New epson scanners on the way

The trick is ignoring the dpi! Saying a scanner is good because it has high resolution is like saying a speaker will sound great just because it has a high power rating, it just doesn't work like that.
But bigger speakers ARE better, right?
 
  • Like
Reactions: PMN
But bigger speakers ARE better, right?

Well, sometimes. This was my 'office radio' the other day - all 3kW of it... :D

iPRMP0O.jpg
 
Last edited:
The trick is ignoring the dpi! Saying a scanner is good because it has high resolution is like saying a speaker will sound great just because it has a high power rating, it just doesn't work like that.

Well yes I see your point, but in my hi'fi days a system was classed excellent because it faithfully reproduced what was on the record. Same explanation as my post #9, so it looks like the more expensive scanners try to achieve this (ref #9) or maybe they don't and a use bit of fiddling in the software (like digital cameras. :eek:....I remember Japanese hi-fi systems in the 60's h'mm great sound but poor quality and this hit the sales of excellent British hi-fi systems...cheap and nasty but Joe public didn't mind.
 
Well yes I see your point, but in my hi'fi days a system was classed excellent because it faithfully reproduced what was on the record. Same explanation as my post #9, so it looks like the more expensive scanners try to achieve this (ref #9) or maybe they don't and a use bit of fiddling in the software (like digital cameras. :eek:....I remember Japanese hi-fi systems in the 60's h'mm great sound but poor quality and this hit the sales of excellent British hi-fi systems...cheap and nasty but Joe public didn't mind.

Faithful reproduction is obviously the aim but forget that, we're not talking about that. I said, specifically, assuming higher resolution scans give you more detail is like assuming a speaker is better because it has a higher power rating. I didn't say anything about what the purpose of hi-fi should be, I was talking purely about the kind of wrong assumptions people make.

Ah balls to it, I give up. There's only so many tangents I can try follow in a 24 hour period before my brain hits it's limit and tells me to drink.
 
But bigger speakers ARE better, right?

Can be to get lower base and reproduce it more faithfully which a small speaker attempts but it ain't the same......my Celestion monitor speakers can shake the house with 15" base units.:eek:
 
Yes, because it's a perfectly valid analogy for the point I was making. You do realise when you scan a frame of film it becomes digital, right?

Jeez, why is this such hard work?

For one, yes, it is a valid analogy. It's the lens that's resolving the source, and having an inferior lens and bigger film doesn't lead to more detail than a superior lens and smaller film, just as scanners with different quality optical paths can resolve different amounts of detail independent of the claimed resolution. For two, I was agreeing with you. For three, I put a winky smiley to denote my comment wasn't entirely serious. For four, I remember why I was ignoring you. Bye bye.
 
For one, yes, it is a valid analogy. It's the lens that's resolving the source, and having an inferior lens and bigger film doesn't lead to more detail than a superior lens and smaller film, just as scanners with different quality optical paths can resolve different amounts of detail independent of the claimed resolution. For two, I was agreeing with you. For three, I put a winky smiley to denote my comment wasn't entirely serious. For four, I remember why I was ignoring you. Bye bye.

Ooh, good rant. I was going to apologise for the misunderstanding but I won't bother now.
 
Faithful reproduction is obviously the aim but forget that, we're not talking about that. I said, specifically, assuming higher resolution scans give you more detail is like assuming a speaker is better because it has a higher power rating. I didn't say anything about what the purpose of hi-fi should be, I was talking purely about the kind of wrong assumptions people make.

Ah balls to it, I give up. There's only so many tangents I can try follow in a 24 hour period before my brain hits it's limit and tells me to drink.

:D well you set the cat amongst the pigeons by mentioning "image quality" ;) IIRC when seeing tests on scanners they e.g. would compare V750 with a Coolscan and it's all about the detail that the scanners can pick up but they don't mention you get better quality images from one compared to another....well there you go :exit:
 
:D well you set the cat amongst the pigeons by mentioning "image quality" ;) IIRC when seeing tests on scanners they e.g. would compare V750 with a Coolscan and it's all about the detail that the scanners can pick up but they don't mention you get better quality images from one compared to another....well there you go :exit:

That detail is the aspect of quality I've been talking about all along. ;)
 
My heads hurting now, I'm gonna have a stiff glass of milk ;)
 
I find it very annoying that they're allowed to claim such high resolutions when it is such a bunch of fibs.
I know what you mean. The marketing department of companies of make electronic gadgets are responsible. They know that some members of the public like gadgets with big numbers in its description.
 
I know what you mean. The marketing department of companies of make electronic gadgets are responsible. They know that some members of the public like gadgets with big numbers in its description.


But high dpi does work sorta in that although the basic true resolution or scanned detail (or whatever the words should be) is say 1600dpi for a scanner, in scanning at 3200 dpi you get less pixel breakup (because of the software fiddle) and an enlarged print looks better.
 
Mmmm, what an interesting thread....:help:

I quite like the results from my scanner. Problem is the images that it is scanning are pants. :D

Not literally pants, I've just realised why I'm not usually allowed on t'interweb when I've been to the pub.
 
  • Like
Reactions: PMN
I pretty sure scanning in your pants contravenes the Geneva convention.
 
well this is going well :) I still dont beleive it for a second though, something that can pack more dots into that inch will give a better result that something that packs less in. Thats just physics and its backed up by endless comparison reviews on endless scanner sites out there. Lower dpi doesnt have as much detail as higher. Im dev'ing some 5x4 tomorrow and i'll scan some in at 2000 and some at 2400 and see if there is a difference.

If i hadnt just spent 3 and a half hours stuck on a train getting home i'd do some now. But i'll defo do it tomrorow.
 
Last edited:
Im dev'ing some 5x4 tomorrow and i'll scan some in at 2000 and some at 2400 and see if there is a difference.

You need to do it on two different devices, one of which has a superior optics, to test the hypothesis in this thread.
 
well this is going well :) I still dont beleive it for a second though, something that can pack more dots into that inch will give a better result that something that packs less in. Thats just physics and its backed up by endless comparison reviews on endless scanner sites out there. Lower dpi doesnt have as much detail as higher. Im dev'ing some 5x4 tomorrow and i'll scan some in at 2000 and some at 2400 and see if there is a difference.

If i hadnt just spent 3 and a half hours stuck on a train getting home i'd do some now. But i'll defo do it tomrorow.
Consider it in a different way. Take a D800 and mount a lens baby on it. Then set up your slide to be scanned so that it's at 1:1 magnification. It will be a super high "resolution" image of 36mp.

Then take an older generation camera, say a d700 but use a top end macro lens on it. It will be "only " 12mp, but I know which image I'd want to have.

The scanner manufacturers are being really tricky by using the word "resolution" in this context. Strictly speaking, the 2400dpi is absolutely not an optical resolution, but a spatial sampling rate. The optical resolution is tied into the combination of sensor and lens used for scanning and will include all the "micro contrast" and sharpness characteristics that good camera lenses reviews include and that make some lenses fantastic for real world use despite having nominally lower resolution than other lenses in their class.
 
  • Like
Reactions: PMN
You need to do it on two different devices, one of which has a superior optics, to test the hypothesis in this thread.

Easy way for quite a few is to compare an Asda scan with your home scanner, do a crop and only judge on the detail (of course as Asda crop shadows don't choose that part).
 
christ, just judge scanners by their results, that's what matters. The shot of the fountain with the good scanner is beautiful btw :)

odd question, but are scanners bayer sensors?
foveon, and hassellblads sensor shift mode do give you around double the detail for a given resolution
 
christ, just judge scanners by their results, that's what matters. The shot of the fountain with the good scanner is beautiful btw :)

You missed the point :rolleyes: Paul knows about good scanners as he has a drum scanner and a Cezanne and he has said the Fuji Frontier used at Asda is a good scanner (well for a new price of about £20,000 it should be). Well do you think the Fuji engineers who set the machine up would cripple the scanner's ability to pick up detail? Might have been a good idea if everyone was using Kodak Brownies but for at least 30 years even cheap compacts for Joe Public had good lenses.

Some results if the Fuji is set up properly to scan shadows:-
http://www.johnnypatience.com/film-is-not-dead/
 
Last edited:
The real Fuji pro scanners were / are the Lanovia and the Celsis.

This link was posted on Large Format info and gives a little more detail on the scanners

http://www.blur-magazine.com/journal/epson-launches-benchmark-flatbed-film-photo-scanners

The only real changes from the V700/V750 seem to be a switch to an LED light source and better designed holders, adjustable height and anti newton inserts.
A lot of people who have the V750/700 have brought "better" holders of one sort or another already so it is doubtful that it would be worth buying the new scanners.

With a high optical density (4.0 Dmax), the scanners reproduce a wide range of tones with extremely high accuracy, especially in dark areas of originals.

A triumph of marketing over engineering as usual a Dmax of 4.0 isn't hard but is utterly meaningless with out reference to Drange which on average is something like 2.0 for consumer level flatbeds.

They have updated the old designs ... if you have one already no real need to upgrade if you don't well if it is not much more than a V750 you could go for it or you could look at the secondhand market in pro scanners were your money would get you a much better albeit much larger machine.
 
well this is going well :) I still dont beleive it for a second though, something that can pack more dots into that inch will give a better result that something that packs less in. Thats just physics and its backed up by endless comparison reviews on endless scanner sites out there. Lower dpi doesnt have as much detail as higher. Im dev'ing some 5x4 tomorrow and i'll scan some in at 2000 and some at 2400 and see if there is a difference.

If i hadnt just spent 3 and a half hours stuck on a train getting home i'd do some now. But i'll defo do it tomrorow.

You'll obviously get better results on the same scanner.... This whole debate has ben about DPI not meaning everything, and from one scanner to another it's more about the actual quality of the scanner more than the DPI it's capable of.
 
  • Like
Reactions: PMN
odd question, but are scanners bayer sensors?
foveon, and hassellblads sensor shift mode do give you around double the detail for a given resolution

Scanners don't need Bayer masks. You need the Bayer/xtrans/foveon layout in a normal camera because you are capturing the whole image at once. In a flatbed scanner you will have a linear array sensor ( either repeated rgbrgb pixels or, if it's very expensive, three whole lines, each dedicated to a single colour). This is scanned across the new stepwise and explains the two dpi figures that are in the full scanner specs. The spacing of the pixels is not necessarily the same as the minimum step the mirrors can make, hence two different dpi numbers.

Drum scanners are even better as they are single pixel devices that take the light from a single point on the meg and then split the colours up to deliver them to 3 single, very sensitive detectors.
 
Back
Top