Scanning backs for landscapes

[lepingzha]

Nobody can be always accurate on everything, but some always
tend to stretch things only and always along their favorite directions.

Look at MR's introduction in which he stated the comparison was
between the scan back and the "4"x5" drum scanned film".
Obviously MR did not really read the Richard's article that is
a great one, since in the article it is clearly stated the film
scan is from a flatbed scanner, not a drim scan, and that he
did not move the 4x5 camera so he did not use the full
4x5 film area but only the central 72x96mm.

It is only a little stretch of course.  But I do not think it need
a scientist to get things straight this time.

LEPING ZHA, Ph.D.
www.lepingzha.com
 

[BernardLanguillier]

I had also considered the scanning back option, but also reached the conclusion that it is impractical most of the time.

There was an interview of Stephen Johnson a couple of issues of the Video Journal back which was very interesting. Stephen is using a Better Light back for landscape applications. However, to be completely honnest, none of the images he showed during the interview were very appealing to me.

The image quality seems to be amazing, as well as the colors, but I wouldn't call the part of his work that was shown really exciting. I guess that giving a look at the huge prints in the flesh would probably change this impression. I might be too influenced by Photosig where the appeal of an image is judged solely based on a 150 pixel tall thumbnail.

My personnal view is that scanning backs are an exercice in style that can be interesting, but that there are other areas of photography that are likely to be much more rewarding at a much lower cost and inconvenience. I guess that pros in fine art feel a strong need to differentiate themselves from the mere amateurs, and that using such gear is an efficient way to do so though.

Regards,
Bernard

[Jonathan Wienke]

Nobody can be always accurate on everything, but some always
tend to stretch things only and always along their favorite directions.

Look at MR's introduction in which he stated the comparison was
between the scan back and the "4"x5" drum scanned film".
Obviously MR did not really read the Richard's article that is
a great one, since in the article it is clearly stated the film
scan is from a flatbed scanner, not a drim scan, and that he
did not move the 4x5 camera so he did not use the full
4x5 film area but only the central 72x96mm.

I don't see the word "drum" anywhere in the introduction.

As to the rest of your criticisms, I fail to see any point whatsoever. Perhaps you missed this statement in the article:

"So, for the record, this is not a review of Better Light backs or a “bench test� between a scanning back, a high-end DSLR, and 4x5 film. I’m merely sharing my experiences based on one shooting experience and sharing the photographic results of this shooting experience for the sole purpose of providing useful information to any photographer interested in learning more about the relative attributes of the three means of capturing a photographic image that were explored and documented."

[Eric Fredine]

Thanks Jack - it's always nice to have a reply that at least addresses the question.

And I think you have gotten to the crux of what always confuses me in these scanning back discussions - How is that your friend is getting scan times of a few minutes in low light when Richard Sexton took nine minutes to get a scan of a reasonably well lit scene?  What are the different set of choices your friend makes (reduced resolution? higher gain or ISO?) to get these within the realms of reasonable scan times and how does this effect image quality?

[Quentin]

So is it viable in the field?  Absolutely.  In fact, it really isn't much different than using 4x5 film, especially when you factor in bracketing, which you don't need to do with the BL.  

It's viable right up to the moment there is a breath of wind, or an animal moves.

That's the scan back's weakness.  Any movement and you see rgb streaks and odd halos.  Great for rocks and stones, but forget moving water or trees waving in the wind.  Its at that point you wonder what on earth you are doing, and wish you have some nice, simple, film  

Quentin

I "quietly" removed the word "drum" as soon as I realized that I'd used it in error.

I am working this month from a location where I can only access the net occasionally, and then only on a 14K dial-up line. It sometimes takes me a while to be able to get back on line and correct my mistakes, even when I recognize them.

Yup, I make 'em. Lots of them. Must be nice to be someone practicing the scientific method, having no biases, and not being one to make mistakes. Being human is tough, but someone's gotta do it.

Michael

[paulbk]

Curious.. where/what is the market for scanning back images?
backlit airport advertising posters?
two meter prints in hotel lobby, corporate head quarters?

Without question, for those limited situations where a scanning back works (static scene) the quality is stunning. But I’m thinking, you need a very discerning customer with a fat wallet to both “see the difference� and pay for it.

[BernardLanguillier]

Being human is tough, but someone's gotta do it.

Michael

Michael,

No worries, I don't think anyone really cares about such details.  Keep up the good work and, as U2 said back then, "don't let the bastards grind you down" (no offense intended to anyone).

As far I am concerned, I found the article to be interesting and completely fair in its conclusions. The gap I see between my high end DSLR and scanned 4*5 film is a bit larger than what was shown in the article, but I scan with an Imacon and use Velvia.

Regards,
Bernard

[collum]

Curious.. where/what is the market for scanning back images?
backlit airport advertising posters?
two meter prints in hotel lobby, corporate head quarters?

Without question, for those limited situations where a scanning back works (static scene) the quality is stunning. But I’m thinking, you need a very discerning customer with a fat wallet to both “see the difference� and pay for it.

Galleries.

If you check the galleries, images by Cramer, Neil, Burkett are all shot with large format (Burkett's larger than the others).. and all exhibit prints up to 40x50.. with normal sizes being 20x24.

The images I sell thru the gallery typcially end up in corporate environments (meeting rooms, hallways, entryways, etc). In that situation, the bigger the better... and companies usually buy in quantities as well

           jim

[collum]

i just checked the exif

ebony 2 slow, 6/30/05, 3:03 PM,  8C, 6000x8000 (0+0), 100%, straight 6 sto,  1/15 s, R27.9, G16.6, B29.7
180mm Schn Apo-Symmar at f22

with 1/15, it's a 9 min scan.

[Eric Fredine]

I've always assumed that scanning backs are largely impractical for many landscape photography situations due to the very long scanning times.  But there seems to be very little data available about the actual scan times required.  Richard Sexton's article seems to have provided some basic data.  

His photographs were taken with an exposure of 1/30 @ f/11 (if I've interpreted the data in the article correctly) which resulted in a scan time of 9 minutes (I'm assuming the view camera was stopped down to f/22 since the B&W exposure was 1/8 of a second).

Well, this is actually a pretty short exposure for many landscape photography situations.  Something like 1s @ f/16 isn't unusual -which is 16 times as long as the test exposure.  I'm assuming the scan time would also increase by a factor of 16 - so the total scan time would be 16 X 9 minutes = 144 minutes which is well over 2 hours in duration - not practical in most (or at least many) situations.  Which I guess is why the landscape samples are always of well lit scenes (no edge of day stuff) or completely stationary subjects (like a pile of rocks).  And even at 9 minutes, I would think that cloud movement would start to create problems.

Am I missing something?

[George Barr]

The maximum (longest)exposure time on a betterlight back is 1/8 second. Your only option to give the equivalent of a longer exposure is to increase the amplification (equiv. to film speed). I have not seen noise examples at 2000+ ei. but know that 1/8 of a second is typically the shortest exposure I use shooting 4X5 with 100 speed film so for me the issue isn't so much how long I have to wait for the scan, its that to get an exposure on each pixel of 1 second, I need to amplify to ei. 800, 2 seconds ei. 1600 etc. This makes a scanning back essentially useless for the kind of landscape photography I do, often well before sunrise or after sunset or in deep deep shade or in a cave or under a bridge or...
This information was given to me by betterlight themselves.

[Quentin]

I tried using a scannng back for a while but the limitations in terms of avoiding any movement while the scanning operation takes place are too resticting for most purposes.

Incidentally I regard the quality from a scan back to be considerably better than any dslr.  The types of subjects shot in the test don't really show up the advantages of the resolution difference.  A building with fine brickwork or distant signs will show up the gulf that exists more clearly.   Bearing in mind these are true rgb pixels, the 6,000 x 8,000 ppi resolution might be equated to 12,000 x 16,000 mosaic (bayer) type pixels (working on the normal 2x foveon assumption).  

Quentin

[Jack Flesher]

A shooting buddy of mine has the Betterlight back and I have shot with him in the field while he uses it, so I feel I am at least partially qualified to offer some insights.

As for scan times, we were shooting together in dense fog along the California coast, so there was not a lot of light.  My 1Ds2 exposures were on the order of 1/4 to 1/15th at f8, ISO 100.  If I had been shooting 4x5 Velvia, my times would have been around 8 sec to 2 sec at f22, accounting for reciprocity and filtration.  By comparison, Jim's were running about 2 minutes per scan.

So is it viable in the field?  Absolutely.  In fact, it really isn't much different than using 4x5 film, especially when you factor in bracketing, which you don't need to do with the BL.  

As far as paraphenalia, the back itself replaces a readyload holder, the black box + battery replaces your box of film (and is smaller than say 10 film holders), which leaves the laptop as the real excess baggage.   Yet with the laptop/BL, you not only get perfect exposure, but can get essentially perfect focus as the software has an edge-contrast measuring feature that displays a numerical readout for whatever edge(s) you select.  Yes, very slick!

And finally, one word about the files: INFREAKINGCREDIBLE!  Essentially a 1.3x crop on a 4x5 sensor at about 140 effective MP with true 14-bit per channel color.

The real question IMO is do you need it?

For me, I am watching and waiting.  I already have the 4x5 cameras and lenses and the laptop computer, so for me the expense quotient isn't all that great.  But the larger point is field convenience and how much better than the 1Ds2 file do I really need?  Heck, right now I can easily get to 24x36 with the 1Ds2 and the prints still look great...  But the BL 24x36's look better.

~~~

EDIT: BTW, I should add that I often will do a flat-stitch with shift lenses on my 1Ds2 to increase resolution.  This requires I capture three frames, a center, left and right, which further requires I counter-shift the camera to avoid parralax.  (While the counter-shifting is imperfect, it is generally close enough that very little parallax needs to be dealt with in post.)  The counter-shifting adds time to the trio of exposures, but I can typically accomplish all three captures in just under a minute.  BUT!  I still need to combine those images later in post, and the time required to do this manually or with software is significant, and far more so than the added capture time required from the BL.  When all is said and done I end up with an image of effectively 30MP at best, so the BL still trumps it there as well.

Heck, I may have just convinced myslef!  
:laugh:  :laugh:  :laugh:

Cheers,

[Jack Flesher]

He sent me a crop from one scan that took under 30 seconds that morning.  He was experimenting with ISO 4000 capture.  It had about as much noise as my 1Ds2 at ISO 640(!), and that was before NEAT.

His "normal" ISO for the field is around 1000 ~ 2000, but even to my critical eye, there is almost no visible noise at 2000 unless you look deep into shadows, and even then it's all but insignificant.

[lepingzha]

I don't see the word "drum" anywhere in the introduction.

As to the rest of your criticisms, I fail to see any point whatsoever. Perhaps you missed this statement in the article:

"So, for the record, this is not a review of Better Light backs or a “bench test� between a scanning back, a high-end DSLR, and 4x5 film. I’m merely sharing my experiences based on one shooting experience and sharing the photographic results of this shooting experience for the sole purpose of providing useful information to any photographer interested in learning more about the relative attributes of the three means of capturing a photographic image that were explored and documented."

1. Michael quietly removed the word "drum" last night.  Anyone
 Â   seeing the introduction yesterday or have a hardcopy can
 Â   testify the word was there.  I did not believe MR monitors
 Â   these forums at daily base and read all the new posts, but
 Â   obviously he does.

 Â   If I were MR I would add a line pointing to the error rather
 Â   than the quiet removal.  I do scientific researches and if
 Â   I do the same there it would be a scandal and my peer would
 Â   be questioning my integrity.

 Â   Another example was MR retrieved his posting on the Nikon
 Â   D2x noise but had never back to the subject to provide us
 Â   the facts.  There are objective measurements everywhere
 Â   else online.  I do not shoot 35mm but MR is well known of
 Â   his associations with Canon.

2. My second point was: as Richard Sexton, the author mentioned
 Â   clearly in his article, since the scan back covers an area smaller
 Â   than the full 4x5 film area, ideally he should have moved the
 Â   camera closer to the Ebony (which I happen to have and use)
 Â   for the same FOV, since obviously he had to move the 1DsII
 Â   much closer for the same purpose.  But since he didn't what
 Â   he actually compared was the scan back and a reduced frame
 Â   large format film not the FF 4x5, that gives the scan back a
 Â   little bit advantage in resolving power.

 Â   The recording area of Kodak Readyload 4x5 film is 92x119mm
 Â   and the area of the Fuji Quickload is 95x120mm, and for the
 Â   typical film holders the area is 96x120mm.  Using the Kodak
 Â   numbers, which is the smallest in the pack, the digital scanback's
 Â   "lens factor" is 92/72=1.28 along the short dimension, and
 Â   119/96=1.24 along the long side.  Thus the 1.25 magnification
 Â   factor quoted in Richard's article is accurate.

I really don't believe the use of 4x5 FF frame (with the associated
possible refocusing errors and DOF changes) and drum scan would
make visible difference on the conclusions.  Personally I found
drum scanned images shows more film grains than the Imacon
FlexTight 848/949 scans I use.  I was only making a comment
that MR's incorrect use of the term "drum scanned".

LEPING ZHA, Ph.D. in Physics
www.lepingzha.com :laugh:

[Jonathan Wienke]

I do not shoot 35mm but MR is well known of
 Â  his associations with Canon.

2. My second point was: as Richard Sexton, the author mentioned
 Â   clearly in his article, since the scan back covers an area smaller
 Â   than the full 4x5 film area, ideally he should have moved the
 Â   camera closer to the Ebony (which I happen to have and use)
 Â   for the same FOV, since obviously he had to move the 1DsII
 Â   much closer for the same purpose.  But since he didn't what
 Â   he actually compared was the scan back and a reduced frame
 Â   large format film not the FF 4x5, that gives the scan back a
 Â   little bit advantage in resolving power.

The methodology of the comparison was clearly described, and a lengthy disclaimer was made that the comparison was not intended as a detailed test-bench resolution contest. The article never claimed the comparison was anything other than what you stated (equal imaging area with same lens and subject distance between the BetterLight and film), so there's no deceit or false claims being made. And if anything, I would say the comparison was an advertisement for BetterLight, not Canon. Take your conspiracy-theory silliness elsewhere.

[Jack Flesher]

Any movement and you see rgb streaks and odd halos.

Absolutely true.  Best if your subject matter remains within the limitations of the medium

[Jack Flesher]

two meter prints in hotel lobby, corporate head quarters?

That's where many of Jim's images end up.

[collum]

His photographs were taken with an exposure of 1/30 @ f/11 (if I've interpreted the data in the article correctly) which resulted in a scan time of 9 minutes (I'm assuming the view camera was stopped down to f/22 since the B&W exposure was 1/8 of a second).

I haven't checked the exif data.. but the total scan time is calculated by the # of pixels scanned (in this case 8000) over the scan time (you said 1/30).. that would give a total scan time of 4 min 26 sec.

I'm the "Jim" that Jack's mentioned, and shoot with the Betterlight (and 1ds mk2). The high ISO settings are *very* useable.. very little noise.. and what is there is easily cleaned up with Neat.  I shoot almost always in the morning before the sun rises.. or on overcast days.. never in direct sunlight. The longest scans for me are 4 min, but typically are about 2 min.

I don't shoot sweeping landscapes.. everything is normally shot at 200 feet or less (and less is up to a couple inches).   In this sense, i find the Betterlight very usable.

I'm in the process of starting a series of articles on the Betterlight from a landscape photographer's perspective. I'll be taking it (along with the 1dsmk2) on a trip to Cambodia and Thailand in October, so will be reporting back on the experience as well.

Quality?.. yup. beats drum scanned Provia. 9-11 stops of latitude (of course, i seldom need that given the type of light i use). Color fidelty? film doesn't even approach.

there are minuses. .yes.. wind really s*cks. but i'm normally at low shutter speeds on the 1dsmk2, so wind is a bother for me regardless of the camera (shooting in the early morning usually removes this worry).

anyway.. i use either one.. depending on the situation. Most of my work is printed at 16x20 or larger.. the Betterlight gives me the ability to easily print at 40x50.

 Â        Jim