Ad
Ad
Ad
Pages: « 1 2 [3]   Bottom of Page
Print
Author Topic: Aptus 75 centerfold issue still not solved  (Read 19229 times)
E_Edwards
Full Member
***
Offline Offline

Posts: 228


« Reply #40 on: December 15, 2006, 02:37:46 PM »
ReplyReply

I personally don't believe any of these stories to do with angle of incidence hitting the imager and all that.

I have two identical brand new Aptus 65 cameras.  I shoot with one either oriented vertically or horizontallly, perfect, no centrefold whatsoever.

Swap to the other identical back, same camera, same lighting, moving nothing, just a straight swap and I get the centrefold on horizontal shooting on the other back.

The problem is evidently with the chip,its internal calibration or whatever. Frankly,  I don't want to know the technicalities or have any excuses. All I know is that one back has the problem and the other one doesn't.

Not that it bothers me much, basically I've stored a custom gain for Vertical shooting and another one for Horizontal shooting and it takes 10 secs to load it prior to the shoot.

I'm sure Leaf will find a solution sooner or later, probably later, when every other camera has already resolved theirs.

Basically I can live with little problems like this so long as the overall experience is positive and the backs save me time and earn me money.

But next year I will upgrade both (if there is something on offer that is better for me) , I will look  carefully at the competition, they all have faults or restrictions. No back is perfect, I know that from experience. If I find an overall better back, I will wave Leaf good bye in an instant and go for whoever offers the best for me. It is also quite possible that I will stick with Leaf, as I believe it is the best choice for me at the moment, despite everything.

Edward
« Last Edit: December 15, 2006, 02:43:13 PM by E_Edwards » Logged
rainer_v
Sr. Member
****
Offline Offline

Posts: 1131


WWW
« Reply #41 on: December 15, 2006, 06:18:47 PM »
ReplyReply

eric,
although i am shooting exterior usually untethered, for interior i am using tethered shooting as well.
 the workflow is the same with the emotion, you either can take stephans tools or capture. i am using stephans programms, which include now also a tethered shooting programm.

the difference between the sinar / brumbaer solution and the leaf gain adjuster is that the cf in the images is corrected directly in the sinar/brumbaer "raw" files  by a specific software algorythm. this is not a patch, its a solution and it is fundamental different than the gain adjuster and its inverting method.
the white reference shots in brumbaer/ capture have nothing to do with the cf issues.
you can do them ( to increase the  grey uniformity ) or you dont do them,- there is no cf issue in any case.


about the reasons and about base info about this issue i suggest to search here in the forum about info stephan and me have already provided.  i dont want to repeat it one time more.
in general it has nothing to do with lenses, exposure times or motifs, this facts just make the issue more or less visible,- further there is also variation between the backs cause the dalsa specs claim 3% tolerances between the sensor zones  which is too much if you have bad luck ( hi tolerance in your sensor ) or which remains invisible if you have good luck with your back ( lo tolerance in your sensor ). so exchanging the backs might bring the solution ( luck ) or not ( not so much luck )
but as i said before sinar/ brumbaer has a definitive solution for that which is not a "cosmetic" patch and this depends not on the dalsa specs,- the cf issue is removed also with "bad" backs ( 3% tolerance ).
i would simply advise for architecture shooting ( using shift lenses ) the emotion 75 or 22, it  seems to me as this would be  the best back at the moment, regarding the image quality, the workflow and even the price.

but i have too less experience with the phase backs, to know how is the workflow with shift lenses using  their lcc shootings and how is the image quality also in terms of fringing,battery use, lcd quality, writing speed and this kind of "details". i was thinking to take the dalsa route cause my experience with the kodak sensors was not that good before, and as i believe they have not changed their sensor design which leaded to many effects as the magenta-green shifts with many lenses and high energy consumption,- which originally the dalsa sensors have shown to a far less degree.  but as i said i havent worked with a p30 or p45.
« Last Edit: December 15, 2006, 07:34:03 PM by rehnniar » Logged

rainer viertlböck
architecture photographer
munich / germany

www.tangential.de
John Camp
Sr. Member
****
Offline Offline

Posts: 1259


« Reply #42 on: December 15, 2006, 07:54:27 PM »
ReplyReply

James,
I don't own a MF format back, but I was one of the first people in the US to get an M8; that M8 is now back in Germany (god willing) to be fixed. If I get lucky, it will be fixed and shipped back before the German Christmas holiday, which, as I understand it, extends from Dec. 23 to Jan. 2. If I get unlucky, I lose the camera for a additional ten days, or five-six weeks total.

The M8 had two major problems. First, it was too sensitive to IR, and therefore dark neutral colors that reflected a lot of IR, usually synthetic fabrics or fabrics with a waterproof coating, became visibly magenta. REALLY magenta. The other fault produced severe banding when over-exposed lights were present in a frame shot in otherwise dark conditions, and a weird green mirror image. In other words, you really can't count on doing low-light street shooting, which is one of the M's reasons for being.

One interesting thing: the **very first person** to take delivery of a commercial M8 (as far as we know) discovered the banding fault **on his first night of shooting.** I think I may have been the first to notice the green mirror images, on the first day that M8s were widely delivered in the US. On the second or third day of shooting, after the first deliveries, somebody shot a concert in which all the players were wearing tuxes, and all the tuxes were...magenta. Nasty.

There was a web-wide Leica freak-out, and what happened then was as interesting as the faults. I honestly believe that Leica simply didn't know what to do. They tried to do the best they could, and IMHO have come up with the most practical solution, but not without a lot of trashing around. These high-tech camera companies just are not good at consumer relations. Nikon is not good, Hasselblad has recently pissed off people by closing the H3 system, Leaf has got this centerfold problem or whatever you call it. To me, the eventual technical responses seem appropriate -- but the communications problem is awful. I just wonder if there is not some cultural thing going on here too; do the European shooters complain as much as Americans? Or is it just us?

JC
Logged
James Russell
Guest
« Reply #43 on: December 15, 2006, 11:35:06 PM »
ReplyReply

Quote
, and a weird green mirror image. In other words, you really can't count on doing low-light street shooting, which is one of the M's reasons for being.
[{POST_SNAPBACK}][/a]


Many of the specialized digital cameras can have issues, but sometimes it makes me wonder if they are really tested and checked for their intended use.

Regardless, I think most of us can accept a problem as long as the manufacturer acknowledges it and gives is a time line for a definate fix  and if a workaround is required, publishes that quickly.

Obviously Lecia did that by recalling the cameras and offering filters.

I would assume most of us here earn our living with these tools and consistent reliability is as important as file quality and look, actually I think more important when it's for commerce.  What seems like a simple issue of IR or WB response can turn a successful project into one that becomes labeled as problematic.

I've had great results with my A-22 as long as I tethered with the older V-8 software and edited/named/processed in third party applications, but even with my own stable workflow, it takes a lot of study and a great deal of  patience to accept these workarounds.

This week I just added a P-30 mostly for the stability and workflow of C-1.  

So far I like the file quality very much but, even if it turns out that image quality is only equal to the Leaf, the stability of a mature software is worth the investment.

So often we all get so deeply involved and invest so much time and effort in learning these cameras we build an unintentional loyalty, almost irrational at times, probably like becoming close to a doctor.

In my case I hope I never get to know the doctors at Phase.

JR
[a href=\"http://ishotit.com/]http://ishotit.com/[/url]
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #44 on: December 16, 2006, 04:35:09 AM »
ReplyReply

The center line is obviously a sensor calibration issue; the sensor has two halves with slightly different sensitivities, and the proper corrections are not being applied to the RAW data prior to conversion. Most likely, 2 different A/D converters are being used to shorten readout times and increase the frame rate, and they are not converting identical voltages off the sensor to identical digital values. A faulty lens would not cause a split-screen type perfectly straight line down the middle of the frame with differing luminance on each side of the line.

One can occasionally see a similar defect in Canon 1Ds-MkII images shot at high ISO and underexposed by two stops or more. The 1D-MkII uses 8 A/D converters to read the data off the sensor, but instead of each converter reading a large rectangular area of the sensor, each converter reads every eighth column of pixels. As a result, what yoiu see is every eighth column of pixels being slightly brighter or darker in parts of the image.

This problem is something that ought to be addressed in the camera firmware.
Logged

ericstaud
Sr. Member
****
Offline Offline

Posts: 384


WWW
« Reply #45 on: December 16, 2006, 09:15:46 AM »
ReplyReply

I did own 4 seperate Aptus 75 backs.  Each one showed centerfold AND colorcasts on my Alpa camera.  None of them showed centerfold or colorcasts on the H1.  So for me, the conversation about what causes centerfold is a little academic.  Even if the centerold issue is solved 100%, shooting a white plexi to correct the color casts would still be a fact of life for the architecture shooter who started this thread.  Because the Custom Gain adjuster would still be run on every shot, there is no point in Leaf fixing the centerfold (except for those costomers like Edward who see it on an SLR camera).  I don't think Leaf is going to start delivery backs that have no colorcasts on these specialized cameras.  So it is, in my opinion, more important for Leaf to continue developing a custom gain solution which works in LC10, works more quickly in LC8, and is more automated when downloading from a card.

Edward, I think your situation with the Aptus 65 is different than with Pixjohn.  Leaf will send you a back without centerfold and you may never shoot a gain diffuser again because you're working with an SLR.  It is a more isolated  incident that youe camera has non-uniformity issues.  For the architecture shooter with an Alpa or a Cambo, EVERY 33MP and 39MP on the market today in not uniform.  As a result of this difference Edward, you are waiting for a hardware solution to be delivered, but Pixjohn is waiting for an improved software solution.  Phase went through this same experience.  A few Phase customers had color casts with the P45 on SLR cameras.  Phase fixed the problem for those people.  Every architecture shooter with a P45 is given a white plexi card and shown how to use it.

There are thousands of Aptus 75, P45, eMotion75, and H3 MFDB's being used by photographers with SLR cameras every day with NO centerfolds or colorcasts.  Edward's is an isolated occurance.  99.9% of those backs would show color casts needing correction if I put them on my Alpa with a 35mm digitar.
Logged
yaya
Guest
« Reply #46 on: December 16, 2006, 10:04:35 AM »
ReplyReply

Quote
The center line is obviously a sensor calibration issue; the sensor has two halves with slightly different sensitivities, and the proper corrections are not being applied to the RAW data prior to conversion. Most likely, 2 different A/D converters are being used to shorten readout times and increase the frame rate, and they are not converting identical voltages off the sensor to identical digital values. A faulty lens would not cause a split-screen type perfectly straight line down the middle of the frame with differing luminance on each side of the line.

One can occasionally see a similar defect in Canon 1Ds-MkII images shot at high ISO and underexposed by two stops or more. The 1D-MkII uses 8 A/D converters to read the data off the sensor, but instead of each converter reading a large rectangular area of the sensor, each converter reads every eighth column of pixels. As a result, what yoiu see is every eighth column of pixels being slightly brighter or darker in parts of the image.

This problem is something that ought to be addressed in the camera firmware.
[a href=\"index.php?act=findpost&pid=90813\"][{POST_SNAPBACK}][/a]

The technical explanation was given several times on this forum, most accurately by Stephan Hess.

The line in the centre is an accute case of colour cast and it appears in the centre as a result of the 6 "panels" the chip is died with. Nothing to do with "two halves" as there are no two halves.

A hardware or firmware solution is the most desired one and sooner or later it will be achieved.

Eric is right about WA non-retrofocal lenses, this is where most sensors will show colour casts and   since there are endless combinations of apertures and shifts, there will always be a need for some sort of gain adjustment, at least with the current lens/ glass design.

Yair

[span style=\'font-size:8pt;line-height:100%\']Yair Shahar | Regional Manager | Leaf EMEA |
mob: +44 77 8992 8199 | yair.shahar@kodak.com | www.leaf-photography.com
[/span]
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #47 on: December 16, 2006, 12:13:05 PM »
ReplyReply

Quote
The line in the centre is an accute case of colour cast and it appears in the centre as a result of the 6 "panels" the chip is died with. Nothing to do with "two halves" as there are no two halves.

The camera is successfully balancing the out the luminance along the edges of the 3 panels on the right, and among the 3 panels on the left, but not along the midsection of the chip. That's got to be a firmware bug; only half of the edge-matching job is being done.
Logged

E_Edwards
Full Member
***
Offline Offline

Posts: 228


« Reply #48 on: December 16, 2006, 01:08:51 PM »
ReplyReply

Quote
Edward, I think your situation with the Aptus 65 is different than with Pixjohn.  Leaf will send you a back without centerfold and you may never shoot a gain diffuser again because you're working with an SLR.  It is a more isolated  incident that youe camera has non-uniformity issues.  For the architecture shooter with an Alpa or a Cambo, EVERY 33MP and 39MP on the market today in not uniform.  As a result of this difference Edward, you are waiting for a hardware solution to be delivered, but Pixjohn is waiting for an improved software solution.  Phase went through this same experience.  A few Phase customers had color casts with the P45 on SLR cameras.  Phase fixed the problem for those people.  Every architecture shooter with a P45 is given a white plexi card and shown how to use it.


Sorry Eric, I don't know where you got that I shoot with SRL's and Aptus. I actually shoot with two Aptus 65, both fitted on Sinar P view cameras and both bought at the same time.

This centrefold occurs when I use the Aptus back mounted on a Sinar horizontally, just the fact I'm turning the mount 90 degrees from portrait to landscape, causes the centrefold. It happens with any lens, from 100mm, 120mm, 135mm 150 and  180mm all digitar, no movements or shift or anything.

I swap the backs, touching nothing, just unclipping from the mount, stick the other back on, and one back has no problem, the other back has. Both backs show absolutely no centrefold shooting vertically, both with the default gain from factory.

If I adjust the gain so that it corrects the centrefold on the "bad" back shooting lanscape orientation, then the centrefold will this time happen on portrait orientation, so the solution for me is to load either of a set of gains (the default for vertical and my own gain for horizontal shooting).

The "good" back doesn't need any of this messing around, it shoots fine in either orientation. I would rather not have to mess changing gain every time I change orientation.

I trust Leaf will find an automatic solution soon, because the way I see it, if every other camera back manufacturer has found a solution and Leaf falls behind, many people, including me, are going to be really pissed off.

Edward
Logged
ericstaud
Sr. Member
****
Offline Offline

Posts: 384


WWW
« Reply #49 on: December 16, 2006, 04:39:46 PM »
ReplyReply

Quote
Sorry Eric, I don't know where you got that I shoot with SRL's and Aptus. I actually shoot with two Aptus 65, both fitted on Sinar P view cameras and both bought at the same time.

Edward
[a href=\"index.php?act=findpost&pid=90859\"][{POST_SNAPBACK}][/a]

My bad Edward.  Just assumed you were using an SLR.  I noticed centerfold much more in horizontal shots because the line would run into the blue sky, which made it very easy to see.  Shooting vetically the subject matter usually hid the effect.  Good luck with a resolution.  I don't think it will be difficult to get two satisfactory backs with those longer lenses.
« Last Edit: December 16, 2006, 04:40:14 PM by ericstaud » Logged
Pages: « 1 2 [3]   Top of Page
Print
Jump to:  

Ad
Ad
Ad