Poor scans - May explain why some switch to Digital

<blockquote>

<p>... could this longer integration time (compared to faster scanners) be a significant factor in producing lower noise scans? My Nikon LS5000ED scans the same 135 portrait in 20 seconds but with vastly more noise ...</p>

</blockquote>

<p>It's easy enough to test this hypothesis. See if noise goes down with different amounts of multipass. If your 5000 is like mine, I don't think you'll see much difference at all.</p>

<blockquote>

<p>Lex Harris wrote:One thing that puzzles me is the "scanner" example you cited (DSLR back on macro copying stand). Is this not a "matrix" type sensor, inferior to the line sensors in scanners? So was this example just tongue-in-cheek or are you seriously proposing that this is a better way of "scanning" than traditional scanners with in-line arrays? Sorry, I can't tell for sure.<br>

My own experience with CCD array sensors and derivatives such as CID arrays in industrial applications over the last 15 years or so is that these devices are limited in dynamic range and no real improvement in dynamic range has occurred over this period. This seems to be consistent which what you are saying so I'm interested to hear more.</p>

</blockquote>

<p>Lex, the moderators of this forum are wiping my posts here, so I will not post. As you are a photographer - if you look carefully on that method and if you look carefully on all those "flare" from modern film scanners you will see the solution.</p>

 

<blockquote>

<a href="../photodb/user?user_id=1599689">Robert Lee</a> <a href="../member-status-icons"><img title="Frequent poster" src="http://static.photo.net/v3graphics/member-status-icons/1roll.gif" alt="" /></a>, Mar 29, 2009; 10:53 a.m.

<p> It's easy enough to test this hypothesis. See if noise goes down with different amounts of multipass. If your 5000 is like mine, I don't think you'll see much difference at all.</p>

</blockquote>

 

<p>Robert you're right, even x16 multipass on my 5000 makes very little difference. But scans from my 5000 are so flooded with flare and ghosting that maybe this is masking any improvement in noise from multipass. Who knows....</p>

 

 

 

<blockquote>

<p ><a href="../photodb/user?user_id=643213">Sergiy Podolyak</a> <a href="../member-status-icons"></a>, Mar 30, 2009; 04:46 a.m.</p>

 

 

 

<p>if you look carefully on all those "flare" from modern film scanners you will see the solution.</p>

 

</blockquote>

<p>My 646 exhibits no flare whatsoever. Imacon scanners have only Rodenstock lens between film and sensor and sensor cover glass is coated both sides. Light sensitive devices on Kodak KLI-8023 sensor are photodiodes, not CCDs, and thus blooming cannot occur either.</p>

 

<p>Hey guys, I have canon 8800f flatbed scanner. here are so many friends with great scanning output with 35mm films, but can I know how can I get the best results from my canon 8800f flatbed, I have posted some scanned samples from my flatbed. Please tell me so that I can get finest results. I think I am not getting what I can get as so many examples are great here so I think I can also get the far better results from a flatbed.</p>

<p>Pankaj,</p>

<p>I'm not sure what makes you think that a Canon 8800f flatbed scanner will give you anything near the quality of scans you are seeing here on this thread.</p>

<p>Follow this link to a resolution test of that scanner: <a href="http://filmscanner.info/en/CanonCanoScan8800F.html">http://filmscanner.info/en/CanonCanoScan8800F.html</a></p>

<p>If these guys did their testing correctly, that's an effective resolution of 1600dpi. That's an effective <strong>3.4MP</strong> . We're talking about <strong>19-20MP</strong> scans from Imacon/Minolta/Nikon scanners. That's nearly an order of magnitude difference.</p>

<p>You can't possibly scan 35mm (or any film, IMHO) with a flatbed while simultaneously caring about image quality.</p>

<p>Sorry :(<br>

-Rishi</p>

<p>Thanks to all who have contributed... I find this a very interesting thread. (Too bad it's gotten so large it takes a long time to load.)<br>

Mauro and Rishi (and Todd) prompted me to want to do an experiment of my own. Take a typical tourist photo with typical equipment and typical processing, and see how Digital compares to scanned film. I went down to the popular tourist spot, Cheers Pub in Boston with a D100 & Nikkor 35mm f/2 along with a Nikon FG & Nikkor 50mm f/1.8 with Fujicolor 200 film. Same shot from the same spot. The lenses gave comparable views. I had Ritz camera process and print the film. Then scanned the negs on a Nikon Coolscan V ED (not 5000, not 9000).<br>

Results below (actual pixels from near center in each case; I don't use the res-up approach)... My conclusions:<br>

1. With these specifics, film out-resolves digital. Read the words under "Adams." <br>

2. The film is MUCH more noisy than the D100. The grey granite is just so much cleaner in the D100 image. How much would we notice the noise in a print? Not at all in a 4x6 print from the mini-lab. Nor on screen. Scanning chromes and negatives, it has been clear to me that we've lived with this noise throughout the film era. We read a lot about noise in our dSLR reviews, but how much does noise matter? <br>

3. Noise reduction on the scan helps a lot, and it reduces the resolving power. Again, the words under "Adams" which seem which seem a little clearer in the noisy scan than after noise reduction. <br>

4. At normal viewing distance, scanned 35mm film vs. a D100 is a toss-up; both are excellent for pleasing consumer prints.<br>

5. The mini-lab print direct from film... I believe the mini-lab prints I get from digital files (D70, D100, D200 with very little adjustment and one-size-fits-all sharpening) are MUCH better than straight-from-film prints. Without doing anything that seems beyond the capability of an automated process, my prints from digital files are better than straight-from-film mini-lab prints. I don't know why. <br>

Here's my comparison...<br>

<img src="http://2under.net/images/Karash-Image035-Cheers-Comparison.jpg" alt="" /></p>

<p> </p>

<p>Hmm... My actual pixels got resized in the forum. Sorry. Here's a link to the actual pixels image:<br>

http://2under.net/images/Karash-Image035-Cheers-Comparison.jpg<br>

And, here's the whole photo to put the above crops in context. This is the scanned Fujicolor 200.<br>

<img src="http://2under.net/images/Karash-Image035-ScannedFilm-Cheers-8A-ScreenOpt.jpg" alt="" /></p>

<p>And, here is the D100 which, even at this size, I like just a little less than the scanned film above. No sharpening in any of these images.<br>

<img src="http://2under.net/images/Karash-DSC_0035-D100-Cheers-ScreenOpt.jpg" alt="" /></p>

<p>Very interesting Richard. Thank you for contributing. The film shot seems a bit out of focus. Where did you focus on and from what distance and aperture?</p>

<p>It'd be interesting to see a comparison using Ektar 100 as well.</p>

<p>The round overhang canopies are good to see the tonal gradation differences.</p>

<p>Mauro, I focused in both cases on the wrought-iron fence, center, lower 1/3 of the image. The big canopy projects out towards the camera, the bricks and granite just a little further away. Distance 100 feet, shot from across Beacon St. Probably f/16 and 125th (both film and D100 were ISO 200). Manual focus in both cases. For the film camera, there's a good DOF scale; everything in the image is way inside the indicated DOF range for f/16. I probably had both lenses focused at infinity.<br>

I see better tonal gradation in the canopies, granite building face is much more lively. The lettering on the big canopy is a little crisper.<br>

Mauro, what makes you think the film might be out of focus?</p>

 

<p>On the film shot it appeared to me that the iron light pole was sharper then the small lettering on the Sam Adams sign (Take pride....). Just wondered - nothing major.</p>

<p>Richard Karash</p>

<p>very nice comparison, and if I may say very good scan results too. May I ask a few questions?</p>

<p>First: there seems to be some texture in the stone work to the upper left of the Samuel Adams sign, in the lit area (forming a kind of triangle shape above the sign). This subtle texture seems to be lost in the negative scan (as indeed the line of the brick work between the lamp post and the sign (just right of the lamp post). Was this subtle pattern / texture present in the brickwork or is it an artifact? If it was then this is an interesting observation. It supports my observations up that I see low contrast details better on digital than on film scans.</p>

<p>Last question: did you need to make more than one scan to be satisfied with the focus and capture of the image?</p>

<p>The D100 is quite similar to my own 10D. I know that my new G1 resolves significantly better than the 10D does. The increase I see is about that given in your comparisons. So based on your findings I feel that my tiddly little G1 will now do as well as my EOS630 with negative. Of course one can not get a better image from ones older digital images unlike film where access to better equipment will reveal better images.</p>

<p>Anyway, I no longer bother with the 35mm as anything more than casual images, (except for the little HIE I have in my freezer) and I use 4x5 and 120 for the "gosh that looks nice" things</p>

<p>Thankyou for your contribution</p>

<p>LANNIE,<br>

in this cropped image of yours,<br>

http://www.photo.net/photo/6649708&size=lg<br>

there is this beam / rod that goes from the brick red wall of the house on the left to the downwards to the other house on the right. this rod also casts a shadow on the red wall. you can see that the edges of the shadow are jagged (staircase type look). <br>

I have two questions. (not just to Lannie, other experts are obviously requested to give opinions if the question is relevant)<br>

1. Would the shadow look jagged if this was shot on slide film and viewed using a projector ?<br>

2. Would the shadow look jagged of this was shot on slide film and then scanned ?<br>

Of course the question is irrelavant if the rod (or whatever it is) itself is having that kind of jagged edge to begin with, in which case, ignore the question. I see that the edges of the other vertical shadow on the red wall does not have jagged edges.<br>

thanks and regards.<br>

p.s.<br>

it is now 4 a.m. i started reading this thread at somewhere around 11:p.m. and its Monday morning tomorrow. Damn you guys @#$#$!!!!</p>

<p><img src="http://2under.net/images/Karash-image035-Cheers-Comparison-TextureInGranite.jpg" alt="" width="705" height="733" /><br>

Yoshio Tanaka<br>

Thank you for your comments. If there is any merit in the scan, it is the Nikon Coolscan V and Nikon Scan software that deserves the credit. <br>

Your questions... First, the subtle texture... I didn't see it at first, but I agree. There is subtle texture in the granite and the D100 did better than the Fuji 200 film in capturing this. But it's certainly subtle. See another 100% crop here.<br>

Second, the scanning. This was the first try. I simply picked the focus area (the "Samuel Adams" sign), clicked for auto-focus, and took the defaults for the scan. Digital ICE was "ON."<br>

I made 8x12 prints of the Fuji 200 and the D100 images. Both prints are excellent in tone, sharpness and detail. It would be hard to pick a winner in a side to side comparison. The texture difference in the granite is hard to see.<br>

Overall, I am pleased that, without extraordinary measures, one can get an excellent scan from 35mm color negative film, one that is close enough to a D100 capture that it's hard to pick a winner. Based on this, I'll start shooting some film again.<br>

About the crop shown here: Fuji 200 on the left, raw scan, no noise reduction, actual pixels. D100 on the right, actual pixels. I see the textured granite in the red box in the D100. In the Fuji 200 scan, the noise eliminates most of that texture.</p>

<p>pankaj,<br>

while this thread is extremely interesting (i spent close to 5 hours on this thread on sunday night), don't take everything said in here in full seriousness. make sure it is relevant to your particular case.<br>

While your Canon scanner may not be top of the line, it may well be enough to satisfy your requirements if you know how to extract the most from it. believe me, i have got scans made only from "professional scanning services", and every single scan has sucked. big time. this was part of the reason i shifted to digital at that time. Had i known what i know now, i would have stuck with film and learned to use the affordable scanners to the best of their ability.<br>

remember, these people are talking about equipment that generate 100+ MB files. what is your requirement really ? if you need to scan your film for posting on web-sites like photo.net or flickr, believe me modest equipment is more than enough. <br>

Case in point :<br>

the above pic has been scanned with a epson perfection v200. something these people will pooh pooh about. but tell me honestly, do you really need more than that ? Also search on flickr for photos with tag "EPSON 4490". You'll find tons of stuff. really good stuff.<br>

Mainly,<br>

1. process your film at a GOOD lab. compare outputs of a few labs around where you live, and choose the best one.<br>

2. When scanning, see if you tweak each scan for that particular photo. Don't use one setting for every photo.<br>

3. Learn Photoshop or GIMP. Every image from a scanner or a DSLR needs basic image editing work. Learn to use these tools to get you the result you want.<br>

Most importantly, don't let anyone discourage you by stating your equipment is mediocre. This particular thread has very long back stopped being about photography. This is about people with points to prove, engineers and Ph.Ds with new scanning devices to build, and parallel universes to create.<br>

This is the wrong thread to ask for the kind of help and advice you are looking for.<br>

Just my suggestion.</p>

<p>Venkat, you made me smile, well said and puts it all into perspective. Pankaj, take Venkat's advice, points 1,2,3 are the basis from which to work, aside from taking good photos in the first place which of course is the fundamental starting point.</p>

 

<p>Mauro-- can you send me the 40D RAW file... I want to try processing with <a href="http://www.raw-photo-processor.com/RPP/Overview.html">RPP</a> ... I just found it leads to dramatic sharpness detail over ACR... so much so that in the future I'm thinking of using Lightroom simply as an asset management tool, what with its pathetic handling of highlights & lack of proofing. Either that or switch to Aperture, but I like Lightroom's selective sharpening.</p>

<p>Thanks,<br>

Rishi</p>

<p> </p>

<p>Oh, I almost forgot. If you want to make good scans from film, just read this scanning primer, and simply follow it:<br /> <a href="http://www.marginalsoftware.com/HowtoScan/tutorial_page_1.htm">A Primer on Image Histograms and Curves -- Taking Voodoo and Black Magic out of Making Quality Scans</a> (updated periodically) <br /> http://www.marginalsoftware.com/Scanner/Scanner_Intro.htm<br>

No really "voodoo" or "magic" there, just science and math.</p>

<p>Mauro, It's time for another superthread. Post something so that we can spend a few more weeks on these issues.</p>

<p>--Lannie</p>

<blockquote>

<p>"LANNIE,<br /> in this cropped image of yours,<br /> <a rel="nofollow" href="../photo/6649708&size=lg">http://www.photo.net/photo/6649708&size=lg</a> <br /> there is this beam / rod that goes from the brick red wall of the house on the left to the downwards to the other house on the right. this rod also casts a shadow on the red wall. you can see that the edges of the shadow are jagged (staircase type look). <br /> I have two questions. (not just to Lannie, other experts are obviously requested to give opinions if the question is relevant)<br /> 1. Would the shadow look jagged if this was shot on slide film and viewed using a projector ?<br /> 2. Would the shadow look jagged of this was shot on slide film and then scanned ?"</p>

 

</blockquote>

<p>Venkat, this could occur with any digitized image, including scanned film.</p>

<p>(I'm sorry I missed this question earlier, but I had no idea this thread was still going.)</p>

<p>--Lannie</p>

<blockquote>

<p>Venkat, this could occur with any digitized image, including scanned film.</p>

 

</blockquote>

<p>Lannie, in my experience that is not true. I have never seen such jagged pixelated lines in a scanned image viewed up to 100%. You would only begin to see such an effect if you zoomed to >100%, to the point where individual pixels become clearly visible.</p>

<p>So to answer Venkat's questions:<br>

1. No<br>

2. No</p>

<p>And the bottom line is?</p>

<p>I just print within the constraints of the medium. So, for me, 100% crops don't really matter. AAfter all, a good photographer can produce a compelling image with either film or digital.</p>

<p>so Lex,<br>

thats a very crucial point in film's favour. this is probably why films give that more liquid smooth feel, perhaps ? in a simple image taken with a digital camera, while the naked eye may not be able to discern such small individual jagged edges to elements in the photo, they might all overall add up when we view the picture as a whole ?</p>