DxO: 24MP Nikon D3200 equals 22MP 5D Mark III

[wickidwombat]

Nikon D3200 Hands-on Review	Small | Large

yay digital rev review of the 3200

[dilbert]

As it happens, DxO publish the full test results for sensors, so that if you want to see what the components are that make up a given score, you can. It's those other numbers that are actually of interest.

Ahhhh, but there's the rub.  Olympic judging is subjective in nature - judges often disagree, there are accusations of bias, etc.  DxOMark purports to be objective, and in fact, their measurements, which are supported by their published methodologies, are objective and useful. But can you point me to a detailed explanation of how the individual measurements are used to generate the DxOMark Sensor Score?  No, you cannot - DxO doesn't disclose that methodology. That makes the Sensor Score as subjective as an Olympic judge giving every performance by someone from their home country a perfect 10.

So just because the formula they use to calculate their sensor score is not disclosed it is therefore subjective?

That's a very very thin argument and you know it.

Maybe they consider that formula to be a trade secret and therefore do not wish to publish it?

But lets face it, even if they did publish it and it was found to be objective, all that would happen then is that everyone would criticise how they calculated it - why does X have a greater weight in the score than Y and so forth.

[briansquibb]

As it happens, DxO publish the full test results for sensors, so that if you want to see what the components are that make up a given score, you can. It's those other numbers that are actually of interest.

Ahhhh, but there's the rub.  Olympic judging is subjective in nature - judges often disagree, there are accusations of bias, etc.  DxOMark purports to be objective, and in fact, their measurements, which are supported by their published methodologies, are objective and useful. But can you point me to a detailed explanation of how the individual measurements are used to generate the DxOMark Sensor Score?  No, you cannot - DxO doesn't disclose that methodology. That makes the Sensor Score as subjective as an Olympic judge giving every performance by someone from their home country a perfect 10.

So just because the formula they use to calculate their sensor score is not disclosed it is therefore subjective?

That's a very very thin argument and you know it.

Maybe they consider that formula to be a trade secret and therefore do not wish to publish it?

But lets face it, even if they did publish it and it was found to be objective, all that would happen then is that everyone would criticise how they calculated it - why does X have a greater weight in the score than Y and so forth.

As we dont know how the numbers are produced then we cant decide how valid they are. Ergo the numbers are unreliable in a comparison.

For all we know there maybe a huge bias for mp and DR at iso 100, and nothing for DR at iso3200. 

[neuroanatomist]

So just because the formula they use to calculate their sensor score is not disclosed it is therefore subjective?

Exactly.  They state it's a 'weighted average' but give no explanation of the weighting.  Therefore, meaningless.

The EPA mandates a method for calculating fuel economy of cars sold in the USA - anyone who's looked at that sticker on the side window then actually checked the fuel economy empirically knows that the sticker value bears little resemblance to real world use.  But...it's an objective measure, and thus useful and menaingful for comparisons.  If car manufacturers, for example, could use different standards for a Hummer vs. a hybrid, it would be completely useless.

I have stated and do maintain that the DxOMark measurements are valid and useful.  DxO themselves state that one should first consider resolution, then consider use case, then look at scores.  Personally, I don't bother with scores - I go straight to the measurements (and not the 8 MP-normalized ones they use to generate the Scores).

[DZY]

Putting too much weight on DXO is a wrong strategy.

Until they favor Canon!

1. I have or use neither N nor C.
2. I was talking about buying a camera in general, no matter what brand.
3. even if DXO would favor C, it is still not smart to take DXO as a major key.
4. follow your attitude, at this moment when N has higher DXO, if any shooter who has a Nikon body but doesn't get good pictures, then the shooter should be the major weak point, not the machine. So blame yourself. Those who have Canon body, can blame the machine and look down the nikon user ... I just joke.

[psolberg]

Until they favor Canon!

yeah. that's the thing. It always happens. All these people with Stockholm syndrom gravitate to whatever site tells then what they want to hear and will try to discredit whatever site tells then what they don't want to hear.

When I decided to switch I did the following:
1) Purchase/Rent/Borrow.
2) shoot.
3) decide.

That's the only test I'll ever believe.

[NotABunny]

We score many things in our life.

We, individually, do NOT score things in the same way. Collectively, as in average, "we" might. Using a single statistical number, humanity eats enough food to survive, yet, individually, people still die of hunger.

If you look at sports. at the Olympics, there are many sports where judges have to watch a performance and score out of 10... Does anyone dispute the way in which judging is done? Not that I'm aware of.

I dispute them. Those score are designed by the Olympics committee or whoever. They aren't mine, nor do I care about them. In some cases, my and their criteria are similar (like crossing the white line); it just happens to be a convergence in some cases.

My type of photography requires specific types of indicators. They can't be represented by a single number, nor by several of DXO's tests. My type of photography requires that I see how the camera behaves in real life. That is done quite simply: show me photos of how the camera behaves in the conditions in which I shoot.

If you really want to encompass the entire camera's behavior in one thing, than simply show me ONE photo. Numbers are worthless to me. Show me the result of what a camera was designed to do: photos.

My brain is much better at interpreting a photo, than are the various DXO scores at representing what matters to me.

[FunkyCamera]

Too many megapixels, will suck like D800. Even Nokia beat Nikon on megapixels lol! So much to learn. Less is best like Canon.

[qurgin]

Re: nikon 3200 beats all other canon cameras dxo mark test!
Too many megapixels, will suck like D800. Even Nokia beat Nikon on megapixels lol! So much to learn. Less is best like Canon.

Yeah, I'm disappointed that the nikon d3200 is just 24 MP. why not 50? or 100? lol 

[Rienzphotoz]

DxO or whatever "xO" results mean nothing to me ... I buy a camera based on its overall performance and not on how many megapixels it has ... DxO can keep all their testing formula secrets to themselves and they can say whatever they want, I am more than happy with my 5D MK III ... no one on this planet is going to take me on a ride saying that D3200 equals or betters 5D MK III.

[tron]

and lest someone says that the above comparison was done at a very high ISO; here's one done at ISO 800 weel with what DxO rate sthe D3200 sensor capable of:

You forgot to add the D3200 camera! However the test seems more relevant (comparing FF cameras).
In addition, I am pleased to see the performance of my 5DMkII beating the rest

[new-z]

Hello,

I'm disagre, you can find a test just here :

http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/801%7C0/(brand)/Nikon/(appareil2)/795%7C0/(brand2)/Canon

new-z

[Rienzphotoz]

Hello,

I'm disagre, you can find a test just here :

http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/801%7C0/(brand)/Nikon/(appareil2)/795%7C0/(brand2)/Canon

new-z

and your point being?

[weixing]

Hi,
    IMHO, DxO is not testing the sensor, but the camera itself... since they use the RAW file output by the camera to perform the calculation. So basically you only can use it to compare camera of same brand (assume all camera of same brand use the same method to produce the RAW file).

   Anyway, it's quite well known that Nikon "RAW" file is not that RAW (had gone through some internal processing before output to the card and this process cannot be turn off) compare to Canon camera "RAW" file, so the comparison between Nikon and Canon camera is invalid.

   By the way, just wonder, did DxO use the same lens on all the camera during the test? I think different lens will produce different result especially testing involve colours.

   Just my $0.02.

   Have a nice day.

[elflord]

For portrait and landscape, the D3200 scores 24.1 and 13.2 respectively whlist the 5D Mark III scored 24.1 and 11.7. Of course the 5D Mark III whips the D3200 in low-light ISO (sports) but otherwise the scores are the same (81) for both cameras.

So it just goes to show that APS-C sensors aren't necessarily as useless, even with higher MP counts. Unless you need that low-light/higher ISO.

As DxO points out, 5 of the top 10 cameras on their website are now made by Nikon.

Got to wonder if Canon will be able to deliver a crop sensor that comes anywhere near that.

If you look at the charts, the Nikon wins on dynamic range at ISO100-200.

At ISO 400 and up, the 5DIII handily spanks the Nikon in every category (and handily beats the Nikon at every ISO in SNR and tonal range)

Dynamic range at low ISO is a well known weakness of Canon sensors, and it seems to be weighted quite heavily by the DxOMark.