Poor scans - May explain why some switch to Digital

[mauro_franic]

<p>Desmond, congrats on the RZ! If you use a meter, cable release, mirror lock-up, f 11-16, Velvia, 50ULD or 110f2.8 lenses, focus properly, etc, (remember you are now capturing over 100MP worth of information so technique is imperative) your 1DsIII won't see many landscapes in the future.<br>

I'll certainly run some scans for you on the Coolscan 9000, just email me when you are ready.</p>

 

[desmond_kidman]

<p>Mauro, thanks, it will only be one or 2 scans. Then I will have some light jet prints made. Strange to re-buy that which I sold, but have to do it to see for myself. </p>

<p>I remember so often thinking "I have to sell the RZ stuff before its value really plummets". Well, it has probably plummetted as far as it's going to go! So it's a no-risk proposition, really. I never even sold the lower-cost items such as viewing screens, compendium shade (now why can't the 35mm makers do one that fits all of their lenses?), extension tubes, prism finder (which I have never really liked on RZs or Hasselblads), etc. I'll feel right at home again! I'll contact you off-line about any other details, I don't want to clog up the thread, but I will post my thoughts in a few weeks as another data point for readers. </p>

[mauro_franic]

<p>Cool beans. I think if they ever stop producing the RZ the units avilable may sky rocket in price. canon took the Elan 7ne from the website a month or so ago and mint units on ebay went from $80 to $150+ almost overnight.</p>

[roger_smith4]

<p>I recommend making profiles for slide film with Noise Ninja- just use a IT8 target. It comes with profiles for some digital cameras. I don't find that it leaves artifacts (unlike Neat Image) and retains the original grain structure, which I prefer to perfect smoothness.</p>

[swilson]

<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; 11:12 a.m.</p>

<blockquote>

 

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

This is the area from the film you requested (replace the 40D at the bottom with the results you processed):<br /><a rel="nofollow" href="http://shutterclick.smugmug.com/gallery/7469988_hodC5#481944852_7SQ5U-O-LB" target="_blank">http://shutterclick.smugmug.com/gallery/7469988_hodC5#481944852_7SQ5U-O-LB</a></p>

 

</blockquote>

<p>I got looking at this with more care, something is off on the scaling between the film shot and the digital, it looks like the digital had a larger FOV, making it scale up more then it should have to match the film shot.<br>

I measured how much I needed to scale up and put that in, only later did that number seem off to me. I should only have to scale up by 145.8% to match the two, and this assume you made full use of the 35mm frame. But I found I had to scale up by 155.3% to get them to match. That seems like a pretty large error in framing to me and is in the film shots advantage.</p>

[mauro_franic]

<p>Could it be cropping on the film holder? I framed them to the closest possible by eye.</p>

[rishij]

<p>Roger & Scott Turner:</p>

<p>I bought NeatImage b/c of its individual control of noise reduction in high vs. mid. vs. low frequency data. I find this allows me to retain more mid-frequency data while getting rid of high-frequency film grain. Noise Ninja has no such provision.</p>

<p>However, Neat Image's artifacts that look like a random isotropic arrangement of crystals are rather distracting & hideous. Even happens with digital images.</p>

<p>So perhaps profiling helps with this? Scott, you haven't been distracted by said artifacts? Roger, I tried some Panasonic LX3 profiles that someone else built... they worked pretty well on my LX3 images... but failed to work on PhotoAcute-treated images... maybe I'd have to build profiles after using PhotoAcute, I don't know.</p>

<p>In generally, though, I have found that the LX3 profiles end up eliminating more noise while preserving detail rather than when I just select one spot for profiling in NeatImage.<br>

-Rishi</p>

[swilson]

<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 23, 2009; 04:42 p.m.<br>

Could it be cropping on the film holder? I framed them to the closest possible by eye.</p>

</blockquote>

<p>My guess is the DSLR's viewfinder is not 100%, check the full size image for FOV, as I said something is off.</p>

<p>If it was cropping on the film holder the error would have been in the other direction.</p>

[laverephoto]

<p>One thing this comparison ignores is that the physical size of the sensor or the film determines how much light is necessary and weather you can get the shot without a tripod.<br>

I'll never understand why people compare a dslr with medium format film, scanned or unscanned. If you want to take the same shot with the same crop, depth of field, and angle of view, you have to stop a medium format camera down at least three stops and then open up the shutter speed to compensate. In real life that means you need a tremendous increase in any artificial light you are using and you will likely need a tripod when you could be hand holding with a dslr. I'm not saying it isn't worth it for some shots, but I am saying the two types of cameras do not handle remotely in the same ball park where speed and mobility are an issue.<br>

A more meaningful comparison would be a modern Phase One back against medium format scanned film. In my area a good scan isn't cheap so the Phase One starts looking pretty reasonable if you shoot any kind of volume. I guess film users would then point to 4X5 film as the gold standard, but now we are back to the days of 30 pound tripods and generators to run a 4800 WS strobe pack. No thanks!</p>

 

[roger_smith4]

<p>"I bought NeatImage b/c of its individual control of noise reduction in high vs. mid. vs. low frequency data. I find this allows me to retain more mid-frequency data while getting rid of high-frequency film grain. Noise Ninja has no such provision."</p>

<p>That may be so but I don't miss the control for my purposes (which are reducing film grain and digital noise too severe for Lightroom to fix). I'd rather not have to fight artifacts. </p>

<p>I use settings of 7 strength 5 smooth and 12 contrast for luminance and color as starting points (depends on the image) and instead of Neat Image's manual "control" I use NN to select areas of the image with visible grain/noise and base its noise reduction on that. I don't really know how it works but do find it to be effective. I also use it locally through masks (for when I want the sky to be smoother than the rest, for example). Others recommend Noiseware.</p>

[scott_turner]

<p>Rishi,</p>

<p>As with all of the profile dependant noise reduction apps, the quality of a Neat Image processed pic is a reflection of the quality of the profile generated. And the best profile is <em>always</em> the most specific one. Generic profiles, such as the one you found for your LX3, are a compromise and done for convenience. Following the steps to make a generic profile for YOUR camera takes things up a notch, while making a profile specific to an individual shot is the best yet, assuming the image has detail-free areas to base it on. </p>

<p>Yes, I occasionally get nasty artifacts, but it's usually a good indication that I've botched the profile. In a well-done image, I don't see what you're describing. But, as noted earlier, I use NI selectively, lightly and only for very grainy images. I don't try to use it to turn film into something that it was never intended to be. I hate grain, but I mostly deal with it by shooting slow film and not pushing print sizes to levels a given format was never intended to support.</p>

<p>And yes, in order to account for what you do with PhotoAcute you'd have to make a profile based on the shot after running it and then apply it. You might be best off to apply NI before running PhotoAcute, but having never used the package, I don't know how that would impact what it does - this may well be a case where the two just don't combine well.</p>

<p>Roger, NI can be used such that it does exactly as you describe with Noise Ninja, leaving the grain structure intact. Like Rishi, I prefer NI's flexibility, especially for film, but the default reduction settings are ridiculously strong and leave many with a poor impression of what the product is capable of producing. It can be used with a very delicate touch.</p>

<p>Scott</p>

[rishij]

<p>Thanks Scott & Roger, you are both always so helpful.</p>

<p>The one weird thing to me about NeatImage, however, is that it only has one box to place in the image when generating the profile. Noise Ninja can sample noise from many parts of the scene. My understanding is that Neat Image does not allow this.</p>

<p>Am I mistaken?<br>

-Rishi</p>

[mauro_franic]

<p>Rishi, NI uses multiple sample areas automatically. You can also fine tune them.</p>

[swilson]

<blockquote>

<p ><a href="http://www.photo.net/photodb/user?user_id=3972491">John LaVere</a> <a href="http://www.photo.net/member-status-icons"><img title="Subscriber" src="http://static.photo.net/v3graphics/member-status-icons/sub2.gif" alt="" /></a>, Mar 23, 2009; 07:26 p.m.</p>

 

<p>One thing this comparison ignores is that the physical size of the sensor or the film determines how much light is necessary and weather you can get the shot without a tripod.<br />I'll never understand why people compare a dslr with medium format film, scanned or unscanned.</p>

 

</blockquote>

<p>

<p>As DLSRs have gone up in resolution the question naturally comes up as to what all they can replace. If a studio photographer has been shooing 6x6 in his/her studio it would be of great interest to them to know if they could replace their film workflow for digit with a relatively cheap DSLR.</p>

<p>From what I have seen DSLRs are being used in many places that only a MF would have been used 10 years ago. This does not mean the DSLR is better then MF but that it is good enough to get the job done.</p>

<p>10 years ago we had some ad shots done on a 6x6 camera and drum scanned, the images were scanned at 2000ppi for an image size of around 20MP. These same images could today be taken with a large number of DSLR cameras and work fine. Could more have been gotten out of the MF frame? Maybe but if you scan at 2000 ppi you can get pretty nice clean pixels, go past that and they start to get not so clean.</p>

<p>For many people it does not matter if a given film format can do better then a DSLR, it is not about using the one that produces the image with the most detail capture. Film is a pain in the ass to deal with, if a DSLR can give you what you need for most pros that is all they care about.</p>

<p>

<p>The points you make are valid is an amateur thinks he can start using a MF camera in the same way he has been using a DSLR and the only difference will be greater resolution with the MF camera. But I don’t think the sites that compare the two are really thinking about the part of the photographic market.</p>

</p>

 

<p>

 

</p>

</p>

 

[rishij]

<p>But I swear I only ever see on box in NI, Mauro... what am I doing wrong?</p>

[mauro_franic]

<p>Nothing. NI only shows the first box.</p>

[mauro_franic]

<p>John, also narrower DOF is sometimes sought for compositions reasons. It is not an ablsolute drawback or advantage. </p>

<p>I personally love the ability to isolate the subject more for people. Landscapes at longer distances it matters less. If I did macro photography, a compact camera like the G10 will outdo a even a DSLR with the ability to include more of the subject in near focus.</p>

[peter_ferling]

<p>What an exaughsting thread, (of course much has to do with the lens, another subject :). <br>

I can tell you that I use both formats for their pros. In the corporate studio, we sold off the cambos and purchased 1Ds to drop the cost of film developing and gain instant gratification that is needed in today's advertising environment. Even more recently we have seen a drop in request for large prints as we are switching to 'digital posters', and displaying large graphics on flat panel LCDs at events, etc.</p>

<p>Personally, In the field I carry both a 40D and a mamiya 645M. I shoot and experiment with the 40D, and when I find a wow shot. I reshoot with the Mam. Even so, there are shots from the 40D that I pick over the MF because they simply look better and/or work for the scene, composition, luck, etc. Then there are a few times I'll just take a T90 or the old FTBn for the simple joy of it. </p>

<p>With the 40D, or any digital, it's bang, bang, bang with almost unlimited ammo. With only 36 film shots... You begin to think a little bit more about the shot before you pull the trigger. I seem to remember that on those days that I only used film, as being more enjoyable.</p>

<p>In regards to scans, I use an Epson 1650XL with the lighted transparency tray, and only for 120. For 35mm I mount and shoot them on a light table with the 40D and 100mm macro lens.</p>

[scott_turner]

<p>Rishi,</p>

<p>Up until the latest version, NI allowed you to do an initial profile, and then virtually unlimited fine tuning in selected areas with additional boxes. They apparently feel that's no longer necessary, as the feature is now gone, at least in the 64-bit plug in I use in CS4. There is still a "fine tune" button, but it purports to use the "whole image" and doesn't seem to do much.</p>

<p>You might also check out NoiseWare from Imagenomic. I was really starting to be impressed by it for film work, but they've yet to do a 64-bit update, so I haven't used it since moving to 64-bit CS4.</p>

<p>Scott</p>

[sergiy_podolyak1]

<p>C. Sharon wrote :"Is a scan done on Heidelberg Tango drum scanner by a professional color lab in Hollywood acceptable to everyone on this thread? Any naysayers?<br /> Here is a link to a test done between Canon 5D, 6x6 MF film and 35mm film. 5D is the winner."<br /> No, nein, НЕТ, НI.<br /> May I ask you: where? As English is not my tongue language and to avoid misunderstanding I will repeat my question again in different wording:<br /> - WHERE is "dizitalie is the winner"? <br /> - Where is it? (© Duke Nukem)<br /> - Really, WHERE?<br /> Look at those photo-pictures:<br /> 1). Dizitalie photo shows leaves made of plastic, not alive. Color is un-natural, artificial, too bright.<br /> 2). Dizitalie leaves' shadows are too bright. It seems that leaves are made of semi-transparent plastic.<br /> 3). Dizitalie vines (body of grape plant) are FLAT, FLAT, FLAT, not rounded. <br /> 4). Digital bricks are not look like bricks, rather like PAINT after nuclear explosion and radiation. <br /> 5). The digital picture looks rather like rendered textures of computer game than picture from real life.<br /> <br /> Is this a weird result of dizitalie click or a masterpiece of documentary photo-art?<br /> <br /> Photos © Ales & Robert Litomisky 2006, La Crescenta, CA<br /> (fair use citation according to Copyright Act of 1976, 17 U.S.C. 107, particularly for the matter of critique or scientic researh or scientific discussion).<P><a href="http://www.ales.litomisky.com/analogversusdigital/5D%20centercrop.jpg"> Click here</a></P>

[sergiy_podolyak1]

<p>Here film scan from 35 mm Nikon:</P><p><a href="http://www.ales.litomisky.com/analogversusdigital/Nikormat%20centercrop.jpg">Click here</a></p>

[James G. Dainis]

<B>Moderator's note: The position on using others' photos on Photo.net is spelled out in the Terms of Use:</B><P>

 

<I>"You agree to upload and post only User Content that you have created yourself." </I><P>

 

<B>I have changed your photos to links. Linking to others' photos is allowed.</B>

[StuartMoxham]

<p>And I guess leaves are made from film grain then.</p>

[mauro_franic]

<p>I would be very nice if people could only post content they created themselves! Thread would be much more educative.</p>

[sergiy_podolyak1]

<p>C. Sharon wrote: "So this thread is still alive. Here is a link to a test whose accompanying text begins with the following<br /> There are loads of tests online comparing slide films with a digital camera. Unfortunately most use a digital scanner to digitize the film, which opens them to the accusation that the scanner is not getting all of the detail from the film. One way to get around this is to examine the film using a microscope, and that is exactly what I have done.<br /> This a comparison between 10 MP Nikon D200 and Velvia 100 without using any scanner at all so the latter is not a factor in image quality. Velvia does not do better."<br /> <br /> http://www.leifgoodwin.co.uk/DigitalVersusFilm/Velvia100-versus-D200.html<br /> <br /> What do you mean saying "film does no do better" ? Do you mean "better" as "close to reality" ? If so, I would like to ask you and your protege on that link several questions:<br /> 1). If that is a "test" where the english word "FLASH" in that text ? Those photos were taken with FILL FLASH (Fig. 1 http://www.leifgoodwin.co.uk/DigitalVersusFilm/_DSC6475%20Test%20Scene%20Edited.jpg). If this is an "expertise" that flash device should be mentioned. In that text of "expertise" it is absent.<br /> If there were no flash, the book "... in 10 minutes" (2-nd from left) should be almost invisible in the shadow of larger book "Framework".<br /> Additionally seems, flash devices on film and dizitalie shots ARE DIFFERENT. How do I know? Because of blicking on top corner of ID card - it is much stronger on dizitalie (Fig. 2).<br /> 2). Why on Fig. 5<br /> http://www.leifgoodwin.co.uk/DigitalVersusFilm/_DSC6475%20digital%20crop%202.jpg<br /> the blue top part of a book "Birds" (sixth from left) IS NOT RED ? If you want to ask me why it should be RED, if it is BLUE, I will tell you BECAUSE there should be a RED REFLECTION from the RED book "Framework 2.0" (third from left).<br /> Where is that red reflection? That reflection should be strong because main lighting is from window on the right. That red reflection is visible on film, is slightly visible on white book "...PRESS", but it is dissapeared on dizitalie shot. Who wiped that red lighting out?<br /> Processor? Or maybe dizitalie sensor?<br /> 3). Why - on dizitalie shot Fig. 5 - the book "... PRESS" (fifth from left) is FLAT ? It is ROUNDED, as seen on film on Fig. 6 and Fig. 7. Who flattened it? Processor or sensor? <br /> Did you just said "Who cares?" <br /> <br /> Photograph copyright Leif Goodwin (fair use ... 17 U.S.C. 107 ... and Bern convention...). <br /> Mauro is right - give us RAW file to research within it, otherwise it is cheap trick.</p>