[falconeye]

Hi Falk,
I’m just citing what Thommo stated, in the link from earlier in the thread :

Thanks Jonson for providing more context.
So, I take this as an educated guess from the Australian distributor. He may be right. But AFAIK, the real cause is under examination as we speak.

An AA filter replacement doesn't cost this much to the manufacturer. But the figure I quoted was to illustrate the amount of work involved. The AA filter cannot be removed without basically disassembling a large part of the camera. It doesn't seem to be accessible from the top.

[ve2vfd]

Falconeye ... Would it be easier to swap out the entire filter/sensor/stabilizer assembly? I haven't seen the K5d schematics so this is just a shot in the dark, but I would expect it to come in one assembled unit.

Pat

[falconeye]

Falconeye ... Would it be easier to swap out the entire filter/sensor/stabilizer assembly? I haven't seen the K5d schematics so this is just a shot in the dark, but I would expect it to come in one assembled unit.

Pat

That doesn't make any difference.

You can replace the SR unit which means another sensor. And the fab would recycle the SR unit later in production.

Or you attach another filter assembly which means you need a (rather) clean room and a cleaning procedure. But it really shouldn't matter. The labour-intense step is to dismantle the camera to get to the SR unit.

What I don't understand is why Pentax doesn't scrutinize entire assembled SR units in an automated fab test (before using it futher down the assembly line). The SR board has a plug which would make this straightforward. The tests would include a pinhole test (against embedded dust and AA imperfections), a hot pixel test after heat, a fixed pattern noise test and a dead pixel test, to enumerate a few. I'm really wondering how they can live without. The isolated SR board makes it straightforward because they don't have to discard an entire mother board.

[Class A]

I'm really wondering how they can live without.

Perhaps they were relying on Sony (or another manufacturer that adds further parts) to do the QC on the sensor?

[johnmflores]

It's amazing that so many forum members were quick to trivialize and/or discredit, often in a snide, mocking tone, those who first reported this. Folks weren't given the benefit of the doubt but instead were treated to chest-beating remarks akin to Live with it Crybaby or Don't be such a Wuss.

If I was part of this group, then I apologize. I sympathize with the folks afflicted with this problem - I'd be returning mine if it had the same problem.

But on the other end of the spectrum, the sheer volume of discussion around this issue (along with some hyperbolic comments by critics) is creating some fear, uncertainty, and doubt among past, present, and potential Pentaxians.

The fact is that the issue mostly affects photographs of featureless surfaces taken at very small apertures. Beyond that, there are hundreds if not thousands of great photos here and on Flickr and elsewhere on the web that show the potential of the camera. I daresay that many have probably been taken with the afflicted cameras, yet we are none the wiser because the above-mentioned shooting conditions didn't exist when the shot was taken.

The K-5 may very well be my next camera. If it is and there is a problem, I thank the forum for bringing this issue to my attention. I'll seek a replacement in due time, but the string-of-pearls won't stop me from taking photographs.

My $.02, non-refundable.

[Tom S.]

That doesn't make any difference.

You can replace the SR unit which means another sensor. And the fab would recycle the SR unit later in production.

Or you attach another filter assembly which means you need a (rather) clean room and a cleaning procedure. But it really shouldn't matter. The labour-intense step is to dismantle the camera to get to the SR unit.

What I don't understand is why Pentax doesn't scrutinize entire assembled SR units in an automated fab test (before using it futher down the assembly line). The SR board has a plug which would make this straightforward. The tests would include a pinhole test (against embedded dust and AA imperfections), a hot pixel test after heat, a fixed pattern noise test and a dead pixel test, to enumerate a few. I'm really wondering how they can live without. The isolated SR board makes it straightforward because they don't have to discard an entire mother board.

That is not the Japanese manufacturing method (which BTW came from an American named Deming). The philosophy is to make the product flawless so you don't need to inspect the final result. If you need inspections, your manufacturing techniques are flawed.

It sound simple, but obviously, Sony has some issues to correct.

[eddie1960]

not the first time sony has had issues to correct many products down the years i was in CE retail
it's really a mass production issue, but this defect rate is off the scale, and i'd like to know why it hasn't affected the d7000 the same way as the pentax. it is the same sensor

[JayBee]

not the first time sony has had issues to correct many products down the years i was in CE retail
it's really a mass production issue, but this defect rate is off the scale, and i'd like to know why it hasn't affected the d7000 the same way as the pentax. it is the same sensor

The last time I had any kind of significant warranty issue was with a SONY projection TV in 2003. The then-famous CRT "flickering" issue. I don't think our Pentax forum members have yet come close to using the amount of bandwidth that problem generated.

Sony did replace the CRT's for me. Twice.

A snip from a post back then:
I will unfortunately never trust Sony again. I believe that they are trying to cover-up this problem. Additionally, their customer service team appears to not have any real concern for the customer and is kept in the dark by corporate Sony. A responsible company would have recalled they affected models.

Deja vu.

[eddie1960]

the first generation samsung DLP were horrendous with rainbow effect, and it wasn't repairable either you could see it or you couldn't (it was an optical effect some people could see) only solution was here's a full refund can I show you something in a non dlp projector

[dgaies]

the first generation samsung DLP were horrendous with rainbow effect, and it wasn't repairable either you could see it or you couldn't (it was an optical effect some people could see) only solution was here's a full refund can I show you something in a non dlp projector

True, but that wasn't as much a QA/QC issue as much the fact that a certain percentage of people happen to be able to see the rainbow effect on slower color wheels. Now, whether they (Samsung) knew that ahead of time or not is another story

[falconeye]

That is not the Japanese manufacturing method (which BTW came from an American named Deming). The philosophy is to make the product flawless so you don't need to inspect the final result. If you need inspections, your manufacturing techniques are flawed.

Interesting aspect.

But I think it is more simple. After every so many manufacturing steps, the probablility of a fault has multiplied enough to be significant. And if you don't inspect at that stage, you risk to produce faulty end products.

The most cost effective was is to inspect when probablilities have risen enough.

And there is no way to get rid of the inspections. Just add enough manufacturing steps ...

E.g., in chip production, yield management is at the core of the process. I.e., you optically search for faults on the wafer before packaging into chips.

I cannot imagine that there is a Japanese and American way to it. The money counts the same both sides of the Pacific

[NaClH2O]

Interesting aspect.

But I think it is more simple. After every so many manufacturing steps, the probablility of a fault has multiplied enough to be significant. And if you don't inspect at that stage, you risk to produce faulty end products.

The most cost effective was is to inspect when probablilities have risen enough.

And there is no way to get rid of the inspections. Just add enough manufacturing steps ...

E.g., in chip production, yield management is at the core of the process. I.e., you optically search for faults on the wafer before packaging into chips.

I cannot imagine that there is a Japanese and American way to it. The money counts the same both sides of the Pacific

I am not familiar with Japanese mfg methods for electronics, but I am well aware of general Japanese mfg, specifically passenger train cars. If Pentax's methods are similar (and I'm not saying they are, there is a world of difference between a subway train and a DSLR) then vendor subassemblies are basically inspected for proper function and installed. The theory is that more rigorous testing has been done by the vendor. Trusted vendors (and I would venture to say that before this problem, Sony was most likely a trusted vendor) are generally given a little more leeway than someone new. I wouldn't be surprised if all Pentax did was make sure the sensors powered up and worked as intended. Exaustive tests can be incredibly expensive and up to now Sony has provided flawless sensors to Pentax. I am wondering if the AA filter is supplied to Sony thru another vendor? It's also possible that assembling the sensor assembly is done by a third party too. Thing is we simply don't know and it is unlikely that we will ever know. I do believe that the problem will quickly be satisfactorly resolved tho.

NaCl(but I refuse to define "quickly" )H2O

[glanglois]

Those who enjoy stopping at auto wrecks may want to take a few moments and ask Boeing about trust in the vendor chain and the schedule of the 787.

In the meantime .... Shewhart (Deming's mentor at Bell Labs) was a statistician, as was Deming although his academic training was focused on physics, I think. Statistics require data which arise from testing (inspection). The secret, of course, is to do enough inspection early to feel comfortable doing less later. It's just another spectrum.

On one end was Rolls Royce, who according to the story, achieved its enviable reputation for quality by employing, on the average, one inspector per assembly worker. Volkswagen reportedly did not find this approach to be applicable to the Beetle and looked elsewhere for quality guidance.

Oddly enough, an early publication of Shewhart's was "A study of the accelerated motion of small drops through a viscous medium."

Perhaps we call these Shewhart drops? (He says with tongue in cheek .... )

[falconeye]

I am not familiar with Japanese mfg methods for electronics, but I am well aware of general Japanese mfg, specifically passenger train cars. If Pentax's methods are similar (and I'm not saying they are, there is a world of difference between a subway train and a DSLR) then vendor subassemblies are basically inspected for proper function and installed. The theory is that more rigorous testing has been done by the vendor. Trusted vendors (and I would venture to say that before this problem, Sony was most likely a trusted vendor) are generally given a little more leeway than someone new. I wouldn't be surprised if all Pentax did was make sure the sensors powered up and worked as intended. Exaustive tests can be incredibly expensive and up to now Sony has provided flawless sensors to Pentax. I am wondering if the AA filter is supplied to Sony thru another vendor? It's also possible that assembling the sensor assembly is done by a third party too. Thing is we simply don't know and it is unlikely that we will ever know. I do believe that the problem will quickly be satisfactorly resolved tho.

NaCl(but I refuse to define "quickly" )H2O

I must speak to experts, glanglois included. I am no expert and wouldn't challenge what you say

Actually though, we don't disagree at all.

There is only one assumption though that may not hold true: That the stains are created by Sony.

I actually assume Pentax creates them when glueing the AA/IR-filter assembly onto the populated sensor PCB. They would have to clean sensor cover glass and AA bottom glass first. And it's probably there where it happens. An imperfect wet cleaning. The stains would be compatible with assuming a contaminated wiper to dry and a two motion wiping, ie. a worker who wipes in two strikes. My best guess with maybe 5% chance to hold true

Anyway, there is no reason to assume that Sony is to blame. Because the sensor chip must be cleaned after ball grid soldiering anyway, that would only be the case if the particles were on the inner side of the cover glass which is the less likely place.

[NaClH2O]

I must speak to experts, glanglois included. I am no expert and wouldn't challenge what you say

Actually though, we don't disagree at all.

There is only one assumption though that may not hold true: That the stains are created by Sony.

I actually assume Pentax creates them when glueing the AA/IR-filter assembly onto the populated sensor PCB. They would have to clean sensor cover glass and AA bottom glass first. And it's probably there where it happens. An imperfect wet cleaning. The stains would be compatible with assuming a contaminated wiper to dry and a two motion wiping, ie. a worker who wipes in two strikes. My best guess with maybe 5% chance to hold true

Anyway, there is no reason to assume that Sony is to blame. Because the sensor chip must be cleaned after ball grid soldiering anyway, that would only be the case if the particles were on the inner side of the cover glass which is the less likely place.

Ah, OK. I was under the assumption that Sony provided the sensor assembly including the AA/IR filters. I agree with your analysis that the stains are most likely on the underside of the AA filter. My mistake was that the whole assembly was provided to Pentax, who then installed it in the camera body. I wasn't aware that Pentax installed the AA filter subassembly onto the sensor subassembly.

NaCl(that puts a different light on things)H2O