[UncleVanya]

Thank you for making my point, Bxf.

They are very similar.

It's possible the 810 is just made from a bigger slice of the K-5 wafer.

Have a look at how close their noise performance is in on DXOMark - go to 'Measurements', 'SNR 18%', 'Screen'.

Nikon D810 vs Pentax K-3 vs Pentax K-5 IIs

Use the 'Screen' tab, not the 'Print' tab, since the latter only involves the software noise reduction when they downsample to their 8Mp reference. (This is simply due to the megapickle count. A K-3 benefits more from this than a FF A7S or Canon 1Dx.)

Sensor size does *not* affect noise performance, whatever you've read other people post in other threads, over many years, I bet.

Numbers and size of pixels, circuitry design and suppression of heat/stray currents do.

That's what I was saying as well. Thank you for clarifying things.

[bxf]

Have a look at how close their noise performance is in on DXOMark - go to 'Measurements', 'SNR 18%', 'Screen'

Now I'm lost.

Given the similarity of the D810:K-5 noise curves, why is the Sports (Low-Light ISO) score so much better? This is the one score that is consistently much higher for FF, and if I'm not mistaken, differences in tech generation come nowhere near closing the gap.

---------- Post added 31-12-15 at 06:48 ----------

It took me years to figure out the nonsense they were posting about "total light".

But the FF sensor is twice the size. An FF sensor collects twice as much light as an APS-c sensor at the same ƒ-stop, and shutter speed and ISO.

Doesn't the second quote essentially mean "total light"?

[Digitalis]

Blame my physics lecturers at university. Density is a general term for a volumetric or area rate that only involves mass in its specific definition within mechanics.

Sorry to jump on you like that.

But whenever I see people mentioning "density of light" it drives me up the wall, just like people who spell lens as "lense"

[clackers]

Now I'm lost.

Given the similarity of the D810:K-5 noise curves, why is the Sports (Low-Light ISO) score so much better?

Now I see your problem.

You're depending on those DXOMark scores.

The 'Sports' score is not just the ISO rating where 30dB is achieved (and which would be unambiguous).

Dxo have decided it's also tied to the Tonal and Dynamic range graphs.

Yet both of these have this Normalization to 8Mp crap!

Please look at the Screen rather than Print stats, and you will find the K-5 is also similar to the mighty 810 in dynamic and tonal ranges. This will be a rude shock to certain forum members.

Then the effects of post-processing are applied and put the 36Mp Nikon ahead.

This bit is a theoretical adjustment done solely on the number of pixels the sensor has, not whether it's FF or APS-C.

If you read all their definitions, you'll understand how they've done it, and how badly misunderstood they are by Full Frame Fanatics.

---

Last edited by clackers; 12-31-2015 at 04:57 AM.

[clackers]

Doesn't the second quote essentially mean "total light"?

Norm points out twice the light falls on a FF sensor, but Norm is also fully aware that light has to be distributed over twice the area, for no net gain.

Of course, that's a simplification, because less light reaches the FF corners and edges due to vignetting, and it's poorer too due to distortions and aberrations.

APS-C uses the very best 50% of all lenses - the centre.

[clackers]

it drives me up the wall, just like people who spell lens as "lense"

It's odd, that.

I've even seen forum members whose primary language is apparently English spell it like that, too.

Weird, because they won't have seen that in a book or website as they learned about this hobby of ours.

[bxf]

Sorry to jump on you like that.

But whenever I see people mentioning "density of light" it drives me up the wall, just like people who spell lens as "lense"

It's odd, that.

I've even seen forum members whose primary language is apparently English spell it like that, too.

Weird, because they won't have seen that in a book or website as they learned about this hobby of ours.

Maybe if several are "lenses" then one must be "lense"

[clackers]

Maybe if several are "lenses" then one must be "lense"

I can see that for non-native speakers, yes.

And centuries ago perhaps it was the standard spelling - there were lots of 'e's on the ends of words that perhaps once were pronounced.

But barring a typo, surely no school, newspaper, magazine or even YouTube video taught any of us to spell it like that.

[Digitalis]

(no) YouTube video taught any of us to spell it like that.

I wouldn't be so sure of that...

[BrianR]

Please look at the Screen rather than Print stats, and you will find the K-5 is also similar to the mighty 810 in dynamic and tonal ranges. This will be a rude shock to certain forum members.

In all honesty, who will this shock? I maintain this is one of the longest ongoing forum arguments over nothing. It will only be solved in the Breakdance Fighting Thunderdome (coming soon in 2016!).

Anyway, it's finally snowing here, so I'm off to go put my snow tyres to use.

[RobA_Oz]

Sorry to jump on you like that.

But whenever I see people mentioning "density of light" it drives me up the wall, just like people who spell lens as "lense"

In my last job, one of the admin types affixed a sign saying "Grammar Police" to my desk, so no need to apologise. Just to finish this side-track, electromagnetic flux density is a common term in physics and some branches of engineering, which of course encompasses light.

And, to round up the spelling bee, the possessive was originally also spelled with an added "es", later abbreviated by replacing the "e" with an apostrophe, which has given grocers and non-native English speakers confusion ever since.

Now, where were we?

Oh yes: Happy New Year, everyone!

[SRD]

Sensor size does *not* affect noise performance, whatever you've read other people post in other threads, over many years, I bet.

Numbers and size of pixels, circuitry design and suppression of heat/stray currents do.

That's only true if you measure noise pixel by pixel. Most people care about final quality of the picture, and that is usually viewed downscaled.

[clackers]

. Most people care about final quality of the picture, and that is usually viewed downscaled.

Downscaling is based only on the number of pixels. :-)

[Fogel70]

Please look at the Screen rather than Print stats, and you will find the K-5 is also similar to the mighty 810 in dynamic and tonal ranges. This will be a rude shock to certain forum members.

Who are you suggesting will be shocked that FI a APS-C crop from Nikon D810 will score the same as the whole sensor in K5?

[clackers]

Who are you suggesting will be shocked that FI a APS-C crop from Nikon D810 will score the same as the whole sensor in K5?

Obviously not you, Fogel ... you have common sense! :-)