Pages: [1]   Go Down

Author Topic: Rodies HR series distortion correction  (Read 6038 times)

DanLindberg

  • Newbie
  • *
  • Offline Offline
  • Posts: 20
    • Dan Lindberg Photography
Rodies HR series distortion correction
« on: June 17, 2011, 05:31:41 pm »

After searching everything I possibly can to educate me on Schneider vs. Rodenstock lenses, it seems that the single most 'not liked' thing about the HR series is that there is complex distortion showing.

Architectural photography with 33x44 sensor and lots of movements is on the agenda! I am looking at the hr23 (still gives me 10mm of rise/fall with the cropped sensor) paired with a hr 40. Or possibly a hr 32 paired with a hr 50.

So for you people that have real life knowledge of needing straight lines with these lenses. Is it doable? Can software correct distortion in a good way produced from these lenses? Any of the four mentioned that is worse than others?

Thanks!
Logged
Alpa FPS & MAX & SWA & TC | SK 28XL | SK 35XL | SK 72L | SK 90N | SK 120N | Leaf Aptus II 5 & Leaf CREDO 60
http://www.danlindberg.com
https://www.facebook.com/danlindbergphotography

A.Garcia

  • Newbie
  • *
  • Offline Offline
  • Posts: 9
Re: Rodies HR series distortion correction
« Reply #1 on: June 20, 2011, 02:54:56 am »

Alpa Lens Corrector software is said to provide the best correction for Rondenstock moustache shaped distortion. And it's for free. :D
Logged

cng

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 57
Re: Rodies HR series distortion correction
« Reply #2 on: June 20, 2011, 08:10:35 am »

IIRC, someone posted a little while ago that the Alpa Lens Corrector is compatible only with CS3.  Having never used the Alpa s/w, I can't comment whether this was/is still the case.  Just thought I'd point it out.
Logged

nazdravanul

  • Full Member
  • ***
  • Offline Offline
  • Posts: 132
    • Stefan Iacob - visual artist
Re: Rodies HR series distortion correction
« Reply #3 on: June 20, 2011, 09:11:33 am »

The Alpa Lens Corrector works up to and including CS5, but in 32 bit mode only. No direct experience with it, though. Just double checked the info on their website :

http://www.alpa.ch/en/news/2010/photoshopcs5

I am also interested in 1st hand experiences with software corrections for Rodenstock lenses, especially (but not limited to) Capture 1 .  
« Last Edit: June 20, 2011, 09:13:46 am by nazdravanul »
Logged

JdeV

  • Full Member
  • ***
  • Offline Offline
  • Posts: 120
    • http://www.jonathandevilliers.com
Re: Rodies HR series distortion correction
« Reply #4 on: June 21, 2011, 10:32:09 am »

Hi,

Having shot a fair bit of architecture with the Rodenstock 40 and 50mm my experience has been that the distortion is generally relatively straightforward to fix manually in Photoshop. However, I tend to do a few images that get a lot of attention lavished on them rather than having to churn out a lot of shots. I also don't tend to do too many very square-on shots.

One very notable feature of these lenses is their extraordinary sharpness wide open. They also have very nice contrast. They are much, much easier to focus and compose with on a ground glass than their Schneider equivalents because of the wider maximum aperture, the increased distance from the ground glass and the relative lack of fall-off. They are very special lenses if you can live with the distortion issue. Having said all that, I have not shot with the 43mm Schneider and would be interested in doing a side by side.

I have been renting the lenses from Linhof and Studio in the UK. If possible you obviously want to rent and try before you buy. I think in this case the costs involved demand it.
Logged

DanLindberg

  • Newbie
  • *
  • Offline Offline
  • Posts: 20
    • Dan Lindberg Photography
Re: Rodies HR series distortion correction
« Reply #5 on: June 21, 2011, 11:42:30 am »

Thank you all. I think I am convinced that it will in fact be Rodies. I am not into stitching, so I think the correction software will be able to take care of my images in a good way.
Logged
Alpa FPS & MAX & SWA & TC | SK 28XL | SK 35XL | SK 72L | SK 90N | SK 120N | Leaf Aptus II 5 & Leaf CREDO 60
http://www.danlindberg.com
https://www.facebook.com/danlindbergphotography

narikin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1372
Re: Rodies HR series distortion correction
« Reply #6 on: June 24, 2011, 12:27:50 pm »

I presume you are talking about HR Digaron-W series?

The HR-Digaron-S series may be even better for your needs if you don't stitch: significantly higher resolution, faster, and almost zero color distortion. 

It's easily overlooked, but in fact the superior lens range, if you are just doing single shot work.

Logged

Christopher

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1499
    • http://www.hauser-photoart.com
Re: Rodies HR series distortion correction
« Reply #7 on: June 25, 2011, 04:30:26 am »

Is it really ? I don't think there is a visible difference between the S and W line at f8-11. Perhaps if you shoot wide open, but otherwise I'm quite sure no one could tell the difference.
Logged
Christopher Hauser
[email=chris@hauser-p

narikin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1372
Re: Rodies HR series distortion correction
« Reply #8 on: June 25, 2011, 07:37:24 am »

Yes there is a clear difference.

Take a look at the charts and see the much higher resolution and chromatic aberration figures:
http://www.rodenstock-photo.com/mediabase/original/e_Rodenstock_Digital_Lenses_3-26__8236.pdf

Compare the 70mm (f5.6) W with 60mm (f4.0) S, for example; or the 90mm W (f5.6) with the 100mm S (f4.0) -
The S range is way ahead in resolution and has much much less CA.  Plus is nearly a stop faster.
(for me, the lack of CA fringing in images alone is worth it, without the speed or extra resolution)

You may not be able to see it in regular sized images, but that is true of 90% of what we fuss about here.
Most people can't see the difference between a Canon and an MF image in a magazine spread.

With 80Mp backs and <5.2micron pixels, if you are not stitching, the Digaron-S range is the way to go.


Logged

Jack Flesher

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2592
    • www.getdpi.com
Re: Rodies HR series distortion correction
« Reply #9 on: June 25, 2011, 09:36:27 am »



With 80Mp backs and <5.2micron pixels, if you are not stitching, the Digaron-S range is the way to go.

Not if you need lens movements -- one of the main reasons I use a tech camera is for the lens movements aiding composition; not just stitching!  

That said, I agree with Christopher -- there really isn't much difference in quality AT THE SAME POSITION ON THE IMAGE CIRCLES.  Your statements are more true if you compare the full length of the curves to the edges of the IC's, but then the W's have much larger IC's so it's not really a comparison...

To wit, take the 80LPmm saggital MTF at f5.6. Your 60 HR-S is about 62% at center of IC and 12% at 35mm from center.  By contrast, the 70 HR-W is 55% at the center, but at 35mm it's about 35% and at 45mm it's still above 20%. Granted, your 60S holds above 50% to 20mm, while the 70W hits 50% at 15mm -- but I'll take the improved edge performance over the center, especially with movements.  So here and for *ME*, I would say the 70W is actually showing BETTER than the 60S.  Falloff is better on the 70W too, especially if you look at the 35mm point on the IC.  Distortion on the 70W starts at 0 and hits -0.2 at about 25mm, while your 60S starts at -0.2 and moves to -0.3 at 25mm. (You need to read the scales and not just look at the slope of the lines!) In fact, it only returns to 0 at the edge of its 35mm IC. At 35mm, the 70W is just hitting -0.3, so arguably better performer across a 70mm IC than the 60S. Yes, beyond where the 60S can cover, the 70W does distort more, but at least I can get there with the 70S to begin with!   CA is arguably better on the 60S though, but frankly very low on both lenses to begin with, and CA is relatively easy to correct.  

My point here is it really depends on what your particular needs are. Your 60S is a great lens if you do not ever intend to shift it on a full-frame MF sensor; the 70W is a great lens if you want to shift freely with a full-frame MF sensor...  
« Last Edit: June 25, 2011, 10:13:44 am by Jack Flesher »
Logged
Jack
[url=http://forum.getdpi.com/forum/

narikin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1372
Re: Rodies HR series distortion correction
« Reply #10 on: June 25, 2011, 09:50:13 pm »

My point here is it really depends on what your particular needs are. Your 60S is a great lens if you do not ever intend to shift it on a full-frame MF sensor; the 70W is a great lens if you want to shift freely with a full-frame MF sensor...  

Jack - I have both lenses, and let me agree/assure you the S range is better for straight shooting.

You chose the more flattering comparison to single out - the 60S vs 70W, and I repeat it is unfair to look at the 60s extreme edges where it is not designed to perform. This lens is all about non stitching performance - pure and outright ultra performance straight single shot resolution - which from what I can tell is what the OP was wanting - no stitching/ tilt/shift/ whatever.  If that is what you want instead, then sure, the W range is right there for you.

You also don't compare the 100S with the 90W where the differences are magnified.  The 90W cannot hold a candle to the 100S in single shot resolution.  The 100S at f8 is off the charts good - really incredible performance.  Even at 80lp/mm it is above 60% almost across the entire frame!

CA is what I was talking about - a very real and big problem with lenses that afflict nearly every image you shoot, and the S range does not have this problem.  'Fixing it' brings in distortion/anomalies, and I would rather have none from the get-go. The two I own (60S and 100s) are the only lenses I have that give no color fringing visible in use.  I also own the 90W, so its not  from here-say I make these comments.

I essentially agree with your last remark - if you want an ultimate sharp lens to walk around with, that is a full stop faster (and lets face it f5.6 is pretty ordinary by most metrics) then the S is the range.  If you want to stitch/ tilt/shift, then there are other very decent options, like the W range.
« Last Edit: June 25, 2011, 09:52:35 pm by narikin »
Logged

Jack Flesher

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2592
    • www.getdpi.com
Re: Rodies HR series distortion correction
« Reply #11 on: June 26, 2011, 02:24:49 am »

  which from what I can tell is what the OP was wanting - no stitching/ tilt/shift/ whatever.  

Maybe I misunderstood this comment from the OP then: "Architectural photography with 33x44 sensor and lots of movements is on the agenda!"

Logged
Jack
[url=http://forum.getdpi.com/forum/

narikin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1372
Re: Rodies HR series distortion correction
« Reply #12 on: June 26, 2011, 02:24:49 pm »

you're right/ my bad.  if OP wants to tilt/shift ("lots") then a W or even more is required.
Logged

DanLindberg

  • Newbie
  • *
  • Offline Offline
  • Posts: 20
    • Dan Lindberg Photography
Re: Rodies HR series distortion correction
« Reply #13 on: June 26, 2011, 03:44:08 pm »

Yes, rise and fall is vital in my quest for the ultimate architectual setup  :)
Logged
Alpa FPS & MAX & SWA & TC | SK 28XL | SK 35XL | SK 72L | SK 90N | SK 120N | Leaf Aptus II 5 & Leaf CREDO 60
http://www.danlindberg.com
https://www.facebook.com/danlindbergphotography

dkaufman

  • Newbie
  • *
  • Offline Offline
  • Posts: 39
Re: Rodies HR series distortion correction
« Reply #14 on: June 26, 2011, 04:57:00 pm »

Andalucia: I have exactly the sort of equipment you may be contemplating: an Arca Swiss M-Line 2 with a Phase One 40+ back and the following lenses: Rod. 32 HR,  Rod 40 HR, Schn 47,  Rod 55 (non-HR), Rod 70 HR, Schn 100 (5.6). My subject is urban architecture, sttreetscapes, etc., usually exterior. I shoot almost everything with two or three vertical (portrait) images in a horizontal stitch yielding large files, sometimes the opposite orientation for vertical images. The minimum requirement for stitching architectural images is a 90mm image circle. The 32mm HR stitched yields a very wide-field image but on the occasions it is most useful for--close in to a building with 15 to 20mm of rise plus horizontal movements for stitching--the 90mm image circle is sometimes not sufficient and the camera has to be tilted up to avoid the image circle impinging on the desired image. Less often but occasionally with the 40 HR as well. Both the 32mm HR and 40HR (less so) require correction distortion after stitching with the Alpa program, mentioned by others, which does a good job.
Stitching makes everything complicated: The LCC (lens cast and fall-off) corrections have to be shot individually and are applied individually before stitching. The applied LCC images, even when shot at the time of picture-taking (which is a must) do not yield perfectly colour  matched images, so images after LCC correction need colour matching (which is tedious and difficult sometimes). Then the stitched image needs a custom Alpa lens correction file (modified from existing provided files with any text program) based on its ultimate dimensions. And then, if the image has been tilted up (not the usual case) perspective correction must be applied (and it cannot be applied earlier in the workflow because of the need to use the Alpa correction on images after stitching). A lot of work for stitched images, but it's the only way to get really wide shots which are often needed in urban and architectural photography. (I thought of getting the Phase IQ180 but have been scared off by the reports of even more difficult issues with wide angle colour casts.)
As for quality of lenses, the 40HR is the best of the bunch. The 32 HR can be nearly as good but quality suffers more than the 40HR as you approach the boundary of the image circle. The Schneider 47mm is excellent at centre not so good as you approach about two-thirds of the way to edge of image circle. The Rod 55, while not an HR, can be excellent at centre, but falls off somewhat in resolution as you move towards image circle boundary. The 70 HR and Schn 100 are generally excellent with the 70mm falling off as you approach image circle. The need for a 90mm image circle at minimum elminates many lenses. The 40HR especially holds its quality almost right up to the image circle boundary.
Hope this was helpful.
Logged

DanLindberg

  • Newbie
  • *
  • Offline Offline
  • Posts: 20
    • Dan Lindberg Photography
Re: Rodies HR series distortion correction
« Reply #15 on: June 27, 2011, 05:08:16 am »

Excellent! Thank you for your detailed info!!! I have decided that stitching will only be an emergency solution because of the same reasons you so clearly lay out. I do not want all that extra work....I will buy a wide that is wide enough for single capture. Should I go for a 33x44 sensor then the HR23 is probably necessary and with a larger senser the interior workhorse would most probably be the HR32.
Logged
Alpa FPS & MAX & SWA & TC | SK 28XL | SK 35XL | SK 72L | SK 90N | SK 120N | Leaf Aptus II 5 & Leaf CREDO 60
http://www.danlindberg.com
https://www.facebook.com/danlindbergphotography

Schwarzzeit

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
    • HIGH-END-SCANS
Re: Rodies HR series distortion correction
« Reply #16 on: June 28, 2011, 01:27:43 am »

You also don't compare the 100S with the 90W where the differences are magnified.  The 90W cannot hold a candle to the 100S in single shot resolution.  The 100S at f8 is off the charts good - really incredible performance.  Even at 80lp/mm it is above 60% almost across the entire frame!
At some point in 2008 or 2009 Rodenstock updated their lens brochure. In their earlier brochure from 2006 they published the MTF diagrams of their HR lenses for 15, 30 and 60 lp/mm. If you compare the 100S's f/8 15, 30, 60 lp/mm chart with the recent 20, 40, 80 lp/mm diagram you'll notice that it's exactly the same curves. Theoretically at f/8 diffraction limited contrast is around 54% at 80 lp/mm and a magnification ratio of 0,05 for 550 nm light. It's about 65% for 60 lp/mm. Even though it's in fact only the 60 lp/mm curve it's still diffraction limited at f8 across the better part of the IC.

-Dominique
Pages: [1]   Go Up