Pages: 1 [2] 3   Go Down

Author Topic: too strong colorfridging with Phase one P65+ shooting architecture  (Read 18599 times)

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #20 on: February 19, 2013, 12:58:06 pm »

that is the reason why they invented technical camera's ...for perspective ...
 
all we need now is a manufacturer that doesn't forget to test their chips for
side way adjustments...

the kodakchip  is better for architecture , simple .

Great consideration is given to the design benefits/drawbacks of microlenses and pixel size at both Dalsa and Phase One as products are developed. The idea that Phase One doesn't "test for sideway adjustment" is just crazy wrong. Tech cameras are in fact a heavy design emphasis for Phase One and it's a heavy part of testing from development all the way past release.

So much so that the tech camera market is dominated by Team Phase One with features like an onboard battery (P/P+/IQ), in-camera application of LCC (Aptus II), custom metadata entry (Aptus II), two axis level (IQ/Credo) including autocorrection in C1 for being off-kilter, no-wakeup required sync (P40/P65/IQ/Credo/Aptus-II), long exposure (P+ excluding 40/60), great 100% review (Aptus II, IQ, Credo), super fast interface for in-camera review (IQ/Credo), ability to review the last 10 images from the back when shooting tethered (IQ/Credo), Live View (with limitations, IQ/Credo).

What you're seeing here is what's thoroughly documented by myself and others on the forum, Phase One on their Knowledge Base, and discussed by any good Phase One dealer with a client that is using (or might consider in the future) these kinds of lenses/bodies/backs: 6 and 5.2 micron backs (P40/P65/IQ140/160/180/Credo) do better with the mild retrofocus design of Rodenstock's lenses. The P65+ works very well with the 43XL on up, and pretty well with the 35XL, and even does very well with the 28XL straight on, but will not sustain 12mm of shift. If you came to me when purchasing and told me that shifting 12mm on a 28XL and using the edge of the frame then I would not sell you a 65+. I would instead send you a sample image of a 32HR with 12mm of shift and explain your options:
- a back with a larger micron size (e.g. P45+)
- or a lens like the 32HR with less extreme angle of light (which provides better sharpness, great microcontrast, very little color cast, but comes at the price of added cost, weight, and distortion).

All that said, I think you are overstating the difference compared to your H4. The images you've posted look of comparable quality to the edge of the H4 frame. The IQ, being a larger sensor, allows you to see further towards the edge of the 28XL image circle where color cast becomes more severe. Likewise if you slapped an IQ140 on there you'd find that there was very little color cast at 12mm of shift, but only because you'd have effectively cropped out the potentially problematic area.
« Last Edit: February 19, 2013, 01:00:58 pm by Doug Peterson »
Logged

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #21 on: February 19, 2013, 01:22:01 pm »

Remember film? Linhofs, Sinars, DeVeres, Deardorffs, Super Angulons...

Just kiddin'!

;-)

Rob C

julienlanoo

  • Full Member
  • ***
  • Offline Offline
  • Posts: 228
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #22 on: February 19, 2013, 04:05:45 pm »

Just to know, not at all bashinh on Dalsas, are there any brands still developping on Kodak sensors?... Just to enrich my knowledge ...

What i ve also remarked is that one could find much more specialized backs i.a multishot and achromatic, based on kodak sensors ( the 39mpix one and the 50mpix one) than on Dalsas, is there a reason for that ?...
Logged

FredBGG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1630
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #23 on: February 19, 2013, 09:15:57 pm »

All that said, I think you are overstating the difference compared to your H4. The images you've posted look of comparable quality to the edge of the H4 frame. The IQ, being a larger sensor, allows you to see further towards the edge of the 28XL image circle where color cast becomes more severe. Likewise if you slapped an IQ140 on there you'd find that there was very little color cast at 12mm of shift, but only because you'd have effectively cropped out the potentially problematic area.

I don't think he has overstated the problem.
Here are the two uncorrected shots overlaid on each other.
It is clear that even if you were to crop them the same the P65+
has significantly more lens cast color artifacting and saturation problems.
Just look at the color of the books. Not even close to comparable to the edge of the H4 50.
The color cast is even visible on the arm chair.
Also the right side is crunched down and darker with the P65+ despite the
exposure being brighter inb the center of the image.
Both shots are without the LCC applied.

« Last Edit: February 19, 2013, 09:18:16 pm by FredBGG »
Logged

siebel

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 59
    • Bryan Siebel Photographer
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #24 on: February 19, 2013, 09:19:36 pm »

Frankly, this whole thread is a little annoying. All the "problems" discussed here are well documented and have been discussed ad-infinitum here and in numerous other places for a long time.
I am a pro architectural shooter who shoots 98% of my work on an Alpa tech cam with an IQ180. It has limitations. I knew about them when I bought this setup because I did my research and did my testing first, as should you have. I had a P65+ before getting my Aptus 12 and the the IQ180 and back then, the problems with the 28mm SK were well known. Frankly, if you are going to have a go at a manufacturer, you should have a chat to SK about the 28mm. They have stuck their head in the sand in their adherance to their lens design methods, when the issues with lenscast on CCD sensors has been there all along. As pixel size decreases, lenscast gets worse. Sensor development was not going to stop because they insist on symmetrical lens designs. The laws of physics don't change because you want them to. You bought your back without doing enough homework. Suck it up and deal with it. Don't blame Dalsa and Phase. It is normal in photography that as you make changes to major items of kit, a few others might need to be changed too. When I went from the P65+ to the two 80MP backs, my SK 35XL went from 'favourite lens' to 'paperweight'. I understood that this would happen, so I dealt with it in the normal way - I sold it and got a 23HR and 40HR Rodenstock. This was expensive, but a consequence of MY decision to upgrade my back.
Personally, I dont see why massive lens-shifts are seen as such a sacred cow. There is more than one way to skin a cat. I shoot more with my 23HR than anything else and it has barely 3mm shift. I've done what photographers have done since the beginning of photography - I adjusted my methods to accommodate the strengths and weaknesses of the gear I choose to work with. Slightly less available shift has not diminished what my clients think of my work one bit, nor has it tied my hands on shoots.
Sorry if this seems a little blunt, but sometimes someone has to speak up and call a spade a spade.

Siebel
Logged
Bryan Siebel

In the end, it's all about the image.
www.bryansiebel.com

ChristopherBarrett

  • Guest
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #25 on: February 19, 2013, 10:16:56 pm »

I guess this was never much of an issue with me because I feel that that much shift on that wide of a lens just looks like crap in the first place.  Honestly, the perspective ends up feeling so exaggerated that I believe that it's a poor imaging choice.  Sure, I used to shoot a 65mm on 4x5 and shift it to the edge of the image circle to get the top of a building in... and I did achieve the composition, but it always looked just awful to me.  I don't carry anything wider than the Schneider 35mm with my P65+ and I don't even use the 35 that often.  The 43 almost always just feels right to me... but so did the 115 on 4x5.

Huh... I'm beginning to sound a bit curmudgeon-y now that I'm in my 40's...
Logged

FredBGG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1630
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #26 on: February 19, 2013, 11:26:09 pm »

I find much of this post very misleading. I don't have enough time or patience to make a complete reply. But suffice to say that implying Dalsa's sensor designs are less photographically oriented because their parent company has many areas of business is complete BS. Do you know how many non-photographic endeavors Kodak, Canon, and Nikon/Minolta-Sony are engaged in??


Not what I said at all. I'm simply pointing out that Dalsa's focus is not on the visual art of photography and that their products are far more focused on
scientific and industrial use. While Kodak comes from a long history of making film that has been used by professionals with tilt shift lenses
Dalsa comes from an industrial, medical, and military "machine vision" background.

Just taking a look at the Dalsa website at their Imaging markets and applications page:
http://teledynedalsa.com/imaging/markets/

Here is what's listed:
Machine Vision & Inspection
Defense & Security
Medical & Life Sciences
Aerospace
Intelligent Transportation Systems
Scientific

or in more detail....

Machine Vision & Inspection
Semiconductor
Electronics
Flat Panel Displays
Solar Panel
Postal and Parcel Sorting
Food and Packaging
Automotive
Pharmaceutical
General Manufacturing
Web Inspection
Industrial X-Ray / NDT
Explore Machine Vision & Inspection
Defense & Security
ISTAR and RSTAR
Threat Detection
Situational Awareness
Night Vision and Enhanced Vision
Explore Defense & Security
Medical & Life Sciences
Orthopedic and Surgical Radiography
Dental Radiography
Mammography
Bone Densitometry
Ophthalmology
Explore Medical & Life Sciences
Intelligent Transportation Systems
Toll Management
Safety Monitoring
Traffic Flow Control
Speed and Red Light Enforcement
Access Control
Road/Rail/Vehicle Inspection
Explore Intelligent Transportation Systems
Scientific
Crystallography
Microscopy
Astronomy
Spectroscopy
Ultra-High Speed
Explore Scientific

Aerospace
Multispectral
Hyperspectral
High Resolution Panchromatic
Explore Aerospace

Not one mention of MF photography. It is obviously not a priority or focus of the company. And that's Dalsa, not even the parent company Teledyne.
« Last Edit: February 19, 2013, 11:44:01 pm by FredBGG »
Logged

yaya

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1254
    • http://yayapro.com
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #27 on: February 20, 2013, 02:10:41 am »

When we started using Philips (now Dalsa) sensors in MF backs in 1998 they've already made a commitment to develop and make sensors for the photographic market. See http://www.teledynedalsa.com/imaging/products/sensors/area-scan/FTF3020C/ and http://www.epi-centre.com/reports/9906cs.html

The first larger than a thumbnail Kodak sensor appeared much later.
Ever since 1998 there were more MF backs produced using Dalsa sensors than any other vendor and right now there are more backs being used on view/ tech cameras that are equipped with a Dalsa sensor of 22 or more megapixels compared to other sensor vendors

The 43x32 17MP chip was a custom design, as were/are the 44x33 (28MP and 40MP), the 56x36 (56MP) and now the 54x42 (60MP and 80MP). They've all first appeared exclusively in a digital back and a few of them became an off-the-shelf product some time after that

Also worth noting that in 2000 we've developed our own 6MP CMOS sensor which at the time was larger than any Kodak/ Sony/ Canon device

Just correcting the perspective that some here try to skew...I'm with Bryan, Doug and Chris. If you want a larger & stronger engine you're going to have to change your tyres...not the other way around...

BR

Yair
« Last Edit: February 20, 2013, 04:05:19 am by yaya »
Logged
Yair Shahar | Product Manager | Phase One - Cultural Heritage
e: ysh@phaseone.com |

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #29 on: February 20, 2013, 08:41:42 am »

Not what I said at all. I'm simply pointing out that Dalsa's focus is not on the visual art of photography and that their products are far more focused on scientific and industrial use.

Hi Fred,

And? What does that imply?

Quote
While Kodak comes from a long history of making film that has been used by professionals with tilt shift lenses Dalsa comes from an industrial, medical, and military "machine vision" background.

I'm not sure what that's supposed to mean, but you do realise that most of Kodak's sensors were sold to the very same target groups as Dalsa's, don't you? I seriously hope that you don't think that creative/professional photography was the main market for Kodak's sensor division.

Quote
It is obviously not a priority or focus of the company.

That's a red herring. Just because they sell more for other uses (some of which with very similar quality goals, as would be useful for creative photography, compared to instrumentation and medical/scientific use)? Some of those fields demand much higher quality (color accuracy / noise performance / resistance to environmental influences / etc.) than creative photography. It's sometimes cheaper to keep some of those superior features than to design them out of the standardized production.

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #30 on: February 20, 2013, 01:35:58 pm »

i feel your pain Bart.

Hi,

I'm not hurting, I don't even use a Phase One back. You seem to have jumped to a wrong conclusion.

Quote
but don't attack the messenger , go talk to Dalsa , that work mostly for people that are pretty angry too , the militairy complex.
and ask them for better wide angle responsiveness .

I'm not attacking anybody, just wondering about the validity of some of the arguments people use.

When people feel the need to use red herrings to support their position, that is often a signal that something may be wrong with the rest of their arguments.

It's unfortunate that apparently some sensor designs are so sensitive to color cast issues, it's a good thing to discuss that. It may help prospective users avoiding a nasty surprise.

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

bcooter

  • Guest
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #31 on: February 20, 2013, 01:38:10 pm »

Hey Yair, Doug, CB,

Don't you know unless your post mentions space exploration and show a photograph of a Squirrel that your responses have no validity.

Get with the program guys and please stop being logical.

In all seriousness, every camera and sensor has a sweet spot for lenses, or lighting, or subject or all three, every camera I've ever used has the perfect storm of issues depending on lighting, subject and/or lenses.

I can take a Canon, Nikon, Leaf, Phase, Leica, RED, Sony and shoot a specific image and prove to you why that one is the best for that scenario.  Can turn around and do the exact opposite.

I'm not dissing the guy with the problem cause that's a drag man, but I know with any camera that costs over $500 get your hands on one and test the hell out of it in every condition you might work in.

When we bought our first RED it was impossible to buy new with an X sensor and searched for nearly a year to find a seller that would actually sell the camera and transfer the ownership.  When we found the right seller,  our Studio Manager went to Chicago, tested the camera in every way possible, then and only then handed over the check.

It might seem like overkill, but it's a lot better than having any issue your not prepared for.


IMO

BC
« Last Edit: February 20, 2013, 02:15:06 pm by bcooter »
Logged

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #32 on: February 20, 2013, 01:42:42 pm »

*quietly deletes my treasured photo of a squirrel in space*

FredBGG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1630
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #33 on: February 22, 2013, 11:29:37 am »

Dalsa's do fine with shift as I posted before.  They are just more dependent on the lenses used.  They don't like the wider Schneiders i.e. the 24mm, 28mm.  If you want 15mm of shift on a wide with a P65+  you have to look to the Rodenstocks preferably the 32mm.  The 60mp Dalsa's from Phase one have quite a bit more leeway with shift once you get to the 43mm Schneider or the 40mm Rodenstock both of these lenses can shift to 15mm.  I often shift my 43mm Schneider to 18mm. 

If you get up to the IQ180, then the list of wide lenses that handle shift well, gets much shorter. 

Paul Caldwell


The important number for architectural photography is NOT the amount of shift in mm, but the angle of correction that the shift produces.

If you have a lens with a retro focus design that positions the lens farther away from the sensor a shift of say 10mm on it will not produce the same amount of
perspective correction as a lens that is closer to the film plane using the same 10mm of shift.

What I am saying is that a retrofocus design will have to be shifted more than a sysmetric design in order to obtain the same correction.
So unless the Rodenstock can be shifted much more than the schneider you really are not changing much.

If one lens is 20mm from the sensor and it is shifted by 20mm the camera will be looking up at 45 degrees.

To acheieve the same with a lens at 40mm from the sensor you would have to shift 40mm... well that's if you actually could.
Logged

FredBGG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1630
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #34 on: February 22, 2013, 11:38:12 am »

Frankly, this whole thread is a little annoying. All the "problems" discussed here are well documented and have been discussed ad-infinitum here and in numerous other places for a long time.
I am a pro architectural shooter who shoots 98% of my work on an Alpa tech cam with an IQ180. It has limitations. I knew about them when I bought this setup because I did my research and did my testing first, as should you have. I had a P65+ before getting my Aptus 12 and the the IQ180 and back then, the problems with the 28mm SK were well known. Frankly, if you are going to have a go at a manufacturer, you should have a chat to SK about the 28mm. They have stuck their head in the sand in their adherance to their lens design methods, when the issues with lenscast on CCD sensors has been there all along. As pixel size decreases, lenscast gets worse. Sensor development was not going to stop because they insist on symmetrical lens designs. The laws of physics don't change because you want them to. You bought your back without doing enough homework. Suck it up and deal with it. Don't blame Dalsa and Phase. It is normal in photography that as you make changes to major items of kit, a few others might need to be changed too. When I went from the P65+ to the two 80MP backs, my SK 35XL went from 'favourite lens' to 'paperweight'. I understood that this would happen, so I dealt with it in the normal way - I sold it and got a 23HR and 40HR Rodenstock. This was expensive, but a consequence of MY decision to upgrade my back.
Personally, I dont see why massive lens-shifts are seen as such a sacred cow. There is more than one way to skin a cat. I shoot more with my 23HR than anything else and it has barely 3mm shift. I've done what photographers have done since the beginning of photography - I adjusted my methods to accommodate the strengths and weaknesses of the gear I choose to work with. Slightly less available shift has not diminished what my clients think of my work one bit, nor has it tied my hands on shoots.
Sorry if this seems a little blunt, but sometimes someone has to speak up and call a spade a spade.

Siebel


So to simplify your response and call a spade a spade.... in moving up to the IQ180 you and your clients have accepted giving up on being able to more freely tilt and shifting a technique that has been the staple of so much professional photography. That is perfectly fine as a personal choice, but hardly representative of photographic equipment advancement.
« Last Edit: February 22, 2013, 11:39:59 am by FredBGG »
Logged

FredBGG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1630
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #35 on: February 22, 2013, 11:44:10 am »

For static architecture and wide angle shots why not use stitching?

With a correctly setup stitching head and a high dynamic range DSLR and a distortion free lens
one can obtain 200mp files or more and use distortion plug-ins in photoshop to correct perspective
distortion. The final files are far superior to single shot MF.

I re shot something recently.... interior TV set. It had been shot with a wide angle and an MF.
The corners were no good one enlarged to 44x180 inches. I re shot it with an 85mm. Came out great.
Funny thing is I was not even there to shoot that, but there to do a portrait.
Client was tickled by how I pulled it off with "toys" and it lead to another project involving many days of shooting.

Similar could be achieved with MF stitching, but just far simpler and quicker to do it with a DSLR and automated robotic heads.
« Last Edit: February 22, 2013, 12:05:56 pm by FredBGG »
Logged

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #36 on: February 22, 2013, 12:37:22 pm »

The important number for architectural photography is NOT the amount of shift in mm, but the angle of correction that the shift produces.

If you have a lens with a retro focus design that positions the lens farther away from the sensor a shift of say 10mm on it will not produce the same amount of
perspective correction as a lens that is closer to the film plane using the same 10mm of shift.

What I am saying is that a retrofocus design will have to be shifted more than a sysmetric design in order to obtain the same correction.
So unless the Rodenstock can be shifted much more than the schneider you really are not changing much.

If one lens is 20mm from the sensor and it is shifted by 20mm the camera will be looking up at 45 degrees.

To acheieve the same with a lens at 40mm from the sensor you would have to shift 40mm... well that's if you actually could.

This is not correct.

20mm of shift results in the same relative frame movement regardless of the design of the lens. Only focal length and amount of shift matter.

FredBGG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1630
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #37 on: February 22, 2013, 01:04:09 pm »

This is not correct.

20mm of shift results in the same relative frame movement regardless of the design of the lens. Only focal length and amount of shift matter.

I think you missed my point. True it is the same relative frame movement, but the change apparent camera tilt is not the same.

Lets say that from the photographers position he has to "look up at 45 degrees in order to get a building in the from top to bottom.
To achieve the same effect of tilting up the tripod head 45 degrees you would need to shift a lens that is 20mm from the sensor up by 20mm.

A retrofocal lens that is say 40mm from the sensor would have to be shifted up by 40mm to achieve the same apparent tripod head tilt.
Logged

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #38 on: February 22, 2013, 01:09:35 pm »

I think you missed my point. True it is the same relative frame movement, but the change apparent camera tilt is not the same.

Lets say that from the photographers position he has to "look up at 45 degrees in order to get a building in the from top to bottom.
To achieve the same effect of tilting up the tripod head 45 degrees you would need to shift a lens that is 20mm from the sensor up by 20mm.

A retrofocal lens that is say 40mm from the sensor would have to be shifted up by 40mm to achieve the same apparent tripod head tilt.

That is not correct.

To achieve the effect of a tripod-head-tilt of XX degrees requires the same amount of rise (in mm) for two lenses of the same focal length - regardless of whether the lenses are symmetrical or retro focus.

This can be confirmed by either actual use of the equipment or geometry taking proper account of the effect of retro focus design on the basic optical principals of a lens.
« Last Edit: February 22, 2013, 01:20:36 pm by Doug Peterson »
Logged

yaya

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1254
    • http://yayapro.com
Re: too strong colorfridging with Phase one P65+ shooting architecture
« Reply #39 on: February 22, 2013, 01:12:49 pm »

That is not correct.

To achieve the effect of a tripod-head-tilt of XX degrees requires the same amount of rise (in mm) for two lenses of the same focal length - regardless of whether the lenses are symmetrical or retro focus.

+7

...and retrofocus design moves the exit pupil, not the lens...
Logged
Yair Shahar | Product Manager | Phase One - Cultural Heritage
e: ysh@phaseone.com |
Pages: 1 [2] 3   Go Up