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!

IQ160 tiling issue

avelpavel

New member
Hello Peter,

I had a similar issue with my P65+ in the past. The back has been recalibrated by Phase One but the issue was still there. The guys at Phase One told me to try this: insert this string in the name of the raw capture folder "P1#1024-OFF#" to make sure that capture 1 wasn't over correcting the images (if your folder name is "capture" use the name "P1#1024-OFF# capture" instead.
For me it worked even though the issue was present only in some low contrast black on black images.
You can give it a try.

Regards

Rob

Roberto Pastrovicchio Photography . Italy . +39 329 9617876 . [email protected]

Thank you very much everyone for your comments.

As said I got a RAW from another IQ160 which doesn't have this issue so I was thinking it could be interesting for me to get some more examples from different backs.

So if some of you have one or two RAW files you can send me via for example Dropbox I would be much appreciated. Preferably images with a lot of sky in it since that's where I mostly have the issue.

My email is pm (at) petermisfeldt . com

Peter
 

Paul2660

Well-known member
Rob.

Were you shooting tethered when you had the problem?. Did it happen when you were not tethered? I assumed this because you mention the "capture" folder.

Thanks
Paul
 

torger

Active member
If you want me to analyze the actual raw file calibration data content I can do it if I get a raw file. I know more than most people outside Phase One about the internals of this format and I should be able to tell if the calibration data seems to be mismatching or not.
 

avelpavel

New member
Paul, it happened both tethered or not. By the way I have used the back a lot in the last months and everything has worked flawlessly, using C1 7.2 at the moment and no issues, however that tip could be useful to try if the problem get back before sending the back to Denmark to recalibrate and using the Canon for the jobs (which I'm not so happy with...:rolleyes:)

Thanks
Rob



Rob.

Were you shooting tethered when you had the problem?. Did it happen when you were not tethered? I assumed this because you mention the "capture" folder.

Thanks
Paul
 

Pemihan

Well-known member
Thanks Rob,

I will try it on a folder and see if it makes a difference.
As I live less that two miles from Phase One HQ in Denmark it's not a big deal for me to send the back in if needed though :)
I would of course prefer to get a back with no issue to start with...

Peter

Paul, it happened both tethered or not. By the way I have used the back a lot in the last months and everything has worked flawlessly, using C1 7.2 at the moment and no issues, however that tip could be useful to try if the problem get back before sending the back to Denmark to recalibrate and using the Canon for the jobs (which I'm not so happy with...:rolleyes:)

Thanks
Rob
 

torger

Active member
Over at Lula I asked Adobe developers if their products (ACR, Lightroom, DNG converter) does additional processing to supress tiling after calibration data has been applied, and Eric (Chan) confirmed that they do, just as I suspected after studying the software's behaviour.

It seems Capture One does not do that, which means that if you want the best tiling supression there is you should currently use Adobe's products. Hopefully Phose One will improve their algorithm though in future C1 releases. Currently it will depend how well the back's calibration data lines up with the actual sensor behavior, which will be different from back to back.

The IIQ format calibration data tags have limited resolution so it's impossible to make a 100% perfect match, and for some sensors the error will be larger than others. For normal use it should be no problem, but in for example black-and-white post-processing there can be very strong contrast increases that will make tiny error visible, and then you probably need some luck to get a back which has a sensor that calibrates well with the current tag format, or use Lightroom and wait until Phase One hopefully improves their algorithm. Not sure how many that runs into a tiling issue in practical use though, I suspect only a few, so those that do probably need to shout loud to get the improvement in there.
 

gerald.d

Well-known member
Over at Lula I asked Adobe developers if their products (ACR, Lightroom, DNG converter) does additional processing to supress tiling after calibration data has been applied, and Eric (Chan) confirmed that they do, just as I suspected after studying the software's behaviour.

It seems Capture One does not do that, which means that if you want the best tiling supression there is you should currently use Adobe's products. Hopefully Phose One will improve their algorithm though in future C1 releases. Currently it will depend how well the back's calibration data lines up with the actual sensor behavior, which will be different from back to back.

The IIQ format calibration data tags have limited resolution so it's impossible to make a 100% perfect match, and for some sensors the error will be larger than others. For normal use it should be no problem, but in for example black-and-white post-processing there can be very strong contrast increases that will make tiny error visible, and then you probably need some luck to get a back which has a sensor that calibrates well with the current tag format, or use Lightroom and wait until Phase One hopefully improves their algorithm. Not sure how many that runs into a tiling issue in practical use though, I suspect only a few, so those that do probably need to shout loud to get the improvement in there.
Interesting - thanks for looking into this.

My IQ180 went back to Copenhagen for recalibration. Came back, and I still have issues in certain situations. Response from Phase One is that it is correctly calibrated, so I'm not entirely happy about the situation.

I'll dig out an old problem file and see how it behaves with Lightroom.

/edit

Wow. No issues in Lightroom - I've tried my best, but cannot force visibility of the tiling no matter how hard I push the file. It's clearly visible in Capture One on default settings, and only goes downhill from there.

It's crazy to think my IQ180 went back to Copenhagen, was recalibrated, and came back still with an issue, when all it (seemingly) takes to fix the problem is software.

Yes - I know software is complicated, but this needs fixing.
 
Last edited:

torger

Active member
Note that if you are using a technical wide angle you'll get slight behavior change, with the light falling in at an angle the sensor tiles will change their pixel vignetting by different amounts. This means that even if calibration data is perfect for perpendicular light (which I assume they calibrate for) it won't match for tech wides.

For a tech wide shot you would be using LCC for the color cast though, and I'd guess that tiling suppression in C1 works differently and possibly better with LCC applied, but I don't know.
 

gerald.d

Well-known member
Note that if you are using a technical wide angle you'll get slight behavior change, with the light falling in at an angle the sensor tiles will change their pixel vignetting by different amounts. This means that even if calibration data is perfect for perpendicular light (which I assume they calibrate for) it won't match for tech wides.

For a tech wide shot you would be using LCC for the color cast though, and I'd guess that tiling suppression in C1 works differently and possibly better with LCC applied, but I don't know.
Edited my post whilst you were writing that I suspect.

The image that I'm looking at was actually shot with the Canon 17mm TS-E - no LCC available.

Happy to share it if you're interested.
 

Pemihan

Well-known member
Yes, thanks Anders for looking further into this. Obviously you were right that Adobe use some kind of tiling suppression besides the calibration data. Phase One need to implement something like that in Capture One as well and better sooner than later.

Most of the time I have no tiling issues when using Lightroom, but instead I often get a thin horisontal line running through the center of the image but that is way easier to get rid of in Photoshop.

I'm also having the issue with other cameras and lenses besides my tech cam such as DF+ with 80LS and 120mm and RZ67 with various lenses.
I got a loaner back which is better but also have the issue albeit to a lesser degree.

Still in the process to get it solved.

Peter
 

torger

Active member
I've had a look at Gerald's problematic TS-E 17 IQ180 file using the development version of RawTherapee where I've recently added IIQ calibration data support (I've reverse-engineered the most important IIQ format calibration tags, sent a patch to Dave Coffin too so hopefully it will get into next version of dcraw and thus into most third-party software in time, any Dalsa-sensor Phase One back needs this).

The strange thing is that even when I really mistreat the file and increase contrast and clarity to obscene levels I can't make the line visible. RawTherapee does not have any additional tiling suppression step, like Lightroom is supposed to have. It just applies the calibration tags and that's it.

This indicates that there is nothing wrong with the file but rather some sort of bug in the current version of Capture One so that calibration data is applied in the wrong way. I can't say that for sure though, because I've only reverse-engineered the most important calibration tags and the ones Lightroom seems to be using. There is additional data too of unknown format that could be just legacy backwards compatibility stuff not used today, or it could be some fine-tuning tags used by Capture One which might make things worse rather than better if that data is not good.

Oh, did I say that this concept of proprietary undocumented file formats suck? :)
 

trond

Member
Dear Torger,

I have seen the tiling issue on my IQ180 using 32HR on my Cambo.

However, I find the problem to be much worse on LR than with C1.

LR is my raw developer of choice, and I only use C1 due to the IQ180 performs better with C1.

In addition, with LR I get a kind of "ripple" problem in blue sky when the HR32 is shifted.

Looks to me like uneven column gain/sensitivity, and is present in the last two tiles in the direction opposite of the shift.

This problem is corrected in C1 once the preview is generated.

If you pm me, I can send you LCC shots from IQ180 and HR32.

Best regards

Trond
 

torger

Active member
PM'd, I'd surely like to look at the files! This issue is puzzling indeed... results seem to be all over the map. So the best current advice if you get bad results in C1 try LR, and if you get bad results in LR try C1 :)

Does the ripple effect look like this?



Then it is a micro-lens issue (it has strongest effect on blue light), more details:
Knowledge base for Phase One and Mamiya Leaf users

Possibly C1 LCC algorithm corrects for that, and LR's maybe not. I guess you need to apply LCC to get rid of the ripple issue?
 

trond

Member
Dear Torger,

Yes, the ripple looks like the example above.

In C1 it is visible only for a few seconds, and goes away once the preview is generated, even without LCC.

In LR, the ripple remains in the processed image.

B R

Trond
 

dougpeterson

Workshop Member
The "Ripple" is corrected in Capture One when using the "Wide Angle" option in the C1's LCC tool.

I do not agree with Torger. I strongly suggest tech camera users should use C1 rather than LR wherever possible. If you are not getting excellent results (i.e. good color, no ripples, no tiling) with supported lenses in C1 with a proper LCC then it's a warranty issue that should be followed up with your dealer. The solution is not to switch to LR to mask the underlying issue, the solution is to eliminate the underlying issue.
 

trond

Member
Dear Doug,

I agree with you, C1 does a fantastic job on the IQ180 and HR32 compared to LR.

The "ripple" is corrected in C1 immediately as the preview is generated, even without the LCC.

B R

Trond
 

gerald.d

Well-known member
The "Ripple" is corrected in Capture One when using the "Wide Angle" option in the C1's LCC tool.

I do not agree with Torger. I strongly suggest tech camera users should use C1 rather than LR wherever possible. If you are not getting excellent results (i.e. good color, no ripples, no tiling) with supported lenses in C1 with a proper LCC then it's a warranty issue that should be followed up with your dealer. The solution is not to switch to LR to mask the underlying issue, the solution is to eliminate the underlying issue.
Where does one get a list of "supported lenses" (presumably this list varies by back)?
 

torger

Active member
I can of course not disagree with that it's desirable that C1 works well in all situations, we'd all like that. But apparently there are issues in some situations, and apparently Phase One has not always been able to solve the issue in cases where it indeed works in LR out of the box.

If some backs have a sensor+calibration data combination that triggers a C1 software bug, which seems likely from my observations so far, it will take time for it to trickle down to developers to get the issue solved, and a photographer may need to deliver now. In that case trying LR and see if it works is a good idea, as there are multiple examples of that working.

I do recommend all owners that see this problem to report it to dealer / Phase One of course, they like any software developer need feedback to find out that the issue exists and raw data to be able to solve it. The more users that report it, the more likely it will be fixed.

From the observations I've made so far it's a bit hard to pin down exactly what the problem is though (it could be more than one problem), and I think Phase One have the same difficulty, that's why some that send in for recalibration can get something back that still doesn't work.
 

torger

Active member
Where does one get a list of "supported lenses" (presumably this list varies by back)?
It's a bit hard to produce such a list for technical wide angles, as it's not an on/off thing, and there's a lot of "it depends". I guess that's why there's no widely available list.

When a tech wide is shifted you get various types of degradations, vignetting, color cast, microlens shading (ripple effect, seen on mainly IQ180, but also on IQ160 to some extent), crosstalk (color desaturation). With software and LCC you can to a large extent compensate for these effects. With improved algorithms, you can get better compensation, ie improve support.

It becomes much a matter of taste, how far one think one can shift and push these compensation algorithms and still be pleased with the result.

The discussion about the "usable image circle" of the early IQ250 tech cam tests was an illustrating example where there were various opinions of how much crosstalk/desaturation one could accept.
 
Top