Before computer design at least, lenses tended to be optimised for either sharpness or coverage - and improving one reduced the other. Within the central 10 degrees or so a Petzval is at least as sharp as any modern lens, but the resolution drops off very rapidly away from the center. Aplanats have somewhat greater coverage, and are equally sharp in the center (my Meyer Aristoplan 270mm f:7.2 seems to me to be diffraction limited at fwide open aperture - dead center). The early Anastigmats improved the evenness of resolution over the image circle at the cost of some peak resolution (yes, a good Aplanat is sharper in the center than a good Protar).
It's not so much that there is a tipping point as that the maximum theoretical resolution is a function of focal length. If you think of resolution not in terms of lines on film but as very small angles between light rays, you will see that two lenses with the same angular resolution but different focal lengths will give different resolution on film. So a 150mm would theoretically be exactly twice as sharp on film as a 300mm lens of the same construction.I only ask because presumably a lens in front of 4x5 that would cover 10x8 would mean 2x the number of lines across the photograph than 4x5? This sort of implies that the best 4x5 lenses should have the smallest coverage which goes against my (admittedly simple) knowlege? Perhaps there is a tipping point around 10x8 that makes lenses a lot harder (or more expensive) to create...
There is still a trade-off between resolution and coverage, but modern design methods have allowed lens designers to make lenses with both good coverage and high resolution across the image circle.
However, there are exceptions. The new Schneider 550 XXL lens, which is a modern multi-coated Dagor design, gives very even coverage over nearly all of its 900+mm circle of illumination. This lens was designed for ULF work and the huge coverage that is needed for these large formats.
> Does this imply that, all else being equal, the bigger the image circle of a lens the lower aerial resolution?
Tim, as a general rule, yes, this is the case, unfortunately. As Ole and Sandy mention, there has been, and always will be a trade-off between resolution/contrast vs. coverage. Hence the continued trend to smaller formats through the years, reversing the "bigger is better" history of photography.
> Perhaps there is a tipping point around 10x8 that makes lenses a lot harder (or more expensive) to create...
I would add some caveats to this statement to make it more accurate. First, it's not just lens design and MTF that limits resolution, it's the apt. you are forced to shoot at using the longer fl lens for the bigger format, to accommodate the DOF in the scene. So with a DOF shot (i.e. not infinity, not flat subject) the larger format lenses often have two strikes against them, smaller apertures required, and often less resolution to begin with, due to the larger coverage. IMO, there is a tipping point in this balancing act, and it falls between 4x5 and 810. If the shot requires some DOF, then 4x5 is the crown jewel. However, if the shot does not require DOF, then additional gains can be appreciated jumping to 810. IMO, this is one of they reasons 4x5 has dominated LF photography for many years. Of course, their is other reasons that are obvious, such as size, no need for prints larger than 4x5 can produce, larger and better lens selection, etc. I also think 4x5 will continue to hold its own in the future due to our excellent films and optics for this format size.
ULF faded many years ago for a very good reason - the advent of higher resolving films which allowed equal resolutions with smaller formats via enlargements.... todays digital trend, (downward in format size) is merely a continuation of the ULF to LF trend 50 - 70 years ago. Both times, it was higher resolution recording media that allowed (or welcomed) the use of shorter fl lenses. Shorter fl's lenses (relative) is one of the few physical laws of optics, that works in our favor. These shorter fl lenses, have the potential to produce much higher resolutions and faster shutter speeds. This same principle is what opened the flood gates for the f2.8 digicams - which are taking over the world. At f2.8, (most common) diffraction limited aerial resolutions are about 530 lp/mm. Compare this to f64 on 8x10, 23 lp/mm. oh yeah, also a "small" gain in shutter speeds, just 8 stops :-)
Last edited by bglick; 17-Jan-2008 at 08:35. Reason: spelling
bglick, splendid info on lens matters. These latest posts get a bit more at the point of the original post.
I've not used any ULF format cameras but have historically done 35 mm. to 8X10. I agree that as the format goes up beyond about 5X7 it becomes more noticeably difficult to capture the plane of focus using movements and my photos became more static. I didn't always mind this depending on the subject matter but I always seemed to gravitate back to the 4X5 format for the flexibility.
By the way bglick you seem to be into lens design and I seem to remember a pioneering company, I think in CT, that specialized in computer optical design in the 70s' or 80s'. They did design work for Leitz and Nikon - I think. I wonder if they are still in business?
Thanks for the kind words Nate
> I agree that as the format goes up beyond about 5X7 it becomes more noticeably difficult to capture the plane of focus using movements and my photos became more static.
Yep, this is another good point which I avoided previously, as the discussion revolved around format / apt. / diffraction / format size. However, as you suggest, another reason 4x5 is such a sweet spot for photography is movements. While, rise/fall corrects distortions or alters composition, it's the tilts that can ease the focus issue and prevent stopping down (degrading the resolution via apt. diffraction) as the only form of capturing the image plane. But to utilize tilt, you must have sufficient image circle, which 4x5 lenses offer. You also will benefit by the relatively shorter fl lenses (vs. 8x10 or larger), as a 300mm fl lens requires 2x the tilt as a 150mm lens. So as you double the format, you eat up the image circle first by the larger format size, then, to make matters worse, you need 2x the degrees in tilt for equal focal plane placement..... quite often, you simply don't have enough enough image circle for all these requirements, so you resort back to stopping down as the only option to attain adequate focus from near to far - which brings on apt. diffraction. Even when you can tilt with the larger formats, it's often the very poor resolving portion of the image circle you are capturing. So due to the over-sized image circles available to 4x5, combined with the shorter fl's which require less degrees in tilt.... 4x5 will very often provide you with an alternative focus means (tilt), allowing you too open up several stops vs. keeping the standards parallel. Yep, another "feather in the cap" for the 4x5 format.
As a general rule, if I have an infinity scene, or a scene with very little depth, I will use 8x10, or the occasional ULF. Otherwise, 4x5 has proven untouchable in versatility, size, weight, etc. Of course, 5x7 is a nice compromise of the two. I have recently acquired a 5x7 back also, so I might eventually settle on one format as most of my lenses have very large coverage, and I really hate carrying so much gear. I think QT figured this out looooong ago!
I am not aware of the optical design firm in CT you refer to...hmmmmm.....
If you want proof that Optics is not a relevant issue when deciding whether or not ULF has utility for you, the potential user, or for those of us already using ULF cameras, just look at a good Albumen Print by Carleton Watkins. The High Museum of Art in Atlanta, Ga. has a print on display that is absolutely amazing--for its superb optical quality(sharp throughout-foreground to background-you can count blades of grass if you want), for its beautiful tonality, and for the complete Craftsmanship in its execution. The one in particular that I am thinking of is a 16x20 contact print made over One Hundred years ago!! I can't remember the name of the print and I can't find it on the High's web site, but if any of you have been there and seen it and really looked at it then you know what I am talking about.
Sudek ambled across my mind one day and took his picture. Only he knows where it is.
Does it need to be emphasized that the photographer who wishes to expose hundreds of rolls of film per day or the equivalent in sheets, that ULF does not have utility for them and the goals of that type of image making? Isn't this obvious?
Sudek ambled across my mind one day and took his picture. Only he knows where it is.
> If you want proof that Optics is not a relevant issue when deciding whether or not ULF has utility for you, the potential user, or for those of us already using ULF cameras, just look at a good Albumen Print by Carleton Watkins.
David, ULF users can maximize the benefits of the big format. For example, an infinity shot on 11x14 with a modern oversized 8x10 LF lens is truly breathtaking. No doubt about it. It will look superb whether it was contact printed, or enlarged. As mentioned in the deleted posts, ULF still has a niche IMO. However, it doesn't change what was discussed above regarding the limits imposed by apt. diffraction.
Even Ansel Adams shot atop a tall ladder, in his later years, as he too was trying to reduce the amount of DOF, so he could limit diffraction effects by opening up more. I shoot atop an RV with 810 and 11x14 to do the same. So there is tricks at every level to maximize a given formats utility.
This thread is about the relevance of using ULF cameras. Someone responded that there is not and proceded to cite all kinds of resolution arguments, MTF tables, that enlarging is better and should be considered and of course the ever present digital.
Well, while all these rationalizations sound reasonable as an attempt to demonstrate there is no relevance for ULF, in the end the proof is in the pudding. Somtehing that those of us who actually USE and photograph with these size have known for a long time.
So in an effort to show and yes discredit all thes rationalizations I am posting pictures and magnifications done with a 12x20 Korona camera, a Nikon Nikkor 450 M lens and Ultrafine 125 film.
I took two shots of this scene, I liked the clouds on negative one better than the other one so I had this extra negative that I do not use and decided to cut for this demonstration.
The first picture is a shot of the finished 12x20 print. All the scans are straight, no manipulations, not even sharpening on photoshop.
Then, I cut up the 12x20 negative and the second shot is the section I cut to put on my scanner. This scan was made at 100 dpi resolution. I then scanned the same cut at 2400 dpi and cut off a part that looked the same as my negative under a 10x loupe. And this is the third picture.
Bear in mind that all the scans have no sharpening in photoshop not even to correct for the loss in resolution when scanning.
So lets examine the arguments that were proposed.
1.Because you have to use small apertures you loose resolution and the resolution of the lenses used is not adequate for these formats.
--- BS. This shot was taken a f/45, for 5 seconds, in windy conditions as you can see by the movement of the branches in the tree.
2. Difficulty to mantain squareness of the camera, film flatness and parallel standards.
--- These are reasons borne out of inexperience and ignorance in the format. These shots were taken with a 90 year old Korona 12x20 camera with S&S holders. A little bit of care, a good tripod and good holders take care of this. If a 90 year old camera is capable of this kind of pictures, the modern ones are far more capable. My Wisner has stops to set parallel standards and I have yet to see a modern ULF camera that is not square both to an horizontal and vertical axis.
3. I can disregard circle of confusion because the lens does not see format size and the depth of focus is unimportant. I want to see the print at any distance I want.
--- specious argument. This person wants to look at a print that is 10 feet x 16.66 feet from a 10 inch distance. If that is the case, so be it...but clearly it is demonstrated in these pictures that even at close examinations the lenses and film are capable of delivering outstanding images.
Conclusion: Theory and tables are good to form an intial impression, but nothing beats experience, knowledge of the material and actually using the cameras to verify that the initial impression obtained form the theory are accurate. IOW, bigger IS better
PS, I forgot to add that the negative under a 10X magnification looks far sharper than the scan.