coolscan 9000 vs Creo iQ smart vs Imacon x5

<p>Hi,<br>

I just received yesterday my scans of a 6x6 diapo from a professional lab.<br>

They scanned the same original slide 6x6 with a coolscan 9000, a creo iqsmart3, and an imacon x5.<br>

The scans are : a) coolscan 9000 @2000 ppi ; b) coolscan 9000 @4000 ppi ; c) Imacon x5 @ 2700 ppi ; d) iqsmart3 @ 2700 ppi.<br>

My goal is to produce a print of 60x60 cm at 245 dpi, contone printer.<br>

Looking at the b,c,d scan files on photoshop at 100% magnification, and (a) at about 150 % to obtain the same view area, I can see very few difference in therms of definition between them. The coolscan @ 2000 ppi ( case "a" ) results a little pixeled. The b,c,d cases result almost identical in therms of definition, with the little advantage of the creo ( case "d"). The overall image best rendering is case d) , followed very closely by case b). I don't like the rendering of imacon ( case "c")<br>

The prices of the scans : a) = € 2,40 . b) = € 4,80 d) = 5,50<br>

do you think is worth to spend € 5,50 for a scan that produce a 60x60 print at 254dpi, than € 2,40 ?<br>

I mean : on the screen , the difference between coolscan 9000 at 2000ppi magnified on ps at 150% , and coolscan 9000 at 4000ppi magnified on ps at 100% is not much evident. The further improvement from coolscan at 4000 dpi to iqsmart3 at 2700ppi is quiet subtle. This appends on screen. Do you think on the print ( 60x60 cm at 254 dpi contone ) there would be a more evident difference ?<br>

Thanks, ciao.<br>

Marco</p>

 

Since the Coolscan is capable for 4000dpi why did you have it scanned at 2000dpi? Is it better to start off with more data.

Hopefully they also scanned to Pro Photo RGB at 16 bits per channel and saved it as a TIFF.

<p>Larger scans often push the resolution of the source image and thus a smaller scan can look sharper and when up sampled, not be much different in appearance to the larger scan except where some pixelization might appear--the larger scan is actually a better starting point. My own experience is that if you can get at least 100 dpi in a scan to the size of the print you want to make, that you will not have pixelization appear even if you upsample to the desired size (my experience is going from a minimum of 100 to 300dpi. Generally, I also find that the print will "pull" together a bit better than what we often see on screen at print size (often, not always).</p>

<p>When you compare the small difference between the cost of the scans to the cost of such a large print, I would suggest getting the largest, best quality scan for images that you plan to print so large unless the quality isn't an issue (wont be seen up close). There have been 20 foot wide billboards done with 4 MP cameras and no one could see the pixelization.</p>

<p>By the way, I' ve got a scan of a 35mm slide from a coolscan 5000 at 2400ppi and a scan of the same slide by a coolscan 9000 at 2000 ppi. At 100% magnification on ps, the scan by coolscan 5000 at 2400ppi looks clearly sharper than the one by coolscan 9000 at 2000ppi. Is it strange ? Do really 400 ppi more do the difference ? Isn't coolscan 9000 supposed to be more capable than coolscan 5000 on 35mm format too ?</p>

<p>If you look at the manufacturers' data sheets, a typical slide film like Fuji Provia 100F will show that the MTF response is down to 50% at about 40 lp/mm and falling fast past that (for subjects of some normal pictorial-range contrast). Well 40 lp/mm corresponds to 2032 ppi. Now what do you suppose is the MTF response of even a very good medium format lens at 40 lp/mm? It isn't 100% or 80%, and is probably considerably lower. Throw in some effect for the limitations of focusing accuracy, depth of field, and slight camera movement (even related to mirror slap or shutter vibration). The simple reality is that a very good scan at 2700 ppi is going to capture the substantial majority of the real detail (as opposed to film grain-related pseudo-detail) in the vast majority of your slides.</p>

<p>The highest-resolution color negative films have manufacturer-rated MTF response curves falling to 50% at about 70 lp/mm (and T-Max 100 will go even higher), the best lenses that only have to cover 35mm can theoretically do a little better than medium format lenses, and at least with subjects that are fairly flat and parallel to the plane of focus, there are some circumstances where a good 4000 ppi scan makes sense. But 4000 ppi corresponds to a <em>system</em> performance of 79 lp/mm, which only the very best combination of technique, lens, and film can capture. As you go to higher and higher scan resolutions, you rapidly get into diminishing returns, because there is less and less real detail on the film for the scanner to capture.</p>

<p>Dave, very interestig report, thanks.<br>

But I'm talking about a comparison of different scans of the same slide. The comparison is between scanners, not lenses, technique, films or whatsover. Just the performans of different scanners and scanning options.</p>

<p>I would add that you have compared only CCD scanners. Why not add a drum scan to your mix?<br>

And before Dave jumps all over me - there are people on the large format forum who talk about MTF and those that don't. We've just had another conversation where the "theroeticals" quoted all kinds of numbers and formulas and those of us that print all day with this stuff argued. The theoreticals were convinced that only so much was available and the photographers who use it more often were not. It's an ongoing argument - or maybe one could say never-ending. Especially when it gets to things like diffraction, which exists, but is so small a factor that everyone agrees it should be mostly ignored.</p>

<p>You'll never know until you try it. Might be best to have a few smaller trial proofs made from each and see how it goes. We, and you, would just be guessing if you don't make some proofs and see w/ your own eyes. What is on your screen is not the same as what is on the print.</p>

<p>Steve, I'll follow your suggestion. I'm goin to tray to print the 3 different scans of the same slide, 60x60 cm at 254 dpi, contone printer.<br>

I didn't make scan by drum scanner because it would be so expansive, at least here in Italy. I guess about € 80 for a single 6x6 slide drum scanned at 200 MB.</p>

<p>As drum scanner, I mean Tango/ Heidelberg</p>

<p>Marco, my point was essentially that the main thing affecting the detail captured in scans of your slides is probably not going to be which high-end scanner is used to scan them--it will be the limitations of their taking (and/or the skill and care of the scanner operator). That's not to say the particular scanner model makes no difference, only that, in terms of detail / resolution, among the models you listed, I would expect the differences to be fairly small. But of course when I want the best quality, I do get drum scans, for the last bit of quality that a good scanner operator can squeeze out of the film.</p>

<p>Steve's suggestion of smaller proofs is a good one. Get, say, A4-size <em>crops</em> of the 60x60cm frame printed from each scan and see how they look. Better yet, have a friend label the backs of the prints and look at them without knowing which is which, to reduce the effect of your personal biases.</p>

<p>Lenny, I agree that a well-run drum scanner is likely to provide the best results. I agree that diffraction is rarely a major factor. But Marco seems at least somewhat interested in the prices of the different scans, and where I've gotten drum scans, the cheapest ones start at $25 US (about € 33) each. Is getting a drum scan instead of an Imacon scan going to give Marco that much better a print, especially considering the cost? Maybe not.</p>

 

<p>I must have been sleepier than I thought--$25 US is about €19. Multiplied when I should have divided.</p>