PDA

View Full Version : Yet another MF/LF scanner comparison question



Oren Grad
6-Jul-2008, 14:25
I have a very specific question for anyone out there who has experience with both of the alternatives I'm going to raise. There has been plenty of discussion here around this general comparison, but nothing that addresses this particular take on it.

If the objective is an 8x10" color inkjet print that will be subjected to outrageously critical, grain-sniffing (dot-sniffing?) inspection, what would you choose to create the source file for printing? Medium format scanned with a Nikon 9000, or 4x5 scanned with an Epson V750 (or Microtek M1)? Assume that the film will be either Portra 160NC or Portra 400NC, and that the pictures would be taken using an appropriately sized view camera on a tripod and first-rate modern glass optimized for the respective format. Also, assume a relatively sharp printing paper, such as the new Harman glossy FB.

Just to get these out of the way: Yes, I've used both 6x9 and 4x5 view cameras in the field and I'm very familiar with the logistical considerations pro and con in either case. No, I'm not asking about larger prints, just about 8x10 prints that will be subjected to unreasonably critical inspection. Also, "use a 12MP FF DSLR", "get a drum scan", and "make contact prints from 8x10 color neg" are interesting alternatives to discuss some other time, but they're not what I'm after here.

Thanks for any thoughts on this.

EDIT: I should add, if the answer turns out to be, "nobody knows because nobody else has been silly enough to worry about a question like that", that's OK too. ;)

Gordon Moat
6-Jul-2008, 14:52
Just for 8x10 prints, if you do not need the movements (swing/tilt/shift), then whatever camera you feel comfortable using should work. I don't consider that a 6x9 view camera would be that much lighter than a 4x5 view camera, though there are exceptions.

Convenience might dictate using a rollfilm back, which would dictate a medium format scanner. The Nikon 9000 has better colour depth (real not theoretical) than either the Epson or the Microtek, so yet another reason to use rollfilm.

Also, you loose a bit of an edge doing inkjet prints, so some of the resolution might be less than you expected after the scan. This is why I suggest using a scanner with better colour ability, since that will be a more critical and more noticeable aspect of the final printed results.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

Lenny Eiger
6-Jul-2008, 15:09
If the objective is an 8x10" color inkjet print that will be subjected to outrageously critical, grain-sniffing (dot-sniffing?) inspection, what would you choose to create the source file for printing? Medium format scanned with a Nikon 9000, or 4x5 scanned with an Epson V750 (or Microtek M1)? Assume that the film will be either Portra 160NC or Portra 400NC, and that the pictures would be taken using an appropriately sized view camera on a tripod and first-rate modern glass optimized for the respective format. Also, assume a relatively sharp printing paper, such as the new Harman glossy FB.

EDIT: I should add, if the answer turns out to be, "nobody knows because nobody else has been silly enough to worry about a question like that", that's OK too. ;)

I think you're asking some of the wrong questions, but to answer it nonetheless, I would go with the film scanner. 4x5 film is fabulous, but the Epson is not, in my opinion. A film scanner may not be a drum scanner, but it is a lot tighter than the Epson.

Lenny

Oren Grad
6-Jul-2008, 15:45
Thanks, guys, those are very helpful responses. I'm glad I asked - the answer may turn out to be clearer than I expected. It will be interesting to see if that's the consensus.

Gordon, I have two reasons for specifying a view camera for medium format. The first is so I can use my favorite medium format lens, which is the 90mm Apo-Sironar-Digital. The second is that once I put the camera on a tripod, I'm happier having movements available.

Lenny - what questions should I be asking?

Sheldon N
6-Jul-2008, 15:52
I believe that both options would be well within their respective capabilities regarding resolution at an 8x10 print. It is highly doubtful that you could see a true difference in resolution. Since the question is not one of resolution (toss print sniffing out the window) you should look at color/highlight/shadow/tonality differences between the two systems.

I think the Nikon 9000 would do better at pulling highlight and shadow detail from color negative film, so you would probably come out ahead shooting 6x9 with contrasty subject matter. Tonality/smoothness is probably a wash given the larger film size vs. better scanner tradeoff.

Overall, probably a question of splitting hairs. Which gear do you own? I'd use whatever you currently have.

Oren Grad
6-Jul-2008, 16:00
Thanks, Sheldon. Although I mentioned dot-sniffing, dynamic range and tonal character are certainly important parts of the equation for me.

I already own the cameras and lenses I would need to go either way, so the question is purely about which approach would better meet my objective.

Cesar Barreto
6-Jul-2008, 17:15
Oren, I also believe that the Nikon scanner would deliver a better result once you get glass masks, using or not wet-mounting. Sharper images to start with, less color fringes and less artifacts due to better stepping motor (I've tested Epson 4890 and have been using Microtek i900 and Nikon 9000).
But I wouldn't be surprised if you could find some slight graininess with 400 ISO film, wich I've never tried myself.
As about software, I think one can achieve equivalent results on either way.

Frank Petronio
6-Jul-2008, 18:30
I don't think the format or scanner will matter as much as how you process the files and how carefully you sharpen them.

I bet you could make either one look better than the other with proper post-processing and I also bet the raw file quality between the 6x9/9000 and the 4x5/750 are nearly the same.

For that matter, I've got some 8x10 inkjets from 12mp dslrs that don't give anything up compared to the 4x5 film based 8x10 inkjets. Of course the digital needs to be more carefully lit or of a lower contrast scene for that to happen.

It's not a silly question at all. I've been thinking about getting a Nikon 9000 and using medium format again for years.

Oren Grad
6-Jul-2008, 18:55
I don't think the format or scanner will matter as much as how you process the files and how carefully you sharpen them.

I bet you could make either one look better than the other with proper post-processing and I also bet the raw file quality between the 6x9/9000 and the 4x5/750 are nearly the same.

Ah, glad you mentioned it. So here's another criterion: for my taste, the less sharpening required by the file out of the scanner, the better. Under most circumstances, the effects of small-radius sharpening start to bother me long before you can see haloes.

So if the raw file quality out of the scanners is the same, that's cool. But if the idea is that you can fix one to look as good as the other through post-processing, that will be a very hard sell for me.

Don Hutton
6-Jul-2008, 19:13
Oren

I've been scanning MF film with a good MF film scanner (a minolta scan multi pro for many years) and scanning LF film on flatbeds for about 5 years. While I don't have a V750, I do own a 4990 Epson and have done comparisons of MF film on the minolta compared to 4x5 on the Epson. For a final print of 8x10, I'd pick the medium format film scanned on a good film scanner every time. Resolution is not going to be an issue on such small print size, but there will be differences in microcontrast and the dedicated film scanner will win the day.

Bjorn Nilsson
7-Jul-2008, 07:17
I fully agree with the 6x9 and Nikon scanner recommendations. (This said by an Epson 700 owner.) One other thing is the fact that 120 film as such (given the same emulsion as on 4x5) resolves a bit better than 4x5 film. At least this is the opinion of the Sinar factory and their research lab. (OK, I read about this in a brochure about the Sinar roll film backs. But I don't think Sinar would publish, even in their own selling material, something fundamentaly wrong.)
As the size difference would make up for the slight advantage of rollfilm, a properly exposed/processed 4x5 sheet would probably come out "on top", but not by much.

//Björn

Kirk Gittings
7-Jul-2008, 09:02
Having owned the Nikon 8000 (and the 9000 is better) and all the Epson's, I would opt for the Nikon. For all the reasons people above have mentioned. One caveat. I would flatten 120m film in a book overnight or so before scanning it to get it falt unless you either do the glass mount or wet mount holder in the Nikon. I did glass mounting and it did a very good job.

Oren Grad
7-Jul-2008, 09:25
Well, it looks like there is indeed a pretty strong consensus. Advice re glass holder for the 9000 duly noted too. Thanks to everyone for the informative and helpful responses!

cotdt
8-Jul-2008, 21:51
I've seen the results for myself and would go for the Nikon 9000. However, I've not seen the 4x5 under Microtek M1. That scanner is supposed to be better than the Epsons.

ivyrob
9-Jul-2008, 10:18
Oren,
This is what I can tell from my own experience with Nikon 8000 & 9000, Epsons and Microtek 1800f (not M1)

I have used them mostly with BW film, Velvias and Kodak 160 VCs.
Nikons always medium format ( 6x4,5 and 6x12 stitching) and Epsons and Microtek with 6x12, readyloads, and 8x10.

I can see that the nikon is clearly in another league. You get sharper, also cleaner images. Color rendition is also different, more accurate, and as it has been said here, you get more details at shadows, and can render better highlighted zones.

You can get good results from either Epson´s or Microtek´s, with more than enough resolution for a 8 x 10 print, but you will be able to apreciate more subtle details in the print, scanned through the nikons. (I insist in not having tested Microtek´s M1, but I own and use 1800f)

It is true you will get a very good file with both scanners, aplying some post-processing. But, in my opinion, if you spend the same time post-proccessing a Nikon scan and an Epson scan, you will get a better image from a 6x9 through the first, than from 4x5 on the rest. Without doubt. (I used to test 6x12 and 6x17 stripes on the Microtek, and scan and stich them at the Nikon when I wanted to get the best)

It will be worthy trying the wet-accesory holder you can get from aztek, which can be used with "kami"- easy liquid solution on the Nikon scanner. It makes good go even better. (I have also used kami´s solution on the Microtek´s glass plate, for scaning my 4x10 and 8x10, and it also improves)

The next question would be a Hasselblad Scanner against the Nikon. This, for me, has been clear as I can scan 6x12, 6x17, 4x5 and 4x10 on the X5. But the way the Nikon 9000 scans 6x9s, the extra amount of money needed is at least questionable to justify.

Hope to have helped further more...

Ken Lee
9-Jul-2008, 11:26
This all presumes no cropping. Once you start cropping the Medium Format images, you are, in effect, starting to enlarge past 8x10. The question becomes, how far past 8x10 can you go, before a penalty is paid ?

By that, I mean, the fact that you are enlarging your original MF image past 5X.

When I formerly used a Minolta Scan Dual (no longer made unfortunately, but equal to the Nikon I believe) I found that the scanner and film were so good, that the weakness of my taking lenses started to become apparent, to the degree that I went past 8x10.

For the record, I now shoot mostly 5x7 and scan with an Epson 4990. There is no problem getting a decent 11x14 that way, and it's cheaper than buying a fancy scanner. I enlarge no more than 3x, so I get to keep 30-40% of the original resolution of my taking lenses - some of which are... vintage. Even so, there's plenty left on the print.

Some of us pay a price in portability, but we pay the price in the beginning. In the end, there's no need to worry about those "outrageously critical, grain-sniffing (dot-sniffing?)" inspectors. :rolleyes:

Scott Kathe
9-Jul-2008, 14:11
At what print size would there be a difference and which would be better when comparing the 4x5 on an epson 4990 and medium format on the Nikon 9000? I guess I'm asking how would 11x14s and 16x20s compare and I know 16x20s are really pushing the 4x5 on an Epson.

Scott

Kuzano
9-Jul-2008, 15:03
With no goal beyond 8X10, and:
1) If you need movements... A Galvin or similar 6X9 monorail with movements and a roll film back. Find a Graflex 2X3 with the extra pin rollers at each end of the opening in the film gate.
2) At a basic investment of $200 for an Epson V500 scanner, plus the price of one of the betterscanning holders for MF with the AN glass
3) Good sturdy tripod

I think you will be surprised at the potential, given good shots and proper use of all the equipment. I think if you shot a comparison using 4X5, you would not see much improvement over the combination stated above as long as your printed output is so small.

However, I am tending toward budget rather than "deep wallet" on these items, as I am prone to do.

The weakest part of this link, but highly satisfactory to me, is the Epson V500 scanner. I believe the dollar value of this scanner is unmatched. I love mine.

If that turns out to be unsatisfactory, I suspect the Nikon 9000 would more than satisfy any criticism in that area.

Roll film handling, cost per shot and overall budget are strong considerations for me, and I don't believe I am really sacrificing any significant quality at 8X10.

Another alternative would be a light (Like Gowland Pocket or Calumet [gowland pocket] or a Calumet Cadet) 4X5 monorail camera, offering movements, 6X9 with a roll film back, and 4X5, and packability all at very budget prices.

In fact, here is a Galvin 2.25X3.25 with sheet film holders and ground glass and a lens currently on ebay.

http://cgi.ebay.com/2-1-4X-3-1-4-GALVIN-MONORAIL-LENS-ROLL-FILM-BACK_W0QQitemZ280244318941QQcmdZViewItem?hash=item280244318941&_trksid=p3286.m14.l1318

Ben Syverson
9-Jul-2008, 15:06
The Epson should be fine for 4x5 well past 16x20, which is only 4x...

I have a 4990, and find that it has a "real" resolution of around 2400 DPI. So if you scan at 2400, or scan at 4800 and resize to 2400 (preferable) you wind up with an image which is ~12000 pixels in the long dimension.

Printed at 20" across, that gets you a 600 DPI image, which is good enough for even the most discerning eyes.

Even if you're (overly) pessimistic and rate the resolution of the 4990 at 1200, you have a 16x20" image @ 300 DPI, which is still great.

But I get what people are saying about the 8000 and 9000 and MF. It makes the difference between 6x9 and 4x5 pretty marginal. The solution for me is to step up to 8x10. Rating the 4990 at 2400 dpi, an 8x10 will yield a 600 DPI 30x40" print... Or a 60x80" at 300 DPI... Try doing that with 6x9. :)

Kirk Gittings
9-Jul-2008, 15:52
Its also not just a question of resolution, but also D-max, lower noise, better highlight and shadow separation etc.