Canon EOS 70D & EOS 7D Mark II Speculation [CR1]

[Tcapp]

The general rule of thumb is your shutter speed should be at LEAST 1/focal length. So if you shoot a 200mm lens, your shutter should be at LEAST 1/200. If you really want to be safe, you should at least double that speed. That is why IS is so important.

Dont forget the crop, so on a crop a 200 it should be a minimum of 1/320.

Perhaps, but I'm not entirely convinced a crop body would matter. The crop is just that: a crop. if you take a photo in photoshop and crop down to the center 60% of the pixels, it doesn't make the photo blurry. So I don't see why cropping in the camera with a smaller sensor would require a faster shutter speed. The crop sensor affects the apparent field of view, not the actual focal length, or the magnification. It's just putting more pixels in a smaller space. Now, if a higher pixel density is what requires a faster shutter speed, that would make sense. Since a smaller movement will be picked up by the smaller pixels. But then that would apply to high MP FF too, not just crop.

So I feel it might not be totally correct to just assume a crop body needs a higher shutter speed. Right? Its not the size that matters, its how you use it!

[Marsu42]

So I feel it might not be totally correct to just assume a crop body needs a higher shutter speed. Right?

No, imho not correct. The "rule of thumb" 1/(shutter speed x crop factor)" is from the good ol' analog days, but for digital and sharpness at pixel level (!) it is important how high the pixel density is, because tighter packed pixels will pick up even the tiniest motion.

So your assumption "a crop camera is like cropping from full frame" would be only correct if the 21/22mp ff sensor had the same pixel density as the 18mp sensor. You are correct if you are only looking at the full picture.

[Tcapp]

So I feel it might not be totally correct to just assume a crop body needs a higher shutter speed. Right?

No, imho not correct. The "rule of thumb" 1/(shutter speed x crop factor)" is from the good ol' analog days, but for digital and sharpness at pixel level (!) it is important how high the pixel density is, because tighter packed pixels will pick up even the tiniest motion.

So your assumption "a crop camera is like cropping from full frame" would be only correct if the 21/22mp ff sensor had the same pixel density as the 18mp sensor. You are correct if you are only looking at the full picture.

Yeah! Thats what I was getting at. 

[DB]

can't we get back to propogating the rumour that the 7DII will be APS-H?

I really like the idea of positioning the 7D as somewhere between APS-C and FF. With the likely discontinuation of the 1D mark IV (once the 1DX actually ships), Canon could pinch some of the other 1DIV features such as the 45-point AF, so as to not cannibalize the 5D3/1DX spec.

If a 7D2 had the larger APS-H sensor, but retained the 18MP resolution, you would still have some extra reach as desired by some professionals, with the added benefit of much improved pixel pitch from 4.3 µm square to about 5.1 µm, coupled with a pair of Digic-5 processors, would likely see at least +1 stop gain in ISO performance (say usable reliably at 3200) or possibly +1.5 to +2 stops resulting in solid 7D2 images @ up to 6400 ISO. Now that would really help sports photographers (wildlife too). A camera like that would probably be able to shoot at 10fps too (using the new dual-Digic 5's).

Then a new 70D could be marketed as the top of the APS-C sensor DSLR's (makes sense as it would be a xxD model, retaining xD as 'Pro' range).

A new 7D2 would be a logical step-up from APS-C users as well as a 2nd complimentary body for Pro's & Enthusiasts' who require a bit of extra reach without breaking the bank on 400mm 500mm or 800mm L lenses. It would also go some way to redress the only real criticism of the current 7D, namely; not great low-light performance.

[briansquibb]

If a 7D2 had the larger APS-H sensor, but retained the 18MP resolution, you would still have some extra reach as desired by some professionals, with the added benefit of much improved pixel pitch from 4.7 µm square to about 5.1 µm, coupled with a pair of Digic-5 processors, would likely see at least +1 stop gain in ISO performance (say usable reliably at 3200) or possibly +1.5 to +2 stops resulting in solid 7D2 images @ up to 6400 ISO.

This would be easily achievable as the 1D4 already is rock solid at iso6400 and does 10fps with twin digic 4s. The extra punch of the Digic 5 would help with the larger image size

[DB]

With all of the advances in gap-less sensor technology, I don't know how cost effective it would be for Canon to produce an 18MP APS-H sensor. If it is readily achievable, then it would be a very obvious 'evolutionary' change for the 7D. This suggested combination of 20% larger pixel size + dual Digic-5 chips could create a speed demon with better IQ, especially when one considers that dual memory card slots is now deemed de rigeur for all Canon xD models. Imagine a 7D with better IQ, improved ISO perf and dare we say 10 to 12 frames per second. Now that would be an award-winner. I'd rather buy a 7D2 with those specs than a budget FF like the old 5D2.

[briansquibb]

With all of the advances in gap-less sensor technology, I don't know how cost effective it would be for Canon to produce an 18MP APS-H sensor. If it is readily achievable, then it would be a very obvious 'evolutionary' change for the 7D. This suggested combination of 20% larger pixel size + dual Digic-5 chips could create a speed demon with better IQ, especially when one considers that dual memory card slots is now deemed de rigeur for all Canon xD models. Imagine a 7D with better IQ, improved ISO perf and dare we say 10 to 12 frames per second. Now that would be an award-winner. I'd rather buy a 7D2 with those specs than a budget FF like the old 5D2.

Absolutely

I think the twin card in the 5DIII is from the 1D4 anyway.

I would also like the same facility as we have on the 1D4 to be able to set the Av and the minimum shutter speed and Auto Iso  + ec - that makes for very easy sports shooting

[KeithR]

It would also go some way to redress the only real criticism of the current 7D, namely; not great low-light performance.

Are you aware of a better-performing cropper in low light?

There isn't one - certainly not the D7000/Pentax K-5.

Bigger pixels do not make for better low light performance - this is uninformed internet hogwash for which there isn't a single Real World example that I'm aware of (although there are innumerable examples that disprove the notion) - and my 7D will happily produce fantastic low light/high ISO images all day long.

If you want better low light performance, a bigger sensor might give you it (despite the fact that the 7D easily kicks the arse of most of Canon and Nikon's older FF cameras, so that's no cut-and-dried), but an APS-H sensor with 18 mps won't make any significant difference whatsoever to noise performance.

Indeed, the 7D is massively better than the APS-H 1D MkIIn that I used for a while a few years back - that only had 8 mps, and comparatively huge pixels.

Converting and processing your files effectively will improve low light results though.

Read this and this from Roger Clark (someone who has forgotten more than most people here will ever know about matters of sensor technology and performance) about the capabilities of the 7D's pixels compared to those of the 1D Mk IV.

He knows what he's talking about, and he tells us quite unequivocally (owning and having studied and tested the pants off each camera) that in fact the 7D's sensor - and therefore its pixels - are more "noise efficient" than those of the 1D Mk IV.

Bigger is better? Not where pixels are concerned. 

[briansquibb]

If you want better low light performance, a bigger sensor might give you it (despite the fact that the 7D easily kicks the arse of most of Canon and Nikon's older FF cameras, so that's no cut-and-dried), but an APS-H sensor with 18 mps won't make any significant difference whatsoever to noise performance.

I am not coninced that the 7D is better than the 5D classic at 800 ISO, certainly not better than the 2007 1DS3 or the 2008 5D2 but there again the 7D sensor isn't much newer.

[Eimajm]

It would also go some way to redress the only real criticism of the current 7D, namely; not great low-light performance.

Are you aware of a better-performing cropper in low light?

There isn't one - certainly not the D7000/Pentax K-5.

Bigger pixels do not make for better low light performance - this is uninformed internet hogwash for which there isn't a single Real World example that I'm aware of (although there are innumerable examples that disprove the notion) - and my 7D will happily produce fantastic low light/high ISO images all day long.

This appears contrary to the opinion of Canon in their white paper on full frame sensors.

"Regardless of format, full-frame sensors are all about image quality. The most
obvious advantage of full-frame sensors is the ability to combine high resolution
with large pixel sizes. Compare two sensors with the same number of pixels, one a
full-frame unit and one smaller. The pixels of the full-frame sensor are larger. Each
larger pixel has a greater surface area available for gathering light. More light
collected means less amplification needs to be applied to the output signal of each
pixel for the purposes of readout and image processing. Less is better here
because magnifying low-level signals inevitably entails picking up and increasing
noise that will then have to be removed as thoroughly as possible in a later step...."

"...In the extreme case of low-light photography and ISO ratings of 800 and above,
high signal-to-noise ratios give full-frame sensors a great advantage. In bright light
with low ISO settings, the abundant charge storage of Canon’s large CMOS pixels
avoids oversaturation."

[RLPhoto]

I still believe canon will pull something out of its sleeve. they need an ACE IN THE HOLE, and i believe the 7D II may be that ace if it has the following specs.

APS-H - 14-18MP
10 FPS
61 point AF
ISO 50-25600

Thats it, and canon would simply drop the High performance 7D APS-C down to the 70D range and state this not to cause confusion among the X0D Range Change. If you want your EF-s lenses still, get a 70D. Nuff said.

[moreorless]

Is ASPH actually "ideal" for most 1D users though? I'd say the main reason for that sensor size is that it was needed to achieve the desired IQ performance. If an ASPC sensor 3 years down the road can achieve something similar then I'd guess alot of 1D users(espeically the bottom end of the market as ther top end have the 1DX) would welcome the extra range that a smaller sensor gives similar lenses.

[dlleno]

lots of engaging reading here guys.  much appreciated.  baring some conspiracy on Canon's part to deliberately code the digic processor to do bad things in low end cameras, some conclusions I draw from recent posts are:

1.  folks read articles and generalize things about pixel pitch that are true, but can't  necessarily  be used by themselves  to predict overall IQ performance of two different cameras.    I could be missing something here, but what I'm seeing presented here recently  is evidence that the 7D should produce better pictures than the 1D4.

2.  Canon must be either oversimplifying or mis-stating  things, or the quote about larger pixels was taken out of context 

3.  There must be factors beyond pixel pitch that contribute to overall IQ.  i.e. the end-to-end system from to LPF to sensor  to Digic processing to  jpg conversion, etc. . SNR of the sensor itself must certainly not be the only  factor, although it appears to be the most engaging and the one assumed to dominate IQ

4.  the preocupation with SNR vs pixel pitch and size causes one to loose sight of technology cycles, manufacturing techniques, and the influence of those other factors mentioned in (3) above  on overall IQ considering the entire IQ "value chain". 

no, I'm not a physicist or a sensor technology expert.  I'm just not following the line of reasoning as it applies to real world cameras in different technology cycles.   

[briansquibb]

Is ASPH actually "ideal" for most 1D users though? I'd say the main reason for that sensor size is that it was needed to achieve the desired IQ performance. If an ASPC sensor 3 years down the road can achieve something similar then I'd guess alot of 1D users(espeically the bottom end of the market as ther top end have the 1DX) would welcome the extra range that a smaller sensor gives similar lenses.

1D4 is already the standard sports/wildlife camera so we are already kitted out for APS-H.

[hyles]

The general rule of thumb is your shutter speed should be at LEAST 1/focal length. So if you shoot a 200mm lens, your shutter should be at LEAST 1/200. If you really want to be safe, you should at least double that speed. That is why IS is so important.

Dont forget the crop, so on a crop a 200 it should be a minimum of 1/320.

Perhaps, but I'm not entirely convinced a crop body would matter. The crop is just that: a crop. if you take a photo in photoshop and crop down to the center 60% of the pixels, it doesn't make the photo blurry. So I don't see why cropping in the camera with a smaller sensor would require a faster shutter speed. The crop sensor affects the apparent field of view, not the actual focal length, or the magnification. It's just putting more pixels in a smaller space. Now, if a higher pixel density is what requires a faster shutter speed, that would make sense. Since a smaller movement will be picked up by the smaller pixels. But then that would apply to high MP FF too, not just crop.

So I feel it might not be totally correct to just assume a crop body needs a higher shutter speed. Right? Its not the size that matters, its how you use it!

Small sensor are more dense, this is one of the reason why you need shorter time comparing with FF. Another reason is that since you are vieving/printing images in the same way, the images of crop sensora are mignified more this mignifies blurr as well.
Diego