Digital capture vs scanning confusion (Know nothing)
Discussion
24 replies
Posted by David Redding on August 12, 2006 at 7:49 PM
It seems that no matter were you go on the web to look for research material on the subject of digital capture vs film scanning, you get tons of anwsers and no one tends to agree on anything.
I was just reading Mark Windoms thread asking for advice of weather to get a 5D or a 645 and scan them with his Nikon9000 scanner. Everyone that anwsered the thread was saying that the 5Ds capture was superior to a scanned 645. That started me searching for no other reason then curiosity, I found that the 5Ds Max resolution is 4368 x 2912. Well the Nikon9000 has a scanning resolution of 4000PPI, so a scan of a 645 negative would give you a file of 24000x18000. Now like was stated in the title of this thread, I'm a still learning know nothing, so my understanding of scanning and resolution is very limited, so I may have this all wrong. But, to back myself up a little more I did a search of film vs digi and found that the list of "What megapixel = what format film" tended to be about the same from site to site. They were saying that you would need a 16MP=35mm Velvia, 51MP=645 Velvia, 240MP=4x5 Velvia. Now all these numbers were takin from the an article available @ http://www.clarkvision.com/imagedetail/film.vs.digital.1.html
Now I know this doesnt take into account the conveniencies of digi and not having to waste time on scanning, this is more a pure quality and enlargement possibilities thing. I am well aware that for the pro that is looking to sell his files to magazines, stock, calenders and so forth, Time+quality wins over pure quality all the time.
Any input into the learning process is highly appreciated
The first thing to keep in mind is that image quality is not just about detail. On a similar note - can you tell the difference between music from a (digital) MP3 file or a (analog) cassette tape? The MP3 file holds less audio detail than the tape, but the sound will usually be more "pure". The tape will likely have hissing sounds and other artifacts that are just part of the medium. By the same token - the 5D capture may be a little less detailed, but the capture is much cleaner and sharper and free of film grain and other artifacts.
Even the finest grained film will begin to show grain when scanned beyond 2000-2400dpi. At 4000dpi you will see very obvious grain. You will also note significant diminishing returns - cranking up the resolution will gradually yield less and less detail, and more and more grain.
Add to that the dynamic range of digital capture which generally exceeds that of film and you have another component of "quality" working in favor of the 5D.
Finally you may want to compare lens resolution for the two configurations you are comparing. The amount of detail is not only a result of the capture medium or size but also of the lens. A lens that can resolve 80lpm on 35mm will be able to capture more detail than a lens resolving say 30lpm on 645.
I'm sure others may jump in with more considerations. Ultimately the conclusion is that finding a ubiquitous set of metrics to allow a 1-to-1 comparison between these different technologies is practically impossible. It comes down to subjective preference, and I would go further to say the comparison should be between the final prints rather than the file. Artifacts in the file will end up with different looking prints, for better or worse.
Guy
Posted by David Redding on August 12, 2006 at 8:51 PM
Thanks for your response Guy, it just gets frustrating when you hear one person say one thing and another say that that last guy doesnt know what he's talking about and starts telling you something completely different.
On the Audio tape vs CD....well yeah I can tell the difference and to me tape (better yet LP) sounds way better.....well more natural, more like the artist is sitting right in front of you.........But that discussion doesnt belong in these forums now does it
As for dynamic range, thats another thing Im getting conflicting answers on. Now in your article in this months ODP, you were mentioning the latitude of the different mediums and I noticed that you put APS-C right up there with colour negitives and Mr Greer mentioned that the D-range of digi beating out film as well on these forums............but then I go to another forum that Im apart of and some people there are saying that digi cant touch certain films when it comes to Dynamic range.
Its just so much information and unless you have had years behind you or a degree in computer science, optical physics and chemistry......its hard to catch up with just reading web articles.
David, you will note I qualified my statement on the 5D as my personal opinion and in the response above stated that the comparison is subjective.
My advice is - try both, process both, print both, and decide for yourself. It will save you a lot of frustration.
Guy
Posted by David Redding on August 12, 2006 at 10:26 PM
I have recieved that advice alot guy........and I plan on taking it to heart......I would guess that its what people prefer......but what Im really looking for is actual information and facts. I am aware that if I only plan on enlarging to max 8x10 then I could get a 8 MP and go nuts not having to worry about MF, LF the digi vs film or anything......but its still good info to know for infos sake.
Its kind of along the lines of am everyday car geek wanting to learn about indie cars.......He will never have a need for the engine alloys of what ever else they use under the hood.....Its just intresting information and facts
Posted by Jerry Greer on August 13, 2006 at 8:20 AM
David,
Test have been done on dynamic rage from the 5D and it has over 7 stops range without any photoshop work, with PS work you can get more that 12 stops in some cases witout using HDR. The only thing that comes close is print film. As for the native resolution, one thing the film guys don't tell you is that you can rez-up a super clean digi file more that 250% without noticable degradation. I've went as much as 400% from image taken with medium focal length lenses. It's all about knowing how to properly work a digi file from the inital capture sharpen to the specific output sharpen.
Just a note, this is the first year that I've had digital images in my calendar. They are mixed with 4 large-format images. I've only had one guy to find the large-format images but he cheated because he knew which were LF from my website. Most pick the digital images as the LF images. This calendar is a 12"x12" wall calendar.
Posted by David Redding on August 13, 2006 at 11:42 PM
Thanks for the info Jerry, I looked up the D-range on that 5D and I found numbers going as high as 8.9 stops at ISO100. The funny thing about the testing where I found the 8.9 was that as the ISO went up, the D-range went down....I would have thought that wide the added light sensitivity that the D-range would have gone up not down. Maybe it starts having problem with the right side of the histogram after you up the ISO
I'll have to read up on this "Rez-up" you are referring to. Is it basically enlargements by interpolation? or something different?
First of all, I'm not sure it makes any sense to compare these cameras, they are totally different beasts. Anyway for the sake of argument I will play along! ;)
Foremost, if scanned properly, a 645 film does indeed capture enormous detail. Unfortunately the web is full of archaic articles comparing digital to scanned film, and most of those are based on last-generation scanners and (frankly) some really awful technique. For example, some people apparently don't realize that if one's intention is to scan, that actually affects how the shot is taken; you know you will be limited by the Dmax of your scanner so you have to adjust exposure a bit. I think that a properly exposed 645 negative will easily, easily compete with a shot from a 5D and will even give a 1Ds a run for its money. But the bottom line is that there is scanning and then there is scanning... it is a skill. And if I ever feel limited by a scan, then I print traditionally, no big deal.
Grain is a consideration, but people should realize that in digital capture there is all manner of noise processing, and there are many noise sources not even present in film, and the camera simply trades detail for lower noise, that's it. Guess what, you can do the exact same thing with a scanned file. A 645 shot will easily generate ~30mp, and if I take that and downsample to 10mp it will look just as clean and glorious as a 5D image. This is well known to those of us who work with medium and large format. MF and LF shooters routinely downsample. Besides all that, noise is kind of anathema in film, we all know that we can shoot ISO 25 or whatever. No big deal. ISO 8? Fine.
On the topic of dynamic range, this is something that digital users really should complain about! Look, there is only one camera on the market right now, the S3, which comes close to how forgiving print film is in the highlights. On the film side, there are many different films with different tone curves, we pick what we want for a certain shot. I simply don't worry about blowing highlights, not even with slide film. Again, proper exposure is the key. It's true in digital and it is true in film. I would urge people to be extremely skeptical of the "this versus that" threads online, because you never know how good the exposure was in the first place. If you really want to compare the capabilities of two different formats, then I am afraid you will need to look at giant enlargements, in person, made by people who really know what they're doing.
P.S. I take extreme exception to the term "rez-up", which I see on all kinds of web forums. Actually there is no such thing as rezzing up! There is interpolation and there is interpolation, that's all. Interpolation does not create information, there is nothing going up except the apparent smoothness of the image. Call it depixelation if you must.
Welcome to NPN, Keith. It sounds like you're an experienced film photographer and skillful at working with scans, but your comments on digital capture would indicate you have not actually used or worked with digital cameras to the same extent.
As I mention above - the proof is in the pudding. We can do all the math in the world but ultimately the prints will be our final judgement. I suggest you don't become too entrenched in your views until you have had a chance to do an objective comparison of prints.
I am no stranger to MF and LF myself. I have worked extensively in all these mediums (and in fact teach workshops on the subject). You can trust me when I say I used to be as skeptical as anyone about the digital "hype", but I was not satisfied with inflamatory statements on photo forums - I decided to give both my best effort and can tell you that even as a die-hard LF user for years, digital capture exceeded my expectations. I still use both.
As an aside - you did not provide a last name, email contact, or web site where your work can be seen. I'll be curious to see some of it.
:) I have two Peltier-cooled CCDs in my semiconductor physics research lab with which I do single-photon counting. My investment in digital technology is, oh, 15 years and counting ;)
Okay, credential-wanking aside... that stuff has nothing to do with real world photography and I agree with Guy that people really need to get out and experiment and not rely on web chatter.
Admittedly, I do not presently use the "high end" digital SLRs, and that is because I like: tilt and shift, rotatable and interchangeable backs, and I especially like shooting b&w. I am also an avid lover of rangefinders, of which there is really only one 6mp entrant on the digital side and it is not at all satisfying. So my feet remain firmly planted in film and for good reasons. I have made several forays into digital over the years and as yet the one camera that lures me is the fuji S3, but unfortunately it lacks the detail that I want. But I may well replace my F100 with one of those soon. I could happily do sports & wildlife on digital. I would love to see something like a 5D with the DR of the S3, that would probably make me forsake film below 6x9 format.
Compared to medium format, I will admit that there are digitals out there that could satisfy me, but not for under, what, $10 or $20k? Therein lies the problem. I am not professional, my day job keeps me far too busy for that. When I am bored with writing papers mayeb I wil retire to photography, who knows. So it's either the car or the digital back, and sorry, I have to go with the car ;) maybe in the future... just a poor physics professor...
I don't have a gallery and I don't intend to sell stuff any time soon, I am far too busy running my lab to get into the biz and I don't want to do a half-assed job. So if you want to see some shots then I'll have to refer you to a few that I posted for discussion purposes:
I guess I must have some 645s somewhere that would be more relevant to this thread, I could hunt for those. But I stopped using 645 about a year ago, now I use mostly 66 through 69.
Very nice gallery, Guy! and thanks for the welcome. I don't guess I'll be hanging around here much, I am already very involved in two other forums and that's alreayd taking too much of my time. I got into this thread per the request of Dave.
Subject matter and print size are sincere factors.
While there have been all kinds of bonuses in field and location shooting, we based our decision to switch from scanned 645 film (NPS160 print) to the "only" 6mp Fuji S2 based upon needs and performance in our portrait studio. It's less an issue with smaller prints, but when you get to 20x30 inch and larger prints where we really start to make money, the trade-off is between an apparent increase in grain with scanned film and loss of detail with digital capture.
In portrait shooting the deck is sharply stacked in favor of digital for its much smoother skin tones due both to a little less detail and no grain. There is also such a thing as "too sharp" in portraits, and we often had to resort to diffusion filters on capture and sometimes even gaussian blurs in post processing with scanned film. From a purely pragmatic standpoint, digital capture on Fuji's 6mp sensor gave us the results we wanted right out of the camera without all the fluff, bother and expense of film. Now, shooting the 12mp Nikon D2X allows us to extend those benefits through much larger print sizes. Combine the extra income from larger prints with the ongoing savings in film and processing (even after factoring in the cost of new DSLRs), and we would only go back to 645 when digital backs drop down to prices competitive with high-end DSLRs.
Posted by Jerry Greer on August 14, 2006 at 6:33 PM
Keith,
The ONLY scan from a 645 that will compare to an image taken from a 5D or 1Ds Mark II will have to come from one of the latest models of drum scanners. For that you will need about $40K to $50K to purchase if you want one at home/office. With that in mind $8K just doesn’t sound that bad. For a desktop scanner, there's just not one that can compare except maybe from a sharpness standpoint but not for dynamic range. Again, as I stated in my first post, only print film is going to match or exceed the dynamic range of the 5D. But, one has to ask oneself, what am I going to shoot the most, color transparency film or print film. Most in our industry shoot transparency film when shooting color so it’s a no-brainer!
As for rezing up, that's a figure of speech that we use and again digital images can be rezed-up (or interpolated if you take extreme offence to it) to as much as 250% to 400% without very much degradation when done in the RAW converter. Something that CANNOT be done with a scanned transparency or negative, it's just a different animal! A scanned image file can’t be compared to a RAW image file from digital capture.
Then there's the noise issue that you pointed out. Well, I’m happy to tell you that with the 5D there's no noise issue up to about 400 ISO. My images from the 5D are as clean at 400 ISO as my 4x5 images on Velvia 50 drum scanned on the latest and greatest ICG drum scanner. Again, we are talking about the 5D not the D2X! So, I shoot at 50, 25 or 8 ISO with, oh maybe, 2 or 3 stops of extra time needed for reciprocity failure just to get an image that’s almost as clean as the image I can take with my 5D. Hummm, I hope the wind doesn't even think about blowing from about 2 to 5 minutes with that film camera.
And to your statement about giant enlargements for a true comparison, I’m sorry but what a load of crap! Size doesn’t matter at all unless all you do is sales of 30” x 40” and larger prints. I really don’t know about you guys but how many of you make a living selling 30” x 40” and larger prints? 98% of my print sales come from 20x30’s and 16x24’s. But print sales are only about 5% of my total income from photography, workshops about 5% and publishing is the remaining 90%!
BTW, just as Guy, I’m also not at all lost when playing around with those big film cameras. Up until 19 months ago I shot exclusively large and medium format color transparency film. Over five years with the 4x5 (6x7cm with a rollfilm back), about a year with 6 x 4.5cm and about 10 years shooting 35mm film. I still spend a lot of time scanning film and now processing digital files for the book and calendar projects that I produce through my publishing company, Mountain Trail Press so I’m not lost when working on a scanner or playing with PhotoShop.
You know, I can remember when I sounded just as sure about the film vers. Digital thing as you do. Man, I’m still eating crow from all my wild statements!
Look, if 10 mp of detail is all you need, then I am happy for you, that is very convenient and inexpensive. I am more than happy to go to the drummer when I get a keeper, and in a year my drum budget is a very small fraction of the cost of a 1DsMk2, even if you throw in a couple hundred 120 rolls of velvia and processing chems. By the way, last time I checked, a brand new drum scanner with Dmax 4.3 costs about $2000 less than a 1DsMk2. For reference velvia has a Dmax of at least 3.8, probably higher in the green channel. You can't scan it on some flatbed and then complain about noise. If you do, I will go borrow a 1Ds and shoot it through a one-stop ND and complain about loss of detail in the shadows just to annoy you.
You want low noise from digital, I want high detail from medium and large format, gee why don't we compromise on a lovely $30,000 H2d? Hmm on second thought I think I'll pass for a while. For a second all this idle chatter made me forget what like most about medium and large format: movements, true b&w, ground glass composition, leaf shutters, rotatable and interchangeable backs, and the priceless satisfaction of knowing that I don't use the same quickly obsoleting plastic box that millions of others use.
As they say in Russia, kazhdi drochit' kak on khochit'.... everybody pleases himself the way he wants (polite translation)....
Posted by Jerry Greer on August 14, 2006 at 10:40 PM
You da man Keith! Us people that are in da bidness of photography are just idiots are we not? Too funny! BTW, the 5D is 12.8MP! Maybe you should stick with the physics gig where attitudes seem to rule.
Again, NPN is a big family and this site is not like the others. Attitudes only flame here every once in a very small while. Also, most of us are pretty easy going folk too. Sometimes we get a knot in our drawers and are short fused but most of the time we’re pretty level headed (I've been known to get mine in a knot every once in a while but I'm getting better, maybe it's the new thong that MG sent me for my birthday sorry Gal's!). So when you pop in I'd say that you shouldn't try to treat all us old guys like a bunch of stupid photographers because lots of us have been doing this a while and we somewhat know what we're doing. This lifestyle of nature photography isn't rocket science or doesn't take a PhD in physics. What it takes is time to learn the craft, a creative mind to design the most pleasing comp, and the personal drive to brave the elements and be in that perfect location when the right light happens (just ask Marc Adamus how its done). To make a business out of the love of photography, well you just need to take it one bigger step.
BTW, $6,000 for a drum scanner well you must be talking about a low-end Imacon, that's a want-a-be drum scanner. A new high-end Imacon will set you back about $20,000. Again, these scanners are really not drum scanners. Hummm, Maybe you should check out a real drum scanner like an ICG. Just a note, a new “real” drum scanner costs on the low end about $19,000 for a 4000 dpi unit and on the high end about $55,000 for an 11,000+ dpi unit. This usually does not include the drums, $1500 - $2500 each, drum mounting stations $1500 - $3000, or software $1500 - $4000. Oh, and the computer big enough to handle it.
NOTE: Today’s price for a new ICG 380, 12000 dpi optical quality resolution, is $43,900 USD brand new and then you must buy all the goodies to go with it. Those, like stated above, will add another $5,000+.
A used Howtek 4000 dpi unit will set you back about $10,000 plus the $4,000 for the extras needed to make your first scan. For a new one you’ll need to more that double that.
Like I said we’re not a bunch of stupid photog’s so lighten up on da family Keith.
Posted by Jerry Greer on August 14, 2006 at 10:57 PM
Oh, and I forgot to mention. I have tilt with rise and fall on my 24mm and tilt/shift on my 45mm and 90mm lenses. So, yes Canon has movements with the 5D using the TS-E lenses.
Posted by Bob Steele on August 14, 2006 at 11:25 PM
I've deleted Keith's last post as it was a little too inflammatory. Let's play nice, fellas.
Posted by Jerry Greer on August 14, 2006 at 11:32 PM
Thanks for policing the thread Bob! Glad you are taking care of da family No seriously...!
Posted by Jim Erhardt on August 15, 2006 at 6:44 AM
Thanks Bob! "Keith" needs to get the feel for the NPN community (ie, this isn't photo.net).
Tell you what. Get the camera and go out. Make images and be happy. I will be happy to get images like GT,JG,TC,MA or TK on any damn format or 2-3 MP cameras instead of wasting time on digital v/s film. So Mr.GT,JG,TC,MA and TK, please please send me your vision so that I can atleast make one fine image in the Alps-outing.
Well actually I don't think my post was any more inflammatory than his. I did not use the word "idiot" for example. Nevertheless I do think that this discussion could benefit by a little more cool-headedness and I will proceed in that spirit.
Yes, I mistyped the res of the 5D as 10mp, big deal. The lateral res of the sensor goes as the square root of the mp count so you know what, my point is rock solid in spite of my silly typo. And you can search around in other forums telling people that I think the 5D is the best 35mm DSLR on the market right now, so kindly put my comments in context.
Foremost there is an implication in Greer's message that I don't' know what I am talking about. Au contraire, I make my living working on improving digital sensor resolution and signal to noise ratio. I am very pro technology and my life is invested in it. So don't dismiss my thoughts because my day job keeps me to busy to have a storefront selling prints. On the other hand, if you do want to dismiss what I say, then go right ahead, no loss for me. I have absolutely no interest in the business side, I took a vow of poverty to go into academia and that is my decision.
I have gotten *spectacular* results from drums on velvia, even mere 120 film, not even large format. These things enlarge to mural size, no doubt about that. If you badly want to see a 645 velvia scan I will go get one for you. But you really don't need to suffer my amateur photography to get the point; this is a simple issue of Dmax. When someone says that scanned velvia is comprable to a 5D at IS 400, I am sorry, that is false on so many levels. That tells me that the scanning technique is entirely incorrect. Now, I don't profess to be a brilliant scanner or even own a great scanner, I go to shop and I pay my $40 and I get my 50mp scan. And the boys at the shop certainly do not have a Tango. If someone is getting a lot of noise from 4x5 velvia scans then they are not scanning on a high enough Dmax instrument and probably not doing enough passes, it's that simple. Don't believe me? Scan provia 100 and compare. Provia will scan very nicely on a flatbed, and it should, it has far less information in the shadows than velvia. Also as I have said, there is an art to exposing for a scan, it is not the same as exposing for a trad'l print, not by a long stretch. The scanner cannot deliver 100% of what velvia has to offer, something has to give.
Now I will assume that there are people who are actually willing to learn something (as opposed to merely blowing smoke), so here are some reasonably good links that compare 4x5 to a ~39mp back. *That* is a perhaps a fair comparison, since people who shoot velvia are no longer likely to do the whole print process themselves, traditionally. But gues what, the results are in, the scanned velvia may be comparable to 39mp back, which wil run you, well you know very well how expensive they are, not to mention the attending gerathat is needed. Suffice it to say you are probably not going to stuff a 39mp back into your pack and go kayaking through the grand canyon. So then what, a scanning back? Please. Things move.
There was another review at LL that compared 4x5 to a 1Ds MK2, and t features some really overreaching claims and give ssome photos that clearly show the superiority of the 4x5 image, and yet the claim is made that the digital is up to par with the 4x5 velvia. I think critical minds can read through this stuff and draw their own conclusions. Yes, you will need to understand logarithms, but it really isn't rocket science. Incorrectly scanning velvia is like sipping champagne from a used styrofoam cup. Don't do it, it is blasphemy. My experience is that nothing less than a drum scan can possibly do velvia justice. (and for those who haven't experienced this, the stated DMax on most scanners is an extreme overstatement; when you realize what Dmax really means, if you get say 3.2 when you thought you were getting 3.8 then with velvia that is a freaking disaster.)
As I said before, this is not merely about pixel counts, to dwell on that rot is to miss the point abut why people use different tools for different tasks. It is about overall capabilities, and for me, an SLR doesn't cut it, even with a zoerk adapter. Been there done that, no thank you sir, t&s are just a few of the things on my long list of requirements. And I also have a good friend who went the TSE route on a Canon, he absolutely went nuts with it and even modified it to extend the motion. Couldn''t get the guy to shut about the TSE. Guess what, a month later he tried and bought a 4x5 and is now giddy with the amount of creativity it permits him. He hasn't shot digital since. You know, people need to learn for themselves and as they say in Russia, kazhdi drochit, kak on khochet....
With that I leave you.
Keith
Posted by Jerry Greer on August 15, 2006 at 5:44 PM
Thank goodness! Bye Bye! Now we can get back to the great NPN group that we are! Thanks Jim!
Posted by David Redding on August 15, 2006 at 9:39 PM
Well I learned alot
Im really didnt mean for this hummbug to happen, all I wanted (like I stated) was some information on some of the things that I have been reading for the sake of knowledge. I wasn't looking for suggestions on what I should be shooting or if I should shoot digital or film. I have already decided to do as Mr Tal had suggest (and what Keith has suggested to me on many occasion in the past) to shoot what intrests me and find out what best suits my needs (currently looking into 8x10 pinhole )
I guess I should have figured it out before hand that a thread that compares film to digital in any way can go no where but bad.....esspecaily when you put two people like Mr Greer and Keith in the same thread
Posted by Jerry Greer on August 15, 2006 at 10:31 PM
David,
First of all, the way I see it you knew and wanted Keith to stir up the pot here on NPN. He's your friend and you knew that he'd spew his rhetoric just as he did. I'm sure he's already preached the digital/film sermon to you many times before. That's why you asked him "to chime in" to this thread. But, to put me in the same league with him I don't think so. My problem with this whole post is that a few of us, like Guy Tal, have been shooting for years with 35mm, 6x4.5cm, 6x7cm and 4x5in. We are well versed in scanning and now with digital capture. Maybe you read Guy's article in OP, if not you should. We've been through the whole film vers. digital thing more than once and it wasn't that long ago that I defended film as though my life depended on it. So we do know a little about what we do and that knowledge came from experience first hand and not from reading articles in print or on the web. Just as Guy noted, he finds that digital capture with the 5D works for him and he stated why and I did the same. Remember, we both were shooting large-format professionally just a little over a year ago. Guy still uses both but I don't. We're not here to force you to use digital we want you to know why we love to shoot the 5D and why we think that it's much more superior in quality than 6x4.5cm and that's all. You can take our advice as you wish. Whether you believe it our not or even care we are here to answer your questions and concerns to the best of OUR knowledge on a specific subject.
Best always,
Posted by David Redding on August 15, 2006 at 11:16 PM
Well, I am sorry Jerry, but how could you know what my intentions were unless you know me......and concidering that I have only been here for a short while, you don't.....So to make a statement like that is more then uncalled for, but I let thing like that pass. My actual request to keith was to read the thread, then get back to me (In a private messege or email), if he decided to chime in, it was his decision and not upon my request.
The reason that I asked him to look at the thread was because I wanted to get his opinions on my questions and opinions of others (instead of me relaying them and more then likely getting them mixed up)
Actually I have read a few of Guys article (OP and Photolife and here), I also know that you have been working with LF as well for a long time.........I have seen and enjoyed not only yours and Tals work but a lot of others on this site, so I thought this would be a good place to start to get my answers.
Now on to the testing and experience.......You have hit the nail on the head there. I have very little experience having only started this hobby of my Last December, so forgive me if I havent had the time and money to try out every known format in the world of photography. I do plan on trying out as many formats as possible (like I have stated 3 times already) but in the mean time.......If I want to get information on one vs the other......I have to rely on others experience. But even in the format wars (well not really a war) there are plan facts, one film handles colours better then the other, digital must handle D-Range better then this film or that film or vice versa. You would think that this information would be easy to come by, but it seems its not..........and that is the reason I originally post this thread.
I guess I'll just have to mark this up as a taboo subject, and move on....I guess if I read enough article and wait to try out different thing later on I'll be able to figure it out myself
Posted by Jerry Greer on August 16, 2006 at 6:58 AM
David,
This is a direct quote from Keith's first post "Alright, Dave asked me to chime in so I will!". So, with this quote I made my assumption that led me to my last post. After rereading Keith’s rants I can see where he would use you as a springboard to jump into this post and begin his sermon. If that is indeed what happened I apologize.
Again, Guy and I are not offering our experiences to you on assumptions. We offered our experiences from true hands on experiences that we both have had by spending months and years with each format. I'm certain, knowing Guy for some time now, that he wasn't trying to conform you or anyone else and I sure wasn't either.
Please spend some quality time here at NPN and you will see that we are not the same old forum like photo.net. We are a true family and everyone here is willing to help by offering up true life experiences in about any genre of photography that you can think of.
Discussion
24 repliesI was just reading Mark Windoms thread asking for advice of weather to get a 5D or a 645 and scan them with his Nikon9000 scanner. Everyone that anwsered the thread was saying that the 5Ds capture was superior to a scanned 645. That started me searching for no other reason then curiosity, I found that the 5Ds Max resolution is 4368 x 2912. Well the Nikon9000 has a scanning resolution of 4000PPI, so a scan of a 645 negative would give you a file of 24000x18000. Now like was stated in the title of this thread, I'm a still learning know nothing, so my understanding of scanning and resolution is very limited, so I may have this all wrong. But, to back myself up a little more I did a search of film vs digi and found that the list of "What megapixel = what format film" tended to be about the same from site to site. They were saying that you would need a 16MP=35mm Velvia, 51MP=645 Velvia, 240MP=4x5 Velvia. Now all these numbers were takin from the an article available @ http://www.clarkvision.com/imagedetail/film.vs.digital.1.html
Now I know this doesnt take into account the conveniencies of digi and not having to waste time on scanning, this is more a pure quality and enlargement possibilities thing. I am well aware that for the pro that is looking to sell his files to magazines, stock, calenders and so forth, Time+quality wins over pure quality all the time.
Any input into the learning process is highly appreciated
Cheers,
David
The first thing to keep in mind is that image quality is not just about detail. On a similar note - can you tell the difference between music from a (digital) MP3 file or a (analog) cassette tape? The MP3 file holds less audio detail than the tape, but the sound will usually be more "pure". The tape will likely have hissing sounds and other artifacts that are just part of the medium. By the same token - the 5D capture may be a little less detailed, but the capture is much cleaner and sharper and free of film grain and other artifacts.
Even the finest grained film will begin to show grain when scanned beyond 2000-2400dpi. At 4000dpi you will see very obvious grain. You will also note significant diminishing returns - cranking up the resolution will gradually yield less and less detail, and more and more grain.
Add to that the dynamic range of digital capture which generally exceeds that of film and you have another component of "quality" working in favor of the 5D.
Finally you may want to compare lens resolution for the two configurations you are comparing. The amount of detail is not only a result of the capture medium or size but also of the lens. A lens that can resolve 80lpm on 35mm will be able to capture more detail than a lens resolving say 30lpm on 645.
I'm sure others may jump in with more considerations. Ultimately the conclusion is that finding a ubiquitous set of metrics to allow a 1-to-1 comparison between these different technologies is practically impossible. It comes down to subjective preference, and I would go further to say the comparison should be between the final prints rather than the file. Artifacts in the file will end up with different looking prints, for better or worse.
Guy
On the Audio tape vs CD....well yeah I can tell the difference and to me tape (better yet LP) sounds way better.....well more natural, more like the artist is sitting right in front of you.........But that discussion doesnt belong in these forums now does it
As for dynamic range, thats another thing Im getting conflicting answers on. Now in your article in this months ODP, you were mentioning the latitude of the different mediums and I noticed that you put APS-C right up there with colour negitives and Mr Greer mentioned that the D-range of digi beating out film as well on these forums............but then I go to another forum that Im apart of and some people there are saying that digi cant touch certain films when it comes to Dynamic range.
Its just so much information and unless you have had years behind you or a degree in computer science, optical physics and chemistry......its hard to catch up with just reading web articles.
I guess this could almost be a rant
My advice is - try both, process both, print both, and decide for yourself. It will save you a lot of frustration.
Guy
Its kind of along the lines of am everyday car geek wanting to learn about indie cars.......He will never have a need for the engine alloys of what ever else they use under the hood.....Its just intresting information and facts
Test have been done on dynamic rage from the 5D and it has over 7 stops range without any photoshop work, with PS work you can get more that 12 stops in some cases witout using HDR. The only thing that comes close is print film. As for the native resolution, one thing the film guys don't tell you is that you can rez-up a super clean digi file more that 250% without noticable degradation. I've went as much as 400% from image taken with medium focal length lenses. It's all about knowing how to properly work a digi file from the inital capture sharpen to the specific output sharpen.
Just a note, this is the first year that I've had digital images in my calendar. They are mixed with 4 large-format images. I've only had one guy to find the large-format images but he cheated because he knew which were LF from my website. Most pick the digital images as the LF images. This calendar is a 12"x12" wall calendar.
I'll have to read up on this "Rez-up" you are referring to. Is it basically enlargements by interpolation? or something different?
First of all, I'm not sure it makes any sense to compare these cameras, they are totally different beasts. Anyway for the sake of argument I will play along! ;)
Foremost, if scanned properly, a 645 film does indeed capture enormous detail. Unfortunately the web is full of archaic articles comparing digital to scanned film, and most of those are based on last-generation scanners and (frankly) some really awful technique. For example, some people apparently don't realize that if one's intention is to scan, that actually affects how the shot is taken; you know you will be limited by the Dmax of your scanner so you have to adjust exposure a bit. I think that a properly exposed 645 negative will easily, easily compete with a shot from a 5D and will even give a 1Ds a run for its money. But the bottom line is that there is scanning and then there is scanning... it is a skill. And if I ever feel limited by a scan, then I print traditionally, no big deal.
Grain is a consideration, but people should realize that in digital capture there is all manner of noise processing, and there are many noise sources not even present in film, and the camera simply trades detail for lower noise, that's it. Guess what, you can do the exact same thing with a scanned file. A 645 shot will easily generate ~30mp, and if I take that and downsample to 10mp it will look just as clean and glorious as a 5D image. This is well known to those of us who work with medium and large format. MF and LF shooters routinely downsample. Besides all that, noise is kind of anathema in film, we all know that we can shoot ISO 25 or whatever. No big deal. ISO 8? Fine.
On the topic of dynamic range, this is something that digital users really should complain about! Look, there is only one camera on the market right now, the S3, which comes close to how forgiving print film is in the highlights. On the film side, there are many different films with different tone curves, we pick what we want for a certain shot. I simply don't worry about blowing highlights, not even with slide film. Again, proper exposure is the key. It's true in digital and it is true in film. I would urge people to be extremely skeptical of the "this versus that" threads online, because you never know how good the exposure was in the first place. If you really want to compare the capabilities of two different formats, then I am afraid you will need to look at giant enlargements, in person, made by people who really know what they're doing.
P.S. I take extreme exception to the term "rez-up", which I see on all kinds of web forums. Actually there is no such thing as rezzing up! There is interpolation and there is interpolation, that's all. Interpolation does not create information, there is nothing going up except the apparent smoothness of the image. Call it depixelation if you must.
As I mention above - the proof is in the pudding. We can do all the math in the world but ultimately the prints will be our final judgement. I suggest you don't become too entrenched in your views until you have had a chance to do an objective comparison of prints.
I am no stranger to MF and LF myself. I have worked extensively in all these mediums (and in fact teach workshops on the subject). You can trust me when I say I used to be as skeptical as anyone about the digital "hype", but I was not satisfied with inflamatory statements on photo forums - I decided to give both my best effort and can tell you that even as a die-hard LF user for years, digital capture exceeded my expectations. I still use both.
As an aside - you did not provide a last name, email contact, or web site where your work can be seen. I'll be curious to see some of it.
Guy
Okay, credential-wanking aside... that stuff has nothing to do with real world photography and I agree with Guy that people really need to get out and experiment and not rely on web chatter.
Admittedly, I do not presently use the "high end" digital SLRs, and that is because I like: tilt and shift, rotatable and interchangeable backs, and I especially like shooting b&w. I am also an avid lover of rangefinders, of which there is really only one 6mp entrant on the digital side and it is not at all satisfying. So my feet remain firmly planted in film and for good reasons. I have made several forays into digital over the years and as yet the one camera that lures me is the fuji S3, but unfortunately it lacks the detail that I want. But I may well replace my F100 with one of those soon. I could happily do sports & wildlife on digital. I would love to see something like a 5D with the DR of the S3, that would probably make me forsake film below 6x9 format.
Compared to medium format, I will admit that there are digitals out there that could satisfy me, but not for under, what, $10 or $20k? Therein lies the problem. I am not professional, my day job keeps me far too busy for that. When I am bored with writing papers mayeb I wil retire to photography, who knows. So it's either the car or the digital back, and sorry, I have to go with the car ;) maybe in the future... just a poor physics professor...
I don't have a gallery and I don't intend to sell stuff any time soon, I am far too busy running my lab to get into the biz and I don't want to do a half-assed job. So if you want to see some shots then I'll have to refer you to a few that I posted for discussion purposes:
http://i25.photobucket.com/albums/c76/keithwms/KY_rollei400_horses_squarec.jpg
(Rollei infrared 400, Mamiya 6 rangefinder, 50mm, 6x6 frame cropped to 645)
http://i25.photobucket.com/albums/c76/keithwms/heron_crop2_bw_sm_redt.jpg
(Also Mamiya 6, but with Delta 400)
I guess I must have some 645s somewhere that would be more relevant to this thread, I could hunt for those. But I stopped using 645 about a year ago, now I use mostly 66 through 69.
Very nice gallery, Guy! and thanks for the welcome. I don't guess I'll be hanging around here much, I am already very involved in two other forums and that's alreayd taking too much of my time. I got into this thread per the request of Dave.
Ta,
Keith
While there have been all kinds of bonuses in field and location shooting, we based our decision to switch from scanned 645 film (NPS160 print) to the "only" 6mp Fuji S2 based upon needs and performance in our portrait studio. It's less an issue with smaller prints, but when you get to 20x30 inch and larger prints where we really start to make money, the trade-off is between an apparent increase in grain with scanned film and loss of detail with digital capture.
In portrait shooting the deck is sharply stacked in favor of digital for its much smoother skin tones due both to a little less detail and no grain. There is also such a thing as "too sharp" in portraits, and we often had to resort to diffusion filters on capture and sometimes even gaussian blurs in post processing with scanned film. From a purely pragmatic standpoint, digital capture on Fuji's 6mp sensor gave us the results we wanted right out of the camera without all the fluff, bother and expense of film. Now, shooting the 12mp Nikon D2X allows us to extend those benefits through much larger print sizes. Combine the extra income from larger prints with the ongoing savings in film and processing (even after factoring in the cost of new DSLRs), and we would only go back to 645 when digital backs drop down to prices competitive with high-end DSLRs.
The ONLY scan from a 645 that will compare to an image taken from a 5D or 1Ds Mark II will have to come from one of the latest models of drum scanners. For that you will need about $40K to $50K to purchase if you want one at home/office. With that in mind $8K just doesn’t sound that bad. For a desktop scanner, there's just not one that can compare except maybe from a sharpness standpoint but not for dynamic range. Again, as I stated in my first post, only print film is going to match or exceed the dynamic range of the 5D. But, one has to ask oneself, what am I going to shoot the most, color transparency film or print film. Most in our industry shoot transparency film when shooting color so it’s a no-brainer!
As for rezing up, that's a figure of speech that we use and again digital images can be rezed-up (or interpolated if you take extreme offence to it) to as much as 250% to 400% without very much degradation when done in the RAW converter. Something that CANNOT be done with a scanned transparency or negative, it's just a different animal! A scanned image file can’t be compared to a RAW image file from digital capture.
Then there's the noise issue that you pointed out. Well, I’m happy to tell you that with the 5D there's no noise issue up to about 400 ISO. My images from the 5D are as clean at 400 ISO as my 4x5 images on Velvia 50 drum scanned on the latest and greatest ICG drum scanner. Again, we are talking about the 5D not the D2X! So, I shoot at 50, 25 or 8 ISO with, oh maybe, 2 or 3 stops of extra time needed for reciprocity failure just to get an image that’s almost as clean as the image I can take with my 5D. Hummm, I hope the wind doesn't even think about blowing from about 2 to 5 minutes with that film camera.
And to your statement about giant enlargements for a true comparison, I’m sorry but what a load of crap! Size doesn’t matter at all unless all you do is sales of 30” x 40” and larger prints. I really don’t know about you guys but how many of you make a living selling 30” x 40” and larger prints? 98% of my print sales come from 20x30’s and 16x24’s. But print sales are only about 5% of my total income from photography, workshops about 5% and publishing is the remaining 90%!
BTW, just as Guy, I’m also not at all lost when playing around with those big film cameras. Up until 19 months ago I shot exclusively large and medium format color transparency film. Over five years with the 4x5 (6x7cm with a rollfilm back), about a year with 6 x 4.5cm and about 10 years shooting 35mm film. I still spend a lot of time scanning film and now processing digital files for the book and calendar projects that I produce through my publishing company, Mountain Trail Press so I’m not lost when working on a scanner or playing with PhotoShop.
You know, I can remember when I sounded just as sure about the film vers. Digital thing as you do. Man, I’m still eating crow from all my wild statements!
You want low noise from digital, I want high detail from medium and large format, gee why don't we compromise on a lovely $30,000 H2d? Hmm on second thought I think I'll pass for a while. For a second all this idle chatter made me forget what like most about medium and large format: movements, true b&w, ground glass composition, leaf shutters, rotatable and interchangeable backs, and the priceless satisfaction of knowing that I don't use the same quickly obsoleting plastic box that millions of others use.
As they say in Russia, kazhdi drochit' kak on khochit'.... everybody pleases himself the way he wants (polite translation)....
Again, NPN is a big family and this site is not like the others. Attitudes only flame here every once in a very small while. Also, most of us are pretty easy going folk too. Sometimes we get a knot in our drawers and are short fused but most of the time we’re pretty level headed (I've been known to get mine in a knot every once in a while but I'm getting better, maybe it's the new thong that MG sent me for my birthday
BTW, $6,000 for a drum scanner well you must be talking about a low-end Imacon, that's a want-a-be drum scanner. A new high-end Imacon will set you back about $20,000. Again, these scanners are really not drum scanners. Hummm, Maybe you should check out a real drum scanner like an ICG. Just a note, a new “real” drum scanner costs on the low end about $19,000 for a 4000 dpi unit and on the high end about $55,000 for an 11,000+ dpi unit. This usually does not include the drums, $1500 - $2500 each, drum mounting stations $1500 - $3000, or software $1500 - $4000. Oh, and the computer big enough to handle it.
NOTE: Today’s price for a new ICG 380, 12000 dpi optical quality resolution, is $43,900 USD brand new and then you must buy all the goodies to go with it. Those, like stated above, will add another $5,000+.
A used Howtek 4000 dpi unit will set you back about $10,000 plus the $4,000 for the extras needed to make your first scan. For a new one you’ll need to more that double that.
Like I said we’re not a bunch of stupid photog’s so lighten up on da family Keith.
Yes, I mistyped the res of the 5D as 10mp, big deal. The lateral res of the sensor goes as the square root of the mp count so you know what, my point is rock solid in spite of my silly typo. And you can search around in other forums telling people that I think the 5D is the best 35mm DSLR on the market right now, so kindly put my comments in context.
Foremost there is an implication in Greer's message that I don't' know what I am talking about. Au contraire, I make my living working on improving digital sensor resolution and signal to noise ratio. I am very pro technology and my life is invested in it. So don't dismiss my thoughts because my day job keeps me to busy to have a storefront selling prints. On the other hand, if you do want to dismiss what I say, then go right ahead, no loss for me. I have absolutely no interest in the business side, I took a vow of poverty to go into academia and that is my decision.
I have gotten *spectacular* results from drums on velvia, even mere 120 film, not even large format. These things enlarge to mural size, no doubt about that. If you badly want to see a 645 velvia scan I will go get one for you. But you really don't need to suffer my amateur photography to get the point; this is a simple issue of Dmax. When someone says that scanned velvia is comprable to a 5D at IS 400, I am sorry, that is false on so many levels. That tells me that the scanning technique is entirely incorrect. Now, I don't profess to be a brilliant scanner or even own a great scanner, I go to shop and I pay my $40 and I get my 50mp scan. And the boys at the shop certainly do not have a Tango. If someone is getting a lot of noise from 4x5 velvia scans then they are not scanning on a high enough Dmax instrument and probably not doing enough passes, it's that simple. Don't believe me? Scan provia 100 and compare. Provia will scan very nicely on a flatbed, and it should, it has far less information in the shadows than velvia. Also as I have said, there is an art to exposing for a scan, it is not the same as exposing for a trad'l print, not by a long stretch. The scanner cannot deliver 100% of what velvia has to offer, something has to give.
Now I will assume that there are people who are actually willing to learn something (as opposed to merely blowing smoke), so here are some reasonably good links that compare 4x5 to a ~39mp back. *That* is a perhaps a fair comparison, since people who shoot velvia are no longer likely to do the whole print process themselves, traditionally. But gues what, the results are in, the scanned velvia may be comparable to 39mp back, which wil run you, well you know very well how expensive they are, not to mention the attending gerathat is needed. Suffice it to say you are probably not going to stuff a 39mp back into your pack and go kayaking through the grand canyon. So then what, a scanning back? Please. Things move.
http://www.clarkvision.com/imagedetail/film.vs.digital.summary1.html
http://www.luminous-landscape.com/essays/Cramer.shtml
http://www.outbackphoto.com/artofraw/raw_28/essay.html
There was another review at LL that compared 4x5 to a 1Ds MK2, and t features some really overreaching claims and give ssome photos that clearly show the superiority of the 4x5 image, and yet the claim is made that the digital is up to par with the 4x5 velvia. I think critical minds can read through this stuff and draw their own conclusions. Yes, you will need to understand logarithms, but it really isn't rocket science. Incorrectly scanning velvia is like sipping champagne from a used styrofoam cup. Don't do it, it is blasphemy. My experience is that nothing less than a drum scan can possibly do velvia justice. (and for those who haven't experienced this, the stated DMax on most scanners is an extreme overstatement; when you realize what Dmax really means, if you get say 3.2 when you thought you were getting 3.8 then with velvia that is a freaking disaster.)
As I said before, this is not merely about pixel counts, to dwell on that rot is to miss the point abut why people use different tools for different tasks. It is about overall capabilities, and for me, an SLR doesn't cut it, even with a zoerk adapter. Been there done that, no thank you sir, t&s are just a few of the things on my long list of requirements. And I also have a good friend who went the TSE route on a Canon, he absolutely went nuts with it and even modified it to extend the motion. Couldn''t get the guy to shut about the TSE. Guess what, a month later he tried and bought a 4x5 and is now giddy with the amount of creativity it permits him. He hasn't shot digital since. You know, people need to learn for themselves and as they say in Russia, kazhdi drochit, kak on khochet....
With that I leave you.
Keith
Im really didnt mean for this hummbug to happen, all I wanted (like I stated) was some information on some of the things that I have been reading for the sake of knowledge. I wasn't looking for suggestions on what I should be shooting or if I should shoot digital or film. I have already decided to do as Mr Tal had suggest (and what Keith has suggested to me on many occasion in the past) to shoot what intrests me and find out what best suits my needs (currently looking into 8x10 pinhole
I guess I should have figured it out before hand that a thread that compares film to digital in any way can go no where but bad.....esspecaily when you put two people like Mr Greer and Keith in the same thread
First of all, the way I see it you knew and wanted Keith to stir up the pot here on NPN. He's your friend and you knew that he'd spew his rhetoric just as he did. I'm sure he's already preached the digital/film sermon to you many times before. That's why you asked him "to chime in" to this thread. But, to put me in the same league with him I don't think so. My problem with this whole post is that a few of us, like Guy Tal, have been shooting for years with 35mm, 6x4.5cm, 6x7cm and 4x5in. We are well versed in scanning and now with digital capture. Maybe you read Guy's article in OP, if not you should. We've been through the whole film vers. digital thing more than once and it wasn't that long ago that I defended film as though my life depended on it. So we do know a little about what we do and that knowledge came from experience first hand and not from reading articles in print or on the web. Just as Guy noted, he finds that digital capture with the 5D works for him and he stated why and I did the same. Remember, we both were shooting large-format professionally just a little over a year ago. Guy still uses both but I don't. We're not here to force you to use digital we want you to know why we love to shoot the 5D and why we think that it's much more superior in quality than 6x4.5cm and that's all. You can take our advice as you wish. Whether you believe it our not or even care we are here to answer your questions and concerns to the best of OUR knowledge on a specific subject.
Best always,
The reason that I asked him to look at the thread was because I wanted to get his opinions on my questions and opinions of others (instead of me relaying them and more then likely getting them mixed up)
Actually I have read a few of Guys article (OP and Photolife and here), I also know that you have been working with LF as well for a long time.........I have seen and enjoyed not only yours and Tals work but a lot of others on this site, so I thought this would be a good place to start to get my answers.
Now on to the testing and experience.......You have hit the nail on the head there. I have very little experience having only started this hobby of my Last December, so forgive me if I havent had the time and money to try out every known format in the world of photography. I do plan on trying out as many formats as possible (like I have stated 3 times already) but in the mean time.......If I want to get information on one vs the other......I have to rely on others experience. But even in the format wars (well not really a war) there are plan facts, one film handles colours better then the other, digital must handle D-Range better then this film or that film or vice versa. You would think that this information would be easy to come by, but it seems its not..........and that is the reason I originally post this thread.
I guess I'll just have to mark this up as a taboo subject, and move on....I guess if I read enough article and wait to try out different thing later on I'll be able to figure it out myself
This is a direct quote from Keith's first post "Alright, Dave asked me to chime in so I will!". So, with this quote I made my assumption that led me to my last post. After rereading Keith’s rants I can see where he would use you as a springboard to jump into this post and begin his sermon. If that is indeed what happened I apologize.
Again, Guy and I are not offering our experiences to you on assumptions. We offered our experiences from true hands on experiences that we both have had by spending months and years with each format. I'm certain, knowing Guy for some time now, that he wasn't trying to conform you or anyone else and I sure wasn't either.
Please spend some quality time here at NPN and you will see that we are not the same old forum like photo.net. We are a true family and everyone here is willing to help by offering up true life experiences in about any genre of photography that you can think of.