22mp vs. Large format?

[mb81]

Well, I know how I feel about my 8x10 camera. I love it, and I

honestly feel that digital will never touch the quality of anything

above 4x5. But I want other opinions.

Today, with traditional photography slowly dying away, less paper,

less film, I cant help but to shudder. Don't get me wrong, digital

has its place, but when ultimate image quality comes in, I just dont

think it will cut it.

Now we have the 20 and 22 megapixil cameras. People all over are

saying goodbye to 4x5 and even 8x10. Crazy???? I think so.

 

Opinions?

[jbq]

Well, us what you feel does the job best for you.

 

Of course the comparison is subjective, but I wouldn't totally discount the recently announced 39MP backs against large format film, and if that one doesn't doesn't do the trick yet we're possibly only a generation away from being there.

 

There's certainly quite some room for growth as far as sensor sizes and pixel sizes go, I almost take it for granted that we'll see 100MP sensors within this decade, in the current form factor, and I wouldn't be surprised to see large gigapixel sensors within the decade.

[rob_landry]

I don't think it'll matter if you can pack 100MP on a small sensor, you'll probably reach a point of diminishing returns, where the lenses just won't be able to resolve the details to get the information faithfully to the chip. Kinda like trying to write an encyclopedia on the back of a postage stamp. In the end, I think square inches will still win out simply because much less enlargement is involved for any given print size. There's just no substitute for square inches. Don't retire your LF cameras just yet.

[mark_tomalty]

Mike

 

I think you might be surprised at the quality that can be achieved with todays crop of

22Mp backs.I shoot multiformats for different applications among them 35 digital with

a 1DsMk2,Contax 645 and 4x5 for large format landscape and 612 pan work.

I scan my own work with an Imacon 848.

In anticipation of the new P45 from Phaseone I had the opportunity to use a 22Mp P25

for a few days recently.

Files were incredible,to say the least and noticeably cleaner with more dynamic range

than my 1DsMk2 but not 5X better than the price differential would suggest.

Test prints in the 24x 36 range were excellent and were not noticeably inferior to

equivalent sized prints I've done with 4x5 velvia. In my opinion there is a slightly

'smoother' look to the digital capture and the impression of a little more ultrafine

detail in the 4x5 film images. Undoubtly there will be measureable differences between

the two mediums but when it comes down to it there is not much to differentiate the

two in actual prints if that is what your end game is. FWIW,carefully processed 1DsMk2

prints at this size did not give up much to either despite the fairly noticeable differences

on screen at 50 and 100%.

Assuming there are no significant problems with the new 39Mp back there should

be no debate that one would be able to achieve 'handholdable' 4x5 quality in a

very compact and portable package. Using a Contax 645 with such a back would yield

unbelieveable results with marco and wildlife imagery where previously the format

could be seen by some as a limitation.

 

The kodak chip in the Phase,and Imacon, back is subject to a color shift from green

to magenta across the frame and requires one to take an additional 'calibration'

frame with a dedicated opaque white balancing disk which one holds over the lens

Very simple but one more step nevertheless. The problem is most acute with wide lenses.

Working with the backs on 4x5 and 6x9 cameras with movements also requires some

carefull research and I would recommend you check the 'medium format digital'

threads at Rob Galbraith to consider a number of issues that are confronting

photographers already working in this area.

 

In my opinion,despite the momentum(and quality) digital capture has at present

the larger film formats still offer ultimate quality at an extremely attractive

pricepoint and a pleasant shooting experience to boot.

 

Mark

[brian_c._miller]

Resolution isn't the only thing to consider. The other part of it is the flexiblity of large format. The Schleimpflug principle does not apply to any box camera.

[kelly_flanigan1]

Here I use to shoot artwork with 4x5 transparency films; and drum or flatbed scan it. After alot of losses; scratches; dealys by local "pro" processing; I got a 35Mpixel 4x5 scan back 1/2 decade ago. Basically local 4x5 trany developing got expensive; senile; drunk; flaky; with delays before dying off. The Phase On scan back was expensive; but did the trick.

[marvt]

Add to the other thoughts the fact that scanners for home use will continue to improve making a blend of digital and traditional even more attractive.

 

I still savor the view on the ground glass and the look and feel of the mahogany and brass. It puts me in a frame of mind that says, "One shot, make it count." Then when I'm back in the digital darkroom I have controls in PS only dreamed about a few years ago.

 

As far as the reduction in papers and film, I have used Tri-X Pro and Zone VI or Ilford VC FB papers for 25 years. I don't need a lot of choices I guess.

[ellis_vener_photography]

The digital backs that make 22mp (and larger) resolution iamges are pretty much

commercial items made for commercial and portrait studios that can amortize the over

$20,000 cost (and that is just for the back and dedicated software) price within a year or

two. And I suppose a few wealthy amateurs are buying them too.

 

A friend of mine describes the situation this way : think of the cost of cameras like the

D2X and the Canon EOS 1Ds mk.2 and the medium format backs as if you bought a truck

load of film. That film has about a two year expiration date -- and you have to think the

same way about these cameras and backs.

[troyammons]

Just my opinion so...

 

When I drumscan an E100G 4x5 with one of my sharpest LF lenses at 2000-2500 dpi it is very digital looking. About like a bayer dslr for edge sharpness (about 2 pixels) and not much noise. It is very clean and smooth. At 2000 dpi that roughly works out to 80 mp. I guess the extra 40 + mp over a 40mp D back might buy you some extra tiny detail, but I would doubt it would be that much. 8x10 is 4 times that so in that case I think it would still be way on top.

 

Who can afford a 40mp back anyway ?? Not me and if I could I would probably be scared to death I would drop it. If I were shooting models all day at 10G per day it would make sense.

 

35mm digital is going to hit a wall oneday. You can only pack on so many sensors, before you have problems with the long wavelengths. I read exactly where that is but for the life of me I cant remember. There are not that many MF camera lenses that can handle ultra high rez chips. I guess Rollei, Contax, Pentax and hasselblad lenses would be up to the task. Maybe Mamiya, but I was not truly impressed with RB/RZ lenses.

 

Also a lot of digital cameras dump due to electronics problems, like 2 of mine did, so I would expect to lay out some serious $ on repairs after the warranty expires, especially if you intend to keep it for 15-20 years.

 

I think its just like the 35mm/MF film vs digital question. If you need a fast workflow go digital. If you are only interested in 20 primo LF photos a year, shoot film. Personally I enjoy looking at LF film. Its tactile and archival, and in 100 years hopefully one of my ancestors will still have it.

 

As far as E6 processing, there are still a few places around here that do 4x5 and MF, and actually MF is so cheap I just let them develop it. Nobody does 8x10 anymore, so I ended up buying a Jobo CPP2 for LF B+W and E6. The CPP2 makes developing very easy. Expensive though, but considering I can do 10 4x5's, or 5 8x10's in a batch or a single 20x24 print it makes sense.

[kelly_flanigan1]

With a digital back one can cheat; ie zero in on the exact focus of a tabletop photo settup. One can do a prescan; of full scan of just a narrow rectangle. This allows the +/- .007 inch or less film holder to film plane to camera tolerance to be real reduced that one can get with film. Here I have found that with several digital trial focus runs that I often get better focus with fast lenses and closeup images. The 4x5 obsolete digital scan back I use is a slow several minute scan; tethered to a computer; good for still life; dicey for landscapes.

[eric_leppanen]

OK, first the speculation....

 

My understanding is that 35mm digital has one more "kicker" coming before it reaches lens resolution limits: the availability of Foveon-like "each pixel sees all colors" sensor capability (i.e., no more Bayer array interpolation, no more anti-aliasing filters). What this will equate to in terms of resolution I have no clue, but presumably 35mm digital and 4x5 will reach equivalence at least at 16x20" print sizes (perhaps much larger). Since 16x20" prints and smaller comprise the majority of the fine art print market, one would seem hard pressed to justify continued use of 4x5 presuming decent tilt/shifts lenses are available (admittedly the 35mm digital vendors would have to invest in new versions of these lenses, which to date they have been loathe to do). Thus digital equivalence to 4x5 will be achieved for most applications without resort to medium format digital backs (with their problematic cost-effectiveness for the average shooter). To really go out on a limb, the guesses I have heard is that Canon/Nikon/etc. will announce such products in late 2007, with production in 2008.

 

From strictly a resolution perspective, MF digital backs would replace 8x10 under this scenario with similarly upgraded sensors. Yet 8x10 film (in theory) would still be far more cost-effective than MF digital for the foreseeable future, as MF sensors are by nature large, and large die-sizes will remain expensive. The big question would be how long 8x10 film would remain available.

 

Of course the problem with all this speculation is that, even when resolution-equivalent digital cameras become available, how many people will stay with film due to price and how many emulsions will remain available? I have been told that Fuji intends to keep making color film until the cows come home, that it is too much of a cash cow (they certainly keep introducing new emulsions, unlike a certain yellow-themed company). The motion picture industry is still using mostly film. There is still no industry standard for archival digital storage (CD's tend to die after a few years). And looking at the LF resale market (which is extremely strong), there is certainly no indication that the marketplace foresees any viable LF replacement for the immediate future.

 

My take is that we'll have to put up with continued consolidation in film emulsions, chemistry, paper, labs, etc., but that analog solutions of some breadth will continue to be available for many years. We may just have to pay a bit more for it.

[mb81]

Being cost effective is a huge problem with Digital today. Most people just cant afford these 22-39Mp backs/cameras. Its just too much.

And what about file size. I have not used these cameras, but the file sizes must be HUGE.

It seems to me that as more digital products come out, the more stuff you need to buy to support the digital products. People say digital is more cost effective, maybe for the consumer, not for the professional (unless your a fashion photographer).

 

I stand by my 8x10 camera. It may take me damn near 15 minutes to take one shot. But those minutes are all incredible.

I just have more fun with film. But who knows what the future will hold.

[michael_briggs2]

LF photography is more than high resolution capture, either film or silicon. It is also about seeing well to make the photo -- a large ground glass to view the image, to compose and judge focus, depth of field, etc. A small viewfinder can't deliver the same view, so no matter how many megapixels are behind the lens you can't make the photo in the same way that you can with a 4x5 or larger camera with ground glass viewing.

 

I agree with the others that 22+ megapixel chips are aimed at a large studios and similar outfits. Between the cost of large silicon and the nature of the target market, I don't expect prices to come down much. I think film with remain available, for amateurs and specialized commercial uses.

[troyammons]

>Eric Leppanen , aug 13, 2005; 12:45 p.m.

>OK, first the speculation....

>My understanding is that 35mm digital has one more "kicker" coming >before it reaches lens resolution limits: the availability of >Foveon-like "each pixel sees all colors" sensor capability (i.e., >no more Bayer array interpolation, no more anti-aliasing filters). >What this will equate to in terms of resolution I have no clue,

 

It will be a lot. I have compared my SD9 to 35mm film shots and to other bayer heavily AA'd cameras, and its roughly about

2-2.5 to 1 foveon to bayer. Foveon 3.4 x 3mp equals roughly 8mp bayer unless its not AA'd like the 14n. Lets just call it 2.25 to 1 for a digital comparison for now. The original research chip that Foveon produced was 16mp x 3 full frame 35mm with a tight pixel pitch so that would roughly equal 36mp bayer. Not too shabby, but you would need better lenses than most sigmas to make use of it. As far as film I feel the 3.4 mp foveon roughly equals a 35mm E100g drumscan, but other people would argue more on the side of Foveon. In the end there should be more rez available with a foveon type chip than bayer within a limited size when all the 35mm dslrs get maxed out. I guess it is possible that someone else might come out with another RBG per well chip one day if they can get around the patents. Seems like the logical way to go to me.

 

>but presumably 35mm digital and 4x5 will reach equivalence at least >at 16x20" print sizes (perhaps much larger). Since 16x20" prints >and smaller comprise the majority of the fine art print market,

 

Actually the SD cameras do pretty well at that size print but personally I like them printed at around 12x18 max. Also i am sure they are working on the next generation chip which should approach MF. I have asked Foveon several times, why dont you stop screwing around and build a 645 super chip and blow everybody away. Of course no reply. Maybe they are working on it.

 

At the current SD9 pixel pitch a full frame 35mm foveon chip would resolve around 54lp/mm and would be 4000x2660 or around

11mp x 3.

 

A 645 Foveon chip at the current pixel pitch would be around 32mp x 3 or just huge. If you figure 32mp x 2.25 factor you end up with an equiv 72 bayer MP. Although a different format, when i drumscan E100G 4x5 at 2000 dpi it is almost a dead ringer for bayer digital and 4x5 at that scan level is around 70-80 mp. At that level foveon digital would match 4x5 film IMO. The new 40mp digital backs should be getting close, or match 4x5, depending on NR and AA.

 

>one would seem hard pressed to justify continued use of 4x5 >presuming decent tilt/shifts lenses are available (admittedly the >35mm digital vendors would have to invest in new versions of these >lenses, which to date they have been loathe to do). Thus digital >equivalence to 4x5 will be achieved for most applications without >resort to medium format digital backs (with their problematic cost->effectiveness for the average shooter). To really go out on a limb, >the guesses I have heard is that Canon/Nikon/etc. will announce >such products in late 2007, with production in 2008.

 

I dont think they can do it in 35mm frame. Not match sharp 4x5. 6x7 but not 35mm. Not enough room. I guess it all depends on enlargement size anyway. If you are only enlarging to 11x17 there is not much reason to shoot 4x5 except for movements, but recently Chinese companies have come out with some pretty neat MF/movement cameras that take Hassy lenses and such.

 

>From strictly a resolution perspective, MF digital backs would >replace 8x10 under this scenario with similarly upgraded sensors. >Yet 8x10 film (in theory) would still be far more cost-effective >than MF digital for the foreseeable future, as MF sensors are by >nature large, and large die-sizes will remain expensive. The big >question would be how long 8x10 film would remain available.

 

I hope film is around for a while. I shot 8x10 the other day and just screwing around I scanned it at 2000 dpi and got a 2 pixel edge. I scanned a crop at 4000 dpi just out of curiosity and got a 4 pixel edge which in my book is very workable, but not practical. Most of my sharpest MF and LF photos drum scan at 2000 dpi with a 2 pixel edge which is about equal to bayer digital. I dont see MF backs approaching 8x10 any time soon. A 8x10 2000dpi scan is just massive and using the full 8x10 for comparison would be 320 mp. With E100G, and my cheap G-claron lens drum scanned at 2000 dpi, I would get a very clean file and a 2 pixel edge. Just a huge sharp clean file. I dont think we will see any MF 320mp chips any time soon. I do think a 4x5 back would come close if not beat it.

 

>Of course the problem with all this speculation is that, even when >resolution-equivalent digital cameras become available, how many >people will stay with film due to price and how many emulsions will >remain available? I have been told that Fuji intends to keep making >color film until the cows come home, that it is too much of a cash >cow (they certainly keep introducing new emulsions, unlike a >certain yellow-themed company). The motion picture industry is >still using mostly film. There is still no industry standard for >archival digital storage (CD's tend to die after a few years). And

 

Big problem as I see it. I have read discussions about some people interested in backing up digital files on film ??? Believe it or not.

 

>looking at the LF resale market (which is extremely strong), there >is certainly no indication that the marketplace foresees any viable >LF replacement for the immediate future.

 

Midwest mentioned they had there best LF year ever last year, and LF cameras are flying out the door.

 

>My take is that we'll have to put up with continued consolidation >in film emulsions, chemistry, paper, labs, etc., but that analog >solutions of some breadth will continue to be available for many >years. We may just have to pay a bit more for it.

 

Probably, but maybe if the big boys drop out the Europeans will pick up the slack like they did with Efke 25. Also I shot some Adox 4x5 microfilm the other day and was pretty much blown away by that. I am still tinkering with it though.

 

[j.w.]

It seems the discussion thus far has been on the subject of resolution, and when the continued advancements made in image sensors will overtake the stagnant film technology of today.

 

Is there any consideration of other image sensor qualities, such as tonal range, and low-light noise levels? Where do the two technologies line up regarding these factors?

 

Where does it appear that electronic images sensors are going regarding tonal range, verses classic Zone System controls available to the traditionalist B/W photographer? That would make an interesting discussion.

 

Technologically, virtually anything is theoretically possible; its marketting forces that determine where actual research dollars go into developing products for predicted future markets. Thus, film scanner technology could (theoretically) continue to advance such that resolution, negative size, tonal range and price-point could keep LF film advantageous in the future, regardless of advancements in electronic imaging.

 

But we all know that won't happen. The prediction of film's demise was made long ago; not by futurists, but by those positioned to profit the most.

[._._z]

<i><blockquote> the fact that scanners for home use will continue to improve

</blockquote> </i><p>

 

Not so much. Scanner sales plateaued, and manufacturers are getting out of the

market. Those who are left are left with a shrinking market and so are not investing in

r&d to improve current home scanners.