000 scanner against a $10,000+ one.

Using larger and larger film formats can help you as the larger the film the smaller the magnification required and therefore the lower the standards and tolerances can be. But once you start making larger prints from the lower quality scans you lose the advantage of having started with a larger negative. Someone shooting MF and scanning with a pro scanner might be able to produce a better 24" or 36" print than some using a LF neg on a prosumer scanner. Why bother then using LF?

One of the things that I have noticed is that when someone upgrades to a professional grade scanner, they tend to rescan all their old work.

[Ted Harris]

Sorry to have been silent so long but I am flat on my back and just starting to feel like posting any thing substantive.

Brian's comments comments are right on point in terms of what is goo enough.

Sandy' last post shows one of the objective differences.

Here's what purely objective studies I have done for scanner tests show:

1) None of the consumer scanners can resolve better than 2400 and that is a bit of a stretch. This is using an AIG T20 test target .... even this doesn't tell the resolution story as the optics in these scanners are mediocre at best. The high end scanners all have measured output, using the same (similar but slightly different substrate on drums) target in a range of 3600 - 8000 depending on the scanner.

2)The DMax, the ability to pull detail out of shadows is severely limited on the consumer scanners. Using a Stouffer step wedge the best any of them do is 2.25. High end scanners range from 3.9 to over 4.2; since this is a log scale that is a huge difference.

All of this is discussed in great detail in other threads which is one reason it hasn't been rehashed here.

[Rob Landry]

I no longer have access to a rez test target but when I tested my Minolta 5400 a few years ago, I got almost 95 lp/mm. BUT, I'll tell you that despite the high resolution that my Minolta is capable of delivering, I just don't see it with real-world images. Do you really think that it's possible to pull a real 5400 ppi of useful image data off a piece of film? As I said before, beyond 3500 ppi (maybe 4000 on a good day), you really don't see a whole lot more image data and that's with 35mm, my best Nikkor glass, tripod, mirror lock-up etc. 4x5 film with all it's "variables" just isn't going to be good for the same absolute resolution.

Now we know that the Epson V series are good for 2200-2400 ppi. In my eyes, this is plenty for 4x5. In testing some real-world 35mm images against my Minolta, at 2700 ppi from the Minolta and 4800 downsampled to 2700 with my Epson, the results are extremely close after careful sharpening. So I'd say that the 2200-2400 ppi for the new Epsons is realistic with careful PP.

So even if the V700 can only deliver half the absolute resolution of my 35mm scanner, it's more than enough for my uses with 4x5.

The thing with the V series scanners is that film holder height is critical, much more so than previous Epsons. Many have found that they can see differences to less than 0.5mm. I haven't tested this far yet but do know that the 0.5mm adjustments of the supplied holder "feet" do make a visible difference.

[Rob Landry]

I think one of the big problems here is that a few seem to be fixated on the raw virgin output of a scanner. I don't know why there's such a fixation on this since it's not realistic. NOBODY is printing images as they are straight off the scanner. It's the same as anybody who's interested in getting the most from their DSLR is not printing their RAW files straight off their cameras. Anyone who reviews cameras knows that sharpening and correct post-processing is paramount to getting the best results. Comparing the raw output from an Epson to that of a dedicated film scanner would be like comparing a JPG from a Nikon DSLR to that of a Canon. Everyone knows that Nikon JPG files straight off the camera suck (save maybe the D3 & D300). Does that make Nikon DSLRs bad. No, just means that you need to do a little more post-processing and shoot RAW to get the most out of the them.

[Ted Harris]

Rob, not the case .... go read all the archives.

[Don Hutton]

I think one of the big problems here is that a few seem to be fixated on the raw virgin output of a scanner. I don't know why there's such a fixation on this since it's not realistic. NOBODY is printing images as they are straight off the scanner. It's the same as anybody who's interested in getting the most from their DSLR is not printing their RAW files straight off their cameras. Anyone who reviews cameras knows that sharpening and correct post-processing is paramount to getting the best results. Comparing the raw output from an Epson to that of a dedicated film scanner would be like comparing a JPG from a Nikon DSLR to that of a Canon. Everyone knows that Nikon JPG files straight off the camera suck (save maybe the D3 & D300). Does that make Nikon DSLRs bad. No, just means that you need to do a little more post-processing and shoot RAW to get the most out of the them.

While you're on your rant, have you thought about comparing a digital P&S with a Nikon D3? They both do what they're designed to do very well, but within reason. Don't expect you P&S to produce output which will make a fabulously detailed 16x24 print. It can make a 16x24 print, sure enough, especially if you spend a day or two working the file over and under in PS, but there's a reason the D3 costs about $4500 more.... Sandy's little crops really do tell the story. Above a certain level, there are very obvious differences in a print - alas, standards of what is or isn't acceptable vary wildly so your expectations (especially based on what you clearly believe on this matter) may be way lower than someone like Sandy's.

I used to use an Epson 4990 for all of my 4x5 and 8x10 scanning. I finally bought a drum scanner and have rescanned most of what I have shot that is worthwhile. The differences in prints made from my 4990 scans and drum scans are visible at just about every level. In smaller prints, they are small nuances like increased texture because of much greater micro contrast; in larger prints the absolute resolution comes into play. Believe it or not, there's a substantial difference in the perception of sharpness when you actually have a sharp scan to start out with compared to when you have to manipulate that perceived sharpness in PS. I actually have compared several of my prints from my "consumer flatbed" 4990 with the same prints from drum scans. Above a 3X or so enlargement, the prints start to look a lot different, and a layman would be able to see the difference upon inspection. Before you break out into yet another rant on how your one-sided experience must be correct, I'd suggest you actually get a couple of good scans made and compare some full sized prints. Most of the folks advocating high end scanners for excellent output above 3x enlargements have actually done the real world comparison.

BTW, the OP was considering shooting LF purely for quality considerations. On that basis, there's little doubt in my mind what the right advice is. Before I owned a drum scanner, I used to shoot quite a bit of roll film in 6x7 or 6x9 with my view camera because I new that for the best quality output, I could do better shooting 6x7 and scanning it with a good dedicated medium format film scanner (I have a Minolta Scan Multi Pro) than I could with shooting 4x5 and scanning it on my Epson 4990 (or 4870 before that, or 2450 before that!)....

[Rob Landry]

While you're on your rant, have you thought about comparing a digital P&S with a Nikon D3? They both do what they're designed to do very well, but within reason.

Oh c'mon seriously now, please tell me you didn't use that argument.

but there's a reason the D3 costs about $4500 more....

There's a reason a D3 sells for 5K and it has nothing to do with its resolution, or lack thereof. Hint, it's a freakishly fast, high-ISO, specialty sports camera.

Sandy's little crops really do tell the story.

Yes, it shows that he's using an older machine with no sharpening and likely hasn't adjusted his holders for optimum height.

I used to use an Epson 4990 for all of my 4x5 and 8x10 scanning. I finally bought a drum scanner and have rescanned most of what I have shot that is worthwhile.

Good for you, but to suggest that others who don't have the desire or ability to own such a "prestigious" piece of equipment can't even make a quality 16 x 20 is misleading. Face it, film in any form is taking a beating and if every time a newbie comes here looking for advice about starting out in LF, we tell him it's useless unless he's willing to fork out big bucks for some esoteric, discontinued, high end scanner, he's likely to tuck tail and run back to his DSLR.

...a layman would be able to see the difference upon inspection.

That's funny, so if any 'ol layman can see these print differences, why have all of them pretty much given up shooting film in any form. As amazing as drum scanned MF and LF film may be, very very few pros still shoot with film anymore.

Before you break out into yet another rant on how your one-sided experience must be correct, I'd suggest you actually get a couple of good scans made and compare some full sized prints. Most of the folks advocating high end scanners for excellent output above 3x enlargements have actually done the real world comparison.

I'm not implying that a high end scanner isn't preferable, all I'm saying is that the 3-4X rule is a little too conservative. Before I bought a scanner, I had to have my 4x5 film scanned by a few local labs; 2 owned drums and the 3rd a Creo flatbed. Well, the 2 that had the drums are no longer in business and the third with the Creo has pretty poor service and primarily scans for marketing brochures and the like; they don't really do much film scanning (at least not in any quality sense). Of the drum scans at 1600-2000 ppi, when compared to the same thing on the Epson (after PP) are not hugely different. So the Epson is sufficient for my needs and I'll be fine printing to 20 x 25, which is as large as I'll ever print.

I could do better shooting 6x7 and scanning it with a good dedicated medium format film scanner (I have a Minolta Scan Multi Pro) than I could with shooting 4x5 and scanning it on my Epson 4990

Here's a reviewer that compared a Minolta Multi Pro, a Canon 1Ds and an Epson V750

http://www.photographical.net/epson_v750.html

(of course someone's gonna pipe up and tell us this one's bogus too)

[Don Hutton]

Rob

Really glad you're happy with your Epson and you like your prints. Clearly any experience which doesn't directly correlate with yours is of no use to any other person reading this forum, so I'll just cut it short....

The fact that you can't see the difference does not mean that everyone else's standards may be quite as undemanding as yours clearly are. The OP was concerned about getting "very high quality" (exhibition purposes) 24x36s (that's a chunk bigger than what you "ever print"). The answer, regardless of how "high quality" you believe your prints to be, is simply no.

[Brian K]

Take a shot with a 35mm film camera, then shoot exactly the same scene, same film type, with a 4x5 film camera, and then make a 4x5" print from both. Results, you will see very little qualitative difference between the 4x5" sized prints done with 35mm and 4x5" film. Then take those same negatives and make 16x20" prints from both, you will now see the significant difference between 35mm and 4x5" film. That is one of the differences between a prosumer scanner and a real pro scanner.

It is preposterous for anyone to say that a $500- $2000 scanner can even come close to the quality produced by a drum scanner or professional flat bed scanner. I certainly would not have paid $20K for a new IQSmart3, if there wasn't this huge difference. I've been scanning film since 1991, I've owned many scanners that were considered the "best" of the prosumer models and none of them came even close to what a drum scanner could do.

My advice is if you are going to scan LF format (8x10), or even better ULF you can get away with 3-5x of enlargement. If you're planning to go bigger than that you will see better results with a pro scanner.

Also sharpening, interpolating and all the other tricks do not add image detail. They add the illusion of image detail, and a lot of artifacts along the way. If your scanner has poor optics, non critical focus, misalignment of any sort, stepping motors with loose tolerances you're simply going to lose the detail on the film. To correct all those mechanical inadequacies costs money. There's just no way around it.