The Economist article on mirrorless cameras

Started 1 month ago | Discussions thread
Erik Magnuson
Forum ProPosts: 11,716Gear list
Like?
Re: More data
In reply to Jerry Fusselman, 1 month ago

Jerry Fusselman wrote:

In particular, the extra dynamic range from the a6000 at ISO 100 is wonderful.

Yes, the base ISO DR advantage is large and readily exploitable in real world photography.  This is a great example of mapping a DxO number to a real world use case.

but I believe it does not pay to take either of these two cameras beyond ISO 1600.

Is there a different camera you would take past ISO 1600 or do you just not shoot past that? But OK, ISO 1600 is what you call low-light performance.

Yes, I looked at the five curves under the Measurements tab, and all five matter, but I didn't want to get bogged down in so much detail. Any subset of that detail can look like cherry picking.

But looking at these graphs, do they support your statement "According to DXOmark, the a7r sensor already beats every Canon camera (I own mostly Canon equipment) in low-light ISO performance."  Four performance graphs at the ISO 1600 line:

  • SNR18%: too close to call where it crosses the 1600 line.
  • DR: The 6D is slightly higher where it crosses the 1600 line.
  • Tonal Range: too close to call but the A7R may be a tiny bit higher.
  • Color Sensitivity: At the 1600 line, Sony about 1-bit higher.

So you look at that and say "beats?" On the right side of the, DxO attempts to indicate the visual significance of those numbers with a color scale. Are any of the differences visually significant? Remember also that this is purely based on the raw data not including a converter. Can you get these same differences (or more or less difference) using your toolchain of choice?

DxOmark needs to go back to graphing school, for the dots at each ISO are way too far too the left relative to the x-axis. That is, the first dots above are for ISO 100 (I'm almost sure), even though they look like they are for ISO 75 or so.

No, you don't understand what they are plotting: DxO plots their measured ISO, not what's set on the camera. For your low light case (set ISO to 1600 but brighten in post), where the lines actually cross the true ISO 1600 x-axis line is probably most meaningful.

the A7r is either slightly better or much better (color sensitivity, which you omitted).

How did you decide that the 1.2 bits (using the camera marked ISO and not actual exposure) was "much" better?  Can you see that difference?

Since the A7r's sensor is better in every way that matters to me in low light, I would say it is false to call it worse, and that's what I have been trying to show in this part of the thread.

I would not call the A7R sensor worse either.  At real ISO 1600-3200 they are likely indistinguishable for any practical photographic use.

-- hide signature --

Erik

 Erik Magnuson's gear list:Erik Magnuson's gear list
Canon EOS 5D Mark II Canon EOS 450D Sigma SD10 Sony Alpha NEX-5 Nikon D3200 +24 more
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow