Customer Discussions > Photography forum

Explaining the problem with DXOMark


Sort: Oldest first | Newest first
Showing 1-25 of 98 posts in this discussion
Initial post: Mar 31, 2012 7:03:37 AM PDT
Last edited by the author on Mar 31, 2012 7:35:44 AM PDT
Tom Martin says:
OK, so I've been trying to figure out a way to describe the problem with DXOMark that even a Nikonian can understand. So I decided to try a car anology.

We take two guys, Soichiro Canon and Carol Nikon, with nearly identical old sports cars. They both have 1 liter engines and they take their cars to two testing services to see which one is better. Now the first testing service Displacement eXamination Organization, measure the displacement of their cars and gives them both a DXO Mark of 1. The second testing service Horse Power Review (hpReview) gives them both a score of 60. So they take their cars out to the track, and they perform almost identically. Now Soichiro Canon's and Carol Nikon's friends all go of course they performed the same look both DXOMark and hpReview gave them the same rating.

So, to Soichiro Canon and Carol Nikon, take their cars and put larger 1.5 liter engines in them, again they both get the same 1.5 DXOMark, and hpReview both rates them at 75, and again on the track they are the same. And the friends all point to the DXOMark and hpReview scores in their blogs and say of course they perform the same.

So, to Soichiro Canon and Carol Nikon, take their cars and put larger 2 liter engines in them, again they both get the same 2 DXOMark, and hpReview both rates them at 100, and again on the track they are the same. And the friends all point to the DXOMark and hpReview scores in their blogs and say of course they perform the same.

So, this time Soichiro Canon leaves the same 2 liter engine in the car, but, optimizes the intake manifold, adds fuel injection, balances the engine, increases the redline to 9000 rpms, and adds a free flowing exhaust. Carol Nikon puts in a 4 liter engine. Now they take their cars in for testing. Canon's car gets the same 2 DXOMark and Carol Nikon gets a 4 DXOMark. hpReview gives Soichiro Canon's car a score of 237 and Carol Nikon's car a score of 210. They take their cars out to the track and Soichiro Canon's car actually performs a little better. But, Carol Nikon's friends all exclaim in their blogs how much better Carol Nikon's car is because it got a DXOMark of 4, but, Soichiro Canon's car only got a DXOMark of 2, and everybody wants a car just like Carol Nikon's. Soichiro Canon's friends blog about how the Soichiro Canon's car was actually faster, but, generally are dismissed because the DXOMark had predicted performance so well in the past, and the hpReview is much more subjective, they also are reminded that Carol Nikon's car has way more torque.

So again Soichiro Canon and Carol Nikon take their cars back and work on them. Soichiro Canon increases the stroke of his existing 2 liter engine, so now it is a 2.2 liter engine, and Soichiro Canon adds a supercharger, an even better intake, and exhaust. Carol Nikon is annoyed that Soichiro Canon's car was better and puts in a 7 liter engine. They again take their cars for testing. Soichiro Canon's car gets a DXOMark of 2.2, and an hpReview of 500. Carol Nikon's car gets a DXOMark of 7 and an hpReview of 500. They take their cars out to the track, and Soichiro Canon's car being lighter wins again. Now, Carol Nikon's friends again blog about how much better Carol Nikon's car is because it got a DXOMark of 7, and Soichiro Canon's car only got a 2.2. Soichiro Canon's friends blog that DXOMark isn't accurate. Carol Nikon's friends retort what are you saying DXOMark is in bed with Carol Nikon? You felt that DXOMark was accurate before. Soichiro Canon's friends point out that Soichiro Canon's car actually won in a real world race. But, again Carol Nikon's friends point to the DXOMark of 7 that Carol Nikon's car got, and the 2.2 DXOMark that Soichiro Canon's car got, and proclaim that Carol Nikon's car is better. Soichiro Canon's friends just shake their head and know that Carol Nikon's friends just aren't going to get it.

Appologies to:
Soichiro Honda and Carol Shelby.

And I'll end this with a quote from Soichiro Honda
`The value of life can be measured by how many times your soul has been deeply stirred.'

Posted on Mar 31, 2012 9:23:03 AM PDT
T. Campbell says:
I can get a fuel-injector and super-charger for my Canon? Why didn't anybody tell me this before? ;-)

Thanks Tom!

Posted on Mar 31, 2012 9:34:05 AM PDT
Last edited by the author on Mar 31, 2012 10:07:48 AM PDT
Gatorowl says:
I want the Canon only if it gives me: gangsta whitewalls, TV antennas in the back, diamond in the back, sunroof top, so that I am diggin' the scene with a gangsta lean!

Apologies and credit to
William DeVaughn

Posted on Mar 31, 2012 11:07:38 AM PDT
k. sandmann says:
Lol
Put some thought into that 1!
:-)
Last night I was thinking along a similar line about the megapixel war in general. Say if a cars top speed is your pixel number. Some people have to buy the car just cos it's faster.

The megapixel right now is like way over 180 miles an hour (or 18 megapixels).
Most people don't go much over 80 or require more than 8 megapixels.

Well what about everything else? Even a professional race car driver looks for much more than just top speed.
If you wanna drag race I guess high speed ( pixels ) is better.

So I drive jeeps..... hmm? My liberty should do 120. I have had my CJ7 up to about 95 miles per hour but that felt all to much like a death wish. Any weak point in the steering & suspension becomes very apparent.... that's why you wanna replace all the old original rubber with polyurethane so you don't crash and burn. Nice wide tires & a lift kit don't help matters though.

Posted on Mar 31, 2012 11:52:43 AM PDT
Gatorowl says:
I'm not sure that I fully comprehend Tom's allegorical yarn, but surely he wasn't alluding to the megapixel wars in criticizing DxoMarks. DxoMarks uses 8MP normalized images for their testing. I think that this means that they down-sample all images to 8MP before running their tests. Thus, increasing MP without a concomitant increase in other measurable sensor attributes should leave DxoMark scores unchanged.

So, I'm still trying to understand what Carol Nikon keeps increasing that produces higher and higher DxoMark scores.

Perhaps, it's dynamic range. :)

In reply to an earlier post on Mar 31, 2012 12:30:06 PM PDT
Last edited by the author on Mar 31, 2012 1:26:16 PM PDT
Tom Martin says:
Gatorowl says: . . . So, I'm still trying to understand what Carol Nikon keeps increasing that produces higher and higher DxoMark scores . . .

He keeps increasing the displacement of the engine, which gives him a higher Displacement eXamination Organization mark.

The point being that while the displacement of an engine can be an indication of the performance of the car, it may not even directly relate to the horsepower output of the engine, and certainly in a complex system like a car, it has only an indirect correlation to the overall performance of the car.

This is the same issue with dSLRs and trying to measure one item an image sensor. DXOMark measures factors of an image sensor that is very comparable to measuring displacement of an engine. And just as displacement does not indicate either the horsepower output of an engine, and certainly not the overall performance of a car. DXOMark scores seem to measure factors that may not fully indicate the actual performance of the sensor and certainly does not directly relate to the performance of the complex system that a dSLR is.

Posted on Mar 31, 2012 4:56:54 PM PDT
Last edited by the author on Mar 31, 2012 7:08:27 PM PDT
Gatorowl says:
So what is it that DXOMarks measures that corresponds to displacement?

Is it sound-to-noise ratio?
Is it dynamic range?
tonal range or color response?
or is it color sensitivity?

Somehow, I think that these attributes are related to IQ. I mean, that I thought that all-else equal, a camera with better SNR for a given ISO value produces a cleaner, lower noise image. The same for the other attributes that the DxoMark measures.

But you did mention "performs better on the track." Did you mean runs faster, corners better, and perhaps runs the slalom course quicker? If so, then perhaps I may understand your analogy. Correct me if I'm wrong, but you're saying that the 5DM3 is like a souped up race or sports car. It will get you there quicker. The ride may be rough and bumpy but it blows away the Lexis that was designed to make the car disappear--super quiet, seat warmers, and a highly refined drive that smooths out bumpy roads--while you drive.

Therefore, what you are saying is the Canon is that souped-up, hot-rodded Shelby, compared to the slower, but sure-footed luxury Lexus?

Bringing it back to photography--I don't know cars, I'm the guy in the boring Prius--, you're saying that the Canon is quicker and more agile than the Nikon. It probably performs a bit better at high-ISO (but the jury is still out on this one) and produces a pretty nice picture to boot. By contrast, the Nikon gives you extra resolution that you may not need, it is definitely slower (but quick enough unless you're shooting sports), and the enhanced DR, color sensitivity, and tonal range don't matter to your photography. Is that a fair translation of your analogy?

If so, I think I understand. Some people drive souped-up sports cars. Speed and agility are everything. Others, prefer luxury and refinement. The luxury car may not be the quickest on a track, and it may not corner like it's glued to the track, but the comfort, the noise-level, and materials make the driving experience an absolute joy. Of course, the guy who loves hearing the roar of a finely-tuned engine running at full throttle would probably have a hard time understanding why anyone would bother, much less, pay for the refinements that go into enhancing the riding experience of a luxury automobile.

So, some photographers go for the Shelby sports car of photography, now the Canon, whereas others go for the Lexus luxury car of photography, the Nikon. Price is not the deciding factor. In fact, the sports car may be more expensive than the luxury car. They're just targeted at different markets. Different traits, different attributes to address the needs/desires of different market segments.

Thus, maybe your allegory isn't so bad after all.

In reply to an earlier post on Mar 31, 2012 5:26:00 PM PDT
DxO, from what I've understood, rips out everything but the sensor itself. They then do some measurement of a rectangular block of pixels extracted from the sensor data.

This is equivalent to the simile of measuring only the engine displacement after ripping off the transmission, water pump, alternator, air intake manifold, exhaust manifold, carburetor/fuel injectors... (Displacement being the sum of the air volume in the cylinders when the piston is down, minus the sum of the air volume in the cylinders when the piston is up). It ignores everything that contributes to the power/torque that actually reaches the road, using a static measurement in place.

But a camera image is what you get when running through the DSP chip in the camera -- The DSP chip is all those things that were removed from the engine. For example, the Canon 7D has dual DIGIC processors -- consider that a fuel injected turbocharger with tuned intake/exhaust manifolds and double overhead cams compared to a normally aspirated engine of the same size sucking air through a carburetor using old-fashioned push rods to activate the valves.

The DSP chip is responsible for taking the data from the sensor (which has a Bayer matrix filter over it) and converting it to RGB pixels, applying demosaicing/moire-reduction, sharpening, etc.

Posted on Mar 31, 2012 6:42:17 PM PDT
JCUKNZ says:
What on earth does this have to do with photography?

In reply to an earlier post on Mar 31, 2012 8:11:57 PM PDT
Last edited by the author on Apr 1, 2012 9:29:32 AM PDT
Gatorowl says:
"What on earth does this have to do with photography?"

Lol, I think that you hit the nail squarely on its head. This thread falls into the category of "trying too hard."

Dennis' response sounds really cool. The only problem is that it is totally wrong. the folks behindd DxoMarks, despite claims to the contrary, do not dissect or disembowel cameras. They actually conduct their tests on a raw file taken with an actual camera and lens. You may have noticed that DxoMarks doesn't review any camera that doesn't produce raw files. So, no, DxoMarks do not rip anything from cameras. It is in fact one of my criticisms of them. They would like readers to believe that they completely isolate sensor performance from the rest of the camera, but they don't. However, given the similar measurements between cameras from different companies that share the same sensor, I think that the marks probably come as close as possible to isolating sensor performance without disassembling cameras.

Frankly, the posters who insist on continuing this tortured camera-automobile analogy are barking up the wrong tree. There is a lot about DxoMarks to criticize, e.g., the one-dimensional final score is incredibly suspect and probably harms their credibility more than it helps it. That number is undoubtedly a marketing device designed to keep fanboys endlessly arguing back and forth on forums like this one.

Forget the final scores. Look at the measurements behind those raw scores. Cameras with high SNR scores do have better noise profiles than cameras with lower scores, and cameras that score well in DR are more pliable in post than cameras with lower scores. However, It's hard to say how much difference in each category is necessary to make visibly perceptible differences.

Therefore, in sum, both new FF cameras in the under $4000 category produce superb images. I suspect that the DxoMark score for the 5DM3 will be much better than the previous generation's score. However, I have been far more impressed with real-world samples that are finding their way onto the web from the D800 than from similar samples from the 5DM3. And it's not just resolution.

These samples support my belief that there is some correlation between DxoMarks and IQ. However, if you don't trust the marks, then ignore them. Instead of spending your time developing awkward metaphors that don't quite work, go out and examine actual samples from these cameras that were taken in the field. If you find that one camera consistently produces better results than the other, then you don't need to prove or disprove the validity of DxoMarks.

In the meantime, I will purchase the best equipment that I can afford (or justify) and try to improve as a photographer.

Posted on Mar 31, 2012 9:12:34 PM PDT
Last edited by the author on Mar 31, 2012 9:31:44 PM PDT
Neo Lee says:
The bottom line is, we don't know how DxOMark analyze and compare the data. Their test methodology is anyone's guess. They can try to explain all they want on their website, and if what they explain cannot be reproduced independently by third parties, the test is not scientifically valid.

I have my doubts. They are testing the sensors, but do all the cameras use the same lens attached to the bodies with special lens adapters? Because if Nikon's or Canon's uses their own respective lenses, the results will be skewed by the lenses too. In signal-to-noise ratio test for instance, if a camera cooks their RAW to average out the Gaussian noises at the cost of image details, how does the software tell if the blurrier image is the result of in-camera RAW cooking or the result of a bad lens?

Different lenses have different characteristics: MTF curve from center to corner and optical aberrations. As long as the camera bodies use different lenses in their tests, the sensor test results are invalid.

Another concern is their resampling method. The sensors don't usually come with equal numbers of pixels, and to compare the images from these different sensors, the data will need to go thru a resampling process to produce an image at a fixed reference resolution. To what degree does this resampling process skew the data? For example, if a sensor produces an 100MP image, will the resampling process give it an advantage or a disadvantage over another image produced in a 10MP sensor?

P.S. What if on an autos forum, two sides of car fanboys want to compare their cars, and they can't explain it in their own car terms so they use DSLR sensors as analogies. Will that work? Does this car-camera analogy makes it easier to understand or is it only complicating?

Posted on Apr 1, 2012 12:43:13 AM PDT
That's why we have DxOMark scores on the one hand and reviews from places like dpreview on the other. The sensor scores at DxOMark may only tell one part of the story, but it's still a significant testing metric that gives us a lot of information about the performance of the sensor itself. I don't really care if someone takes that number alone as a reason to blog that their camera is "better"; I don't read those blogs anyway because that's a game for amateurs and tech-geeks, not for real photographers. The most important part of the camera is the person behind it. But I really enjoy following the evolution of camera sensors at the DxO site and I hope they continue to do the testing.

I agree with Gatorowl who said "this (post) falls into the category of trying to hard.' "

Posted on Apr 1, 2012 3:10:48 AM PDT
Tom Martin says:
This analogy was meant to show how testing something that seems to relate to performance can as technology changes become meaningless. And further how it can appear to favor one persons (companies) way of doing things over another's and that doesn't mean they're in bed with them.

There are many misstatements in the dissenting posts above, which I will respond to later.

One that I will respond to now is that almost all automotive forums do have a photography thread.

For now I'm off to Autocross.

In reply to an earlier post on Apr 1, 2012 5:12:51 AM PDT
Last edited by the author on Apr 1, 2012 2:05:27 PM PDT
Yeah Tom I understood all that from the initial post. I appreciate the insight but I just don't buy that sensor ratings are completely irrelevant. The low light, dynamic range, and color depth scores correlate well with my experience having owned 5 DLSRs.

To use your analogy, I might consider buying 3 different sportscars and base my choice on everything from the body style, the engine size and horsepower, the dashboard, and the sunroof to the heated seats and the cupholder. And just because I know that one of the cars won a real world race with a 2.2 liter engine doesn't mean that's the car I'll choose--I'm not competing in races and I live in a cold climate, so maybe I prefer the one with the bigger engine and the heated seats.

After all, it's the driver behind the wheel that won the race, just as it's the photographer behind the camera that makes the art.

Posted on Apr 1, 2012 9:25:55 AM PDT
k. sandmann says:
Well from the start I took this thread to be slightly whimsical.
:-)

There are a lot of good comparisons to be made. As both are complex machines that are operated m guided by humans.

Getting back to DXO-
I always try to read between the lines. It would be wonderful if I could take something as gospel - but perhaps I was just born to cynical.
As far as I know they do testing with lenses on.

In reply to an earlier post on Apr 1, 2012 11:19:45 PM PDT
"....but I just don't buy that sensor ratings are completely irrelevant."

Conversely, it isn't entirely clear just how relevant the sensor rating is in terms of real-world performance. Even more so considering DxOMark uses their own RAW conversion software and most of us aren't.

Posted on Apr 2, 2012 8:56:39 AM PDT
EdM says:
I would suggest that everyone should check out one position by a pro at Luminous Landscape. Most in this thread are not taking a scientific look at this, because the results are not to their liking. Compare with the post from March 24, 2012:

http://www.luminous-landscape.com/whatsnew/

"Artistic License & D800 Rating"

"DxOMark ratings are controversial. Some people (myself included) have expressed concerns ..."

"Regardless of the weight that one gives to DxOMark test results and numerical analysis, they are rigorous and consistent. Therefore if read properly they can provide comparative insights into a camera's sensor performance."

This, by a person not without concern. If one views the big picture, DxO has no interest in making Canon fans mad; it would like to sell Canon fans [as well as all other raw shooters] their raw converter, which [raw conversion] is their business. It is clear that pretty much any DSLR sensor [in use in a camera of today] would wipe the socks off sensors from only a decade ago in most respects.

Qualitative differences do exist in sensor ability. Some sensors are better than others. Some scientists are better at innovation than others, or perhaps just lucky. If there were no gradual improvement, we'd be shooting with the sensors of 2002, or in my case, I'd still be shooting film. IMO, there would also not be the market penetration in every area of photography, including in cell phones, no less, that today exists.

One tactic that some people use when the results do not match desires, is to tear down the test used. Some people here appear to be doing just that. I would remind everyone that in any contest, there will be a ranking. That your entry is in the lower half or even last, does not mean that your DSLR is worthless. If anyone compared any of today's camera sensors to those from the first days of DxOMark, even the best from then, some perspective would be obtained, especially since many superlative images using those sensors of say 5-8 years ago [haven't looked up how far back DxO goes] have been sold and published in the finest photography magazines, etc.

DxOMark ratings are not meaningless, IMO, but they do not determine how good your or my images are: that's up to the person in the mirror. Perhaps more Canon people shoot JPEGs. For them, raw ratings mean less. OTOH, compare your images from today's DSLR bodies with those you got in the early to mid 2000s. If there is improvement, then I'd suggest that sensor ability has something to do with that. Even the least of us benefit by having better tools to bring forth our personal vision.

I suggest that competition among camera makers is and has been good for all of us, regardless of whether it is measured in MPx, Dmax, DxO sensor rating numbers, or something else. Rank order of sensor used, does not define how good a shooter's photos are. Ask your self - would you want today's camera to use the sensors available in 2004, e.g.? I'd suggest that any scientific test which measures properties of our technological gear is good as a driver for competition, towards better products in the future. If Canon or Canon users don't like the results from DxOMark, why haven't they come up with a test that yields better, more accurate results?

Posted on Apr 2, 2012 11:41:07 AM PDT
G. LO says:
I am a Canon user and primarily shoot raw. I would take the DXOMark results with a grain of salt. It may sound interesting and give us some interesting data points to look at. But I don't think it is an absolute reference point that describe IQ differences among various cameras.

In reply to an earlier post on Apr 2, 2012 3:19:53 PM PDT
EdM says:
There must be a Forrest Gump-ism that applies to this statement:

"But I don't think it is an absolute reference point that describe IQ differences among various cameras."

Neither DxOMark, nor anyone else knowledgeable that I know of says what YOU EVIDENTLY MISUNDERSTAND. A DSLR sensor rating is NOT the same as the DSLR's ultimate IQ, and the DxO people have always been very careful about that. Otherwise the IQ of a shot using Canon's pretty bad 75-300 zoom would be exactly the same as the IQ of a shot with the same DSLR using the best L type lens. What you said is totally off base as far as understanding DxOMark and what its numbers or rankings mean.

EVERYONE should agree with your quoted statement; that's what DxO itself says.

Consider: the overall DxOMark sensor rating of this year's G1 X is 60 (designed for high IQ and with a sensor almost as large as Canon's APSc sensor), last years T3 [EOS 1100D] is 62, and the Canon 1D Mk IV (introduced in 2009) is 74. Are you offering an opinion that the 1D IV is somehow worse than the T3 of last year, or the G1X of this year? [The new 1D X rating would be a further interesting data point, but it's not yet tested.] Somehow, I doubt that anyone would argue very much with the relative ranking achieved by DxOMark, for those _sensors_. Canon does not give away its best IQ sensors in its low end DSLRs, and that's a fact. [Neither does Nikon, for that matter.] I'm doubtful anyone would argue with the relative ranking in this by DxO.

The error that you make is similar to the error made by some others arguing here, on what the tests actually mean. There is another factor that many have not noted. There is a considerable correspondence of some of the measured abilities of sensors with the physical size of the sensor broadly, and the pixel pitch specifically. Canon, with its 1.6 APSc crop factor, has a built in disadvantage in these tests, compared to a hypothetical, same tech sensor with a 1.5 crop factor, due solely to the physical aspects of pixel pitch, having the same number of pixels but on a larger surface, for larger pixels.

However, Canon and Canon users correspondingly have an advantage over 1.5 crop factor sensors, because the sensor is the single most costly item in any DSLR. Thus, a smaller sensor provides a slight to significant price advantage for Canon DSLRs. So, as usual, there is always a tradeoff when choosing a camera, lens, etc. Further, same brand FF DSLRs essentially always have an IQ advantage over same brand APSc DSLRs. This is due to the larger sensors and their cost, whereby higher priced gear has more room (on a % of cost basis) for better parts than does the cheap stuff. This is plain economics 101.

So, argue what you will, but just remember that "life is like a bunch of chocolates ..."

In reply to an earlier post on Apr 2, 2012 4:17:01 PM PDT
Tom Martin says:
While DXOMark may appropriately provide an accurate relative ranking of sensors from a particular manufacturer, DXOMark does not appear to provide accurate relative performance ranking between sensors from different manufactures.

In reply to an earlier post on Apr 2, 2012 4:33:20 PM PDT
"Canon, with its 1.6 APSc crop factor, has a built in disadvantage in these tests, compared to a hypothetical, same tech sensor with a 1.5 crop factor, due solely to the physical aspects of pixel pitch, having the same number of pixels but on a larger surface, for larger pixels."

You're talking about a 12% difference in pixel density for the same MP count. That would be the same as comparing a 10 MP to a 11.2 MP camera with identically sized sensors. I would not expect such a difference in pixel density to cause any visible (not to humans) differences whatsoever in real-world results. If DxOMark really finds a 12% difference in pixel density to be significant then differences in scores aren't very meaningful in the field.

Posted on Apr 2, 2012 6:12:31 PM PDT
Gatorowl says:
In polite company never discuss religion, politics, or...DXOMarks!

In reply to an earlier post on Apr 2, 2012 7:52:57 PM PDT
G. LO says:
I hear you. I am not sure what people are discussing here. For me, I am not paying much attention to DxOMark anyway

Posted on Apr 2, 2012 8:45:37 PM PDT
Last edited by the author on Apr 2, 2012 8:49:09 PM PDT
Neo Lee says:
Do people shoot photos for discerning eyes, or for scientific algorithmic analysis? If for the former, you don't need DxO or whatever because you're shooting for the eyeballs, for people to see. If the difference doesn't matter to the eyeballs, it doesn't matter in most real-life applications. The only technical sensor test I read is the dynamic range test which is done simply by shooting at Stouffer step wedge backlit by a calibrated light source. Such a test is reproduced by many since it's not as secretive as DxO methodology, and you could fact check on multiple sources. Even then I won't base my purchase choice on the dynamic range alone.

The ultimate comparison test in my opinion is by selecting enough random people to vote on photos they see as 'better'. To avoid predetermined bias during the evaluation, the photos must not be labeled with the specific camera name so as to deter people from voting for their favorite brand. Comparison photos are framed identically under controlled, referenced lighting. The cameras must use the same lens, even if custom lens adapters must be used. When done, the tester will get to see cameras of their votes and other people's. If you're a fanboy for instance, you may be surprised when the results say you voted for the other guy's DSLR more. This is similar to the comparison tool on imaging-resource.com but they don't have voting out there, and I could see their photos are not consistent; they could have done better.

Posted on Apr 2, 2012 10:55:24 PM PDT
Even though "It's not the size of the boat, it's the motion of the ocean that counts", guys will always want to compare the size of their dicks. It's what we do.

Arguing about it is pointless.
‹ Previous 1 2 3 4 Next ›
[Add comment]
Add your own message to the discussion
To insert a product link use the format: [[ASIN:ASIN product-title]] (What's this?)
Prompts for sign-in
 


 

This discussion

Discussion in:  Photography forum
Participants:  15
Total posts:  98
Initial post:  Mar 31, 2012
Latest post:  Apr 23, 2012

New! Receive e-mail when new posts are made.
Tracked by 5 customers

Search Customer Discussions