S
Stanton
Guest
I had a Minolta Dimage A1 fail. There was recall on the defect. Dave
How many of your digital cameras have failed?
- Thread starter Tuolumne
- Start date
- Latest activity Latest activity:
- Replies: Replies 96
- Views: Views 6K
Not true, if by "many situations" you're trying to imply "most situations" or even "a sizable number of situations".
The Gaussian distribution fails in certain specific arenas (as any misapplied tool will), and LTCM's models of economics had a lot less to do with the Gaussian distribution than wishful thinking and imaginative conning; and their failure had a lot less to do with their mathematical models than running a smoke-and-mirrors pyramid scheme via leverage.
This is all beside the point. Taleb's theories are interesting but mostly not yet rigorously demonstrated; and his Black Swans have very little, if any, bearing on comparative failure rates of established consumer products.
The failure rate of desktop computers, according to Gartner, has fallen below 5% first-year, peaking at just over 10% at year four. I cannot, unfortunately, find at the moment comparable stats for film cameras, but the last I remember seeing in the late 90s (involving, like Gartner's computer stats, the sector as a whole, not just high- or low-end) were about the same or slightly better.
So, I am going to drop the assertion that computer failure rates are less. But the start here was that you were trying to make a point that computers fail all the time, and quickly (you asserted that they must be replaced every three years to avoid impending doom, which is simply a falsity), and therefor link that to digital cameras being less reliable than film cameras.
I'd like to see you retract the one total falsehood, and then make your argument, if there is one.
Roger Hicks
Mentor
'Sizable number of situations': yup, I'm comfortable with that. As Taleb points out, how many of the important events in your own life are Black Swans: the person you married, where you live, your job...?
'Mostly not yet rigorously demonstrated': well, that's his point, really. What definition of 'rigour' were you looking for?
'his Black Swans have very little, if any, bearing on comparative failure rates of established consumer products': this was cheerfully conceded in my last post, with the rider that I'd like to see your figures before conceding that computers are more reliable than cameras -- figures you admit you do not have.
'the start here was that you were trying to make a point that computers fail all the time, and quickly (you asserted that they must be replaced every three years to avoid impending doom, which is simply a falsity),'
What I actually said was, 'I don't trust a computer beyond three years, or five at the outside.' This is not the same as saying that they must be replaced every three years. I cheerfully accept that some may last a good deal longer. I'm just saying that I don't trust them to do so, and with your own figures of 5% in year 1 and 10% in year 4, I'd say this is a rational level of mistrust, given the possible consequences of failure. I'll even accept that the distribution of failures is almost certainly Gaussian. So?
What is the falsehood you would like me to withdraw?
And if you haven't seen the argument so far, where do you think I have failed?
Cheers,
Roger
Tuolumne
Mentor
I think you need to compare the failure rate of embedded systems to digital cameras, not general purpose computers. Embedded systems probably have a much lower failure rate than a general purpose computer. At least they don't have any rotating media which is usually the first thing to fail. For example, when was the last time a computer in your car, microwave or washing machine failed? I would guess it is not a very frequent event, as these systems tend to run trouble free for years.
/T
/T
Agreed, which is part of where I was heading vis a vis the silliness of Roger's "But computers break all the time!"
Roger Hicks
Mentor
"Silliness" . . . "Deliberate distortion" . . . come on, let's have an answer to the points I actually made, not to your straw men.
My original point was not drawing a parallel between computers and cameras. That was your interpretation, and Tuolumne's (I assume: he is now on my ignore list, and I am successfully resisting the temptation to see what new insults he has devised). I apologize for not making myself more clear.
My intended point was that computers (without which a digital camera is of very limited usefulness) are not very reliable. I cheerfully concede what I take to be Tuolomne's point about embedded computers, that they are indeed vastly more reliable than what one might call 'general purpose' computers.
I still await any rebuttal of what I have actually said, rather than ad hominem attacks, which I actually quite welcome: it shows there is no answer to the arguments, and that the attacker has no alternative but to go for the person, rather than the statement.
Cheers,
Roger
Either this is still silly, or you're having a hard time writing today, because you just claimed that you weren't trying to draw the parallel and then drew it again.
Yes, yes, yes. We've all heard this one. I called the argument silliness. Not an ad-hominem, as you well know. As for the distortion... well, I'll refer you to what I said in my previous paragraph, as it falls into the same category. Either you are communicating very poorly (a possibility you admitted to your credit above), or are engaged in glib distortions. I'll leave open the possibility that it's the former.
As for answering the points you made, your only solid assertion so far is about computers having to be replaced "constantly, on a very short time-base [implied after to be three years] before they break", which is, again, obviously not true. What is your point, otherwise? I can't figure it out, since the only sensible and obvious interpretation is apparently not what you were going for....?
Roger Hicks
Mentor
I was less trying to win anything than figure out what he thought his point was, but oh well.
Roger Hicks
Mentor
I was less trying to win anything than figure out what he thought his point was, but oh well.
Ah, my apologies. There were multiple points, and I was reasonably clear in my own mind about what they were, though apparently I failed to communicate this to you. Please allow me to attempt to clarify.
My first point was that computers are not outstandingly reliable, and that after 3 to 5 years one may reasonably expect a failure, possibly catastrophic. This is not the same as saying that it is certain; merely that one would be extremely unwise to take account of it, given the magnitude of possible consequences. Addendum: Of course, my definition of 'outstandingly reliable' may differ from yours -- though on your own figures, I'd stick by my assertion that they aren't.
With sufficient data storage redundancy, the effects of this failure can be minimised. I rather favour Drobo.
My second point (conceded in view of your arguments) was that 'statistical profiles' may indeed be adequate to predict equipment failures, with the rider that you failed to adduce the necessary statistics to prove that computers were more reliable than cameras: a point you in turn conceded.
This is a separate point from conflating the reliability of digital cameras and computers.
My third point is that for many kinds of 'broad assertions' -- quite possibly the majority of important ones, in real life -- Gaussian statistical analyses are of limited usefulness.
Which of the above do you find objectionable?
Once again, I must apologize for not distinguishing the three points more clearly, but at that, I'd say that dismissing any (or all) of them as merely 'silly' does not advance your own arguments a great deal.
If the above amounts to being arrogant and opinionated, I cheerfully put my hand up.
Cheers,
Roger
Last edited:
Dogman
Mentor
I've only owned one digital camera and I've only had it for less than 18 months--Canon 30D. It works fine. I hated it at first but now I like it well enough that I've ordered a second digital. This time a Canon G10.
Computers? In an ill-advised swap, I got a refurbished HP that crashed in about 9 months. I've only had three computers so I've had a 1/3 failure rate.
I've owned about 50 film cameras in the past 35 years. Of those I've had numerous mechanical and meter problems but only two cameras that simply were total crap and failed to function--a Nikon F3 and a Nikon FM. I'm not picking on Nikon. I've owned 8 or 9 Nikon SLRs over the years and these were the only ones that were unreliable.
Computers? In an ill-advised swap, I got a refurbished HP that crashed in about 9 months. I've only had three computers so I've had a 1/3 failure rate.
I've owned about 50 film cameras in the past 35 years. Of those I've had numerous mechanical and meter problems but only two cameras that simply were total crap and failed to function--a Nikon F3 and a Nikon FM. I'm not picking on Nikon. I've owned 8 or 9 Nikon SLRs over the years and these were the only ones that were unreliable.
Benjamin Marks
Mentor
Here's a data point from my personal experience: All of the computers I have owned, with the exception of my original 1985 IBM PC luggable (no HD, 256K of RAM and an orange, 9-inch built in screen) a Toshiba laptop I used in the late '90's and sold before it broke have failed catastrophically. Rough estimate: 7 machines. One of the five digital cameras I have owned failed (a 1997 Sony Mavica). Of all of the film cameras I have used, only three have partially or completely failed: an M6 and an R4, both as a result of electronics that suddenly went poof and an M2, whose shutter curtain had the temerity to rip after 50 years of service. I'm not sure what to make of all of this, but I will tell you that I keep three separate sets of HDs these dayes, each with a relatively complete set of back up data. Resume argument, please.
Ben Marks
Ben Marks
gavinlg
Mentor
If your computers are breaking - buy a mac. Even if at some point something goes wrong with it, you'll love it so much that you won't mind forking out to fix it.
Seriously.
Seriously.
Migracer
"MigRacer&amp ;qu ot; AKA Miguel
Time Before Failure. Obsolence Before Failure.
Time Before Failure. Obsolence Before Failure.
As en engineer we adopted the test criteria of Time Before Failure. I worked on both electrical and mechanical in the aircraft and automotive industries. All specifications for components are rated this way.
The embedded components do have a very low failure rate, heat is their enemy, so operational time is how they are rated in general they become obsolete before failure.
Time Before Failure for rotating masses (as pointed out by Climbing Vine) have a much shorter TBF. Components on critical mission (aircraft) track every component and require replacement with in a safe window before the stated TBF.
My first home computer was a Comodore 12 followed very shortly after by the Comodore 20. We got our first IBM PC in 1981 I just recently disassembled that one for recycling, it become obsolete after just two years, but we kept it working long past that point with hard drive upgrades, power supply replacements and several CPU and memory upgrades. Since then with eight children I was kept busy doing up grades and maintenance on the family and business computer fleet. The only failures on the Computers I have had have been hard drives, power supplies and CD/DVD drives and they all gave me ample warnings that I ignored. My home computers I do not consider critical and they have been for the most part the ones we removed from our shop where the mission is critical. ALL OF YOU IN THIS DISCUSSION ARE RIGHT :bang:
Time Before Failure. Obsolence Before Failure.
As en engineer we adopted the test criteria of Time Before Failure. I worked on both electrical and mechanical in the aircraft and automotive industries. All specifications for components are rated this way.
The embedded components do have a very low failure rate, heat is their enemy, so operational time is how they are rated in general they become obsolete before failure.
Time Before Failure for rotating masses (as pointed out by Climbing Vine) have a much shorter TBF. Components on critical mission (aircraft) track every component and require replacement with in a safe window before the stated TBF.
My first home computer was a Comodore 12 followed very shortly after by the Comodore 20. We got our first IBM PC in 1981 I just recently disassembled that one for recycling, it become obsolete after just two years, but we kept it working long past that point with hard drive upgrades, power supply replacements and several CPU and memory upgrades. Since then with eight children I was kept busy doing up grades and maintenance on the family and business computer fleet. The only failures on the Computers I have had have been hard drives, power supplies and CD/DVD drives and they all gave me ample warnings that I ignored. My home computers I do not consider critical and they have been for the most part the ones we removed from our shop where the mission is critical. ALL OF YOU IN THIS DISCUSSION ARE RIGHT :bang:
Roger Hicks
Mentor
As well as MTBF, I always remember something my father told me decades ago. In the event of failure, look at the electromechanical interface first: switchgear, plugs, soldered joints...
Cheers,
R.
Migracer
"MigRacer&amp ;qu ot; AKA Miguel
Many jump to conclusions
Many jump to conclusions
I have acquired over the years everything from cameras to aircraft at bargain price due to mis diagnosis, by either the owner or a tech, more often than not it was a simple interface failure not a catastrophic mechanical failure.
Many jump to conclusions
I have acquired over the years everything from cameras to aircraft at bargain price due to mis diagnosis, by either the owner or a tech, more often than not it was a simple interface failure not a catastrophic mechanical failure.
andydg
Established
Several have failed while in use. All were known issues with the particular camera.
Nikon D2h, metering problem fixed under warranty.
Nikon D2x, focusing problem fixed under warranty.
Olympus E1, sensor grunge. Olympus ended up replacing the sensor on this camera as it could no longer clean itself.
Olympus E3, went in for focusing adjustment. While in service they damaged my lens that was sent in along with the camera. Took them three return trips to correct the problems they caused. (This was for a 150mm f/2 lens)
Funny, but all my 'consumer grade' cameras, D70, E510, various point and shoots never have given me any problems.
Nikon D2h, metering problem fixed under warranty.
Nikon D2x, focusing problem fixed under warranty.
Olympus E1, sensor grunge. Olympus ended up replacing the sensor on this camera as it could no longer clean itself.
Olympus E3, went in for focusing adjustment. While in service they damaged my lens that was sent in along with the camera. Took them three return trips to correct the problems they caused. (This was for a 150mm f/2 lens)
Funny, but all my 'consumer grade' cameras, D70, E510, various point and shoots never have given me any problems.