Negative Scanning Resolutions

[nicholas_riggs]

<p>A 35mm negative has a physical size of ~840mm/sq, and a 6x7 Medium-Format negative has a physical size of ~3920mm/sq. Yet in researching scan-sizes at various labs, I am stumped as to why Medium-Format scans are so proportionately small compared to 35mm scans. The Darkroom will provide "Super" scans of 35mm at 4492x6774 px, but the pixel dimensions of their "Super" scans of a 6x7 Medium-Format negative that has 4.5 times the physical image area as the 35mm is actually SMALLER, at 4814x5902px.</p>

<p>I am coming across similar results at other labs -</p>

<ul>

<li>35mm Res vs. 6x7 Res. - Lab Name</li>

<li>4492x6774 vs. 4814x5902px - The Darkroom</li>

<li>6955x4668 vs. 7175x8793 - IndieFilm Lab</li>

<li>3339x5035 vs. 4815x5902 - North Coast Photo</li>

<li>4492×6774 vs. 3533×4824 - Old School Photo Lab</li>

<li>Etc.</li>

</ul>

<p>Only one of these labs offers Medium-Format scans that are larger than their scans of 35mm. Even a $16,000 Hasselblad Flextight X1 Negative Scanner lists it's optical resolution for 35mm as 6300dpi, and 3200dpi for Medium-Format scans - which negates any resolution-advantage of a larger negative.<br>

<br>

#1 - How is it possible that every lab and scanner commercially available (except one), can/will not scan a MUCH larger negative at a proportionately high resolution as 35mm?<br>

<br>

#2 - Why even shoot Medium-Format if it is impossible to get scans that will do such a large negative justice?</p>

[User_6502147]

<p>There are many places that will do drum scans and MF will really shine. The negative does not have to work as hard as 35mm to get to the same size of an enlargement. I've seen a photo out of Yashi 124G and it totally blew me away....thinking it was a drum scanner.....a 750 Epson was used. So much for surprises.<br>

Much depends how much knowledge/experience the scan operator has. I suggest you do cross comparison (test both formats) and see why the MF results are better. Hmmm, resolution is only part of the story....<br>

Les</p>

<p> </p>

[nicholas_riggs]

Oops, Les you are correct, North Coast

does scan MF at a higher res. I have

contacted Indie Lab, but it's like

midnight there, and I always like to

hear from as many people as possible.

The more experience and opinion I can

draw from, the better.

 

I'm mainly horrified because I was on

the brink of buying a Pentax 6x7 until I

found out about all this. Based solely

off the resolutions I was reading, my

Nikon F and Canon AE-1 will give me

roughly the same bang-for-my-buck

when printing, EPECIALLY when you

factor in film/processing/scanning

costs.

[peterbcarter]

<p>If you want control of your workflow, you either need your own staff or do it yourself.</p>

[nicholas_riggs]

Peter, I'm not sure I understand your response. I already pointed out that a Hasselblad scanner that costs one year's

worth of minimum-wage wages still, by design, scans Medium-Format negatives at %50 of the resolution of a 35mm.

We're talking about scanning negatives here, I cannot build a scanner capable of scanning at 1dpi, let alone 6500. It's not

a matter of being picky, or controlling workflow, it seems to be an industry-wide consensus that Medium-Format negatives

must (or need?) never to be scanned at the same resolution as a 35mm frame. Or maybe you just had one too many

wines and replied to the wrong thread.

[peterbcarter]

<p>My mistake. I thought you were discussing scanning services offered by processing companies.</p>

 

[peterbcarter]

<p>As a programmer, it's probably a hardware limitation. Not optics per se, but scale of data processed internally before it goes through the pipe. The smaller scale scanners (consumer/prosumer) don't seem to have that limitation by format.</p>

[peterbcarter]

<blockquote>

<p>#2 - Why even shoot Medium-Format if it is impossible to get scans that will do such a large negative justice?</p>

</blockquote>

<p>It's about the lens not the film......99% of the time. Generally lens are much better in the larger formats, for a large amount of reasons. Sometimes size is everything. You have to have a really great lens (35mm) to compete with a mediocre one in a larger format.</p>

[thirteenthumbs]

<p>Several years ago when The Darkroom was operating in my area the lab manager said that if I brought a scan in to be printed that she would resize the scan to print resolution of 300ppi for the size print requested if the scan was larger.<br>

So using the lab print resolution of 300ppi your table of 35 vs 6x7 print sizes in inches are:<br>

Darkroom 16x22; 16x20<br>

Indiefilm 15.6x 23; 24x29<br>

North Coast 11.3x 15.7; 16x 19.7<br>

Old School 15x22.6; 11.8x16</p>

<p>They may even speculate that they can upscale for larger or print at 150ppi for larger sizes and get acceptable results.</p>

[glen_h]

<p>I believe it is somewhat optics, and hardware.</p>

<p>A nice small, and relatively high quality, lens will focus a 24mm line on a line CCD sensor.<br>

The scanner then move the slide or negative with a stepper motor, nice and simple.</p>

<p>Flatbed scanners like the V700 have to image a larger line, using a larger lens. Also, most use a moving mirror instead of moving the film, complicating the optical path, making it harder to keep the focus optimal. </p>

<p>In both case, one should ask what the actual optical resolution is. That is, how sharp the image is on the sensor. There is not much point in digitizing at a much higher resolution than the optics allows. <br>

�</p>

[James G. Dainis]

The question is,

 

Why do they scan 35mm film at 6300 ppi but scan medium format film at 3200 ppi? Why can't (or won't) they scan medium format film at 6300 ppi?

[bgelfand]

<p>There is resolution and then there is resolution. <grin></p>

<p>In scanning there is optical resolution of the scanner and then there is an interpolated resolution when the scanner "manufactures" a pixel from the values of adjacent pixels. The higher resolution scans are usually quoting interpolated resolution figures.</p>

<p>Since Medium Format negatives have much more information, scans are usually made at or close to optical resolution (for most scanners it's about 2400 dpi). It gives a truer scan.</p>

[Robin Smith]

<blockquote>

<p>Pentax 6x7 until I found out about all this</p>

</blockquote>

<p>Well I think this is indeed a good reason to be wary of MF film. Unless you have a scanner yourself and can scan at equivalent resolution to 35mm then you are indeed wasting much of the advantage of MF. This is one of the reasons why I gave up MF and went to 35mm digital. You can pretty well achieve scanning parity with 35mm using a lab such as IndieLab, but it is not cheap and it is out of your control, plus there is all the hassle of mailing negs etc. etc.</p>

<p>Personally, I think that for most of us (not all though) you will only really see MF quality when you are printing in your own darkroom (or at least in someone's real darkroom), or, I suppose, you could pay someone to print it for you. To pay for someone else is also a hassle, as their interpretation may not be yours, so there may be cycles of revision involved. This may be not so much of an issue if you live nearby to a good lab, but most of us nowadays don't, so there remain the time/communication problems.</p>

<p>There is an argument that MF just looks different due to the smaller DOF for a given perspective and the different tonality of film over a sensor. There is something in it, but unless you are prepared to pay quite a bit and engage mailing to places often a long way away, this is why I for one gave it up. My feeling is, if you want to try MF film, then do it for fun, but do not expect to see much (if any, but I won't go there) advantage over 35mm if you are working in a digital environment.</p>

[wblynch]

<p>Most scanners have to 'zoom out' to get the MF film into the frame, thus the limited MP sizes. Plus, 10 or 20 years ago computer storage was tiny and you certainly couldn't manage scan files that were multi-gigabytes in size.</p>

<p>And the real problem at that time was getting digital photo files to the printer as most were still optical. Later it all reversed when digital printers took over.</p>

<p>Scanners were invented to digitize film for transportation, library storage, editing and printing. They weren't invented for pixel peeping.</p>

 

<p>It's great fun to zoom in and in and in on a scan to look for tiny microscopic features or to see what detail there is hiding in there. But think about typical print size and you realize you don't need such huge scans. </p>

<p>Scanning technology pretty much halted in 2003. We are frozen in time.</p>

 

[jamie_robertson2]

<p>If you look at a 30Mb scan from a MF neg and compare the result to a 30Mb scan from a 35mm neg, the image quality of the MF scan will blow the 35mm away. The 35mm scan will show all the grain of the film. The MF scan will not show any grain at all (unless you're shooting very high ISO film) so the sharpness and tonality will be far superior.</p>

[peterbcarter]

<p>Jamie, simply not true. If you are magnifying to the same level, grain is exactly the same size. That being said, grain is not evil; just a factor in the look....fine or otherwise.<br>

robin:</p>

<blockquote>

<p>Well I think this is indeed a good reason to be wary of MF film. Unless you have a scanner yourself and can scan at equivalent resolution to 35mm then you are indeed wasting much of the advantage of MF.</p>

</blockquote>

<p>Not true. If you are scanning for the printed size, you require a lower dpi from the neg. This means the flaws become physically smaller and probably unnoticeable. That alone may be a significant gain. Another would be what I mentioned before.....the lenses generally are much better in that medium.</p>

<p> </p>

[James G. Dainis]

As Charles Monday points out above, the scanner resolutions are set so if the scanned files are then printed at 300ppi, the 35mm and the MF would both result in ~16x20 inch prints.

 

Perhaps it is arrogance on the part of the lab people. "A 16x20 inch print is the most that anyone should need."

 

People used medium format and large format so they could make large prints with less grain and a fine gradation of tone. Scanning the 6x7 film at 6300ppi would get a 14868 x 17388 file which printed at 300ppi would get a ~49.5 x 58 inch print that would be good for a life size portrait of that cute little blond.

[glen_h]

<blockquote>

<p>Why do they scan 35mm film at 6300 ppi but scan medium format film at 3200 ppi? Why can't (or won't) they scan medium format film at 6300 ppi?<br>

</p>

</blockquote>

<p>For 35mm at 6300 ppi requires (about) a 6300 pixels wide line (one dimensional) CCD. (Usually lines for R, G, and B.) There is a good supply of those.<br>

<br>

For 60mm, at 6300 pip requires many more pixels across, and sufficient optics to use it.<br>

<br>

Now, why can't you get 6300 pixels across a 6cm frame? That is different.<br>

<br>

Why do lenses for medium format cameras cost so much more than the same focal length for 35mm cameras? They are harder to make! </p>

[jamie_robertson2]

<blockquote>

<p>Jamie, simply not true. If you are magnifying to the same level, grain is exactly the same size.</p>

 

</blockquote>

<p>You misunderstood. I said "If you look at a <strong>30Mb scan</strong> from a MF neg and compare the result to a <strong>30Mb scan</strong> from a 35mm neg"<br>

<br>

A 30Mb scan from a MF neg will have been scanned at a much lower dpi than a 30Mb scan from a 35mm negative. If they were scanned at the same dpi the MF scan would be much larger than 30Mb. You won't see the grain on a 30Mb scan of a MF negative.</p>

[Alan Johnson]

<p>For C41 prints my lab requests files at 254 dpi resolution (they may be sharpened after downsizing to this). For a 16 x 20 print that is 4064 x 5080 = 20.6 mpx. They said there is no point in sending scans that give higher than 254 dpi resolution (the exact figure will vary between labs) as they are resampled to this for printing. My 6x9 film camera is interesting to use but I believe the resolution on C41 prints is limited by the printing process, not by the lens, film or scan.</p>

[glen_h]

<p>Hmm, 254 dots/inch, or 10 dots/mm. </p>

<p>The math for resampling is well understood, but I am not sure that any printers do it right.</p>

<p>That is, I am not completely sure that higher resolution doesn't mean more accurate resampling. </p>

<p>Once in a while, Shutterfly will send me an offer for a free 16x20 print. (Not actually free, as they still charge shipping.) The submit page will complain if your file doesn't have enough resolution, though I am not so sure what the number is. </p>

<p>Since I have the page out for another discussion, Ektar 100 MTF goes out to about 80 cycles/mm, or 3840x5760 pixels for a 35mm frame. Pretty close to what you say for 16x20.</p>

<p>http://imaging.kodakalaris.com/sites/prod/files/files/products/e4046_ektar_100.pdf</p>

<p>The scanner I have (ScanDual IV) goes to about 3200 pixels across a 35mm frame, so a little worse than that. That is the pixel resolution, I don't know about the optical quality at all. </p>

[glen_h]

<blockquote>

<p>You won't see the grain on a 30Mb scan of a MF negative.<br>

</p>

</blockquote>

<p>Well, you shouldn't see grain on any color film, but maybe you meant dye clouds instead. <br>

(I forget now if we are discussing color or black and white.)<br>

<br>

But yes, as above, Ektar 100 claims MTF out to about 80 cycles/mm, which is better than many scanners.<br>

<br>

Silver grains have much finer structure than the size of the grain. With high speed (large grain) films, you might see that in a scan. Otherwise, a microscope or good enlarging lens and full magnification. </p>

[James G. Dainis]

"640K ought to be enough for anybody."- Bill Gates (supposedly)

 

"16x20 is big enough for anybody". - Printing labs

 

Before Photoshop people used to do cropping in the darkroom. Raise the enlarger head to 11x14 image and crop out that annoying "No Parking" sign to get a nice 8x10 photo. I would rather scan once to a maximum resolution and then have a choice of getting 8x10, 11x14, 16x20, 20x30 prints made from that one scan than have to adjust the scanner resolution for each scan for each size print I want. Down sampling should not be a problem for the lab.

 

I might want my medium format negative scanned at 6300 ppi not because I want to make a 49.5 x 58 inch life size print of a cute little blond but because I want to crop out a life size head and shoulders 11 x 14 inch print. That should look very impressive almost like a contact print from an 11x14 inch negative.

[wblynch]

<p>Okay think about the scanner. It has a sensor of a certain size, say one inch. And say there are 3200 points on the sensor. So that's 3200 points per inch. </p>

<p>Now a 35mm film negative is about one inch across (24mm, not quite an inch...) so that scanner can scan 35mm at 3200 ppi.</p>

<p>And to scan a 2 1/4" 120 film negative the lens has to zoom out to fit 2 1/4 inches on a one inch sensor. So that means it scans 120 film at about 1400 points per inch. (2.25/3200).</p>

<p>This is why it scans medium format at a lower "PPI" than 35mm.</p>

 

[AlanKlein]

<p>Bill I don't think that's how a scanner works. It's scanning, not taking a picture with a fixed sensor. So there will be more pixels that are scanned from a MF shot than from a 35mm. </p>

<p>Regarding grain sizes, while these are the same size on both MF and 35mm of the same type film, the 35mm format film has to be enlarged over three times more than the MF film for the same picture area that was photographed. So the grain will be enlarged more than three times the size in the 35mm print or electronic file than the grains in a MF enlargement. </p>

<p>There are also more than three time the total quantity of grains in the MF shot than the 35mm shot for the same photograph angle. This is why you get more resolution in MF and also the graduation of the tones are better and more gradual than the 35mm film. Again because of the enlargement factor.</p>