DXOmark sensor Ratings..Should I take them with a pinch of salt ?

I think ergonomics, handling menu systems and other intangibles are more important than some slight differences in sensor measurements.

 

From what I have been able to gather, the DXO "scores" are based on somewhat subjective interpretations on the sensor data. Also, they only look at the sensor, which is only part of the overall system that drives IQ.

I can't find the thread now, of course, but Amin had an excellent analysis of the DXO results for the new Nikon.

Also, another interesting recent thread on IQ here:

https://www.mu-43.com/f92/what-image-quality-17052/

 

The DXOmark scores are good for engineers to compare sensor performance in different cameras and it gives you a "ballpark" measurement for comparison ... it's the "ballpark" to keep in mind.

If one camera scores a 51 and another scores 58 and another scores 59 their sensors deliver roughly the same image quality in terms of the signal-to-noise ratio, color accuracy, etc.. If one camera scores an 89 and another camera scores 50 then that is something you might want to pay attention to when considering sensor performance.

 

No not at all, lol.

This is nothing new to me... I've been seeing Dx0Mark give much lower rankings to cameras which are taking much higher quality images than other cameras which Dx0Mark ranks a lot lower for a long, long time now. This great discrepancy and irrelevancy of what Dx0Mark measures to the production quality of an image has always been there... but it's just never been so apparent now as with the comparison of the Nikon 1 and Micro Four-Thirds.

 

If you find the thread on the Nikon DXO results, specifically, it will show that NR impacted their results. And, again, this is just the sensor... does not take into account lens, IBIS, interface, handling, etc. etc.

 

I usually take them with a margarita...with a pinch of salt.

Seriously, I consider them informative but not definitive. A 10% difference between the scores of two cameras does not mean that one is 10% better, but rather there are certain conditions (i.e., low light/high ISO) where that 10% difference starts to show. All the other times the performance of the cameras will be much closer than the scores indicate.

 

I don't have my own E-P3 to compare with, but I've poured over images taken with both E-PL2 and E-P3 and I have to say that they are actually very similar but there is a very slight edge to the E-P3. The E-PL2 holds its own, but the E-P3 would not be a downgrade in any way, no matter what Dx0Mark says.  

 

DxOMark cannot even spell "lens"

Go to the DxOMark tab marked lenses. The first bold heading reads:

Latest camera lense reviews​

I emailed them to let them know they were embarrassing themselves, but nothing has changed.

 

DxOMark uses a scientific method to rate cameras against each other. It can measure various aspects that make up what we call "image quality" and deliver a score based on that. What it can't do is measure the "look" a particular camera may give. It can't explain why I might prefer the look of a lower-ranked camera over a higher-ranked camera. It can't guarantee a higher-ranked camera will be the best camera to use all of the time for all applications.

My personal opinion is that DxOMark should be used to compare sensor performance, NOT image quality.

 

I don't know if this is a recent phenomenon or not, but DxO's overall scores are a load of crap. I'm not sure how they get them, but they seem almost arbitrary and disregard a lot of important elements like resolution/sharpness. The tests themselves can be useful, but their approach to ranking sensors isn't worth the bytes it's saved on.

 

This is the big thing. The E-5, E-PL2, E-P3, E-PL3, and E-PM1 for instance can capture a lot of lens resolution and fine detail (which all equates to sharpness) at all ISO levels which the majority of competitors can't. I would think that would be one of the primary elements of high image quality from a camera system, no?   Yet the internet is so keen on downing these cameras for having lower "Dx0Mark scores" without even knowing what the heck that means to them.

 

I equate the DxO ratings to the "Tale of the Tape" in boxing. For me, just numbers.

One competitor has the advantage in weight and reach, the other, height and age, etc. Meaningful? Yes!

The younger, taller fighter always wins.

NOT!

See how ridiculous that sounded?

Just because a sensor's specifications rated higher does not necessarily mean the camera it is installed in will capture better images.

 

Good boxing analogy. I was going to use horse racing (which is big here in HK) as an analogy. Everyine would be a winner and Jockey Club would be broke if the actual result is based on measurements...

 

+1

I like a high resolving sensor. I like a sharp image; something that has a bit of "bite" to it. More often than not it suits the style of image I am trying to achieve. And yet, when I see cameras I know I like compared against the DxOMark kings that are so widely lauded, all I see is this slightly cleaner, slightly mushier output. That's why I do take a lot of the usual image quality conjecture with a grain of salt, and know that everytime an opinion is offerred on the matter it is just that: an opinion, specific to that person's tastes. Even DxOMark's scores are essentially just an opinion based on the parameters that their equipment is programmed to measure.

BTW...this post was just an opinion.