Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: CastorScan on September 08, 2013, 04:34:09 pm
-
I wish to share this link that shows some crops of a 3 Gb drum scan from a very good Kodak Portra 160 8x10" negative.
It shows how much quality and resolution are achievable from a large format color negative. As you can see, it's possible to extract up to 3 Gb of true and crisp detail from a 8x10" shot (crops published at 100% of magnification).
http://www.flickr.com/photos/castorscan/9696267924/
http://www.flickr.com/photos/castorscan/9693023737/sizes/l/in/photostream/
http://www.flickr.com/photos/castorscan/9693037335/sizes/l/in/photostream/
-
Impressive.
What are the pixel dimensions of the 3GB file? Would you mind also sharing how much a scan like that set you back?
Kind regards,
Gerald.
-
Wow! That really is impressive!
-
Wow, that's really impressive!
-
It is indeed, and an informative commentary/advert on the Flickr link ;)
Can I, and do take this as tongue in cheek please, argue against 10 x 8 being here ;D
The discussions against discussing 35mm crop up all the time as the forum is "Medium Format/Film/Digital Backs." So should 10 x 8 be allowed?
I suppose it creeps in as "large sensor" photography ;D ;D
-
Does Andreas Gursky use your services for scanning or is that confidential information? :)
-
That's confidential.
I can say that the Grieger lab in Dusseldorf (where Thomas Struth, Andreas Gursky, Thomas Demand, Elger Esser, Candida Hofer, etc.. print and mount all their works ) argues that my scans are the best scans they've ever seen and probably the best scans on the international market.
-
The original on that page was something like 6100x6000 pixels "only". About the same as a D800 frame. Already in that size the grain shows quite clearly, while in a 36 MPix digital camera file it does not. Low contrast also.
Quite frankly I was not impressed at all. And yes, where does that 3 GB come from? 36 MPix as a 16 bit TIFF is only 36x2x3 = 216 MB.
-
The original on that page was something like 6100x6000 pixels "only". About the same as a D800 frame. Already in that size the grain shows quite clearly, while in a 36 MPix digital camera file it does not. Low contrast also.
Quite frankly I was not impressed at all. And yes, where does that 3 GB come from? 36 MPix as a 16 bit TIFF is only 36x2x3 = 216 MB.
The "original" is original of a crop, ie not the full frame. 3GB inidicates 3 * 1024 * 1024 * 1024 / (3 * 2) = 536 megapixels, or about 2500 ppi and 25000x20000 pixels original size, but it would be nice if CastorScan posted the full details. If you look at the first link you can see from where the crop of the second link is taken. Taking measurements from that also gives a result of about 25000x20000 pixels original file.
Scanning at a high enough resolution so you get low contrast pixels I think is key to make a huge print look good up close. Medium format digital with it's lack of AA filters and thus jagged closeup look is not great to print over-sized. If film grain is visible up close is not a problem I think, it gives the print a nice photographic look. I'm allergic to pixel structures or digital-looking upsizing effects though.
Related: here's a great side-by-side resolution comparison with IQ180 vs 8x10" scanned at 2000 ppi: http://static.timparkin.co.uk/static/tmp/cameratest-2/800px.html (among other cameras).
-
OK, hasty conclusions by me. Still the look is a bit anemic to my taste.
How much dynamic range can they squeeze from these negatives?
-
In principle you can get huge dynamic range from negative film, Portra 400 has about 19 stops or so, but funky stuff happens to color response when you push the limits so it depends on how you see it. Film makes non-linear compression of highlights which make them strong in dynamic range. Negative film is also "easy" to scan as the negative is quite low contrast (ie those captured stops are put in a much smaller contrast range in the negative, and then you expand it again in post) unlike slide film which is dense and contrasty. You can use HDR techniques when scanning I guess to minimize noise in the scanning, but I don't think that is necessary for color negative in most cases. For slide film it can be though, especially if you need to push in post. If you choose to use slide film you don't do it to make huge contrast adjustments in post though, for that color negative is much better.
The scans I've seen from Castorscan have not been examples of tough lighting conditions though so I don't know what their scanning hardware can do.
-
I remember the discussion here when IQ180 vs 8x10 "showdown" article has been posted at LuLa - obviously biased, uninformed and unjust. I have no idea how could anyone think a digital capture from IQ180 could match the quality of a well executed/well scanned 8x10 frame.
-
Hi,
I don't think the article was biased, more like erroneous. It is hard to do a proper test. Tim Parkin and some friends made a similar test with very much different results.
https://www.onlandscape.co.uk/2011/12/big-camera-comparison/
I did some comparisons between film (6x7 Velvia and Ektar 100) and 24 MP DSLR and found the DSLR to be better in almost all aspects, but that result may have originated in my technique, although I tried to be careful.
Best regards
Erik
I remember the discussion here when IQ180 vs 8x10 "showdown" article has been posted at LuLa - obviously biased, uninformed and unjust. I have no idea how could anyone think a digital capture from IQ180 could match the quality of a well executed/well scanned 8x10 frame.
-
I still own a howtek drumscanner which I didn´t use for years. Whilst I always liked the impressive color depth I never liked the amount of grain that showed up on everything from 35mm to 6x7cm. Wet scanning of course, and scanning tricks included like scanning with a larger aperture to make grain less apparent.
I got one b/w scan from 35mm made on a Hasselblad X5 last year and I was impressed by excellent sharpness and very little grain.
Bernd
-
Erik:
Well, that is of course my own opinion but for me it read that way. To even try making such a test with no LF experience/knowledge whatsoever is not a good idea..
As for the Tim's test- I'm of course familiar with it and it tells a whole different story.
-
Nice image quality, seems pretty close to what a 12 frames (3x4 with 30% overlap) stitch with the D800 delivers.
Cheers,
Bernard
-
Hi,
Digital images at low ISO have very little noise. Also, it seems that digital can extract more low contrast detail than film.
The samples here may illustrate this: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/59-sony-alpha-900-vs-67-analogue-round-2?start=1
Pentax 67 and Velvia scanned with film scanner and drum scanner compared with 24 MP DSLR. The DSLR picture has some shaking (IS not disabled), unfortunately.
Best regards
Erik
8x10 images usually look "more true". Why? No idea. People seem to look more natural. With digital they look more staged.
BUT sometimes the digital image just is the better image because there is something special I often can't explain with words.
In my opinion whether an image is good depends not on the resolution.
But I would love to have a scanner with the drum scan quality and the ease of use and the size of an Epson V750. And someone who carries my photo cases. And someone who removes the dust from the scans. And what I would like most would be better se(x*) and more money.
Best,
Johannes
___________
x*: the x is only a placeholder. You can add a letter of your choice other than a "x".
-
Nice image quality, seems pretty close to what a 12 frames (3x4 with 30% overlap) stitch with the D800 delivers.
Cheers,
Bernard
Really glad I don't have to throw that stitch at my computer, woow! :D
-
Really glad I don't have to throw that stitch at my computer, woow! :D
12 frames from a D800 is nothing. Try a few hundred from an IQ180 if you really want to have some fun :D
-
12 frames from a D800 is nothing. Try a few hundred from an IQ180 if you really want to have some fun :D
Exactly, my 6 years old Mac Pro processes such smallish stitches in a few mins with PTgui or Autopano pro.
For an image like to one shown, shooting would take less than 2 mins once the tripod and pano head is set (which still takes less than setting up an 8x10 camera), stitching would be exactly a 16 clicks operations done in less than 3 mins end to end... voila.
This being said I am in the process of purchasing a used Betterlight super 6K back for my Ebony... so not everything is about the end result. ;)
Cheers,
Bernard
-
Exactly, my 6 years old Mac Pro processes such smallish stitches in a few mins with PTgui or Autopano pro.
For an image like to one shown, shooting would take less than 2 mins once the tripod and pano head is set (which still takes less than setting up an 8x10 camera), stitching would be exactly a 16 clicks operations done in less than 3 mins end to end... voila.
This being said I am in the process of purchasing a used Betterlight super 6K back for my Ebony... so not everything is about the end result. ;)
Cheers,
Bernard
Having said that though, I would LOVE to be able to capture a 1GP image in a single shot with 8x10 film. Problem is round these parts it's impossible to get the film developed and scanned :(
-
If film grain is visible up close is not a problem I think, it gives the print a nice photographic look.
Aha - this is where you show your true colors! Once you switch over to digital in your perception you begin to see film grain as nothing but a mushy blur over your sharp image. Photography is not about grain, anymore than music is about the tape hiss of early recordings. It becomes hard to look at grainy film images the same way once you see how it is without it. (though I agree with you about upsampling too much - it is not nice)
This argument has been done to death. I know all the photographers you mention and nearly all have Phase One/Alpa in their equipment packs, so... stitching MF digital seems to be where high end is at right now, for them at least.
-
If you refer to Andreas Gursky and other highly regarded artists- I think most of them still shoot both analog and digital
-
Interesting Tim Parkin is mentioned. Tim did recently some scans for me and we spoke for some time about our shared thoughts on digital capture compared to film. All i can say is he is not exactly a digital capture lover.
-
I collaborate with many international photographers shooting both analog (LF) and digital (Phase One or Alpa bodies + P65 or IQ180 ecc).
A good 8x10" negative perfectly drum scanned exceeds by far the quality delivered by the last and most expensive digital photographic systems, in terms of three-dimensionality, realism, color and density nuances, detail.
Also, the rendition of details (edges, etc) is much more pleasant and natural.
The amount of "crisp and true" information contained in a good 8x10" negative is usually between 1,6 and 3 Gb
-
A good 8x10" negative perfectly drum scanned exceeds by far the quality delivered by the last and most expensive digital photographic systems, in terms of three-dimensionality, realism, color and density nuances, detail.
Sorry but I respectfully disagree with you, from decades of experience (I operated high end scanners too - Heidelberg Tango drum and flatbed). You could scan a 6x7cm negative at 12,000dpi and get a multi GB file, but its just grain /mush you are scanning at that level. I simply cannot believe it when I look at my old LF negatives how much grain ruins the image.
Also have to point out that you are in the business of making money from this fact being what you wish it to be, and have come here cold -just 3 posts ever - drumming up business for yourself, on that basis. It doesn't look good.
Yes 8x10" is great, and if you want to use it - please go right ahead, its absolutely everyone's choice.
-
If you say that it means that very probably you never saw any really good drum scan before.
This is a old FUJI REALA 6x7 cm drum scanned:
whole frame: http://www.flickr.com/photos/castorscan/8291295942/
crop 1: http://www.flickr.com/photos/castorscan/8291390058/sizes/k/in/photostream/
crop 2: http://www.flickr.com/photos/castorscan/8291392122/sizes/k/in/photostream/
crop 3: http://www.flickr.com/photos/castorscan/8291285426/sizes/k/in/photostream/
And you should know that the Tango is far worse than a Dainippon 8060 in terms of true optical resolution, color gamut, noise, grain rendition.
-
Sorry but I respectfully disagree with you, from decades of experience (I operated high end scanners too - Heidelberg Tango drum and flatbed). You could scan a 6x7cm negative at 12,000dpi and get a multi GB file, but its just grain /mush you are scanning at that level. I simply cannot believe it when I look at my old LF negatives how much grain ruins the image.
Also have to point out that you are in the business of making money from this fact being what you wish it to be, and have come here cold -just 3 posts ever - drumming up business for yourself, on that basis. It doesn't look good.
Yes 8x10" is great, and if you want to use it - please go right ahead, its absolutely everyone's choice.
+1
That is absolutely my own experience, using Howtek Drum Scanners with 4000 and 5000 dpi. I went through this because I wanted to go the hybrid way with photographing on film and then drum scanning. Every word you say is true, I found it out loosing years; and loads of money. Don´t believe the scan hype, 10.000 by 12.000 pixel or so- film is dead! Scanning today is only good (and important!) for archive purposes.
And I do also have the impression that the post is about making business rather than sharing experience.
Bernd
-
You should know that scanning at the highest scanning resolutions is foundamental to dramatically reduce grain and noise.
So 12.000 dpi are extremely usefull in any case. Detail of the negative apart.
-
The links clearly show the benefits of scanning a 6x7cm color negative at 10.000 dpi:
Whole frame:
http://www.flickr.com/photos/castorscan/8291295942/
crop 1: http://www.flickr.com/photos/castorscan/8291390058/sizes/k/in/photostream/
crop 2: http://www.flickr.com/photos/castorscan/8291392122/sizes/k/in/photostream/
crop 3: http://www.flickr.com/photos/castorscan/8291285426/sizes/k/in/photostream/
Of course the output provided by an old and "entry-level" Howtek limited at 4000-5000 dpi would be completely different...
-
CastorScan, i believe you are wasting your breath on this website, it is so digitally biased you will never convince anyone. For what it's worth, from my point of view you are 100% correct, and quite honestly a good 6x7 Portra drumscan is much nicer in every way than any digital capture i have seen, let alone 10x8.
-
If you say that it means that very probably you never saw any really good drum scan before
How could I have possibly seen a good scan before? Nobody can make a great scan but you. We are all rubbish, but you and your equipment is the best.
It is amazing that we survived till you came here to correct our ignorance and get some business for yourself. :P
I'm out of here.
-
I agree with you here.
CastorScan, i believe you are wasting your breath on this website, it is so digitally biased you will never convince anyone. For what it's worth, from my point of view you are 100% correct, and quite honestly a good 6x7 Portra drumscan is much nicer in every way than any digital capture i have seen, let alone 10x8.
-
Hi,
I downloaded the image and downscaled to 20% Photoshop using bicubic and applied some smart sharpen, see below. That would correspond to something like 24 MP digital capture. In my view it is quite noisy and not very sharp.
On the other hand, I agree with Anders Torger about digital artifacts being ugly, see the sailboat rig below. Fortunately, digital artifacts are normally not very visible in normal captures, see third example.
The digital examples are taken from a P45+ back, 39 MP. I also added a 24 MP capture from an OLP filtered DSLR.
Enclosed images:
1) 10000 PPI scan downsampled to 2000 PPI
2) 39 MP P45+ (shows severe aliasing)
3) 39 MP P45+ (no obvious aliasing)
4) 24 MP DSLR (with Optical Low Pass filtering)
Best regards
Erik
The links clearly show the benefits of scanning a 6x7cm color negative at 10.000 dpi:
Whole frame:
http://www.flickr.com/photos/castorscan/8291295942/
crop 1: http://www.flickr.com/photos/castorscan/8291390058/sizes/k/in/photostream/
crop 2: http://www.flickr.com/photos/castorscan/8291392122/sizes/k/in/photostream/
crop 3: http://www.flickr.com/photos/castorscan/8291285426/sizes/k/in/photostream/
Of course the output provided by an old and "entry-level" Howtek limited at 4000-5000 dpi would be completely different...
-
Just a brief point about film: it looks different. It acts different. It smells funny. I like the way it looks for what I shoot. I like how grain adds texture, I like how film looks printed on chromogentic paper, I like how film looks printed on a web press. I like how the larger formats render. I don't shoot scenics, but the ones I like are generally shot on film, even 35mm film. Its different.
Film is an impression of reality, rather than an exact reproduction. The exact reproductions that mark digital bore me and most digital post is vulgar.
Not that digital doesn't have a place and can't look really good, it just doesn't look really good for me, for how I see things.
So lets all keep it in our collective pants and realize that one man's grain free crisp shadow detail at 100% is another man's soap opera video still. Neither opinion is wrong for the person holding it.
-
Just a brief point about film: it looks different. It acts different. It smells funny. I like the way it looks for what I shoot. I like how grain adds texture, I like how film looks printed on chromogentic paper, I like how film looks printed on a web press. I like how the larger formats render. I don't shoot scenics, but the ones I like are generally shot on film, even 35mm film. Its different.
Film is an impression of reality, rather than an exact reproduction. The exact reproductions that mark digital bore me and most digital post is vulgar.
Not that digital doesn't have a place and can't look really good, it just doesn't look really good for me, for how I see things.
So lets all keep it in our collective pants and realize that one man's grain free crisp shadow detail at 100% is another man's soap opera video still. Neither opinion is wrong for the person holding it.
+1
-
I still have 3 unopened boxes of 8x10 Provia that cost me a fortune, and are now worth even more! Haven't had the chance to see what 8x10 can really do for me personally.
My favorite image quality in the world right now is from Sigma. So you're all wrong! :)
-
It's funny, I looked at these scans and thought they were pretty great, but I didn't look at the grain or detail really but instead the chrome on the chair, the text on the books where the text color was really close to the binder color, the globe with attention to color.
I've been doing my own comparison of film and digital shot on the same camera with the same lenses for the last year. The biggest differences are seen when an analog print is made from the negative and put next to a digital print made from a digital capture. btw - I haven't been doing the comparison to see which is better, I've been doing it to see when to use what.
-
I stopped worrying about film/digital when I got my first digital SLR, Canon EOS-1D. In theory those 4.7 MPix were much less resolution than with slide film (100 ASA Astia or Provia mostly), so I did some testing. Using the same zoom lens and framing it turned out that there was just as much detail in the digital image as there was in the slide (60 MB professional scan). Good enough for me, and cameras have getting only better all the time like 3 times the resolution and twice as many stops of dynamic range with amazing low light capabilities: me not complain.
-
Hi,
For me it was about being good enough. I started moving into digital because I had problems with the wet darkroom. The chemicals were not kind on my respiratory systems and so on. Digital printing came as a rescue. That meant scanning, and scanning was no fun. Michael Reichmann made a test back in 2003, comparing the Canon 1Ds with the Pentax 67 and Velvia: http://www.luminous-landscape.com/reviews/shootout.shtml. Michael used an Imacon scanner but he also went to drum scanning and his findings still held. He found that the 1Ds surpassed the Pentax 67 in image quality. I made similar tests with my Sony Alpha 900 and had similar results.
Film based photography is very different from digital photography:
- Digital photography is linear, until we manipulate it.
- In most cases digital images have very little noise, compared with film.
- Digital images are limited in resolution by sensor pixels, we have aliasing and moiré, although small pixels and OLP filtering helps. Anyway at certain magnification a digital image falls apart.
- Analog photography is very nonlinear. Film has curves built in. In the printing process the paper also has curves built in.
- Analog photography has grain.
- Analog photography can use larger formats. We still don't have a full size 6x6 sensor or a full size 6x7 sensor, except scanning backs.
The discussion tends to be polarised, because analogue digital photography often has different aims. Both can be used to achieve excellent results and both can produce garbage.
I still feel that there is something special about projected Velvia 67.
Best regards
Erik
I stopped worrying about film/digital when I got my first digital SLR, Canon EOS-1D. In theory those 4.7 MPix were much less resolution than with slide film (100 ASA Astia or Provia mostly), so I did some testing. Using the same zoom lens and framing it turned out that there was just as much detail in the digital image as there was in the slide (60 MB professional scan). Good enough for me, and cameras have getting only better all the time like 3 times the resolution and twice as many stops of dynamic range with amazing low light capabilities: me not complain.
-
I think, Marco, that it should be pointed out that it doesn't require anything close to the the 8060's maximum claimed resolution of 12,000 dpi to produce the results you post here, and that this was approximately a 2600 ppi scan. That is of course plenty for an 8x10 film. There are a number of drum scanner models that can achieve this, including the Aztek Premier you mention, and it's wrong to suggest otherwise as you do in your promotional blurb (not to mention the claim, made second-hand, that this particular practitioner makes scans of quality unique in the world).
As to the larger point of the capability of large format film, this is amply demonstrated here, and the Tim Parkin site mentioned gives parallel results. Those of us who shoot in various digital and film formats don't find these results in the least surprising. I would only add that these discussions are incomplete, centering exclusively on spatial resolution and dynamic range. However, equally important to apparent image quality is tonal resolution--the ability to separate tiny differences in tone--as this is responsible for the sense of richness in an image, which large format film so excels in. While high spatial resolution is required for high tonal resolution, it is not by itself sufficient. This is often confused with dynamic range, which only defines the extremes of useful exposure scale, but not how many steps of tone are recorded within that range. I have no technical data to confirm this, but digital camera images always look a bit deficient in this area, and I wonder if the large amounts of sharpening applied, both in the camera and later in Photoshop, to achieve adequate spatial resolution, does not force close tones together enough to obliterate fine gradations. The effect is certainly visible in high magnification, and it seems reasonable that this accounts for a certain brittle look to many digital camera images, even at normal print sizes, that I find a bit unnatural. Of course this may be a cultural effect, as I did not grow up with digital imagery...
-
I looked at the price list. But at 123 Euros for the first 500mb, and 2.5gb to go... That Phase IQ260 starts looking like a screaming deal.
I wish to share this link that shows some crops of a 3 Gb drum scan from a very good Kodak Portra 160 8x10" negative.
It shows how much quality and resolution are achievable from a large format color negative. As you can see, it's possible to extract up to 3 Gb of true and crisp detail from a 8x10" shot (crops published at 100% of magnification).
http://www.flickr.com/photos/castorscan/9696267924/
http://www.flickr.com/photos/castorscan/9693023737/sizes/l/in/photostream/
http://www.flickr.com/photos/castorscan/9693037335/sizes/l/in/photostream/
-
http://www.idea-digital.com/index.php/scanning
It's nothing new if we talk about price. Equipment cost a lot of money, years of practice cost as well.
It is for people who really love photography...
-
I agree that film and digital both have their merits. It would be nearly impossible for me to handle film developing in my current circumstances, so I shoot 100% digital now. But if film were still easy to buy and process, I'd still be shooting 6X6 and 4X5 as well.
A lot of the commentary about the differences ends up concentrating on considerations of grain/noise and ultimate resolution. But I think a lot of what I like about the film "look" results from way older lenses rendered. There's no way any modern digital camera/lens combination is going to produce an image that has the same character as a classic film lens. That doesn't make one choice necessarily "better" or "worse." They're just very different, regardless of grain or resolution.
That's why I really enjoy this kind of project:
http://www.imaging-resource.com/news/2013/07/25/lomography-resurrects-the-petzval-lens-over-kickstarter
-
It would be nearly impossible for me to handle film developing in my current circumstances
Unless you shoot B&W film. It only needs a small tank, 2 clean bottles and can be done in a standard bathroom.
-
Unless you shoot B&W film. It only needs a small tank, 2 clean bottles and can be done in a standard bathroom.
Where are you located that you cannot access a film processing lab? Nearly all labs process film that is mailed to them. And the quality:cost calculation usually works out very favorably for medium- or large-format film when compared with a MF digital back. It is true that drum scans can quickly shift that calculation in the other direction, but some fairly good scans can be made with inexpensive scanners, especially if one is scanning only negatives and the enlargement is not enormous. The quality of a decent scan of a Hasselblad, Rollei, or Mamiya MF negative, even when scanned on a modestly priced flatbed scanner is simply astonishing and cannot be equaled by even the best 35mm DSLR. And 4x5 and 8x10 are much, much better still.
-
Unless you shoot B&W film. It only needs a small tank, 2 clean bottles and can be done in a standard bathroom.
Sorry to get back to this thread months later...
Yes, what you say is true, but 4X5 is a whole lot messier. And then you have to deal with chemical disposal and storage. I live in a part of the country that gets very, very hot in the summer, and water is scarce. One of the issues with that is that things delivered by mail have a habit of suffering thermal damage. I still have my 4X5 enlarger and all of my other film gear, but I never seem to find the time to get back into it.
-
Where are you located that you cannot access a film processing lab? Nearly all labs process film that is mailed to them. And the quality:cost calculation usually works out very favorably for medium- or large-format film when compared with a MF digital back. It is true that drum scans can quickly shift that calculation in the other direction, but some fairly good scans can be made with inexpensive scanners, especially if one is scanning only negatives and the enlargement is not enormous. The quality of a decent scan of a Hasselblad, Rollei, or Mamiya MF negative, even when scanned on a modestly priced flatbed scanner is simply astonishing and cannot be equaled by even the best 35mm DSLR. And 4x5 and 8x10 are much, much better still.
I've been through a couple of flatbeds, and found them all wanting, esp. for 4X5. Too many problems with newton rings or uneven sagging problems. I may make another run at it again though. I still have thousands of excellent negatives and transparencies.
-
I've been through a couple of flatbeds, and found them all wanting, esp. for 4X5. Too many problems with newton rings or uneven sagging problems. I may make another run at it again though. I still have thousands of excellent negatives and transparencies.
Scan wet and get rid of Newton rings forever.
No need for complicated Mylar layers around your film.
Just fluid between the glass and the film is enough.
I do that with my Nikon scanner and MF film and get good results.
The really interesting thing about film is the old glass which gives images a different rendering and the color rendition of film, especially in the yellows and reds.
Sunrise/sunset with film is something completely different than digital.
And a lot of the problems with films (not all, of course) can be eliminated by working hybrid in an analog/digital workflow.
Large inkjet prints from good film scans are a really interesting option if you can live with the restrictions, which are obviously there - I don't want to argue that.
What I would really love to see is a quad color Bayer pattern sensor with Red+Green+Blue+Yellow instead of Red+2xGreen+Blue.
In the moment there is no digital equivalent to a multilayered film with different spectral sensitivities on each layer and super robust long time exposure capabilities.
Digital still has to learn some lessons from film.
I strongly believe shooting both, film an digital is THE way to give you all the options.
Giving up one simply is self castration.
Cheers
~Chris
-
What I would really love to see is a quad color Bayer pattern sensor with Red+Green+Blue+Yellow instead of Red+2xGreen+Blue.
Chris,
I agree with you about film - and that's an interesting idea for sensors. So much of the latest chip designs are about maximizing luminosity detail at the expense of color detail, so this is nice to see someone thinking about improving color response. But maybe this works better when pixel density gets higher?
Eric
-
Chris,
I agree with you about film - and that's an interesting idea for sensors. So much of the latest chip designs are about maximizing luminosity detail at the expense of color detail, so this is nice to see someone thinking about improving color response. But maybe this works better when pixel density gets higher?
Eric
I don't think replacing one green pixel with a yellow one would decrease luminosity resolution - yellow is a pretty common color in almost any image.
I (wildly) guess changing the Bayer pattern to a 3x3 pattern with 9 different colors, e.g. additional violet or deep reds near IR spectrum should work - but the details would be up to the engineers - I'm only a physician without too detailed engineering knowledge.
But it is clear - every pixel samples luminosity - so why not increase the number of colors to get a better sampling of the light spectrum too?
-
Scan wet and get rid of Newton rings forever.
No need for complicated Mylar layers around your film.
Just fluid between the glass and the film is enough.
I do that with my Nikon scanner and MF film and get good results.
+1 - you have to wet mount to get a decent scan.
And as this self-promoting thread has sprung zombie-like back to life (!) I have to point out that I saw the Thomas Struth exhibition at Marianne Goodman Gallery in midtown Manhattan right now, and was struck by how poor the quality was on those huge prints. A stitched digital MF back would give much better results than the scanned LF film he was using. Sharpness and color. The image softened into mushy grain as you moved closer.
If OP made those scans, like he hints, its a slam dunk against his claims.
-
+1 - you have to wet mount to get a decent scan.
And as this self-promoting thread has sprung zombie-like back to life (!) I have to point out that I saw the Thomas Struth exhibition at Marianne Goodman Gallery in midtown Manhattan right now, and was struck by how poor the quality was on those huge prints. A stitched digital MF back would give much better results than the scanned LF film he was using. Sharpness and color. The image softened into mushy grain as you moved closer.
If OP made those scans, like he hints, its a slam dunk against his claims.
That is not the way it works. We have at the beginning of the thread "proof" of the quality you can get from 8x10 scanned. If somebody else gets a shitty result that does not throw out reality. They botched it, or they had an equipment problem. It takes far more skill to get the film right in 1 go than when you can check a digital capture. It's also basically hearsay. There is nothing presented to show a side by side of the 2 methods. The "slam dunk" is into the round file.
I am not saying film is better. These days it is near impossible to do with film what a person determined to get the best from high end digital can do. Multi-shot stitch, exposure bracketing, focus stacking, etc. Try to do it with film and the light has changed before you are done.
-
That is not the way it works. We have at the beginning of the thread "proof" of the quality you can get from 8x10 scanned. If somebody else gets a shitty result that does not throw out reality. They botched it, or they had an equipment problem. It takes far more skill to get the film right in 1 go than when you can check a digital capture. It's also basically hearsay. There is nothing presented to show a side by side of the 2 methods. The "slam dunk" is into the round file.
I am not saying film is better. These days it is near impossible to do with film what a person determined to get the best from high end digital can do. Multi-shot stitch, exposure bracketing, focus stacking, etc. Try to do it with film and the light has changed before you are done.
Sorry, but I respectfully disagree in the case of this thread's OP statements. His imputation that he makes all these amazing scans for the high end German art photographers, that knock digital out of the ring altogether, then we are entitled to look at the 'pudding' and see if it bears out the 'proof'. It did not.
Yes of course, someone looking at a bad digital photograph cannot then conclude that all digital is rubbish, and likewise for film, but - at this level of craftsmanship, one is right to expect the highest of standards by a mature renowned photographer, and his production process. There are many steps in that of course, but it wasn't that the print (by Grieger) was bad, but the grainy XL image was clearly softer than it could be, and did not hold to closer inspection - it only worked viewed from a significant distance. I have seen large digital captures that greatly exceed this result, in less esteemed hands. The conclusion I would draw from this is quite clear - but ymmv, of course!
-
I've noticed in several shows recently of older famous film photographers that a lot of the work was soft, but the key thing is not all of them - and of course from shooting my own films, I also know that to be true. My point? Since the advent of digital cameras and 100% view, a whole lot more people became more obsessed with sharpness than in the past and get lots of sharp images (with hopefully not less attention on the other aspects of the images). But to go from a number of soft film images and conclude that its not possible to get a sharp detailed file from scanning is wrong. Tim Parkin showed us that even some MF film cameras can out resolve MFDB when properly handled for sharp images. It may be pointless to go to the effort for the best scans unless you have a truly great negative though.
-
I feel the whole film vs digital debate gets pointless when discussing in terms of "better" or "worse".
"Strengths" and "Weaknesses" or "Usability" I think help better to discuss the differences if desired.
Seeing many of these discussions derailing into a war of zealots is plain boring - one of the greatest sins in art - to bore your viewers, IMO.
What we are discussing here in these last posts is a subset of problems connected to scanning and to color rendition, nothing more.
No one wants to say film is better or worse.
I love the rendering of sunrise and sunset how it is done by film and I love the features and new use cases my digital camera gives me.
I'd also love to see a sensor design with better color rendering in certain situations.
There are many photographers out who know very well both media and from what I hear from the majority of the ones I'm in contact with is basically:
"Horses for courses".
Sharpness, color rendition, microcontrast - all these technical parameters are only tools, nothing more.
There's a reason why there are Softars and why some people like old glass.
There are reasons why some people still shoot film, even professionally.
Please lets not open this can of worms again.
Its pointless.
Cheers
~Chris
-
Sorry to get back to this thread months later...
Yes, what you say is true, but 4X5 is a whole lot messier. And then you have to deal with chemical disposal and storage. I live in a part of the country that gets very, very hot in the summer, and water is scarce. One of the issues with that is that things delivered by mail have a habit of suffering thermal damage. I still have my 4X5 enlarger and all of my other film gear, but I never seem to find the time to get back into it.
My advice was about film only and only about B&W. Developing B&W film, be it 120 film or 4x5 is relatively easy and does not need too much water or produce loads of chemical waste. It is actually competitive with sending out the film to a laboratory far away to have it processed.
Developing colour film is a lot more difficult and labs may be easier to find, so home processing is less an option. Enlarging is also a lot more involved, even in B&W, if only because of the space necessary.
My advice was thus to develop film at home and to scan it and print it. It is a good compromise between quality and convenience. It may not be the way to achieve the highest quality possible, but that was not the question.
As to quality, anyone who has seen exhibition prints enlarged optically from sheet film knows that it is spectacular. Hiroshi Sugimoto comes to mind. At this level of quality my feeling is that endless discussion as to which technology gives the best results is pointless: it hangs in museum and the visitors are baffled, so it is good enough.
-
There is a guy in the uk that does great drum scanning for way less. Here is his link -
http://cheapdrumscanning.com/
I looked at the price list. But at 123 Euros for the first 500mb, and 2.5gb to go... That Phase IQ260 starts looking like a screaming deal.
-
There is a guy in the uk that does great drum scanning for way less.
You don't always need drum scans.
A simple Epson V700 scan gives you 80+ MP from a 4x5" - thats much more than needed to get gorgeous prints at huge size.
Scanning different exposures and doing an HDR with them gives you additional dynamic range.
Sure - A drum scan still is the nonplusultra - no discussion about that.
But if you know your way around you can get great results with a lot less change.
Cheers
~Chris
-
I'm doing most of my 'scanning' now with a 528cf multishot back and schneider macro symmar lens over a light table. Black and white is really easy, and I made myself a blue filter to neutralize the orange carrier of the color negative film I shoot. I then use color perfect plugin to get the colors right. I side step the whole scanning issue sort of and this is much faster.
btw - the blue filter is pretty easy to make - a least how I did it - I took a shot of the film put it in photoshop, sampled the orange carrier, inverted it and filled a whole layer with that color and just printed it out to transparency film. Ideally you'd do this with the light source you will use for your 'scanning'. For 4x5 and smaller there is a neat older device called a Bowens illumitran. This was made for copy work. It's basically a flash box with a slide holder on top. The flash has probably a better CRI than the fluorescent light tables - though I do use the eVision executive:HF copy stand with 40khz ballast fluorescent light box for negatives larger than 4x5.
-
+1
That is absolutely my own experience, using Howtek Drum Scanners with 4000 and 5000 dpi. I went through this because I wanted to go the hybrid way with photographing on film and then drum scanning. Every word you say is true, I found it out loosing years; and loads of money. Don´t believe the scan hype, 10.000 by 12.000 pixel or so- film is dead! Scanning today is only good (and important!) for archive purposes.
And I do also have the impression that the post is about making business rather than sharing experience.
Bernd
And for fun, don't forget the fun. Having a film based camera, doing the lab and then scanning can be a lot of fun :) .
Of course for fun one buys a used scanner, and never ever pixel peep the pictures.
Best regards,
J. Duncan
-
I tried the v700 but found it terrible with medium format. Ended up buying a Minolta multi pro dedicated film scanner. Not quite a drum scan but close enough for me.
You don't always need drum scans.
A simple Epson V700 scan gives you 80+ MP from a 4x5" - thats much more than needed to get gorgeous prints at huge size.
Scanning different exposures and doing an HDR with them gives you additional dynamic range.
Sure - A drum scan still is the nonplusultra - no discussion about that.
But if you know your way around you can get great results with a lot less change.
Cheers
~Chris
-
INTO THE WHITE - MASSIMO VITALI & ERICH LINDENBERG.
Marco Campanini_CastorScan with Massimo Vitali talking of his new LightJet prints. 180 x 220 cm LightJet prints on Kodak Endura paper mounted on DIASEC made at Grieger lab, Düsseldorf. DRUM SCANS BY CASTORSCAN.
www.castorscan.com
https://www.youtube.com/watch?v=Q1yVLrRfgcQ
-
INTO THE WHITE - MASSIMO VITALI & ERICH LINDENBERG.
Marco Campanini_CastorScan with Massimo Vitali talking of his new LightJet prints. 180 x 220 cm LightJet prints on Kodak Endura paper mounted on DIASEC made at Grieger lab, Düsseldorf. DRUM SCANS BY CASTORSCAN.
www.castorscan.com
https://www.youtube.com/watch?v=Q1yVLrRfgcQ
Wow, great work. Thanks for posting.
-
+1
Wow, great work. Thanks for posting.
-
Post this over at the Large Format forum. They may be more interested in 8 x 10.
-
Hi,Michael Reichmann made a test back in 2003, comparing the Canon 1Ds with the Pentax 67 and Velvia: http://www.luminous-landscape.com/reviews/shootout.shtml. Michael used an Imacon scanner but he also went to drum scanning and his findings still held. He found that the 1Ds surpassed the Pentax 67 in image quality. I made similar tests with my Sony Alpha 900 and had similar results
I'm sorry but that comparison was frankly laughable...
Here's a comparison between a D800E and a Mamiya 7
(http://static.timparkin.co.uk/static/tmp/D800E-vs-Mamiya7-including-colour.jpg)
And that was scanned on a flatbed!!
As for Castorscans scans - yes they're very good but I don't think they're visibly better than a good scan from a Heidelberg Primescan. It downsamples to 2000dpi without losing detail and 2000dpi is about perfect for scanning 10x8 (although in the centre with a good lens at f/22 you can get 3000dpi's worth probably - rarely).
Anyway - he does have one of the best drum scanners in the world but some of his claims are, shall we say, loosely grounded. e.g. the 12,000dpi scan of Ektar downsamples to about 5000dpi with no loss of detail or grain.
I'd love to find time to do a comparison (which he has offered to help with).
If you want a good comparison of film vs digital, look here..
https://www.onlandscape.co.uk/2011/12/big-camera-comparison/
This was done with the cooperation of Phase and a dedicated Phase photographer who used to shoot film. It's about as unbiased as you'll get.
The end result? It's a damn sight easier to get good results from digital than it is from film but if you do persevere with film the results are as follows
perceived sharpness gives 5x4 equivalent to an IQ260 - 10x8 is about 120-150 megapixels
In terms of resolution and raw detail - the the 5x4 exceeds the IQ280 for colour resolution and just about matches it for fine detail.
In other words the IQ280 is equivalent to 5x4 up to about 20x24 when the IQ280 looks better. At about 30x40 they start to be similar. Then the 4x5 starts to look better because the fine detail carries on (albeit with grain and low contrast).
for most purposes medium format film drum scanned just about beats 24-36Mp DSLR's but you have to have great lenses and technique (plus a camera that holds the film flat!).
I haven't a clue how Micheal Reichman got he 1ds2 to beat a Pentax 67 - actually I do ... try this sentence on for size
"But since there clearly isn't really any significant amount of additional real information in the drum scan, ressing up the digital file will essentially accomplish the same thing."
OK so he's got a 12,000 pixel by 15,000 pixel file and he prints it at 13" x 19". 13" at 360dpi is 4680px
Put another way he converted the 5330dpi drum scan into a 2000dpi scan and says "well there is only 2000dpi worth of detail in the film shot..
I know for a fact I get more detail out of film when I scan at 5000dpi over 4000dpi ....
He wasn't biased at all :-)
Tim
-
Hi,Michael Reichmann made a test back in 2003, comparing the Canon 1Ds with the Pentax 67 and Velvia: http://www.luminous-landscape.com/reviews/shootout.shtml (http://www.luminous-landscape.com/reviews/shootout.shtml). Michael used an Imacon scanner but he also went to drum scanning and his findings still held. He found that the 1Ds surpassed the Pentax 67 in image quality. I made similar tests with my Sony Alpha 900 and had similar results
I'm sorry but that comparison was frankly laughable...
Not that much laughable. The same web site you cited also compared the A900 with the Mamiya 7. You'll find the results here: http://static.timparkin.co.uk/static/tmp/cameratest-2/800px.html (http://static.timparkin.co.uk/static/tmp/cameratest-2/800px.html). Quite frankly, I find very little difference between the Mamiya 7 and the Sony A900, at least for colour pictures.
That would mean that a 6x7 camera is similar to a 20-25 mpix digital camera, which is also what Michael Reichman found. The 6x7 camera was a Pentax and the 20-25 mpix camera a Canon in his test, that is all.
These tests have been running for years and, interestingly, they all find roughly the same results. It is just that on the base of these same results, film lovers and digital users argue different conclusions.
The results appear to be that 6x7 rollfilm is about equivalent to a 20-25 mpix digital camera, 4x5 inches sheet film is about equivalent to a 50-60 mpix digital camera (and it is about twice as big as 6x7) and 8x10 inches is indeed unmatched by single shot digital cameras. This is valid for colour film, some B&W film add some extra resolution.
-
I'm sorry but that comparison was frankly laughable...
Not that much laughable. The same web site you cited also compared the A900 with the Mamiya 7. You'll find the results here: http://static.timparkin.co.uk/static/tmp/cameratest-2/800px.html (http://static.timparkin.co.uk/static/tmp/cameratest-2/800px.html). Quite frankly, I find very little difference between the Mamiya 7 and the Sony A900, at least for colour pictures.
That would mean that a 6x7 camera is similar to a 20-25 mpix digital camera, which is also what Michael Reichman found. The 6x7 camera was a Pentax and the 20-25 mpix camera a Canon in his test, that is all.
These tests have been running for years and, interestingly, they all find roughly the same results. It is just that on the base of these same results, film lovers and digital users argue different conclusions.
The results appear to be that 6x7 rollfilm is about equivalent to a 20-25 mpix digital camera, 4x5 inches sheet film is about equivalent to a 50-60 mpix digital camera (and it is about twice as big as 6x7) and 8x10 inches is indeed unmatched by single shot digital cameras. This is valid for colour film, some B&W film add some extra resolution.
Hi Jerome,
I'll retract laughable as they probably didn't know how to get the best out of the film and a drum scan isn't just a drum scan (I own three drum scanners and two high end flatbeds now) as they all are good at different things and sadly many operators don't have a clue. Given this (and the downscaling of their scan to 2000dpi) I won't say it's laughable - just disspointing.
And for the comparison - Yes that's my website - in the narrative you would also read that it isn't a great comparison for the Mamiya 7 as I didn't have the appropriately scaled lenses. The Mamiya had too short a lens for the test (i.e. for the outdoor test we used a 50mm on the A900 and an 80mm on the Mamiya 7 - these are way off to the point that it gives about half the resolution for the Mamiya 7).
The initial test we made was primarily for medium format digital vs large format film where we were within 4 or 5% on comparative lens focal lengths but we thought we'd throw in DSLRs and Medium format film to see what would happen.
Because the tests weren't satisfactory for MF film and DSLRs and the D800 had just come out we went back the next month and shot the comparison you can see above which is a D800E vs a Mamiya 7 using high resolution black and white film and Velvia 50.
I also bought a new scanner as my previous Howtek scanner wasn't as good as it could be - I now use a Heidelberg Primescan and a Screen Cezanne Elite Pro. Interestingly the results above were done on the flatbed Screen Cezanne (a dark horse to say the least).
So in the test above the we shot the Mamiya 7 with the 50mm lens and the D800 with the 25mm distagon.
The comparison shows that the Mamiya 7 is getting about 30% more linear resolution with Velvia 50 (so about 55 megapixels) and about twice the linear resolution with high resolution black and white (about 90 megapixels).
We tried comparing the Mamiya 7 against an IQ180 to see what the difference would be and they were fairly close. The Mamiya 7 was very noisy with Portra 160 film but it resolved as much as the IQ180. In terms of print quality it was probably about the same as the Velvia at about 50Mp.
(http://static.timparkin.co.uk/static/tmp/iq180-mamiya7.jpg)
The high resolution film (Adox CMS20) was a different story. The Mamiya 7 blew everything out of the water and beat 4x5 colour film. Adox CMS20 is just ridiculous and the Mamiya 7 has the best lenses ever sold in a consumer camera in my opinion.
(http://static.timparkin.co.uk/static/tmp/iq180-mamiya7-adoxcms20.jpg)
The camera had a different lens on it but we were at the right distance and the Hasselblad logo tells most of the story.
Tim
p.s. the 1Ds he used was 11mp
-
The results appear to be that 6x7 rollfilm is about equivalent to a 20-25 mpix digital camera
I should also add that Michael Reichman's choice of lens was incorrect. He prints both images to a fixed width and hence we should be scaling the focal length by the short side of the film and sense.
The short side of the 1Ds is 24mm and the short side of 120 roll film is 56mm
24/56 = 2.333
Hence if the DSLR was using a 100mm lens the Pentax 67 should have been using a 233mm lens. As it was a 200mm lens was used. This gives the DSLR a 36% megapixel advantage (i.e. 233/200 is the edge difference hence the square of this is the megapixel count difference).
The test was poor by any count..
Tim
-
Hi,
I have made some tests of my own and spent a lot of time studying Tim's tests, which I feel are really good.
What I have seen in my test was that my Pentax 67 on Velvia resolved better than my Alpha 99 for high contrast detail but the Alpha 900 images were much easier to work with. I also think the Sony was better at extracting low contrast detail
Recently I made a 70x100 cm print from a Sony Alpha 99 image that I also shot on Pentax 67 many years ago and printed 70x100. The prints were very similar.
I would say that film has more resolution for high contrast detail, while CMOS sensors have low noise and are very good at extracting low contrast detail. Another factor may be that Pentax 67 lenses may be a bit weak. I feel that my Hasselblad V series lenses may be a bit better. The Mamiya seven was famous for it lenses, contrary to some belief that Mamiya lenses are not really good.
Tim's results are better than mine, I guess, he uses a high quality drum scanner and he knows what he is doing :-).
Best regards
Erik
-
p.s. the 1Ds he used was 11mp
Oops. Did not notice that detail. I would agree that the comparison was a bit too optimistic for digital then.
And for the comparison - Yes that's my website - in the narrative you would also read that it isn't a great comparison for the Mamiya 7 as I didn't have the appropriately scaled lenses. The Mamiya had too short a lens for the test (i.e. for the outdoor test we used a 50mm on the A900 and an 80mm on the Mamiya 7 - these are way off to the point that it gives about half the resolution for the Mamiya 7).
Maybe. Still: the results of the Mamiya 7 are consistent with the results of other analog cameras, so they cannot be entirely off either.
We tried comparing the Mamiya 7 against an IQ180 to see what the difference would be and they were fairly close. The Mamiya 7 was very noisy with Portra 160 film but it resolved as much as the IQ180.
Come on. The Mamiya 7 is a fine camera and certainly still competitive in today's world, but you are pulling our collective leg here or possibly fooling yourself. In any case, I don't see that it resolves "as much as the IQ180" in your example. The high contrast text is indeed well rendered, but that is an artefact of the sharpening procedure. The lower contrast details on the camera itself are much better on the IQ.
The high resolution film (Adox CMS20) was a different story. The Mamiya 7 blew everything out of the water and beat 4x5 colour film.
Well, yes, of course. If one is prepared to live with the inconveniences of microfilm emulsions (I believe that Adox CMS20 is a remake of the old Kodak Technical Pan), one gets higher resolution.
Quite frankly, I don't really see the point of these analog versus digital fights. The medium are so different that comparisons will never be accurate anyway, so the best we could do is to say "this analog camera with colour film is roughly equivalent to so many pixels give or take 25%" and these 25% is what people fight about. If one needs to be convinced of the capacities of large format film, one should visit an exhibition with, for example, prints by Hiroshi Sugimoto. 8x10 film, B&W, analog enlargements. It is good.
-
Come on. The Mamiya 7 is a fine camera and certainly still competitive in today's world, but you are pulling our collective leg here or possibly fooling yourself. In any case, I don't see that it resolves "as much as the IQ180" in your example. The high contrast text is indeed well rendered, but that is an artefact of the sharpening procedure. The lower contrast details on the camera itself are much better on the IQ.
With high resolution film it does - I'd put good money on that. In fact if you want the raw proof, here's the resolution target 'trumpet' section.
Well, yes, of course. If one is prepared to live with the inconveniences of microfilm emulsions (I believe that Adox CMS20 is a remake of the old Kodak Technical Pan), one gets higher resolution.
Agreed - which is why I've never said that the Mamiya 7 is "better than" the IQ180 (for a start, one is a camera, the other is a sensor).
Quite frankly, I don't really see the point of these analog versus digital fights. The medium are so different that comparisons will never be accurate anyway, so the best we could do is to say "this analog camera with colour film is roughly equivalent to so many pixels give or take 25%" and these 25% is what people fight about. If one needs to be convinced of the capacities of large format film, one should visit an exhibition with, for example, prints by Hiroshi Sugimoto. 8x10 film, B&W, analog enlargements. It is good.
Agreed - I know for a fact that I can get results from my Mamiya 7 with my scanner that beat anything I could get with a D800E in terms of raw resolution.
However I don't shoot medium format (apart from on a roll film back). I shoot 4x5 and 10x8 and I know for a fact that I can get better results than any digital bar the IQ180 with my 4x5 and that I can get better resolution from the 10x8 than anything full stop.
However, that isn't the reason I shoot film - I shoot it because it looks better than digital for me and I also prefer how the cameras work. i.e. Fully corrected lenses that render perfectly to the corners with full movements and the colour of Velvia 50 and Portra which are beyond anything a digital sensor can produce. That and I can afford them and won't have to upgrade ever...
For my work this is more important than any of films downsides.
Tim
p.s. I also shoot 10x8 for commercial jobs where the client has tried the IQ Phase backs and found them not producing what they wanted (battalion photography for the Army).
-
Recently I made a 70x100 cm print from a Sony Alpha 99 image that I also shot on Pentax 67 many years ago and printed 70x100. The prints were very similar.
I would say that film has more resolution for high contrast detail, while CMOS sensors have low noise and are very good at extracting low contrast detail. Another factor may be that Pentax 67 lenses may be a bit weak. I feel that my Hasselblad V series lenses may be a bit better. The Mamiya seven was famous for it lenses, contrary to some belief that Mamiya lenses are not really good.
Tim's results are better than mine, I guess, he uses a high quality drum scanner and he knows what he is doing :-).
Agreed with everything you've said. I've also spent far too much time working out the best way of scanning various film types :-)
Tim
-
I scan 4x5" and 6x7cm neg and pos film with an Imacon 949. I use a Linhof Techno with "digital" Rodenstock lenses and Mamiya 7ii with a full set if lenses. Comparisons I've made echo Tim P's examples. The Imacon doesn't scan at the same res as his Howtec, but the tone and detail on those scans is amazing. In my opinion, they print better, by my subjective measures, than files I've shot on D800 and 5diii cameras. The digi native images just seem to fall apart past a certain size but the film scans keep going.
-
I scan 4x5" and 6x7cm neg and pos film with an Imacon 949. I use a Linhof Techno with "digital" Rodenstock lenses and Mamiya 7ii with a full set if lenses. Comparisons I've made echo Tim P's examples. The Imacon doesn't scan at the same res as his Howtec, but the tone and detail on those scans is amazing. In my opinion, they print better, by my subjective measures, than files I've shot on D800 and 5diii cameras. The digi native images just seem to fall apart past a certain size but the film scans keep going.
Hi TJ,
Are you using a digital back too? I'd be interested in which one if so (I presume you aren't using a film back on the Techno - although that would be as impressive as the Mamiya I imagine).
If you want any of your shots scanning on my Screen Cezanne feel or Heidelberg Primescan feel free to send them over - I'd be interested in how they compete with a good 949.
Tim
-
If you want a good comparison of film vs digital, look here..
https://www.onlandscape.co.uk/2011/12/big-camera-comparison/
This was done with the cooperation of Phase and a dedicated Phase photographer who used to shoot film. It's about as unbiased as you'll get.
The end result? It's a damn sight easier to get good results from digital than it is from film but if you do persevere with film the results are as follows
perceived sharpness gives 5x4 equivalent to an IQ260 - 10x8 is about 120-150 megapixels
In terms of resolution and raw detail - the the 5x4 exceeds the IQ280 for colour resolution and just about matches it for fine detail.
In other words the IQ280 is equivalent to 5x4 up to about 20x24 when the IQ280 looks better. At about 30x40 they start to be similar. Then the 4x5 starts to look better because the fine detail carries on (albeit with grain and low contrast).
for most purposes medium format film drum scanned just about beats 24-36Mp DSLR's but you have to have great lenses and technique (plus a camera that holds the film flat!).
Sounds right to me. You also have to factor in the advancements in stitching that make the limit on digital close to the angular resolution of your lenses. Add the ability to check to ensure you have the shot as you like it.
Big sheet film is still a competitive system for the craftsperson. The color, as you say, is excellent, the system price is very low. The archival durability is also a significant advantage. You don't have to transfer everything to another purchased hard drive as the warranty on the old one runs out every 3-5 years. That is another big cost of digital.
-
Tim,
Ever since seeing your test, I started shooting a lot more film! :) I think your tests were the best conducted ones I've seen and do not seem to have any built in biases. That there are so many different results only show how difficult the testing can be. When the shot is controlled and the film well handled and scanned the results are quite impressive. That said, I had wondered about those mamiya 7 portra 400 images - mentioned a few threads up - when I saw them.
Comparing film to digital is probably tricky business and I'm glad to leave that you and other experts. But I do make my own casual observations in my own shooting as my Hy6 takes both film and digital backs, I have had the opportunity to make some interesting comparisons using the same lenses. My conclusions are both have their advantages under different conditions. For shear detail, though I'd have to think the 80mp digital back is ahead under most circumstances when shot between ISO 50 and 200 for sure and probably at 400. After that its starts to be unclear. But there's so much more than just detail. The look, the tonality, etc. I don't like to shoot my AFi-ii 12 back at ISO 800 and really not 400 either, but I'm happy shooting film at 400 and 800 and even much higher.
I like to use black and white film in the camera when I'm shooting a high contrast scene as I feel I get nicer results that include some detail in shadows and highlights. I also like to shoot film when I'm going to be above ISO 400 really with the digital backs as they loose a lot going away from base ISO - - its an area where film just seems to work out better. Film also manages to show nicer tonality and can be really forgiving on models skin - saving a lot of retouching time. I think taking the film all the way to print saves some of its goodness as well.
I think also that the larger formats offer a look that can't be replicated with smaller formats and in this area film stands alone. I don't think we'll see 8x10 sensors anytime soon.
btw - back to the original topic - I do think the sample images and the Castorscan work are both very good.
-
Hi Tim,
I am indeed shooting film on my Techno. I use a Horseman 6x7cm back. Film flatness is sometimes a problem due to the transport mechanism, but as long as I shoot each film within a day of loading it things are generally okay. In contrast, the Mamiya 7 keeps things dead flat all the time.
I bought the Techno as part of a slow transition to digital capture, hence don't currently own a digital back. I'm eying up an IQ160 for the future.
The Rodenstock lenses I use are amazing, particularly my 90mm HR W (blue band). I must say though that the Mamiya 7 lenses, particularly the 50mm, are every bit their equal, all be it without the possibilities of movements. Man, it'd be awesome to have a digital Mamiya 7 system!
I live in New Zealand, so a comparison set of scans might be a problem. It's a generous offer and I bet a done properly a wet scan, particularly using your Heidelberg, would be in an other league than the 949. I absolutely LOVE my 949, but it ain't a drum scanner, despite the marketing material that goes along with it. I will say though that it's absolutely foolproof. There is no one I can send my film to here in NZ that does an even halfway okay job at drum scanning. People just don't seem to take pride in their work, or at least they don't seem to know how to get the best out of their machines. In contrast, your scans look beyond amazing!
TJV
Hi TJ,
Are you using a digital back too? I'd be interested in which one if so (I presume you aren't using a film back on the Techno - although that would be as impressive as the Mamiya I imagine).
If you want any of your shots scanning on my Screen Cezanne feel or Heidelberg Primescan feel free to send them over - I'd be interested in how they compete with a good 949.
Tim
-
the Dainippon Screen 8060 Mark II is definitely in another league than any Heidelberg drum scanner, in terms of optical resolution, color transitions, noise-grain reduction.
The Heidelberg drum scanners are even surpassed by many other scanners, including ICG 380, Aztek Premier etc...
-
the Dainippon Screen 8060 Mark II is definitely in another league than any Heidelberg drum scanner, in terms of optical resolution, color transitions, noise-grain reduction.
The Heidelberg drum scanners are even surpassed by many other scanners, including ICG 380, Aztek Premier etc...
In the past 3 years of searching, I've NEVER come across a used 8060P MkII scanner on the used market. NOT ONE. I know they were not produced to the quantity level of Heidelberg, Howtek, or even Aztek units, but still... Maybe they're just 'too good' to scrap out, or they were sold to scrap guys rather than going for sale, publicly.
I have an Aztek DPL8000, and it works very well. Pretty much the same as a Premier, optically/PMT wise. I've not had any scans made on a Screen, but from what I've seen of your work, the Screen seems like a very capable scanner.
However, there is ONE BIG THING: the Screen is HUGE. My DPL8000 is 2'x3' approximately, and two guys can easily put it into the back of a volvo station wagon for transport. I don't believe that is possible with the 8060SP mkII? I know, I know, these machines don't get moved very often(or ever!), but the sheer size of the physical machine is a huge difference. Folks like myself, who don't have loads of space, but still want to produce their own scans, like these smaller machines that can fit on a desk(a solid one!), or tucked into an office closet sitting atop a coffee table, etc... The Screen is big.
I'm only shooting 5x7 and smaller formats now. I had an 11x14 camera last year, but have unfortunately sold it. I'd have loved to shoot it for a portrait project I'd like to start in the next year, but looking at the benefits for the print sizes I'll be making(20x24 approx and smaller), I could easily get the quality level I need out of a 4x5 even. For someone like yourself, who scans for ULF and LF shooters(I've seen some of those pictures of 12x20" color negatives :) ), I can understand your needing a machine that can handle film that large.
However, if time, money and SPACE were allowing, I'd love to get one of these Screen units like you have, simply due to the sheer mounting potential over the smaller drums that my Aztek scanner uses(8x10 is a tight squeeze!)...
-Dan
-
You can get a Howtek 6500/7500 for a song these days and they will scan up to 18.5x24 with 2500 ppi on the 8" diameter drum. Plus, you can still get part and service.
-
I gave a screen (1045 I think that was it's number ) away a few years back. The oil mounting and scanning time was more than I could bother with. Plus getting a scan from colour neg was a real hard task, even though I had the negascan software.
To me shooting with film is not about resolving power, I still have not seen a digital file that has the look of Portra film. Portra does bright colours in a none over saturated way, they can be colourful and have gradation. They can be subtle but not bland. TV is awash with digital footage that has a film look added in post. To me it looks like digital with a film look added and not like film. I don't get why quality is judged by zooming into an image and declaring a winner by the amount of pimples you can see. Why not just look at a print as it should be viewed and decide which moves you the most.
I was viewing a photographers Horse portraits this weekend, printed about 40x30 and shot with a 5dII, really lovely images, seeing more hairs would not of made them even better by any amount. As long as the mechanics of noise/grain artefacts are not getting in the way of the image as you like to present them what does any of that matter?
It's got to be the overall look of the finished image that counts, not component parts taken separately.