What are the chances of getting a good lens?

Canon Bob

Canon Bob

Loves the Enemy
Messages
9,754
Name
Bob
Edit My Images
Yes
Lensplay ( www.lensplay.com ) have been surveying buyers to see how reliable lenses are from the different manufacturers out there. Full details are here but this is the summary information.
(It's useful info so please don't turn it into a "chest beating" thread :nono:)
Bob

Canon lenses - 11923 with 888 defects
The probability of getting a good lens is 93 %
The probability of getting 5 good Canon lenses in a row is 68 %
Sigma lenses - 2516 with 563 defects
The probability of getting a good lens is 78 %
The probability of getting 5 good Sigma lenses in a row is 28 %
Tamron lenses - 1262 with 193 defects
The probability of getting a good lens is 85 %
The probability of getting 5 good Tamron lenses in a row is 44 %
Tokina lenses - 420 with 69 defects
The probability of getting a good lens is 84 %
The probability of getting 5 good Tokina lenses in a row is 41 %
Nikon lenses - 2164 with 171 defects
The probability of getting a good lens is 92 %
The probability of getting 5 good Nikon lenses in a row is 66 %
Pentax lenses - 616 with 67 defects
The probability of getting a good lens is 89 %
The probability of getting 5 good Pentax lenses in a row is 56 %
Minolta/Konica/Sony lenses - 578 with 60 defects
The probability of getting a good lens is 90 %
The probability of getting 5 good Minolta/Sony lenses in a row is 58 % Other lenses - 814 with 86 defects
The probability of getting a good lens is 89 %
The probability of getting 5 good Other lenses in a row is 57 %
 
Its quite interesting, it just shows that buying the Canon lens over Sigma and Tamron are worth it. 93% is very high for Canon.
 
Oh ****, All 6 of my current lineup of Canon lenses seem very good, and the 7 I have previously owned seemed equally as good, so that makes 13 in total, I have just bought another a 17L TSE, not had a chance to try it yet, but on those odds I must be due a lemon!
 
The data is pretty meaningless without an understanding of the caveats listed on the originating page.
 
Better sell most of mine then..........:D
 
That's interesting. And it feels about right.

I guess we can perhaps say that more people are likely to report faults than just post to say their lens is perfect, so maybe the true figures are actually a bit higher? In the last five years I've either bought or carefully tested about 20 lenses, mostly Canon but including a few Sigmas, and they have all been fine. And I haven't filled out that form!

And the sample size for several brands is really too low to draw any firm conclusions, but having said all that, it's hard to ignore the relatively poor showing from Sigma.
 
Very interesting.

I used to 'asume' people with expensive cameras and expensive lenses must know their stuff.

There seems to be a culture where the value of your opinion is measured by the value of your kit but I've figured out that this isn't always the case.

I'd like to know what percentage of folk can tell a bad lens from a good one.
 
Originally Posted by dinners View Post
I'd like to know what percentage of folk can tell a bad lens from a good one.

Indeed, and that may be compounded by people deliberately posting wrong info.
 
arclight said:
Originally Posted by dinners View Post
I'd like to know what percentage of folk can tell a bad lens from a good one.

Indeed, and that may be compounded by people deliberately posting wrong info.
Click to expand...

The surveys definition ...
"Defective means that there was something wrong with the lens that the manufacturer fixed,...."

This is pretty clear, if those surveyed stuck rigerously to this than I would say the data is good. Consumers are pretty good arbitrars of quality in this respect - it was not working, I sent it back, they fixed it, ... therefore it must have been defective.

But then they add ..
"... or something that wasn't wrong with the next lens you got if you sent it back to the vendor for exchange."

:thinking::thinking::thinking::thinking::thinking:

The surveyors try to give guidance that ...
" If you just didn't like the lens or you though it wasn't worth the price you paid for it, that doesn't count as a defect. "

Of course you can challenge the way individuals have interpretted these statements, but if there are enough returns the data would be robust, particularly if it is the relative differences between manufacturers you are interested in.

What I would like to know is how many returns did the surveyors receive.
 
I find these data very hard to take seriously. I've bought 183 new Canon lenses over the last 2 years or so, and I've only had two defectives: a 14mm f/2.8L with a misaligned element, and a 15mm f/2.8 fisheye with AF failure. If the failure data are correct, then the chances of such a good experience are only 0.0087% (1 in 11,464).

Normally in statistical analysis one rejects hypotheses that have only a 1 in 20 chance of being true by chance. For that level of probablility, the Canon defect rate would have to be more like 97% instead of the 92.5% in this survey. Or to put it another way, the survey would be over-reporting defective lenses by a factor of 3.
 
StewartR said:
... I've bought 183 new Canon lenses over the last 2 years or so....
Click to expand...

Thats a whole lotta lenses!!
:D

If you don't mind commenting could you tell us how many Sigma and Nikon lenses you've bought and hw many you've had to send back?
 
StewartR said:
....I've bought 183 new Canon lenses over the last 2 years or so, and I've only had two defectives......
Click to expand...

I suspect that your purchases are likely to skew the data somewhat as you're more likely to be buying higher up the range and not the 50/1.8, 55-250, 75-300 type offerings. It's a fair bet that QC or component quality is higher with pricier lenses.

Whilst the numbers may be based on the reports of people not all truly in a good position to have judged the quality, it is reasonable to assume that the relationship between the various marques would not be unduly affected by this.

The other possibilty is that 15 out of your next 17 purchases will be, as Ed puts it, "Lemons"

Bob
 
I have enjoyed photography since childhood ( i am now 53). I have learn't over the years that it i worth saving up to get the best quality products. My Nikon 24-70 afs is a classic example of this. I could have bought a sigma for a lot less but I will be keeping the lens for a long time so I feel spending the extra money to get a top quality product well worth it.
 
I reckon I had a dud Canon 75-300 USM Mk 3, it was very soft....its since been given to a cousin on permanent loan...if she likes it I get £135. :D
 
So what if you do happen to get a lemon? You send it back, you get it fixed or replaced. Ok, so it's a bit of a hassle, but you still end up with a working lens.

If you buy Canon rather than Sigma, you stand a slightly higher chance of getting a working lens first time. You also pay twice as much.
 
LizardMan said:
So what if you do happen to get a lemon? You send it back, you get it fixed or replaced. Ok, so it's a bit of a hassle, but you still end up with a working lens.

If you buy Canon rather than Sigma, you stand a slightly higher chance of getting a working lens first time. You also pay twice as much.
Click to expand...

Interesting point there.

People often say things like 'if you get a good copy it's 90% as good as the Canon' or 'As good except for the auto-focus, mind you depends if you get a good copy'. But in theory you will get that 90% of the Canon lens. Just not always 1st time.

Is it the manufacturers or suppliers responsibilty when it comes to replacing/fixing a bad copy?
 
StewartR said:
I've bought 183 new Canon lenses over the last 2 years or so, and I've only had two defectives
Click to expand...

Canon Bob said:
I suspect that your purchases are likely to skew the data somewhat as you're more likely to be buying higher up the range and not the 50/1.8, 55-250, 75-300 type offerings. It's a fair bet that QC or component quality is higher with pricier lenses.
Click to expand...

Good point, well made.

About 150 of those 183 are "L" series lenses, and if the QC is better on them than on consumer-grade lenses then that could account for my experience.
 
StewartR said:
Good point, well made.

About 150 of those 183 are "L" series lenses, and if the QC is better on them than on consumer-grade lenses then that could account for my experience.
Click to expand...

I think there is something in your earlier post Stewart, and also in Bob's comment. People are much more likely to report failures, and it stands to reason that an L lens will have maximum attention paid to every detail.

From the data, I think it's impossible to draw too many specific conclusions other than Sigma's quality control is not as good as most others, and you're perhaps twice as likely to get a dud Sigma as you are a Canon lens.

However, the likelyhood of getting a dud of either brand, or any other brand for that matter, is almost certainly far lower than stated.
 
HoppyUK said:
From the data, I think it's impossible to draw too many specific conclusions other than ....
Click to expand...
OK, I'm going to put my statistician's hat on now. (I have one, but I try not to wear it too often because it does tend to kill off interesting debates.)

If you don't know how the data were collected, you can't draw ANY conclusions. End of. It's as simple as that.
 
Personally, I'm more interested in how long the lens will last. I'm no expert and probably couldn't tell if there was any difference in optical quality, so if I pay twice as much for a lens, will it last twice as long?
 
StewartR said:
OK, I'm going to put my statistician's hat on now. (I have one, but I try not to wear it too often because it does tend to kill off interesting debates.)

If you don't know how the data were collected, you can't draw ANY conclusions. End of. It's as simple as that.
Click to expand...

Was it you that said something about "lies, damn lies, and statistics"? :LOL:

Gotta agree with you though and it's a very dangerous business. What I am assuming is that the absolute figures are meaningless but I think we can deduce that all the data was collected in a similar way, so at least the errors apply equally to those brands with over 1,000 examples - statistically a fairly robust sample, I would say.

And from that, I'm seeing Sigma showing significantly lower than Canon. Then taking an entirely subjective view, I cannot belive that their total overall failure rate is anything like that high (they'd be out of business if it was). I'd be surprised if it was actually much higher than three or four percent.
 
You must log in or register to reply here.
Back
Top