Lets get the 70D/7D Mark II rumors flowing!

[jrista]

(and this is indicated by the current Sony 24mp APS-C sensor, which also has very low read noise...but is noisy as hell.)

This is incorrect information. The Sony 24MP APC-C sensor has the same noise performance as their 16MP sensor. Both outperform 7D at high iso noise.
According to DXOMARK the difference between NEX-5n and NEX-7 is less than 0.1 stop. This has also been confirmed by several others by subjective IQ tests.

I wasn't actually referring to read noise, which is superb in Sony's sensors. I was referring to photon noise, which tends to make up the bulk of noise anyway. At such a high density, as a matter of physics, low-ISO photon shot noise is going to be higher (well, it will be higher at all ISO levels, but Sony wins out on the high end due to their extremely low read noise levels.) Sony APS-C sensors are a little larger than Canon APS-C, so its not going to be as noisy as a 24mp Canon APS-C would be, but it doesn't really matter who manufactures them...if we keep pushing APS-C sensor resolution, things are going to get noisier.

 

[D_Rochat]

My prediction is that the 7D falls off the face of the earth leaving the 70D as the King of the Crops. 70D goes back to the magnesium body, retains 8fps, new sensor still at 18mp, updated AF system, CF/SD slots, same body layout as the 5DIII but with pop up flash and an included 7D II decal for those who are obsessed with owning a XD series rather than a XXD. Price? $2000.

Are you on drugs or what?! 7D is one of Canon's best selling and most popular cameras. 5000 Pro's were surveyed so that Canon could come up with the specs for the 7D

I think the 5D3 already shows where the 7D2 is headed. After all, these cameras have probably been developed mostly in parallel and the 5D3 already borrows a lot of stuff from the original 7D design.

Therefore I expect:
- Pretty much a 5D3 body with an APS-C sensor and a pop-up flash
- All the new software features from the 5D Mark III
- New sensor, but staying at 18mp (there were rumors about that a while ago and I hope they are true)

I'm not sure if they'll give it a new AF system or update the current one (low-light AF performance could be improved but apart from that, the 7D's AF is pretty nice).


Overall, the 7D Mark II will probably be more of an evolutional upgrade. Canon's APS-C line already had it's "revolution" when the original 7D was released, and in terms of features, that camera is still pretty much up-to-date compared to the competition.

Now this guy knows what he's talking about. 7D is up-to-date and in September Canon will anounce a MK2 version that will cost a whole lot more

Please link these statistics. I'd like to see them.

 

[Tuggem]

(and this is indicated by the current Sony 24mp APS-C sensor, which also has very low read noise...but is noisy as hell.)

This is incorrect information. The Sony 24MP APC-C sensor has the same noise performance as their 16MP sensor. Both outperform 7D at high iso noise.
According to DXOMARK the difference between NEX-5n and NEX-7 is less than 0.1 stop. This has also been confirmed by several others by subjective IQ tests.

I wasn't actually referring to read noise, which is superb in Sony's sensors. I was referring to photon noise, which tends to make up the bulk of noise anyway. At such a high density, as a matter of physics, low-ISO photon shot noise is going to be higher (well, it will be higher at all ISO levels, but Sony wins out on the high end due to their extremely low read noise levels.) Sony APS-C sensors are a little larger than Canon APS-C, so its not going to be as noisy as a 24mp Canon APS-C would be, but it doesn't really matter who manufactures them...if we keep pushing APS-C sensor resolution, things are going to get noisier.

Since we are only talking about image level the photon noise has nothing to do with pixel size. This is also why sonys 16MP and 24MP sensors have the same performance.
I think one possibillity to actually crank some more noise improvement of the Bayer sensor is to significantly increase the number of pixels. I expect that a 200MP, if it was possible to manucature, would outperform 1DX in low noise at high ISO in final image. This is because NR algorithms are more efficient than pure averaging. The same sensor noise but a huge improvement in the possibility to perform NR.

 

[sawsedge]

(and this is indicated by the current Sony 24mp APS-C sensor, which also has very low read noise...but is noisy as hell.)

This is incorrect information. The Sony 24MP APC-C sensor has the same noise performance as their 16MP sensor. Both outperform 7D at high iso noise.
According to DXOMARK the difference between NEX-5n and NEX-7 is less than 0.1 stop. This has also been confirmed by several others by subjective IQ tests.

I wasn't actually referring to read noise, which is superb in Sony's sensors. I was referring to photon noise, which tends to make up the bulk of noise anyway. At such a high density, as a matter of physics, low-ISO photon shot noise is going to be higher (well, it will be higher at all ISO levels, but Sony wins out on the high end due to their extremely low read noise levels.) Sony APS-C sensors are a little larger than Canon APS-C, so its not going to be as noisy as a 24mp Canon APS-C would be, but it doesn't really matter who manufactures them...if we keep pushing APS-C sensor resolution, things are going to get noisier.

Since we are only talking about image level the photon noise has nothing to do with pixel size. This is also why sonys 16MP and 24MP sensors have the same performance.
I think one possibillity to actually crank some more noise improvement of the Bayer sensor is to significantly increase the number of pixels. I expect that a 200MP, if it was possible to manucature, would outperform 1DX in low noise at high ISO in final image. This is because NR algorithms are more efficient than pure averaging. The same sensor noise but a huge improvement in the possibility to perform NR.

Using my own eyes and looking at NEX-7 RAW files, the Sony 24 mpix sensor looks noisier than my 50D at 1600. I haven't a clue how DXO measures anything, but I seem to disagree with their measurements. At 100, that sensor is beautiful, but it gets too noisy rather quickly IMHO.

So I hope Canon sticks with 18 mpix, and improves DR and noise as they have done with the 1D X and 5D3.

 

[BRNexus6]

Yes, 18 MP's is fine on crop sensor bodies. I want to be able to shoot clean video at ISO 1600 and hopefully a new 18 MP sensor with the DIGIC 5+ processor will make that a possibility. If you need megapixels buy a Hasselblad.

 

[Tuggem]

(and this is indicated by the current Sony 24mp APS-C sensor, which also has very low read noise...but is noisy as hell.)

This is incorrect information. The Sony 24MP APC-C sensor has the same noise performance as their 16MP sensor. Both outperform 7D at high iso noise.
According to DXOMARK the difference between NEX-5n and NEX-7 is less than 0.1 stop. This has also been confirmed by several others by subjective IQ tests.

I wasn't actually referring to read noise, which is superb in Sony's sensors. I was referring to photon noise, which tends to make up the bulk of noise anyway. At such a high density, as a matter of physics, low-ISO photon shot noise is going to be higher (well, it will be higher at all ISO levels, but Sony wins out on the high end due to their extremely low read noise levels.) Sony APS-C sensors are a little larger than Canon APS-C, so its not going to be as noisy as a 24mp Canon APS-C would be, but it doesn't really matter who manufactures them...if we keep pushing APS-C sensor resolution, things are going to get noisier.

Since we are only talking about image level the photon noise has nothing to do with pixel size. This is also why sonys 16MP and 24MP sensors have the same performance.
I think one possibillity to actually crank some more noise improvement of the Bayer sensor is to significantly increase the number of pixels. I expect that a 200MP, if it was possible to manucature, would outperform 1DX in low noise at high ISO in final image. This is because NR algorithms are more efficient than pure averaging. The same sensor noise but a huge improvement in the possibility to perform NR.

Using my own eyes and looking at NEX-7 RAW files, the Sony 24 mpix sensor looks noisier than my 50D at 1600. I haven't a clue how DXO measures anything, but I seem to disagree with their measurements. At 100, that sensor is beautiful, but it gets too noisy rather quickly IMHO.

So I hope Canon sticks with 18 mpix, and improves DR and noise as they have done with the 1D X and 5D3.

Noone else that has made serious test on NEX-7 seems to have disagreed with DXOMARK. Therefor its obvious you are doing something wrong. I would guess you are comparing them at different sizes making it a pointless comparison.

 

[sawsedge]

Since we are only talking about image level the photon noise has nothing to do with pixel size. This is also why sonys 16MP and 24MP sensors have the same performance.
I think one possibillity to actually crank some more noise improvement of the Bayer sensor is to significantly increase the number of pixels. I expect that a 200MP, if it was possible to manucature, would outperform 1DX in low noise at high ISO in final image. This is because NR algorithms are more efficient than pure averaging. The same sensor noise but a huge improvement in the possibility to perform NR.

Using my own eyes and looking at NEX-7 RAW files, the Sony 24 mpix sensor looks noisier than my 50D at 1600. I haven't a clue how DXO measures anything, but I seem to disagree with their measurements. At 100, that sensor is beautiful, but it gets too noisy rather quickly IMHO.

So I hope Canon sticks with 18 mpix, and improves DR and noise as they have done with the 1D X and 5D3.

Noone else that has made serious test on NEX-7 seems to have disagreed with DXOMARK. Therefor its obvious you are doing something wrong. I would guess you are comparing them at different sizes making it a pointless comparison.

You would be guessing wrong then.

About three years ago, when I was looking at the 50D, I did a lot of research on it. Lots of reputable sites and reviewers gave it poor/mediocre marks. I thought, I trust Canon more than these reviews. And I looked closely at the reviews. And I found lots of flaws in these tests. Misplaced focus, sub-optimal processing, using beta software, etc.

And aside from the surprise of the 7D coming out less than a year later, it was a good choice. I only say that because the 7D is better, but oh well, I pulled the trigger, thinking Canon would only ever do incremental upgrades. Now I'm waiting to see what the 7D2 brings before I upgrade. I should note that IQ isn't the reason I'll upgrade. I have beautiful large prints on the wall. I just want better AF at this point. More DR and better ISO is always welcome. I don't believe the camera matters anywhere near as much as the light and composition and vision of the person controlling it.

And, I trust my eyes more than some algorithm. Anyone is free to disagree. When I see the NR applied by Sony on the NEX-7 images, I see a tremendous loss of detail. When I see RAW NEX-7 images with no NR, I see more noise than my 50D.