Is a dedicated film scanner THAT much better than flatbed?

[Nick R.]

Here's an example from a slide scanned on my Epson 4870 flatbed and a Dimage Scan Elite II dedicated film scanner. The spec'd resolution of the Dimage is 11Mpixels and the Epson is spec'd at 46Mpixels. I've only included one slide because I spent a couple of hours scanning with the Dimage at the wrong settings. I'll revisit those later. The full shot and the blowups are unprocessed in any way except to convert to web-sized jpg's. The first two are the flatbed.

 

[T_om]

Tom, "objectively' is a fraudulent concept. No scientist believes there's such a thing as objectivity. You're announcing what is and what isn't, and you're wrong, even about your own intentions.

Soooo, setting aside the semantics and red herrings about objectivity and scientists, do you have an example of a "wrong" statement made by me in this thread? Let's see it.

I am always open to correcting my errors. 😀

Tom

 

[raid]

I have been using a flatbed scanner and a photo scanner, and recently I simply asked the photo finishing deprartment to directly scan my rolls of film to a CD. I can see many more details from the scanned negatives via a photo scanner or directly done by a Fuji Frontier machine compared to scanning a print on a flatbed scanner.

 

[Nick R.]

At the risk of opening up old wounds, here's a couple of more shots from my flatbed/dedicated scanner experiment. One thing I've learned is that scanning and post processing is very boring.
The first two shots are from the Epson 4870 the next two are from a Minolta Dimage Scan Elite II. The blowup is from the black threaded rod in the middle of the frame.

 

[djon]

Tom, faulty semantics are at the root of many deceptions and self-deceptions.

 

[T_om]

Tom, faulty semantics are at the root of many deceptions and self-deceptions.

So I take it you don't have any examples?

Tom

 

[djon]

Nick, your comparisons are interesting...what are the file sizes?

It's well known that flatbed plane of focus isn't necessarily accurate at the film holder..sometimes sharpness increases if the emulsion side is pressed directly to the glass.

As well, flatbeds have issues with haze on the glass, external and even internal. While my old Epson 3200 isn't nearly as sharp as my Nikon V, I'd expect better results with that B&W image, even though it's nominally a lot less sharp than your 4870.

The color images are vastly different in density and contrast, are hard to compare, though the flag is meaningful.

Another factor: older Minoltas usually want to sharpen scans by default, rather than leaving it to post processing. I wonder how your results would compare if you turned off all sharpening on the Minolta and if you did sharpen the results of the Epson ?

I also wonder how the two machines would compare if you used Vuescan for both...

 

[Nick R.]

Sorry, I should have added that I used Vuescan for both.

 

[Nick R.]

I should add that I used each machine to the best of its ability. That is, I used the focus command on the Dimage but not on the Epson since it doesn't allow it. I used two passes on each scan. I used auto settings for color balance on the slides. I felt a slide would give the Epson a fighting chance since it holds the negative flatter than the film holder.

I did not manipulate these images in any way except to size them for the web.

 

[doubs43]

Let me first state that I have NO experience with a dedicated film scanner. None, zip, nada. My present scanner is an Epson 4180 that seems to do a reasonable job but I'll also concede that the dedicated scanners are likely much better.

Microtek has just released the i900 flatbed film scanner and initial reports are that it thinks it's a dedicated film scanner. One review I read had untouched maximum size images from the i900 and a KM Elite 5400II (IIRC) and the differences were there but not as great as one might expect. The i900 will do 35mm, MF and 4x5.

I'm not arguing one way or the other but just throwing out what a new model flatbed is apparently capable of doing.

Walker

 

[T_om]

I'm not arguing one way or the other but just throwing out what a new model flatbed is apparently capable of doing.

Walker

Walker, there is no question the flatbed technology is there to produce good output (I have repeatedly said the Linotype Hell-Heidelberg machines are capable of astounding output). Instead, what has been the case is that consumer grade flatbeds have never been able to match dedicated film scanners.

Even the Hell-Heidelberg units cannot match, say, a drum scan.

What we are talking about though, is various levels of implementation.

It would be great if the top "pro machine" technology were suddenly available at Wal-Mart for $75.00 bucks, but that is not the case at present. It may be true someday and I'll be the first to say "Happy Day" indeed.

When the tests are done on this new machine, perhaps it will deliver the goods. I will have to see some pretty extensive testing though. Here's why... most people think the important part of a film scanner is the resolution. That is not the case. Once you get to about 4000dpi, resolution falls off as the driving factor in a pro scan. dMax is a whole 'nother ball game. You can view scan output of two scanners of equal resolution and the one with the greater dMax will be the better scan every time.

Tom

ETA: I just went and read some reviews and commentary on this scanner. The "Happy Day" ain't here yet. 😉

 

[Socke]

A friend of mine is working on her diploma in arts, one part is a book project with pictures taken by her in Argentina and Brazil on 135 ans 120 slides.

She started to scan her 135 slides on my old Canon FS2710 and bought a Canon 9900f flatbed for her 6x4.5 slides. Since the flatbed offers batch scanning possibilities my slide scanner hasn't, she scanned a couple 135 slides on it as well.

This is far from a scientific test, but the flatbed scans from 135 slides didn't look as good to our eyes as those from the dedicated scanner.
To some degree, this could be solved with postprocessing and it was certainly enough for the book she had to make, but there was a visible difference.

BTW, the difference between 6x4.5cm and 24x36mm was WAY bigger!

 

[KoNickon]

Well, this has gotten a little out of hand, but it's worthwhile reading nonetheless. I have been contemplating getting a scanner for a while now, and all the indications are that flatbed scanners can be "pretty darn good" but that dedicated film scanners are still better. Flatbeds work "well enough" for medium format, I gather, presumably because of the lower enlargement factor. But, what options are there for a dedicated film scanner that can handle both 35mm and medium format? Do such machines exist that are reasonably (say, under $800) priced?

 

[peter_n]

Not under $800 AFAIK. 🙁 I understand that the Nikon 9000 ED is a well regarded multi-format film scanner that sells at a very reasonable price for the functionality it offers. The street price is ~$1800...

 

[T_om]

Well, this has gotten a little out of hand, but it's worthwhile reading nonetheless. I have been contemplating getting a scanner for a while now, and all the indications are that flatbed scanners can be "pretty darn good" but that dedicated film scanners are still better. Flatbeds work "well enough" for medium format, I gather, presumably because of the lower enlargement factor. But, what options are there for a dedicated film scanner that can handle both 35mm and medium format? Do such machines exist that are reasonably (say, under $800) priced?

Used Nikon LS8000 scanners are available at right around the $1k mark. Some as low as your $800.00 figure but you have to wait for a 'deal'.

They are capable of professional quality output.

Tom

 

[RObert Budding]

Wrong!

Wrong!

Tom, "objectively' is a fraudulent concept. No scientist believes there's such a thing as objectivity. You're announcing what is and what isn't, and you're wrong, even about your own intentions.

Well, I was educated as a Chemist (practicing Finance now), and I believe there is such a thing as objectivity. Your hypothesis has been disproved.

Robert

 

[djon]

Robert, "belief' is not scientific. It's more appropriate for finance than chemistry.

The act of observation itself destroys objectivity. My hypothesis has not been disproved by the belief of a formerly technical person.

 

[djon]

Socke, good observations.

The difference between 645 and 35 is real in a flatbed, but 6X7 and 6X9 can make 645 look almost the same as 35. I don't think 645 has any future at all, if only due to the excellence of current film scanners.

I think your friend would find a dedicated film scanner would nearly eliminate the difference between good 35 and 645, at least in common inkjet sizes like 12X18"...35mm is impacted geometrically Vs larger sizes by flatbeds, seemingly not so much by film scanners.

 

[x-ray]

T_OM is correct about flatbed vs dedicated. There are varying needs for different levels of quality and different people have different perceptions of what quality is. A cheap dedicated film machine might not be asgood as a mid level flatbed. Flatbeds vary greatly in quality and so do dedicated film machines. I just sold a epson 1680 tht's at the top of epsons quality and I also just sold my Imacon. The imacon stomped the pants off the epson but that's to be expected. I had a 4990 for a very very short time and wasn't impressed with performance or build. I now have a fuji finescan 5000 flatbed that is slightly better than even the imacon. I've had thousands of images from 35mm to 11x14 chromes scanned on top of the line drums and know what they can do in a real pros hands. The imacon comes extremely close to drum scans for most work and the finescan is even closet to a top end drum. The catch is there are varying levels of drums that won't equal the imacon or fuji and there are drums that exceed them. The person operating the machine has alot to do with it too. A great drum scanner run by a poor technician will put out poor scans. The short answer is you have to spend big money to get the quality that approaches the high end drums.

 

[RObert Budding]

Robert, "belief' is not scientific. It's more appropriate for finance than chemistry.

The act of observation itself destroys objectivity. My hypothesis has not been disproved by the belief of a formerly technical person.

So, you criticize me for being unscientific? Based on what - your freshman philosophy class? You clearly know nothing about science, objectivity, or hypothesis testing.

As for observation destroying objectivity - a gross misapplication of the Heisenberg Uncertaint Principle.

I think you simply like to argue. And to offend.

Robert