You react here claiming you have tested 70Ds (plural!) and draw conclusions based on your testing regarding the difference in noise between the 70D and the 7D.
That might be useful information. When I ask about your test results you come up with a set of lens cap shots pushed 4 stops of a single 70D that most likely never left the shop you tried it at and no 7d results to compare them with.
That’s a bit meager, don’t you think?
for the price?
if you visited my website, you will also find a page dedicated to comparing read noise from a large number of cameras, including the 7D
I've got about 10 other bodies, 70D included, which are still waiting for me to get the time to prepare and add to that page, including some older models.
I have enough data on the 7D and 70D to satisfy my curiosity and I believe I've adequately summarized what I've found, if subjectively, in that CR post.
From all those who claimed to have a 7D (or 5D2) that does not have a pattern noise problem, none have offered or supplied a requested raw file for me to compare.
And the only way those bodies leave the shop is if I pay for them. That kind of research doesn't come cheap.
If someone wants to sponsor more testing, I'll consider it. I do enough on my own dime & time.
The web site so many love to hate, DxOmark, provides plenty of basic data for everyone to review and interpret.www.dxomark.com/Cameras/Canon/EOS-70D---Measurementswww.dxomark.com/Cameras/Canon/EOS-7D---Measurements
Have a look at the FULL SNR tab and compare for yourself.
There really isn't much of an obvious difference between the two bodies, especially in the low ISO shadow region.
But DxOmark does not publish anything on the nature of the measured noise; whether it's random, or patterned.
Visualizing the noise the way I present it provides an opportunity to decide whether the camera's pattern noise characteristic might be an issue for some users.
Compare it to audio; what's easier to tolerate in a quiet music passage, a faint white noise or a discrete 60 Hz hum?
your conclusions are derived from the one thing Canon falls short at and that is low ISO dynamic range...no matter how good the rest of the performance is.
I'll give Canon an approving pat on the head for everything else they've accomplished. Truly, I enjoy using their gear and still retain some.
But I really hope they can pull up their socks a little more on the low ISO
PATTERN NOISE PROBLEM.
DR is NOT what I've been complaining about, PATTERN NOISE is what I've consistently had problems with from Canon's sensors and specifically, DIGIC 4 systems.
Fixed pattern noise is a direct factor in DR measurements, but they are different problems to deal with when producing a final image.
It seems you only like to focus on the low iso part as far as image quality is concerned. I can even understand for some highly specialized photographers having more than 12 stops DR at low iso is important. If so, don’t waste your time on the current offerings from Canon. Buy a D800 or an A7R if you are on a budget or a good medium format camera if low ISO performance is crucial.
I did exactly that - because every Nikon, Pentax and now Fuji and Olympus body I've added to my inventory lacks the one thing I get only from a digic-4 Canon.
FIXED PATTERN NOISE at low ISO .. that fouls up shadow (and sometimes even midtone) areas and limits manipulation of those areas in post without requiring inordinate amounts of time and compromises to correct.
I should not have to waste time in post to fix problems that are innate to the camera; I prefer tools that don't require me to work harder to achieve something so I choose those that don't have deficiencies where it matters to me.
The newer digic 5 products have improved on the FPN issue, just not yet enough to convince me they will outperform my current tools, as either a sensor or a system.
The 70D's sensor performance is no longer much of a limitation compared to the competition, unlike previous digic-4-based canon bodies. And I say that based on the improved FPN performance, not overall DR.
The low ISO DR is still lower than competitors but, with the shadow noise now more random, it's not as much of a post-processing problem as it was when the shadows looked like a picket fence of stripes if pushed, a'la 7D.
Then I visited the website .. you don’t even have to pay me for this great idea.
I'm fairly sure "Canon Plaid" or "Canon Stripe" will not catch on as wall art... But I've been wrong about mass market appeal before.
Do you think Nikon, Sony, Pentax, Fuji, film or Oly fans would buy large prints of that? After all, they can't produce that kind of art with their cameras.
Hmmm... Maybe you're on to something after all.