ben_lazarus

Are you saying that an EXIF-aware image tool won't tell you the lens type and aperture / focal length in plain english for an image shot with a D70? I thought it would.

I'm currently looking at the D70 and 10D and not terribly excited with either, in large part because I really like manually focusing with a split-prism screen. If I do go the D70 route, sign me up as someone who'd love to buy a split-prism screen from you or whomever.

Whoa - I had searched these forums previously looking for information about swapping out the focus screen on the 10D, and I was pretty sure that I'd come up empty. I have an Aria and love it, but I also love working digitally (and don't love scanning), so I've been looking for something I can live with that costs less than a new automobile. If I could put an Aria's focus screen on a 10D, that might be it. Do you have any more information on how this is done, who can do it, and what the effects on AF and metering might be? I assume the warranty is history, but two 10Ds are still cheaper than one 1D Mark II.

 

By the way, I don't have any of the issues the above poster had with Zeiss lenses' build quality. I only have a 50/1.4, but it is the most precise, solid-feeling lens I've ever owned. I bought it used, have knocked it around a little, and it still handles absolutely perfectly, with no "play" or whatever.

Bob -

 

Thanks for the explanation about sensor sizes. I guess that makes sense, if you're talking about sensor per wafer yield, as opposed to defect count. If we're talking CMOS, I don't see why you'd get appreciably more defects per wafer for larger image sensors, given a fixed pixel count(in fact, I'd imagine you'd get fewer), but yes, of course you'll get fewer chips from a given wafer, as the chips get larger, and a given defect means you'll throw away more silicon.

 

If that is what is driving the price difference, though, then I'm not sure how rapidly the price on larger sensors will fall. I'd guess not very, since by now we must be at the shallower end of the curve as far as silicon handling and doping is concerned. The major advancements in CMOS fab have seemed to revolve around lithography and masking improvements resulting in smaller processes - which I don't think will help us get larger image sensors. What do you see as the trend that will drive down the price of larger image sensors? Do you think we'll make some advancements in how we get from unprocessed silicon to wafers ready for lithography? Or, do you think it's just a simple demand problem?

 

As to manual focus dSLRs - I'd have hoped it was clear from my post that I'm not holding my breath by any means - I guess it wasn't.

 

Also, yes, I mentioned material costs. Note though, that I also said "amateur." $10 / roll seems about right, but I'd still have to shoot between 100 and 150 rolls, in this (highly personal and unscientific) comparison, depending on the new/used variable, which equates to a 2-3 year ROI, at the rate I currently shoot. I hope this isn't coming off like a film vs. digital debate; it isn't - all I meant to say is that I still see a big gap in the amateur SLR market between film and digital, and I personally expect to see that gap close quickly - evidence that there should still be major movements to come for dSLR on the price/performance curve.

 

Thanks again for your thoughts.

I really doubt that either sensor size or price (or pixel count, for that matter) are anywhere close to stabilized in the dSLR market. With digital, the upgrade/price curve that we've long associated with computers has now found its way into photography, which to me is both exciting and annoying. In one sense it's fun, because better and better things are always coming along for less and less, but it's also frustrating because it means that nothing you buy will hold a shred of resale value, and for those people who always want to have relatively current technology, it means constant upgrading.

 

Considering that the closest digital EOS match I can find to my Contax Aria and Contax 50/1.4 combo ($925 new - I paid $500 used) is $1920 for a 300D and 35/1.4L, dSLR still has a ways to fall on price/performance before we can say that it's filled the market space film now occupies for amateurs. Note that I would be making several serious personal compromises in the 300D system that I'm not yet prepared to make, either (e.g. no manual focus screens), but I'll ignore that for this comparison, in light of all the digital benefits that we all know about (e.g. no scanning, instant feedback, material costs, etc). Yes, I'm comparing a manual focus film system to an auto-focus digital system, but please save your breath, because that's the point.

 

I wonder if there will ever be a serious digital SLR offering for those of us who still prefer manual focusing - maybe in several years, when digital is the norm, but maybe manual focus will get lost altogether, and we'll be left with LEDs as manual focus aids, and focus rings on lens that are short and light enough for fast AF, but feel horrible for manual focusing.

 

I do hope that Canon and the other vendors don't standardize on different sensor sizes for different target markets in the dSLR lines. The thing I hate about Nikon and Canon's film offerings is that (for example), I have to spend $1000 on a body before I can get a focus screen I like, but the nice thing about them is that you can use the same glass across their product lines. If they standardize on 1.6x for consumers and 1.3x or 1x for "prosumer" or pro, then the lens lines will also diverge, which I don't think is best for the consumer population as a whole.

 

As an aside (and hopefully not too off-topic), does anyone know why it's so difficult/expensive to make a full-frame sensor? It seems that in general, when it comes to semiconductors, SMALLER is harder/pricier, while bigger is usually easier/cheaper. Intuitively, it would seem that putting the same number of pixels onto a bigger sensor would be easier and cheaper. Does anyone know why this isn't the case?

I've been following WINE for almost tens years - it strikes me that the programs it is most capable of running very well are usually a version or two out of date. This has been the case for as long as I can remember. Great tool, but it has an extrememly difficult task to perform - it speaks volumes that it is even capable of doing what it does. Disclaimer: I haven't checked on it lately.

 

As to Linux for photography, I don't know. I've used Linux (as well as other Unix or "Unix-like" operating systems) exclusively in my home for ten years. Meaning, I used ONLY Linux (or OpenBSD, or Solaris, etc). No Windows. What finally brought me to installing XP on a PC a little over a year ago was a need for Photoshop, color management, and support for printers, scanners, digicams, and the ability to burn DVDs. Windows strikes me as the right tool for the job, if you've got an investment in PC hardware, no interest in using Linux for its own sake, and no budget for an Apple.

 

Funny that we are going in different direction - you, having used Windows, and considering switching to Linux to do image work, and me, having used Linux for ten years, and finally installing Windows to do the same thing :)

 

As the guy who alluded to a Windows vs. Linux holy war, I hadn't seen anything of the kind in this thread, until he brought it up.

 

Anyway, in case I wasn't clear above, I suspect (especially if you are command-line-adverse) that XP (or OS X) is probably the right tool for you, at this point. Yes, the distributions get more and more polished, but I've still yet to install one and NOT need to hand-edit XF86Config to get things just right (if you don't know what this means, then just trust that it's something you really, really, don't want to do).

 

By the way, you've got nothing to lose checking out Linux and seeing for yourself. Even if you don't have the disk space to do a full install, there are a variety on small distibutions you can burn onto a CDR, which you can in turn boot off of and play with at your convenience, without committing to anything like buying another disk or deleting Windows. One such distribution is <a href="http://www.knoppix.org">KNOPPIX</a>.

I haven't seen this in my SD IV scans, although I have had <a href="http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=008nAE">other problems</a> with mine, and will be returning it shortly. I would try disabling all processing in the scanner and its software. Just get a raw scan without any color correction, sharpening, compression, "pixel polish," etc, and see if there's any change.

Wow, thanks again for all the comments and suggestions. I guess I'm still dubious about it purely being the film grain itself. I happened to also just receive my new Epson R800, so I did some tests making prints. I can very clearly see (with the naked eye, held close to arm's-length) the grain/aliasing/noise/whatever-it-is effect in even a 4x6" print (with borders) from a E100VS scan. Again, it's most strikingly visible in the smooth-toned areas, but from zooming and paging around in Photoshop, I believe it's consistent across the entire image. The best way I can think of to describe what the effect looks like in print is if anyone remembers what looking through a Pentax K1000 viewfinder was like (probably other dark focus screens are similar, but this is what I'm familiar with) - the focus screen always added a consistent layer of dark 'grit' to the scene.

 

I just can't believe that what I'm seeing is FILM GRAIN, in a 4x6 bordered print held at arm's length, made from a 3200dpi scan of an ISO 100 slide film.

 

I also tried the Neat Image demo, as several suggested, with very good results. I had to bump the noise reduction to 100% in both channels to eliminate the effect, but it did eliminate it. I did an 8x10 print of the image filtered with Neat Image, and another of the unfiltered image, and I cannot discern any loss in sharpness or image quality in the filtered print, while it at the same time, it did completely remove the gritty/grainy effect I've been talking about.

 

So where does that leave me. I still do not really know what it is - if my scanner is slightly malfunctioning, if something else in my setup is contributing to the effect, if it's film grain, grain aliasing, or pure scanner noise. I am still within the return policy window the DS IV, and considering returning it for a Coolscan V. I know that I don't have any assurance that I wouldn't see the exact same effect with the Nikon, but maybe Digital-GEM is another solution to the problem. The price difference between the DS IV and the Coolscan V isn't as much once I've had to buy Neat Image Pro to be able to get satisfactory results from the DS IV. Also, on an unrelated note, I completely underestimated how much of a pain dust control would be. Even scanning these slides and negatives that I received from the pro lab the same day, tediously applying canned air and a camel's-hair brush, while wearing darkroom gloves, I still have a lot of work with healing brush waiting for me to eliminate all the dust. I think the rest of the difference in price may be worth it just for the Digital ICE.

 

I went with the DS-IV, because I was considering this to be a stop-gap solution for a year or two, which is when I expect to invest in a dSLR system, but if it's going to be this much of a pain, it seems almost worth it to spend the extra couple hundred for a Coolscan V (assuming I don't get the same results with that scanner), if not bite the bullet and jump to dSLR now.

 

Thanks once more for all the contributions.

Rick - I don't see the grain/noise/whatever-it-is in your images as I do in mine. I see some jpeg artifacts, but not big chunks of mottled ugliness like I see in mine. Am I being especially critical of my own images, or crazy, or uh.

 

Mendel - Thanks for the suggestions. Are you serious that you can't see the effect I'm trying to describe in the sky of my picture? It is very visible to me, even in the jpeg. I can see it in the jpeg when viewed on both my Trinitron CRT, as well as the dim old LCD on which I'm writing this. As to your favorite suggestion - do you really think this is film grain? I would not expect film grain from a modern ISO 100 slide film to be so strikingly resolved by a $250 3200dpi scanner.

 

As to your other suggestions:

 

1. Experiment with different film - I've used both an ISO 400 C41 film and an ISO 100 E6 film and see the same effect. Do you think I should try more?

 

2. Trying a different scanner - this isn't very easy for me to do, but I'll keep it in mind.

 

3. Trying noise filters - I went to their websites and looked around. Actually, this seems to make the most sense, in terms of what might be causing it. Again, I have no experience with film scanning, so I didn't consider that it might be digital noise, but THIS I might expect from a $250 3200dpi scanner. Thinking along those lines, I did another scan of a 400UC shot, but with the scanner moved far away from the monitor, and 16x multisampling. No change.

 

4. Backing the focus off - tried it; no change.

 

5. No, I can't get used to it :) As to whether it's film grain or noise, I'm attaching another picture, this is a crop at 100% of a 3200dpi scan, this time of 400UC. It's some twilight blue sky, and part of a pile of some type of crushed stone. Maybe someone with more experience than me can call this out for what it is.

 

Thanks again for the suggestions.

I just received my new Minolta DS IV, and I'm seeing some really

disappointing results. I've scanned about a half dozen E100VS and

Portra 400UC shots, all of which were taken with a Contax Zeiss 50mm

1.4. All the scans were at 3200ppi, and I've tried a variety of

options, including 8bit, 16bit, 16bit linear, with and without

multisampling, with and without "digital grain dissolver," with and

without "auto dust brush," and with and without "pixel polish." I've

also tried a couple of scans with Vuescan.

 

All the slides and negatives were processed at a pro lab, put into

archival pages immediately, and removed only long enough to scan them.

 

In all of them, I am seeing what looks like huge, prominent grain -

especially in smooth-toned areas like blue sky. Sizing down reduces

the effect of course, but it's still visible, even sizing down to 4x6"

300dpi. Did I get a lemon? Is it just a serious coating of dust? I

do not see this effect on the light table under my makeshift loupe (an

inverted Pentax 50mm). Am I doing something wrong? Expecting

something I shouldn't be expecting? This is my first film scanner, so

I have nothing to compare these results to, but they are not at all

what I was expecting. Any comments or suggestions will be really

appreciated.

 

I'm going to try to attach an example. It's E100VS, scanned at 16bit

with no scanner processing or multisampling at all enabled. It's been

cropped down to the film frame size, and I've manually removed dust

with Photoshop healing brush, but no other adjustments have been made,

save converting to 8bit, and making it into a photo.net-friendly jpeg.<div></div>

I also like this film (Portra 400UC) a lot. My guess is that Kodak removed it from the "Portra" line and created the "Ultra Color" line purely as a marketing/branding move, because it seems pretty clear that they are trying to market the Portra (NC & VC) line as (duh) portrait films, with an emphasis on "natural skin tone." 400UC may be a little too wild to fit that bill, and so they created the "Ultra Color" brand to cover it, because obviously they have no interest in discontinuing it, but at the same time, want to protect the "Portra" name.

I am new to the ideas of color management, but I've been reading Michael Kieran's book _Photoshop Color Correction_ and I'm a little confused by some of Scott's comments on profiling and color management (was also confused by some of Kieran's). That's my disclaimer, so bear with me, please.

 

First, what does "open loop" mean? Also, isn't the point of profiling and calibrating all of your input and output devices so that you can achieve a more predictable and flexible version of your approach, Scott? Meaning, you've hand/eye calibrated your monitor to the point that you feel it matches Frontier output. But, if you'd calibrated your monitor to a known reference, then profiled it, and if you also had a good profile for the Frontier/paper combination you're printing to, wouldn't you accomplish the same with a quick View->Proof Colors? This way, your Lord of the Rings and Halo are unaffected, and you can just as easily soft-proof any printer/paper combination for which you have a profile. If you find yourself using another monitor for whatever reason, then no worries, as long as it's calibrated and profiled. Am I understanding correctly?

 

Calibrating your monitor (whichever way it's done) will not make its gamut identical to the Frontier's, but aren't you in a way doing all of your corrections directly to the Frontier? I thought another idea behind profiling everything was that you could (with a profiled input device, and a calibrated and profiled monitor) do all of your corrections in an output-device independent colorspace, and then save off a 'master' corrected file, so that days or years from now, if you want to output to another device entirely with a completely different gamut, you don't need to do all of your retouching and correction work again.

 

Lastly, don't you need to calibrate and profile your monitor to compenstate for what Kieran calls "nonlinear gray reproduction of the monitor," or is this somehow taken care of in the way you adjust your monitor's setting to match a Frontier print?

 

Hope I didn't derail this thread into a general help-the-newbie-understand-color-management thing, but it seemed like a good time to ask. Thanks,

 

Ben

I know this is an old thread, but I wanted to post for two reasons; first, to thank the poster above who provided the Fuji Frontier magic decoder ring. That is very useful information. Second, to weigh in with a test of sorts I just conducted.

 

I took three rolls of Porta 400UC to a local pro lab (Procolor on Hennepin, for those in Minneapolis), which has a Frontier and prints on Crystal Archive, and got 4x6 prints. I then took the same negatives to a Ritz (in IDS Center, relocated from City Center) that has a Frontier and prints on the same paper, and ordered 4x6 reprints of about 35 of the frames.

 

I casually compared the prints, examining each side-by-side for no more than a minute each. The prints are the same size, from the same model machine, on the same paper, and made from the same negatives. Objectivity ends there, of course, because this test is nothing more than determing which prints I subjectively liked more.

 

For many (15 or so) of the prints, there is no real difference that I can see, but for several of them, they may as well be different photographs. In all cases where there is a difference, the prints from the pro lab are very much better, where "better" is defined as above - I liked them more, for reasons that vary which each comparison, but almost always, a color present in one print that I feel is missing from the other. Or, a vividness (sorry for the limited vocabulary) that is missing in one, but punches you in the face in the other, as I remember it having punched me in the face in the original scene.

 

I then (after reading this thread), compared the codes on the back. All of the Ritz prints are (unsurprisingly) "NNN 0" - a handful of the prints from the pro lab (5 or 6 of the 35) were tweaked, but there are many prints that look much better from the pro lab that are also "NNN 0"

 

I should also mention that the majority of the handful that were tweaked by the pro lab, I feel I had overexposed these by a full stop or more. These are my first rolls back into photography after ten years away, and with a new camera, so in some of the frames, both labs were working with flawed source material.

 

Draw your own conclusions. Mine is what I guess we all already knew: it takes more than a certain machine and certain paper to make good prints, and 'garbage in' will by default result in 'garbage out', but with some work, it doesn't always have to.

 

P.S. I also got scans from Ritz, and I have never in my life seen such shoddy work. The scans are *filthy*, with enormous pieces of dust and hair and god knows what scarring every single frame, and all kinds of other problems that I have neither the technical vocabulary nor time to describe. Remember that these are pristine negatives never out of their sleeves, and only 24 hours from the lab that processed them. I have to go back and check them now to make sure they weren't permanently damaged.