Canon 7D mk2 Thread

D.P. said:
DXOmark have their 7DMK2 sensor analysis up. Slight improvement over the original in terms of high ISO improvement (about 1/3rd of a stop), but no change at all in terms of Dynamic range or colour depth which is disappointing. Still a long way behind Sony, Nikon, Toshiba, Aptina and Samsung sensors in regards DR and colour depth but high ISO is catching up.

To be fair at higher ISO the sensor does quite well, which again points out the biggest issue is Canon's ADC which affects the noise floor. Being a sports camera then high ISO usage is going to be common so the poor overall sensor performance is probably less of an issue than the headlines would indicate.

What is kind of ironic is the Canon G7X which uses a sony 1inch sensor actually get better DR and colour depth at base ISO. The sensor despite being half the size provides twice the DR! Also interesting is the G7X outperforms the latest sony RX100 so either sony has done more tweaks or Canon has tweaked the sensor themselves.
Click to expand...

I'm more disappointed in how long it took you to post once the results were up :D

1/3 to 2/3 improvement, meh. They must have awesome parties in the Canon sensor development team, because they're not doing a lot else! :D If I stay with Canon I'll pick up one eventually as the feature set is good.

I'm interested to see the results from this chap's review, as well as the D800E. Not that it makes any difference to the sensor performance but it highlights where the wall is for Canon:

http://www.naturescapes.net/forums/viewtopic.php?f=57&t=249565&start=112
 
Well they do say that the 7d2 sensor is as good or better than other crop sensors at higher ISOs so I guess that's something... Guess they took a break from the partying to get that done :D

It fits with the whole idea of the camera, if you're doing action/sports then you'll probably need the higher ISOs more.
 
Rojin said:
I'm more disappointed in how long it took you to post once the results were up :D

1/3 to 2/3 improvement, meh. They must have awesome parties in the Canon sensor development team, because they're not doing a lot else! :D If I stay with Canon I'll pick up one eventually as the feature set is good.

I'm interested to see the results from this chap's review, as well as the D800E. Not that it makes any difference to the sensor performance but it highlights where the wall is for Canon:

http://www.naturescapes.net/forums/viewtopic.php?f=57&t=249565&start=112
Click to expand...


I didn't interest me at all and I just stumbled across it. I wasn't expecting any change despite some of the excited rumours. Canon and Nikon have both publicized 1stop improvements in high ISO images but this is actually just due to more powerful CPUs allowing better noise reduction in the jpegs which isn't that exciting. Saying that there are plenty of sports pros that shoot in jpeg so that they can get a huge buffer and quickly send off photos to the head office for publication.

Edit: reading that link and i think the OP is missing something and making some unsubstantiated claims. On many sensors the very lowest values are set to zero because these are effectively more or less just photon noise with very low signal. When you raise shadows then you are effectively raising zero instead of noise so in theory it looks cleaner but actually lowers theoretical max Dynamic range contrary to what he believes (because if there was signal in that noise you've thrown it out). The main reason to do it really just to save file space because there is little need to save noise, so the 12 or 14bits per pixel can be assigned to more useful ranges. Most Nikon and Sony sensor did this until recently. Nikon increased the full well capacity of the D810 and stopped cutting the noise values to zero. This has resulted in 0.5 stops higher DR. So whatever the OP in that thread was thinking has been disproven because this theory doesn't apply to the D810 (and the other new sensors like D750 and D5300). Beyond that it is easy to see with empirical evidence, all modern Sony and Nikon (from D7000 and 800 onwards) sensors have had the ability to pull shadows 5 stops that the canon sensors just couldn't do, regardless of what post processing you or whatever that OP thinks the cameras are doing the results speak for them selves. DXO numbers are observable in the real world and are replicated elsewhere, but you do have to look in details to see what they mean and not just take the headlines "DXOmark" numbers which are useless.

The 7Dmk2 has a perfectly adequate sensor for high ISO work, which is the main use of that type of camera. It is the best crop DSLR for sports work where high speed and deep buffer is needed.It wouldn't be had for Nikon to best it though, the D7200 is pretty close and really just needs to shove in 3x the amount of ram and put it in an updated D300 body and the job is done - keep your eyes pen at CES...
 
Last edited:
Phal said:
Not sure why people would buy a 1Dx to do only landscape and weddings unless they've got plenty of cash and then I guess they might as well... Like you say, it's the ultimate Canon camera

To me the 1Dx is a sports/action camera because of the features that it has like the great AF and silly fast FPS. That's not to say you can't use it for anything else but I'm not sure people would upgrade from a 5d3 to a 1Dx for better IQ or a builtin battery grip and an RJ45 port? :E
Click to expand...

Well for wedding photography you want the best IQ and ISO performance possible and that's the 1DX. Granted it doesn't offer that much over the 5D3 for twice the price so the latter obviously is better value.

For me the crop factor is convenient but isn't a factor that would stop me getting something like a 1Dx. The better resolution available from a FF sensor would let me crop a fair bit so I guess the difference might not be as much as it seems and I'd be able to take advantage of a lovely FF sensor.

7d2 will suit me well and it's cheaper so I'll 'make do' with that for a while lol.
Click to expand...

Each to their own but that's the bit I don't understand tbh. If the crop factor isn't essential to you (and with 560mm available I'd agree it's not), then I'd get a 5D3 over a 7D2 any day. You have a £9k lens - give it the full frame IQ it deserves! :D
 
The crop factor can certainly be helpful but there are times when I wouldn't mind the perks of the 5d3 with its FF sensor. The problem is the FPS and Buffer are a bit sucky in comparison and it seems like the 7d2 AF is better too which is a winner since AF is one of my most important features.


Ideally I'd have them all :D but alas that's too much for even me to justify lol. The 200-400mm is the best allrounder lens for me and the 7d2 is looking like the best allrounder body for me too even if the screen doesn't flip out.
 
The 7D2's AF may be technically better but the 5D3's system is hardly shabby! :p

Whether the higher FPS is an important consideration is obviously subjective.

My advice would be to rent/borrow/beg/steal a 5D3 to at least try out before plumping for a 7D2 :)
 
Vertigo1 said:
The 7D2's AF may be technically better but the 5D3's system is hardly shabby! :p

Whether the higher FPS is an important consideration is obviously subjective.

My advice would be to rent/borrow/beg/steal a 5D3 to at least try out before plumping for a 7D2 :)
Click to expand...

Plenty of comparisons out there. It's only when the ISO gets really high that the crop loses it's benefit. Otherwise it produces better results than the 5D3 cropped. This was with the 7D, so should be slightly better with the 7D2.

If I was looking for a full frame for this sort of use I'd get the Nikon D810, as in DX mode it's still 15MP. The 5D3 is something like 8MP. (if someone would just give me all the money for my gear, rather than having to go through all the hassle of selling, I'd have a D810 quite happily)
 
Last edited:
Rojin said:
Plenty of comparisons out there. It's only when the ISO gets really high that the crop loses it's benefit. Otherwise it produces better results than the 5D3 cropped. This was with the 7D, so should be slightly better with the 7D2.
Click to expand...

Of course you'll get better quality and more pixels per duck from a crop sensor than you will cropping a 5D3 image down to the same field of view - that's why I said that, if the crop factor is important the 7D2 may well be the best option.

If the crop isn't that important then the 5D3 will always produce better images than the 7D2 when the full frame of the sensor is used.
 
Plus the AF is more on a level with the 1DX apparently, not too mention the points cover much more of the frame, also you get 40% more shots in a burst. So it's quite easy to see why action and wildlife tog's are liking it. Even as a second body to a 1DX.

I still want a 1D4 more though :D I have gotten over my initial G.A.S. though so will wait until next year... probably.
 
Rojin said:
Plenty of comparisons out there. It's only when the ISO gets really high that the crop loses it's benefit. Otherwise it produces better results than the 5D3 cropped. This was with the 7D, so should be slightly better with the 7D2.

If I was looking for a full frame for this sort of use I'd get the Nikon D810, as in DX mode it's still 15MP. The 5D3 is something like 8MP. (if someone would just give me all the money for my gear, rather than having to go through all the hassle of selling, I'd have a D810 quite happily)
Click to expand...

apples and oranges
 
For a lot of sports work you will only benefit from the larger sensor if you have longer lenses to make up for the crop factor - this quickly gets very impractical, expensive and heavy. E.g., you might use a 70-200mm f/2.8 quite effectively for many sports. Going FF would require a 300mm f/2.8 (or sigma 120-300 f/2.8) to maintain similar subject magnification in order to benefit form the FF sensor. In which case it is less about the sensor, more the fact that you have now attached a friggin huge lens to the front of your camera that is making the difference - the physical aperture is now twice the area giving you the extra stop of light.
 
^^^ nothing new there at all! what we've been saying for a long time. The overall dxomark is pretty meaningless, the value is in the measurements across ISO. How the measurements are defined is clearly described on the website.


DXO measurements are well supported across several other testers along with empirical proof.

Yeah, DXO use engineering DR which is why the number are high but that makes the numbers impartial. Engineering DR has a a signal to nose ratio of 1:1 as the lowest DR, you can be that 2:1 or 4:1 and loose 1 or 2 stops if want but the same difference is apparent across cameras. It is then entirely subjective which SNR you choose. You can require 4:1 at base and then claim the canon has 9.7stops DR and the Nikon 12.7DR


Here is a great alternative to dxo, unsurprisingly the results support dxo
http://home.comcast.net/~nikond70/Charts/PDR.htm
 
Last edited:
Well they announced a few other lenses at the same time like the 400mm DO mk2 and the 24mm STM

I guess this lens is a big enough thing to warrant its own fanfare since people have been waiting for ages for it!
 
You must log in or register to reply here.
Back
Top Bottom