The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

32HR Lens Corrections in Capture One 9.0.1

dougpeterson

Workshop Member
As a small token of holiday appreciation for our tech camera users throughout the world, DT worked with Phase One to get the 32HR fully profiled for Capture One 9.0.1.

32HR.jpg

[if you're on v8 and looking to upgrade to v9, DT it is free when you attend our Capture One Masters Program training]
Happy Holidays!

Doug
 

jagsiva

Active member
Doug,

How does the profile work when the lens is shifted? Does it take the amount of shift into account? Cheers.
 

JeRuFo

Active member
I see there are profiles for the 35,23 and 70mm too. Does that mean there is no real reason left to buy an a-series? I can imagine some people would rather substitute the mandatory 35mm with a 40 or 32 HR. Or the TC with an STC, or whatever substitution.
 

dougpeterson

Workshop Member
The Lens Corrections for the 32HR (and other tech camera lenses) include corrections for images captured with movements (rise/fall/shift). In the Lens Correction tab select "Movements" and enter the relative location of the digital back. For instance if you shifted the back 15mm to the right in order to "see" to the left then enter (positive) 15mm. Another example: if you shifted the back 20mm to the left to "see" to the right then enter -20mm.

The A-Series offers built-in LCC correction for color cast; easy to provide given that the A-Series does not provide movements like rise/fall/shift. "Lens Corrections" in Capture One does not account for color cast; just chromatic aberration and distortion. In general we sell far more Arca and Cambo tech cameras than A-Series systems.
 

JeRuFo

Active member
The Lens Corrections for the 32HR (and other tech camera lenses) include corrections for images captured with movements (rise/fall/shift). In the Lens Correction tab select "Movements" and enter the relative location of the digital back. For instance if you shifted the back 15mm to the right in order to "see" to the left then enter (positive) 15mm. Another example: if you shifted the back 20mm to the left to "see" to the right then enter -20mm.

The A-Series offers built-in LCC correction for color cast; easy to provide given that the A-Series does not provide movements like rise/fall/shift. "Lens Corrections" in Capture One does not account for color cast; just chromatic aberration and distortion. In general we sell far more Arca and Cambo tech cameras than A-Series systems.
Right, sensor issues remain unique ofcourse. I jumped to conclusions a bit too fast there.
I was discussing the a-series with someone recently and it just seems confusing to me if you combine that system with other lenses or an stc and use it with movements sometimes. You would probably forget to make an LCC half the time. I see the merit of the a-series mostly in the wide angles where you probably don't need tilt and with the 23mm can't shift either.

These corrections show how mature tech cams are getting. The lenses already made a name for themselves, but with more and more pieces of the capture process getting automated, these systems get more and more appealing to a wider audience.
 
For instance if you shifted the back 15mm to the right in order to "see" to the left then enter (positive) 15mm. Another example: if you shifted the back 20mm to the left to "see" to the right then enter -20mm.
Any suggestion about how to note down the amount of movements in the field? Will it be a possible to associate this data in RAW file itself (may be via user interface in the DB), instead of noting down into a piece of paper or in any other device?

Thanks,

Subrata
 

Jamgolf

Member
I just now updated to C1 v9.0.1 and applied the newly added 32mm lens correction profile to an experiment from yesterday ...

Full frame: 32HR @f9 No CF, -12mm shift (i.e. shifted the back -12mm to the left to "see" to the right)


Crop with no lens correction profile applied in C1


Crop with the new 32mm lens correction profile with "Shift x" = -12mm & "Aperture (f)" = 9 in the Movement tab


Side by side (left=no correction, right=32mm profile applied):


So it does not seem to perform miracles but does make a noticeable improvement.

Cheers!
 
Last edited:

stngoldberg

Well-known member
Hi Paul,
My copy of the 32mm Rodenstork does not exhibit the mustache distortion that you continually refer to whenever anyone writes about wide tech camera lenses.
My 32mm Rodie with its large image circle and excellent resolution is the best lens that I have ever used.
Stanley
 

Paul2660

Well-known member
I just now updated to C1 v9.0.1 and applied the newly added 32mm lens correction profile to an experiment from yesterday ...

Full frame: 32HR @f9 No CF, -12mm shift (i.e. shifted the back -12mm to the left to "see" to the right)


Crop with no lens correction profile applied in C1


Crop with the new 32mm lens correction profile with "Shift x" = -12mm & "Aperture (f)" = 9 in the Movement tab


Side by side (left=no correction, right=32mm profile applied):


So it does not seem to perform miracles but does make a noticeable improvement.

Cheers!
Thanks for the images. It does indeed help.

Paul
 

Paul2660

Well-known member
Hi Paul,
My copy of the 32mm Rodenstork does not exhibit the mustache distortion that you continually refer to whenever anyone writes about wide tech camera lenses.
My 32mm Rodie with its large image circle and excellent resolution is the best lens that I have ever used.
Stanley

Hi Stanley

Not trying to be a hater, but my experience with the 32mm was just not a great experience.

Jamgolf's post shows a very good example of the distortion created by the 32mm. Sadly mine is a bit worse than his at 12mm of shift.
With the fire hydrant it's very easy to see the elongation and squatting of the hydrant in the lower right. The C1 correction does help for sure and hopefully Phase can tweak it even more over time. It's easy to miss this sometimes in natural subjects, like trees or rocks, but objects like cars or fire hydrants ;) show it very well.

No doubt it's possible to get a lens with more or less of this distortion, and I am not trying to sound like a broken record, however based on the all the positive praise on the 32mm, I was a bit taken back by just how bad the distortion is (at least on mine). I sent files to both Rodenstock and Arca, and was politely told, "this type and amount of distortion is a known issue of the 32mm and your lens is within spec". Live and learn, I should have made a better acquisition decision.

Paul C
 

Pelorus

Member
Hi Stanley

Not trying to be a hater, but my experience with the 32mm was just not a great experience.

Jamgolf's post shows a very good example of the distortion created by the 32mm. Sadly mine is a bit worse than his at 12mm of shift.
With the fire hydrant it's very easy to see the elongation and squatting of the hydrant in the lower right. The C1 correction does help for sure and hopefully Phase can tweak it even more over time. It's easy to miss this sometimes in natural subjects, like trees or rocks, but objects like cars or fire hydrants ;) show it very well.

No doubt it's possible to get a lens with more or less of this distortion, and I am not trying to sound like a broken record, however based on the all the positive praise on the 32mm, I was a bit taken back by just how bad the distortion is (at least on mine). I sent files to both Rodenstock and Arca, and was politely told, "this type and amount of distortion is a known issue of the 32mm and your lens is within spec". Live and learn, I should have made a better acquisition decision.

Paul C
The issue Paul is raising is clearly shown in this document from Alpa. I can't work out how to upload the file here so please don't all hit my bandwidth :). If somebody knows how to post the file, please do so if it's of interest.
 

GrahamWelland

Subscriber & Workshop Member
It would be interesting to compare the Alpa lens corrections available for PS and the C1 Rodie 32HR profile. I don't have many images with lateral shifts with the 32HR but I'll take a look.
 

Pemihan

Well-known member
Any suggestion about how to note down the amount of movements in the field? Will it be a possible to associate this data in RAW file itself (may be via user interface in the DB), instead of noting down into a piece of paper or in any other device?

Thanks,

Subrata
I have been asking for that for years. In the Leaf Aptus backs you can input which lens you use and it shouldn't be any problem to implement that and more in the IQ/Credo firmware.

Peter
 

GrahamWelland

Subscriber & Workshop Member
Given the processing power of the IQ backs I'm still surprised that they haven't offered the option of recording the shifts with EXIF data via the LCD screen. They definitely could but have chosen not to.
 

dougpeterson

Workshop Member
Distortion does not vary any meaningful amount from one sample of a lens to another. It is a characteristic of the lens design and is a well understood compromise to accommodate improvement in other areas. For instance allowing slight distortion may allow increased corner sharpness. The only variation will be whether someone notices the distortion, cares about that distortion and what lenses they have previously worked with (and therefore will be comparing to consciously or otherwise). One photographer may shoot mostly abstract landscape (where it's hard to see the distortion even if you look carefully) while another may shoot rigorously composed highly-structured architecture shots (where even the slightest hint of distortion can be easily evident). One may be comparing to previous dSLR wides they've used where the distortion was more severe while another may be comparing to a Schneider Digitar wide angle (which had near zero distortion)*. So, in other words, you're both right!

Anyway, distortion can be easily mapped out digitally with very minimal effect on image sharpness (since the changes being made are pretty small; unlike, for instance, perspective correction where pixels can get stretched to upwards of two hundred percent of their original size). However it does require a complete mapping of the lens to be present in the raw processor. Up until now with the 32HR the only mapping was in Alpa's Photoshop-based tool. Now it can be done within Capture One (for those using Team Phase One digital backs).

Also note: perspective distortion and geometric distortion are two entirely different things. Everything above has been referring to geometric distortion. People sometimes confuse them, especially with very wide lenses. The 32HR coupled with a single capture from a digital back is not excessively wide, but if you do a two-shot stitch (increasing the effective size of the sensor) then it can be an ultra wide lens, and in such a case perspective distortion will be quite severe - this has nothing to do with image quality or lens quality, but is rather an inherent part of capturing a very wide-angle image.

The 32HR is one of the (maybe even the) strongest wide angles ever made for any platform and a full profile in Capture One is icing on the cake.

*I'm not using the two of you as specific examples here; the point is generalized

- - - Updated - - -

Given the processing power of the IQ backs I'm still surprised that they haven't offered the option of recording the shifts with EXIF data via the LCD screen. They definitely could but have chosen not to.
Baby steps have been made. With the latest firmware you can enter copyright information via an on screen keyboard, and on the A Series you can select the lens from the list of A-Series lenses.

Hopefully it's just a matter of time before a more complete solution is provided to tech camera users. It also bugs me to see a raw file with a [--] where the focal length is supposed to be listed :(.
 

jagsiva

Active member
I have been asking for that for years. In the Leaf Aptus backs you can input which lens you use and it shouldn't be any problem to implement that and more in the IQ/Credo firmware.

Peter
Agreed that this should not be a big deal to implement. Having said that, back when I had the Aptus II 12, I tried using it, but given all the other things we need in tech cam use, it quickly became tiresome.

A quick/simple workaround is using the rating system. For example, * means 5mm shifts, ** 10mm shifts etc. Not perfect, but a quick reminder. As for FL, I can eyeball this without too much trouble.
 
Last edited:

Paul2660

Well-known member
Just some examples of the issues I am referring to.

The shot below is from a 12mm shift from a 32mm Rodie. F12. LCC applied, but no distortion corrections applied as I don't yet have C1 9 up and running.

The effect of this amount of distortion on "known" shapes would make use of this lens on certain type of shots problematic since the distortion has effected and misshapen object that the human brain knows the shape of.

Examples:

1. The 55 gallon drum barrels, there are 4 of them, the 2 further left show the effect the worse, and the left most barrel worse than the one next to it. Right here, you would more than likely say, "something is wrong" as all 55 gallon drums are the same size and shape.

2. The SUV, suburban in the distance, is elongated and squat, and it's only in a 5mm of the shift

3. The shelters/pavilions are all elongated, but these may not get caught since someone may not have a "mental" idea of their known shape.

4. The powerlines, and poles are out of whack

5. The rock structure in the foreground is way out of shape as it's normally much taller and less squat.

I can see that I have created friction on this site, the thought that the 32mm can have bad press. However I use this site as a 'real world" test ground as the fact presented tend to be very clear and straight forward. I would hope that before someone made a purchase of this lens, they were aware of just how bad this issue can be on "certain" lenses. I say certain as I realize most people don't complain about it. I can clearly state that I have read out mustache distortion for year, never really understanding the effect until I was able to use the 32mm. It was evident immediately for me. As I stated earlier, I did go back to review many of my shifted series with the 40mm Rodie and now realize the distortion is there also, just not as severe as with what I found with the 32mm.

Paul C
 

Attachments

jagsiva

Active member
Paul, I see what you are saying regarding the barrels. The one on the far left is squished! I have never really looked at this lens this way, since most landscape stuff I do has very little absolute references like these.

Having said that, this is the extreme left edge of a 80mmx40mm frame. The elongation of the SUV and buildings look like wide angle perspective distortion to me. The two barrels on the left are a different story.

For an extreme edge, that is one sharp crop though :)

Some samples....these are shifted 15mm to either side so approx 85mmx40mm effective sensor.

BTW, thanks Doug! Now I need to go reprocess all my files again:)





 

dougpeterson

Workshop Member
Paul: all of that is perspective distortion based on wideness of view.

For instance with a full frame back shifting 20mm (whether to stitch or not) is seeing the same perspective in the corner of the frame as a 14mm on a full frame in the corner of the frame. In other words: really stinking wide. At the edge of that perspective the amount of perspective distortion

This has nothing to do with quality of lens, and has nothing to do with geometric distortion (pincushion/barrel/mustache).

Anytime you have a rectilinear projection that wide you will see fairly extreme perspective distortion. This is true whether by shooting a long lens and using Photoshop to merge them into a rectilinear stitch, or by an ultrawide on an 6x17 camera or an ultra wide on a digital back, or a moderate-wide with lots of image circle stitching, or any other method. The only "solution" is to use a different kind of projection than rectilinear; for instance you can do nodal-point-stitching and then stitch in Photoshop using a Cylindrical projection (I guess another "solution" is just not shoot as wide of an angle of view).

I highly recommend this article on projections and the pros/cons of each:
Panoramic Image Projections

Look about halfway down that article to the graphic under the phrase "The first example demonstrates how a rectilinear image projection would be rendered in a photo stitch of the above three photographs." This graphic clearly illustrates the axial stretch that occurs in any rectilinear image as you get wider in angle of view.

The slight loss of straightness of the light poles is the geometric distortion that is the topic of this thread. That slight bend in the poles can now be mapped out using the Lens Correction tool in Capture One 9.0.1.
 
Top