cineski wrote:
I just can't see how a 16mm focal length will give the same distortion as a 45mm with regards to any aspect.
I'm guessing that you may have a 16mm lens uncounsciously associated with 'fisheye' types of lenses and is thereby influencing your concept of what's happening. Just leave the numerical value out of the equation.
A side note: When I shot 4x5 with my Schneider 47mmXL Super Angulon, the rear element was barely 1/2 inch away from the film. So close in fact, that you would get a hot spot, so the lens was sold with a center neutral density grad filter for the front of the lens. (Held the center area back a stop or so. I forget exactly.) Schneider advertises that lens as 'almost symmentrical' in design. It does barrel a tiny bit. Before I was well versed in linear distortion, I thought the "bow" in any window frames I shot was due to curled film in the holder, or curled film in the glass-less scanner tray.
That's okay, I'm home sick with some stupid sore throat and being over sensitive Sometimes we can't hear what we type.
RalphJ wrote:
Not if you feel bad I don't. I didn't mean to come on so strongly; it was just a slow day at the office and it was this or work! I'm sorry about that; I'll temper my remarks.
JJL- When I mean 16mm on 35 vs 47 on 4x5, to me, a 16mm simply distorts a lot. Whereas a 47mm focal length should distort much less. When you put the two lenses on their designed for system to get the same field of view (due to crop factor) the distortion of the 16mm would be much worse. Am I wrong or right no this?
Jim Roof wrote:
.....since most 35mm WA lenses were not designed to the same exacting standards as LF lenses. BTW, MF lenses barrel distort too....
Jim- actually the opposite is the case. Since 4x5 get enlarged very little compared to a 35mm size, 4x5 lenses didn't need to be that good at all. The barrel distortion is because lenses for reflex cameras (cameras with mirrors and pentaprisms) are of the "retro-focus" design school. That means they are planted further away from the film plane than a rangefinder or large format lens is designed to sit (in order to clear the mirror.)
The Hasselblad 38mm Biogon was a symmetrical lens design, and hence was sold permanently mounted on a very thin Hasselblad body that did not have a mirror (or focus screen on top). It had very little noticeable linear distortion (barreling) as a result. Compare that to its closest cousin that gets used on a conventional 500 series Hasselblad body, the Hasselblad 40mm Distagon, and that lens shows lots of barreling. Same thing when you compare the Zeiss 21mm Biogon for the M-mount rangefider (very little linear distortion) against the Zeiss 21mm Distagon for the SLR (more barreling again).
cineski wrote:
JJL- When I mean 16mm on 35 vs 47 on 4x5, to me, a 16mm simply distorts a lot. Whereas a 47mm focal length should distort much less. When you put the two lenses on their designed for system to get the same field of view (due to crop factor) the distortion of the 16mm would be much worse. Am I wrong or right no this?
The 16mm on a 35mm system would show barreling (more linear distortion) because of the deficiencies of retrofocus design principles, not because it is for a smaller format. If you used a Zeiss 15mm M-mount rangefinder lens and compared it to the same 16mm Canon SLR lens, the SLR lens would show notable barreling, the rangefinder lens would show much less. And both are on a FF 35mm frame.
Psychic - There is a small amount of barrel distortion even in the image you posted. If you look closely at the window frame on the left, you can see it curve slightly. You've also got a fair amount of keystoning too, indicating either not a level camera or careless postproduction. Just out of curiosity, what is the intended color space of that image, as it's untagged and looks too dark in anything but Apple RGB with its 1.8 gamma.
24 T/S's do show a small amount of barrel distortion, more apparent near the edges, but it's pretty easy to fix even in Ps.
Here's the admittedly basic MTF taken from the Nikon site for the PC-E 24mm... any of the chart reading experts care to comment on real life differences?
In the Canon 30 LPM graph, are we seeing a dip in resolution in the "zone B" range, yet it picks up again towards the edges? The Nikkor in contrast seems to gradually drop off towards the edges.
Psychic...that image is like the perfect example of what a tilt shift can do. That image is absolutely impossible to get without a tilt shift. It looks to me like this is both tilted and shifted, right? With a conventional lens, you'd have the camera pointed down, and not have nearly enough depth of field, with expanding verticals as you go up...this is wonderful. Now having this ability with a 17mm is awesome. If only I could afford it!
jjlphoto wrote:
Jim- actually the opposite is the case. Since 4x5 get enlarged very little compared to a 35mm size, 4x5 lenses didn't need to be that good at all. The barrel distortion is because lenses for reflex cameras (cameras with mirrors and pentaprisms) are of the "retro-focus" design school. That means they are planted further away from the film plane than a rangefinder or large format lens is designed to sit (in order to clear the mirror.)
The Hasselblad 38mm Biogon was a symmetrical lens design, and hence was sold permanently mounted on a very thin Hasselblad body that did not have a mirror (or focus screen on top). It had very little noticeable linear distortion (barreling) as a result. Compare that to its closest cousin that gets used on a conventional 500 series Hasselblad body, the Hasselblad 40mm Distagon, and that lens shows lots of barreling. Same thing when you compare the Zeiss 21mm Biogon for the M-mount rangefider (very little linear distortion) against the Zeiss 21mm Distagon for the SLR (more barreling again).
I am not a lens designer so I may be off on the history of lens design for various formats, but I have used a lot of LF lenses (I used to own ten different focal lengths for my work). Suffice it to say that this might be a cart and horse, or perhaps horse and cart issue as my point was that 35mm lenses did indeed exhibit barrel distortion almost without exception whereas LF lenses did not, almost without exception. In fact, I know of no view lenses that were not straight as arrows though my Nikkor f/4 65 would LOOK like it distorted at times because the film holders could not keep film perfectly flat all the time.
As for MF, I shot a Pentax 6x7 and as I recall my 45mm Takumar was not that bad but it was not as straight as the view lenses.
My point was this, 35mm wides almost always have barrel distortion. Prior to any digital manipulation in the work flow they were almost useless for shooting serious architecture. For this reason 4x5 was the camera of choice. It also gave nice sharp and detailed images that were relatively grain-free. Sure, the lenses were of a simpler design (you brought this out) but this in no way negates my claim that they were made to exacting standards. No doubt a decent 35mm lens, say at a 75mm focal length, would out-resolve a 75mm Super Angulon per square mm on the film, but the 75mm 35mm optic is not going to have an angle of coverage that almost covers a 5x7. From the standpoint of precision, whether it is the result of 'exacting standards' or not, the LF wides out-performed anything when it came to distortion. BTW, I have bought enough them also to realize that, absent AF motors, circuit boards, and internal focusing mechanics, the price would lead one to believe that they were precision instruments as well.
Jim Roof wrote:
I am not a lens designer so I may be off on the history of lens design for various formats, but I have used a lot of LF lenses (I used to own ten different focal lengths for my work). Suffice it to say that this might be a cart and horse, or perhaps horse and cart issue as my point was that 35mm lenses did indeed exhibit barrel distortion almost without exception whereas LF lenses did not, almost without exception.
Jim- just to clarify- Lenses for reflex cameras, especially wide angles, are retrofocus designed lenses. Retrofocus designed lenses exhibit barreling as a result of it being a 'retrofocus' lens. It was Zeiss that developed this design principle in the 1930's, but it was French lens designer Pierre Angénieux who coined the term 'retrofocus', and brought that lens design into the mainstream. http://en.wikipedia.org/wiki/Angenieux_retrofocus
Before that time, lenses were designed via the 'symmetrical' design principle. Symmetrical designed lenses exhibit little or no barrel distortion. Large format lenses, enlarger lenses, process camera lenses, etc, etc, are examples of symmetrical design lenses.
I feel it is better and more accurate to state that reflex camera lenses exhibit barreling because of the retrofocus design principle rather than stating that most 35mm WA lenses were not designed to the same exacting standards as LF lenses, as the latter is simply not true. In fact, the opposite is more true. It is much more difficult to design and manufacture a retrofocus lens than a symmetrical lens.
* I've owned many LF lenses myself. Schneiders, Rodenstocks, Fujinons, Nikkors, from 360mm's all the way down to 47mm, on everything from Sinar 8" x 10" cameras to 6cm x 9cm Arca Swiss cameras. When I switched over to the Canon 1Ds five or six years ago, I was instantly bummed out by all the barreling I was getting. But this is the way the industry is going unfortunately.
At €2499 for the 17mm and €2299 for the 24mm they are not cheap lenses.
Perfect! now I have the reasons to apply for an Bail Out from Obama's government for my business!!!
Question for the TS-E users. If they can make the 45mm & 90mm with bigger aperture (f/2.8), why are they making the 17mm & 24mm with f/4 & f/3.5 respectively?
That brings me to the next question. If they can only make the wide angle TS-E with smaller aperture, why do they cost a lot more than the 45mm & 90mm? With the telephoto lens, the price seem to increase exponentially as the aperture gets bigger, the 300mm f/2.8L & f/4L for example.
ShaneEngelking wrote:
I know this is entirely speculative, but anyone think the 17mm will be a good performer with a 1.4x converter?
Your guess is as good as any, but being a prime and judging from results others have gotten from the 24, 45 and 90 TS-E's beforehand I'd say there is a pretty good chance the combo would perform well. IIRC I don't think Canon TC's are compatible.
M Vers wrote:
Your guess is as good as any, but being a prime and judging from results others have gotten from the 24, 45 and 90 TS-E's beforehand I'd say there is a pretty good chance the combo would perform well. IIRC I don't think Canon TC's are compatible.
Thanks, that gives a lot of options with one lens-FF, 1.6 crop, and 1.4x, gives you a lot of different focal lengths; 17, 24, 27, and 38mmequiv. For me, that covers pretty much the range that I would want from TS-E.
I am just psyched. I shoot a lot of interiors and the 16-35 II just does not leave me satisfied, even though the lens is so convenient to use. I have used the older 24 TSE lens and I did notice some softness in the corners and other aberrations.
I have rented the 90 TSE and now own the 45 TSE and love both lenses. The 45 is a marvelous tool.
But I really am hoping that the new 17 TSE will deliver. With this economy who wants to spend $2500? But this lens will probably serve me well for years and if it makes my interior shots look better then its a no-brainer...
ShaneEngelking wrote:
Then the lens should have a tripod mount.
Actually, there are manufactures of something similar. Zoerk makes a PanoShift Adapter that has a tripod mount. However, it is designed for those who use medium format lenses on a DSLR body.
Sometimes we can't hear what we type.