Poor scans - May explain why some switch to Digital

<p>here is downsized original (cca 4600x3100px)</p>

<div></div>

<p>for illustration, this is unresized 512x256 cutout<br />yes, 35mm film cannot compete at high isos, but you can get very good home made results even with cheap hardware<br />and if you can see this slide projected - its just another world :-)<br>

it's handheld shot at 1/125, f4 or 5.6, so the DOF is just few mm, but surely, it's not film what is limiting resolution here :)</p><div></div>

<p>Rishi</p>

<p>you comparison posts are nice pictures but it would be better if you had taken exactly the same thing at the same time to compare. Comparing different things is hard.</p>

<p>While it is true that scans limit the ability to deliver the best from film, for the past I know this is important, but for today I think that digital offers the best for streamined and enhanced procesing. For example I recently took an interior for a client and I was stuck using difficult interior light. I used both a 35mm Canon EOS and a 20D for the image. I used HDRI to cope with the existing non-supplimented interior lighting, and I hoped that the negative film would provide good details and cope with the situation. The HDRI enabled me to capture enough to substantially cope with the 2450 deg K lighting temperature and bring the image out.</p>

<p>Perhaps you could get better details from the film than I could, but the processing ability of the RAW files allowed me to get in : take my images : get out : deliver a product to the client all in the same day.</p>

<p>Perhaps if I had used my 120 roll and a 6x12 I could get a better image ... perhaps not as I am certain that digital holds better subtle tonal variations than film does. My experience is that film is excellent when local contrast ratios are high, but where micro contrast is low details may be lost (particularly in shadows or hilights).</p><div></div>

<p>looking at a 100% segment the tonals in both hilight and shadow area are well superior than I could do with film (and I think I can work with it better than many film users).</p>

<p>Some times I prefer to use a single frame of negative (over anything else), but other times it is convenient (and better to have the right tool), sometimes digital is that tool</p>

<p>so it is not all 'marketing demons selling us short'</p><div></div>

<p>This is 35mm film vs a Canon 20D <br /><img src="http://www.pbase.com/mark_antony/image/82451759.jpg" alt="" /><br /> The film was Acros 100 in Rodinal the scanner a Minolta Dual III lots left in the neg at that mag. Not the 20D!</p>

<p><em>Perhaps you could get better details from the film than I could, but the processing ability of the RAW files allowed me to get in : take my images : get out : deliver a product to the client all in the same day</em><br>

Yoshio, it's true, and your work, it's exactly what is digital used for<br>

the goal of this thread is, that for some categories of photographers, film is fully suitable when properly scanned, and that earlier conclusions about 6-8mpix limit of 35mm film, or 10-12mpix from 6x4.5, these are just myths :)<br>

no doubt, for you is far better to shoot digitally, but many others can get comparable or far better results with film gear instead of digital<br>

you don't need to buy mf digital back for landscape shooting, as you can get stellar results with some cheap 6x7 system ... of course, the best drum scans are pricey, but how many "best" shots are good enough to be drum scanned ... so as with 35mm, you don't need to buy ff dslr at 10x - 20x higher price, compared to film slr<br>

and you still have ff dof, can use ff lenses, and can simply achieve 14-18mpix (and maybe more if drum scanned) resolution<br>

and you have much more than digital shooter - slides, with it's unbeatable look, look which can not be achieved by any s-xga or hdtv :)</p>

 

<blockquote>

<p ><a href="http://www.photo.net/photodb/user?user_id=423641">Landrum Kelly</a> <a href="http://www.photo.net/member-status-icons"><img title="Subscriber" src="http://static.photo.net/v3graphics/member-status-icons/sub8.gif" alt="" /><img title="Frequent poster" src="http://static.photo.net/v3graphics/member-status-icons/1roll.gif" alt="" /></a>, Mar 20, 2009; 10:39 p.m.</p>

 

 

<blockquote>

<p>"With digital I can preview my shots, use 1600 iso and still get low grain, and color balance in the raw conversion." --Steve J. Murray</p>

</blockquote>

<p>Yes, it does seem that most of the comparisons in which 35mm film is able to match or exceed digital, the operative assumption seems to be that one is using low ISO film. In other words, the idea of this thread and others like it seems to be "What is the very best that I can get from film if I do everything right, including scanning, at low ISO?"</p>

 

</blockquote>

<p>

<p>When I get a bit more energy I think I will start a new thread in the casual photo conversation that deals with what format is best suited for what kind of photography.</p>

<p>In this thread I think it is valid to use the best conditions for film. My one concern is that people don’t run out and buy a $2000 scanner thinking that every scan they get will be as good as some of the scans we have seen in this thread.</p>

<p>In going through my scans I was reminded how many skies I had to remove grain from to get the photo to look good enough for even displaying on the computer screen. If I show someone the best scan out of a roll they could well think that film works great, if I show them all the scans from the roll they would be less sure. For someone thinking about investing heavily into a scanner I would advise shooting a few rolls of film and having the best few shots professionally scanned to get a good idea of how well it would work from them.</p>

<p>I have attached a low-res scan of what I see a lot in my scans, the snow looks fine, the rocks are hard to tell but the sky looks like crap.</p>

</p>

<p> </p>

<p>Hm, the photo did not seem to attach to the last post, I will try again.</p><div></div>

<blockquote>

<p ><a href="../photodb/user?user_id=2163626">Mark Smith</a> <a href="../member-status-icons"></a>, Mar 21, 2009; 07:17 a.m.</p>

 

<p>This is 35mm film vs a Canon 20D <br /><img src="http://www.pbase.com/mark_antony/image/82451759.jpg" alt="" /><br />The film was Acros 100 in Rodinal the scanner a Minolta Dual III lots left in the neg at that mag. Not the 20D!</p>

 

</blockquote>

<p>Something looks way off here, I have a Minolta Dual III and a 20D, with the scanner you can get about 10MP and with the 20D about 8MP, so how is it that the pixels from the scanner look to be about half the size as from the 20D?</p>

<p><em>In going through my scans I was reminded how many skies I had to remove grain from to get the photo to look good enough for even displaying on the computer screen.</em><br>

Scott, i think visible grain (or invisible grain) is critical for displaying on computer screen, as what you see is 70-100dpi image, for prints it's much more acceptable, at 300 or even more dpi<br>

no doubt, generally film has more grain than has digital noise, and it's easy getting worse (for film) at higher isos<br>

but even 35mm film is capable to catch details, fully comparable with ff digital, of course for many people is film grain unacceptable and disturbing</p>

<blockquote>Something looks way off here, I have a Minolta Dual III and a 20D, with the scanner you can get about 10MP and with the 20D about 8MP, so how is it that the pixels from the scanner look to be about half the size as from the 20D?</blockquote>

<p>The Dual III is over 11mp, I think that the film scan is representative of what I get. The 20D shot was blown up to match the size then I took a screenie.<br>

My point being the 20D is never going to match the film. Both cameras are from the same VP with 50mm FL</p>

<blockquote>

<p ><a href="http://www.photo.net/photodb/user?user_id=4848492">Radek Pohl</a> <a href="http://www.photo.net/member-status-icons"></a>, Mar 21, 2009; 07:48 a.m.</p>

 

<p><em>In going through my scans I was reminded how many skies I had to remove grain from to get the photo to look good enough for even displaying on the computer screen.</em><br />Scott, i think visible grain (or invisible grain) is critical for displaying on computer screen, as what you see is 70-100dpi image, for prints it's much more acceptable, at 300 or even more dpi<br />no doubt, generally film has more grain than has digital noise, and it's easy getting worse (for film) at higher isos<br />but even 35mm film is capable to catch details, fully comparable with ff digital, of course for many people is film grain unacceptable and disturbing</p>

 

</blockquote>

<p>

<p>I have a very low tolerance for gain, when I got my first digital (3MP) that could make a good 4x6 inch print I was amazed at how much better then looked then any 4x6 inch print I ever got form film. At that time if I was printing at 8x12 size I had to choose between high grain from film or soft prints from digital. Now I just shoot digital and have left grain behind me. I should add it is not just the grain in skies that bugs me, it seems to show up on a lot of surfaces that show not show that kind of texture i.e. a smooth surface should look smooth IMO.</p>

</p>

<blockquote>

<p ><a href="http://www.photo.net/photodb/user?user_id=2163626">Mark Smith</a> <a href="http://www.photo.net/member-status-icons"></a>, Mar 21, 2009; 08:05 a.m.</p>

 

 

<blockquote>Something looks way off here, I have a Minolta Dual III and a 20D, with the scanner you can get about 10MP and with the 20D about 8MP, so how is it that the pixels from the scanner look to be about half the size as from the 20D?</blockquote>

<p>The Dual III is over 11mp, I think that the film scan is representative of what I get. The 20D shot was blown up to match the size then I took a screenie.<br />My point being the 20D is never going to match the film. Both cameras are from the same VP with 50mm FL</p>

 

 

</blockquote>

<p>If you can scan the whole 24x36mm area you get 10.8 MP, I rarely get that many pixels from my scan using the Dual III</p>

<p>Your 20D shot looks like it was blown up but a whole lot, between 8 to 10.8 mp it should only have been scaled up 16%, something is way wrong with the images you posted.</p>

<p>Scott, this is the raw file you requested. Please post your best upsized processed results for comparison against this if you would. </p>

<p><a href="http://www.yousendit.com/download/UmNLb3BBYTJPSHhjR0E9PQ" >http://www.yousendit.com/download/UmNLb3BBYTJPSHhjR0E9PQ</a></p>

<div></div>

<p>"Your 20D shot looks like it was blown up but a whole lot, between 8 to 10.8 mp it should only have been scaled up 16%, something is way wrong with the images you posted".<br>

Same FOV same focal length taken within 30 seconds of each other, the film has WAY further to go, how much more the 20D has to go.<br>

Fact is the 20D can match the resolution of the film all things being equal.<br>

I'll post 1000% views later</p>

<p><br /></p>

<p ><a href="../photodb/user?user_id=359171">David Littleboy (Tokyo, Japan)</a> <a href="../member-status-icons"><img title="Subscriber" src="http://static.photo.net/v3graphics/member-status-icons/sub8.gif" alt="" title="Subscriber" /> </a> , Mar 21, 2009; 02:11 a.m.</p>

 

<p>"why do you lump 645 film and 12MP FF digital together "<br>

Because they both make prints that look a lot better than 35mm.</p>

<p>David, if your prints from 35mm film do not look better than a 12MP DSLR I believe you. But the problem is not the film. Go out and shoot two identical scenes with 35mm Velvia film and your 12MP DSLR, same lens and all, and post the results....</p>

 

<blockquote>

<p ><a href="../photodb/user?user_id=3995956">Mauro Franic</a> <a href="../member-status-icons"><img title="Frequent poster" src="http://static.photo.net/v3graphics/member-status-icons/3rolls.gif" alt="" /></a>, Mar 21, 2009; 08:22 a.m.</p>

 

<p>Scott, this is the raw file you requested. Please post your best upsized processed results for comparison against this if you would.</p>

 

<p><a rel="nofollow" href="http://www.yousendit.com/download/UmNLb3BBYTJPSHhjR0E9PQ" target="_blank">http://www.yousendit.com/download/UmNLb3BBYTJPSHhjR0E9PQ</a></p>

 

</blockquote>

<p>It would seem I got the wrong raw file, this is what I got from it.</p><div></div>

<p>Oops. Give me 5 mins.</p>

<p>Please try this:<br>

<a rel="nofollow" href="http://download.yousendit.com/TTZrYlJhV3JtNEkwTVE9PQ" target="_blank">http://download.yousendit.com/TTZrYlJhV3JtNEkwTVE9PQ </a></p>

<p>"But the problem is not the film."</p>

<p>I've shot and scanned enough Velvia 50 to be quite familiar with it, thank you. Here, it looks very much like the images at the various test sites referenced (and rudely dismissed) in this thread: unacceptably grainy when enlarged the 13x required to get 35mm to 12x18. I'm not finding anything that differs from what lots of other people find. Meanwhile, the 5D produces very nice 12x18 prints every time.</p>

 

<p>Grainy Velvia at 13X? <br>

Once again, my prints/results are different than yours but if you want to remove grain you can do so with software. You can't add detail with software though.</p>

<p>If you have a microscope look at Velvia under it at 40x and tell me how much grain you see...</p>

<p> </p>

<blockquote>

<p ><a href="http://www.photo.net/photodb/user?user_id=3995956">Mauro Franic</a> <a href="http://www.photo.net/member-status-icons"><img title="Frequent poster" src="http://static.photo.net/v3graphics/member-status-icons/3rolls.gif" alt="" /></a>, Mar 21, 2009; 09:04 a.m.</p>

 

<p>Please try this:<br /><a rel="nofollow" href="http://download.yousendit.com/TTZrYlJhV3JtNEkwTVE9PQ " target="_blank">http://download.yousendit.com/TTZrYlJhV3JtNEkwTVE9PQ </a></p>

 

</blockquote>

<p>Ok this is what I got, I turned of any color noise reduction, which help get the color back into the DOV scale on the lens.</p>

<p>If you could post about the same crop area from the film shot I would be interesting in printing them both out. I would like to see how the film did at some of the lower contrast areas of the photo, like where the paper has a kink in it on the far right side.</p><div></div>

<blockquote>

<p ><a href="http://www.photo.net/photodb/user?user_id=3995956">Mauro Franic</a> <a href="http://www.photo.net/member-status-icons"><img title="Frequent poster" src="http://static.photo.net/v3graphics/member-status-icons/3rolls.gif" alt="" /></a>, Mar 21, 2009; 09:46 a.m.</p>

 

<p>Grainy Velvia at 13X? <br />Once again, my prints/results are different than yours but if you want to remove grain you can do so with software. You can't add detail with software though.<br>

If you have a microscope look at Velvia under it at 40x and tell me how much grain you see...</p>

 

</blockquote>

<p>

<p>I get lots of grain from Velvia. In the scans attached the one on the left is from my Minolta Dual III and the one on the right is from the Nikon V ED sized to match my scan. I sent the slide to a friend who had the Nikon V to seem how much detail my scanner was missing, along the way it got some dust on it which is why the scan on the right looks dirty. But both scans show a lot of grain.</p>

</p><div></div>

<p>I have added to the scan images what the car really looks like (far right image), IMO smooth should look smooth, I don't want any added texture to my photos. Some people complain that digital looks plastic, well a photo of plastic should look like plastic.</p><div></div>

<p>Scott, try upsampling your results and place them at the same size next to the four extracts of the image above.</p>

<p>This is the area from the film you requested (replace the 40D at the bottom with the results you processed):<br>

http://shutterclick.smugmug.com/gallery/7469988_hodC5#481944852_7SQ5U-O-LB</p>