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

[tashley]

Thank you to all who have kindly taken the time to reply, often in such detail. I am extremely grateful.

It would appear that there is no clear-cut answer. As is so often the case in recherché issues such as this.

A number of possibilities present themselves:

1) that consumer flatbeds have recently come up to scratch but that some very experienced users have not been able to test everything all the time and are therefore a picosecond behind the times

2) that those who believe the consumer flatbeds are up to the task I set them are in some way less fussy or discerning than those that disagree with them

3) that there are so many angels on the pinhead that it's very hard to count them accurately

I will purchase (and I suppose more or less at random!) a 750 or M1/F1 and use it for proofing scans - but of course I'll also try to print to my target size from the files, having played around with various ways of slicing and dicing post processing. And then I'll know what the truth is for me... with the niggling doubt that I'm doing something wrong/should have made a different purchase/have dodgy eyesight etc etc!

Kirk, I had a look through your site. Gorgeous, amazing work. I know Santa Fe really well, once spent a month there but unfortunately only took shots with a Canon G4. However, some of them are printed quite large and look fairly good.. to me...

:-)

Tim

[sanking]

People who hold that consumer flatbeds are only good for 3X4X base their opinion primarily on the difference in resolving power, or ability to capture fine detail, of these scanners compared to drum scanners and professional flatbeds. Dynamic range is also much greater in drum scanners and high-end flatbeds, but that usually only comes into play with transparencies. If detail is not an important consideration, as it is not in many photographs, and/or the print is be viewed from a distance, then there is no practical limit to the size image you can make. For years I had a 30X50" B&W print that was made from a 35mm TRI-X negative hanging in my house. From a distance of 10-15 feet it looked great, but when you walked up to it and viewed the print from ten inches or so, it looked terrible.

I have an Epson 4990 and a Scitex/Creo EverSmart. If you compare the two with scans of a high resolution target the EverSmart beats the 4990 by a lot in terms of pure resolution. A scan on the EverSmart at 3175 spi gives an effective resolution of about 60 lines/mm. The Epson 4990, scanning at 4800 spi, gives only about 35 lines/mm. That is a huge difference, and however I chose to sharpen or rez up files from these scanners the ones from the EverSmart will have an advantage. And you cannot add real detail in sharpening or rezing up the file.

Here is a test you might consider. A large print from a scan should be as sharp as one of the same size made with an enlarger. Have one made with a scan from an Epson flatbed at the size desired, and one made with an enlarger and compare the two prints. If the one made with the enlarger is superior in terms of detail, at the distance you want the print to be viewed, then the consumer flatbed is inadequate for the job and you need a drum scan or one with a professional flatbed. If the print made with a scan from the consumer flatbed scan looks just as good, or better, then you have the answer to you question.

Sandy King

[Kirk Gittings]

Tashley, Thanks for the kind words.

Rob, What I am talking about is real life experience with all those scanners I mentioned making prints for all the group/solo shows and print sales I have had over the last few years since doing digital printing. With years of experience and thousands of scans on the consumer scanners (for many years I did all my commercial work on the consumer scanners-to the tune of about 80 scans a week) mentioned I could never get a scan that was acceptable to me from a 4x5 at 16x20 PERIOD! The difference at a 16x20 print with a drum or proflatbed scan is the difference to me of an embarrassment vs. a fine print. It is allot about resolution and not having to oversharpen files for prints but it is also about shadow/highlight detail and minimizing noise. I have done hundreds of side by side comparisons and the differences are tactile in prints. I can't get what I want from a consumer flatbed at 16x20, which is almost identical in resolution to an enlarged silver print. If the consumer scanners meet your needs great.

[Rob Landry]

Kirk, I respect your opinion but at this point, that's all it is. At least I included some sample scans to illustrate my point.

Out of curiosity, I just used an 8X loupe on the cemetery image and I really can't see any additional detail. Even with the loupe, it's pretty hard to read the writing on the headstone. Perhaps a 10-12X loupe would be better but seriously, at this point it's getting pretty silly.

Honestly, I think much beyond 2500-3000 ppi, we're giving 4x5 film more credit than it deserves considering that many (including myself) are using wooden field cameras and off-the-shelf film holders. The real benefit to higher PPI is to allow for overscanning to reduce grain aliasing and noise.

As I showed previously, I reworked the 4990 crops to closely match the Cezanne and if you look closely, neither is resolving any more real data than the other, in fact the Cezanne is showing some obvious artifacting. Of course with the 4990, it is slightly softer but I was working from an already compressed JPG from an older generation scanner that was unlikely adjusted for optimum film holder height. Someone wanna send me the original to scan on my machine?

With 35mm and my Minolta 5400, if I'm honest with myself, I'd have to admit that much beyond 3500 ppi yields very little additional detail, mostly just more grain. It's debatable if this minute extra amount would be visible in print. Now my 5400 can resolve 85-90 lp/mm so has the ability to eek out pretty much anything the film has to offer. I know this because I've tested it with a resolution target. Unfortunately a resolution target is never representative of real world images.

Here's a quote from Ellis Vener's review of the V750 in PP Mag, January 2007

"How good are the wet-mount scans on the Epson Perfection V750-M Pro? Mac Holbert of digital fine art specialists Nash Editions (www.nasheditions.com) has compared scans of medium-format film made with their $51,000 Kodak (nee Creo) EverSmart Supreme II and the V750-M Pro + FMT combination. He says he can't see $50,000-worth of difference between the two scans, especially after making high-end prints from them. My experience making 3,200ppi scans matches Holbert's. The one downside is that we both see a little color fringing with the Epson FMT scan."

So if Nash Editions couldn't find much of a difference, who can? Even though there's differences in the various crops in our very own scanner comparison here, they're not huge, are largely academic. In fact, some of the higher end scans are oversharpened and very grainy, some have visible artifacting and others are plain soft. Again, this illustrates my point that at this stage of the game, the operator makes just as much difference as the hardware. Of course this is a common concept with many things in life from motor racing to sailing to painting, you name it.

[sanking]

As I showed previously, I reworked the 4990 crops to closely match the Cezanne and if you look closely, neither is resolving any more real data than the other, in fact the Cezanne is showing some obvious artifacting. Of course with the 4990, it is slightly softer but I was working from an already compressed JPG from an older generation scanner that was unlikely adjusted for optimum film holder height. Someone wanna send me the original to scan on my machine?

I looked at your comparison of the Cezanne to the re-worked 4990 and the scan from the Cezanne is to my eye vastly superior in terms of resolution. On the files you uploaded I am able to read in the text about 12-13 lines down with the Cezanne scan, and only to 7-8 with the re-worked 4990. That is what I see, and if what I see is real, then the Cezanne clearly gives much higher resolving power.

This kind of comparison is not valid IMO, but even by the files you posted the Cezanne file looks much superior to me. Maybe someone else might weigh in on this as to their evaluation of the two files. Just try, as objectively as possible, to read down as far as you can.

Sandy King

[Ken Lee]

This subject is subjective, a matter of personal taste and judgment. It reminds me of cars and houses.

Size is often over-rated, unless you're trying to sell something, where it's often promoted as a cheap substitute for... quality.

The difference between a $100,000 car, and a $10,000 car may be worth the expense to some - especially to those who can easily afford to pay. On the other hand, it's a challenge to prove that it's worth exactly 10 times the price.

We all know that price often rises steeply - disproportionately - as we approach the best. But there are reasons for this.

I don't much like large photos, even when they are very detailed. For me, 12x15 is plenty big enough. I also don't like huge houses - or huge cars. I'll take a charming cottage by the sea any time, over an imposing mansion.

[Rob Landry]

I looked at your comparison of the Cezanne to the re-worked 4990 and the scan from the Cezanne is to my eye vastly superior in terms of resolution. On the files you uploaded I am able to read in the text about 12-13 lines down with the Cezanne scan, and only to 7-8 with the re-worked 4990. That is what I see, and if what I see is real, then the Cezanne clearly gives much higher resolving power.

Sandy King

I don't know what you're looking at but I can clearly read the entire text in both crops. You clicked on the cops to enlarge them right? The only difference is that the Epson is slightly softer and the Cezanne is showing artifacts. I've examined them on 2 CRTs and an LCD.

[Rob Landry]

Rob, your comparing the Cezanne and Epson on your monitor at 72ppi. A $75 scanner looks good on the web at this resolution (except for differences in Dmax). Get real scans from the Cezanne and make PRINTS from the two. I know what the answer is. Don't forget, we're all trying to save a buck, we all have considered the Epson thing, we all own one, and all use it for editing through our images. But in the end were willing to buy a pro drum scanner/flatbed or pay $100+ on pro scans for our bigger prints. Ted owns both the Cezanne, Creo IQ3, Epson 4990 (teaches scanning) and has found 3x for enlarging to be about right. He is looking at real scans, real prints, your opinions are from viewing a monitor. Have you ever had a drum scan done, seen prints from one?

Nice try, but if there aren't glaringly obvious differences on screen, thery're not going to magically appear in print. And BTW, the 72dpi/ppi rule for monitors is a myth; your monitor will display whatever you tell it to display. If you display an image with pixel dimensions of 3000x2000 on a monitor set for 1280x1024, it will display it. You will have to zoom in to see the pixel level detail, but you will see it. The video system has no concept of dpi whatsoever, it has no meaning. It simply does not matter to the screen what the image dpi value says. Digital images are dimensioned in pixels and the only one thing that matters at all is the size of the image, in pixels. It's only when sending a file to an output device do we need to concern ourselves with PPI or DPI. In fact, the crops we're referencing, when sent to a printer at 300 dpi would print to a measly size of 1.5 x 1 inch in size. Do you really think that you'd be able to see a huge difference between them at this size?

Let me clarify my position. I'm not saying that a "consumer flatbed" is better than a multi-thousand dollar machine, all I'm trying to debunk is the notion that they're not even good enough for a 16x20 from 4 x 5. Now, if these so called high-end scanners are the pancea they're made out to be, why are most of the manufacturers out of business? Of the ones still remaining, why do they sell so few units per year? I'm pretty sure Epson and Microtek sell a ton of their scanners and not all are sold to "consumers". Many local bureaus use 'em as do the majority of large format shooters, not all of whom are rank amateurs. Perhaps they know there's a difference but not enough of one to make a tangeable difference in their work. In fact last weekend, I was in at my local pro shop and there was a nice 24x30 B&W print hanging in the gallery area and when I remarked how nice it was to the shop owner, he told me it belonged to a local artist and was printed on their newly acquired Epson 9880. When I asked how it was scanned, he told me from a 4x5 neg on their Epson V750. It was sharp, it was very nice and it was a print I'd be proud to call my own.

Perhaps high-end scanners have been killed off by the massive switch to digital. Wow, that worse! Yes it's true, many former medium and large format pros have traded in their gear for FF DLRS. If you ask them, they'll tell you that in an absolute sense, a quality drum scan from MF & LF still holds an edge but they'll also tell you in the next breath that for their work, the DSLRs are good enough. So perhaps, the high-end scanners aren't selling because most realize that dropping 5 figures on a scanner is just not worth the slight increase in quality. Either way you slice it, the so-called high end scanner days have all but come to an end.

[Rob Landry]

This subject is subjective, a matter of personal taste and judgment. It reminds me of cars and houses.

Size is often over-rated, unless you're trying to sell something, where it's often promoted as a cheap substitute for... quality.

I don't much like large photos, even when they are very detailed. For me, 12x15 is plenty big enough. I also don't like huge houses - or huge cars. I'll take a charming cottage by the sea any time, over an imposing mansion.

I agree, I'm not much for huge photos either. 16 x 20 is as large as I've ever printed and it's unlikely that'll change. Oh, an I'll take that cottage by the sea as well. In fact I'm pretty much decided to sell my current house for something smaller and less maintenance.

The difference between a $100,000 car, and a $10,000 car may be worth the expense to some - especially to those who can easily afford to pay. On the other hand, it's a challenge to prove that it's worth exactly 10 times the price.

Yea, what's really funny is the guys who show up at the local track with their Dodge Vipers only to get smoked in the 1/4 mile by some kid with a 12 year old Toyota Supra and a turbo and tranny upgrade.

[sanking]

I don't know what you're looking at but I can clearly read the entire text in both crops. You clicked on the cops to enlarge them right? The only difference is that the Epson is slightly softer and the Cezanne is showing artifacts. I've examined them on 2 CRTs and an LCD.

I am looking at the crops you uploaded, on the monitor of my 24" iMac. Yes, I clicked on the crops. The Epson crop is a lot softer than the Cezanne to my eye. Neither of the two texts are legible on my monitor all the way to the end, but the crop of the Cezanne is legible many more lines down than the crop from the 4990. The Cezanne crop is clearly superior to my eye.

But no matter. This is not a good methodology for comparing these two scanners, IMO.

Sandy King