Reading mode: Light Dark

Would a Over $1200-1500 Body only 7D MKII make sense.....

Go to last post
bobn2 said:
Buffering is a very simple operation with respect to processing, all they have to do is find somewhere to save the image until it's written.
Click to expand...
It's not half as easy as you write it, not if you want to make optimal use of the available memory. Else you have to calculate everything by worst case, so how big can the RAW file get (worst case), how big can the corresponding JPEG get (worst case). The V1 firmware probably did do just that. With the buffer management advancements coming from the 1Dx development they probably have switched to a dynamic allocation strategy - thus making much better use of the available memory.
bobn2 said:
which is contradicted by the shot above).
Click to expand...
The image posted looks very much like a simplified version but digging through the Magic Lantern information seems to confirm it's information. But it also clarifies that not both Digic do work in parallel on the image data. They speak of the architecture of the 7D as a master/slave configuration - which would mean that one Digic does handle the UI, metering, LiveView (and for that has it's connections to the A/D converter bank, which probably are only active for the video feed - so for image capture they are not used in parallel) while the other does all the image processing including the writing to the CF card.
 
BarnET2 said:
Is this a problem? just get this lens instead of the 70-200mm

Loading…

www.bhphotovideo.com
Click to expand...
Sure, that's an option, as is the Canon 300/2.8LIS, which I have. But neither is portable. And with a 70-200, even the f/2.8, you can still look like a soccer dad. Whipping out the 300 at my kids' soccer games, I get some funny looks and questions. Which I actually kind of enjoy, but others may not. But it would be nice not to have to lug it around, and it would be nice to be able to sell it, buy a 7D2, and pocket a couple grand.
 
  • Like
Reactions: PV2
Karl Gnter Wnsch said:
bobn2 said:
Buffering is a very simple operation with respect to processing, all they have to do is find somewhere to save the image until it's written.
Click to expand...
It's not half as easy as you write it, not if you want to make optimal use of the available memory. Else you have to calculate everything by worst case, so how big can the RAW file get (worst case), how big can the corresponding JPEG get (worst case). The V1 firmware probably did do just that. With the buffer management advancements coming from the 1Dx development they probably have switched to a dynamic allocation strategy - thus making much better use of the available memory.
Karl Gnter Wnsch said:
which is contradicted by the shot above).
Click to expand...
The image posted looks very much like a simplified version but digging through the Magic Lantern information seems to confirm it's information.
Click to expand...
Plus the TI datasheets, I looked at all those before I started writing about how it would work.
Karl Gnter Wnsch said:
But it also clarifies that not both Digic do work in parallel on the image data. They speak of the architecture of the 7D as a master/slave configuration - which would mean that one Digic does handle the UI, metering, LiveView (and for that has it's connections to the A/D converter bank, which probably are only active for the video feed - so for image capture they are not used in parallel) while the other does all the image processing including the writing to the CF card.
Click to expand...
I doubt that they share it up like that, because each chip already has specialist pipelines to do most of those things - basically it can already do the UI, the pre-processing, the JPEG encoding and the writing to the image card all in parallel it has co-processors and intelligent peripherals to do all of those things.
 
bobn2 said:
Karl Gnter Wnsch said:
bobn2 said:
The 'buffer depth' which you refer to applies only to processed images waiting to be written to the card - and most likely the increase in capacity in terms of number of shots has not occurred through any sophisticated buffer management but through tweaking the buffering and file writing to work more concurrently.
Click to expand...
Then a slower card would stall that - which is not true, the difference in the buffer depth between slow and fast cards is small.
Click to expand...
How small?
bobn2 said:
bobn2 said:
The picture you paint of the image processing being split with such a fine granularity that 'all coherency problems do apply' would be madness, when all the image processing needed partition so nicely by raster anyway. Simply, you split the image up into two pieces (looking at the way its arranged, in pairs of lines so that each processor gets full bayer quads to work with)
Click to expand...
I suggest that you look at bayer demosaicing algorithms - having a full bayer quad is insufficient for all but the most basic algorithms which would produce tell tale artifacts - which are not visible. So a more sophisticated demosaicing is run in the camera which is not as localized...
Click to expand...
Those basic methods don't produce 'tell tale artifacts' - they simply don't produce the same resolution as the cleverer ones - in fact, the clever ones produce more artifacts. But, in principle you are right, it would seem better to partition in larger chunks than two line pairs, except that Canon doesn't seem to be doing it that way - this from their explanation on the 1D X

dual_digic_5plus_sensor-to-processor__hero.jpg


--
Bob
Click to expand...
That diagram does not seem to agree with your previous statement:

bobn2 wrote:

"7D has two DIGIC 4. 70D has a DIGIC 5+. 1D X has two DIGIC 5+. Each DIGIC is a different model of TI applications processor, so the 'architecture' of the 1D X is more similar to the 70D than the 7D - except that it has two of them."

If you eliminate one DIGIC 5+ in the 70D, where does the extra data go? Or does the 70D have only four signal paths to DIGIC 5+? In either case, it is not same 'architecture', unless you are talking only about the DIGIC 5+. The diagram of the 7D is more similar to the 1Dx except it has half as many signal paths.
 
Last edited:
altair8800 said:
bobn2 said:
Karl Gnter Wnsch said:
bobn2 said:
The 'buffer depth' which you refer to applies only to processed images waiting to be written to the card - and most likely the increase in capacity in terms of number of shots has not occurred through any sophisticated buffer management but through tweaking the buffering and file writing to work more concurrently.
Click to expand...
Then a slower card would stall that - which is not true, the difference in the buffer depth between slow and fast cards is small.
Click to expand...
How small?
altair8800 said:
The picture you paint of the image processing being split with such a fine granularity that 'all coherency problems do apply' would be madness said:

Loading…

www.learn.usa.canon.com

--
Bob
Click to expand...
That diagram does not seem to agree with your previous statement:

bobn2, post: 54325666, member: 519878"]
"7D has two DIGIC 4. 70D has a DIGIC 5+. 1D X has two DIGIC 5+. Each DIGIC is a different model of TI applications processor, so the 'architecture' of the 1D X is more similar to the 70D than the 7D - except that it has two of them."

If you eliminate one DIGIC 5+ in the 70D, where does the extra data go? Or does the 70D have only four signal paths to DIGIC 5+? In either case, it is not same 'architecture', unless you are talking only about the DIGIC 5+. The diagram of the 7D is more similar to the 1Dx except it has half as many signal paths.
Click to expand...
The 1D X has 16 sensor readout channels channels. The 7D and 70D have 8. In any case, the ADC's feed into one DIGIC, not two. Besides, the discussion about similarity concerned the signal processing and memory usage, not the ADC connection.

--
Bob
Click to expand...
Click to expand...
 
Last edited:
bobn2 said:
altair8800 said:
bobn2 said:
Karl Gnter Wnsch said:
bobn2 said:
The 'buffer depth' which you refer to applies only to processed images waiting to be written to the card - and most likely the increase in capacity in terms of number of shots has not occurred through any sophisticated buffer management but through tweaking the buffering and file writing to work more concurrently.
Click to expand...
Then a slower card would stall that - which is not true, the difference in the buffer depth between slow and fast cards is small.
Click to expand...
How small?
bobn2 said:
The picture you paint of the image processing being split with such a fine granularity that 'all coherency problems do apply' would be madness said:

Loading…

www.learn.usa.canon.com

--
Bob
Click to expand...
That diagram does not seem to agree with your previous statement:

bobn2, post: 54326337, member: 507819"]
"7D has two DIGIC 4. 70D has a DIGIC 5+. 1D X has two DIGIC 5+. Each DIGIC is a different model of TI applications processor, so the 'architecture' of the 1D X is more similar to the 70D than the 7D - except that it has two of them."

If you eliminate one DIGIC 5+ in the 70D, where does the extra data go? Or does the 70D have only four signal paths to DIGIC 5+? In either case, it is not same 'architecture', unless you are talking only about the DIGIC 5+. The diagram of the 7D is more similar to the 1Dx except it has half as many signal paths.
Click to expand...
The 1D X has 16 sensor readout channels channels. The 7D and 70D have 8. In any case, the ADC's feed into one DIGIC, not two. Besides, the discussion about similarity concerned the signal processing and memory usage, not the ADC connection.

--
Bob
Click to expand...
The diagram of the 7D here:

http://cpn.canon-europe.com/content/education/technical/eos7d.do

Clearly shows each ADC feeds into two DIGICs just as does the 1Dx. The 70D has only one DIGIC and is not similar at all.
Click to expand...
Click to expand...
 
bobn2 said:
joyclick said:
no matter what upgrades featured in it?
Click to expand...
The official price of the 70D is $1,200. The official price of the 6D is $1,900

So, $1,200 makes no sense. Can't see a 7D being less than $500 more than the 70D and less than $500 less than the 6D. The hole in the price range is between the $1,900 6D and the $3,400 5DIII, maybe around $2,500 to $2,700 and that is where the new camera will most likely be.

--

Bob
Click to expand...
Great guess, Bob! I see you got 5 thumbs up for that!
 
now it is out, it makes sense only at $1,200. looks like my old 7D will have to extend 2 more years before retirement.
 
altair8800 said:
bobn2 said:
joyclick said:
no matter what upgrades featured in it?
Click to expand...
The official price of the 70D is $1,200. The official price of the 6D is $1,900

So, $1,200 makes no sense. Can't see a 7D being less than $500 more than the 70D and less than $500 less than the 6D. The hole in the price range is between the $1,900 6D and the $3,400 5DIII, maybe around $2,500 to $2,700 and that is where the new camera will most likely be.

--

Bob
Click to expand...
Great guess, Bob! I see you got 5 thumbs up for that!
Click to expand...
Great guess? It was announced at $1799.
 
Albert Silver said:
altair8800 said:
bobn2 said:
joyclick said:
no matter what upgrades featured in it?
Click to expand...
The official price of the 70D is $1,200. The official price of the 6D is $1,900

So, $1,200 makes no sense. Can't see a 7D being less than $500 more than the 70D and less than $500 less than the 6D. The hole in the price range is between the $1,900 6D and the $3,400 5DIII, maybe around $2,500 to $2,700 and that is where the new camera will most likely be.

--

Bob
Click to expand...
Great guess, Bob! I see you got 5 thumbs up for that!
Click to expand...
Great guess? It was announced at $1799.
Click to expand...
Maybe altair8800 was being sarcastic. Maybe he was just feeling bad for Bob - who procalimed that there would never be a 7D successor as well as the 5D III was the successor - and just wanted him to feel better.
 
I'd only pay $20 for the 7DII.

Boy did Canon surely screw the pooch on this one.
 
PalmettoFellow said:
I'd only pay $20 for the 7DII.

Boy did Canon surely screw the pooch on this one.
Click to expand...
Go away, landscaper, this is a sports/wildlife camera.

--
As the length of a dpr thread approaches 150, the probability of an equivalence debate approaches 1.
 
Last edited:
Or at the very least, your sarcasm meter is WAY OVERDUE for a calibration.

LOL
 
Mark B. said:
Albert Silver said:
altair8800 said:
bobn2 said:
joyclick said:
no matter what upgrades featured in it?
Click to expand...
The official price of the 70D is $1,200. The official price of the 6D is $1,900

So, $1,200 makes no sense. Can't see a 7D being less than $500 more than the 70D and less than $500 less than the 6D. The hole in the price range is between the $1,900 6D and the $3,400 5DIII, maybe around $2,500 to $2,700 and that is where the new camera will most likely be.

--

Bob
Click to expand...
Great guess, Bob! I see you got 5 thumbs up for that!
Click to expand...
Great guess? It was announced at $1799.
Click to expand...
Maybe altair8800 was being sarcastic. Maybe he was just feeling bad for Bob - who procalimed that there would never be a 7D successor as well as the 5D III was the successor - and just wanted him to feel better.
Click to expand...
Yes, I was being sarcastic. He thought price was the main thing that counted in a purchase. I and others pointed out that the 6D and 70D was no competition to the 7DII for our purpose. While I had hoped for a higher res sensor, I have placed a pre-order for it. The 28Mp sensor in the Samsung NX1 is more what I wanted, but it is mirrorless and limited lenses. I am now looking for 30Mp 7DIII (2018?).
 
altair8800 said:
bobn2 said:
altair8800 said:
bobn2 said:
Karl Gnter Wnsch said:
bobn2 said:
The 'buffer depth' which you refer to applies only to processed images waiting to be written to the card - and most likely the increase in capacity in terms of number of shots has not occurred through any sophisticated buffer management but through tweaking the buffering and file writing to work more concurrently.
Click to expand...
Then a slower card would stall that - which is not true, the difference in the buffer depth between slow and fast cards is small.
Click to expand...
How small?
altair8800 said:
The picture you paint of the image processing being split with such a fine granularity that 'all coherency problems do apply' would be madness said:

Loading…

www.learn.usa.canon.com

--
Bob
Click to expand...
That diagram does not seem to agree with your previous statement:

bobn2, post: 54329016, member: 519878"]
"7D has two DIGIC 4. 70D has a DIGIC 5+. 1D X has two DIGIC 5+. Each DIGIC is a different model of TI applications processor, so the 'architecture' of the 1D X is more similar to the 70D than the 7D - except that it has two of them."

If you eliminate one DIGIC 5+ in the 70D, where does the extra data go? Or does the 70D have only four signal paths to DIGIC 5+? In either case, it is not same 'architecture', unless you are talking only about the DIGIC 5+. The diagram of the 7D is more similar to the 1Dx except it has half as many signal paths.
Click to expand...
The 1D X has 16 sensor readout channels channels. The 7D and 70D have 8. In any case, the ADC's feed into one DIGIC, not two. Besides, the discussion about similarity concerned the signal processing and memory usage, not the ADC connection.

--
Bob
Click to expand...
The diagram of the 7D here:

http://cpn.canon-europe.com/content/education/technical/eos7d.do

Clearly shows each ADC feeds into two DIGICs just as does the 1Dx. The 70D has only one DIGIC and is not similar at all.
Click to expand...
And on the 7DII, each ADC may feed into two DIGICs just as does the 1Dx, but there is another processor between them, so not clear about path. It may be for sorting the dual pixel AF data.
Click to expand...
Click to expand...
 
Last edited:
altair8800 said:
altair8800 said:
The diagram of the 7D here:

http://cpn.canon-europe.com/content/education/technical/eos7d.do

Clearly shows each ADC feeds into two DIGICs just as does the 1Dx. The 70D has only one DIGIC and is not similar at all.
Click to expand...
And on the 7DII, each ADC may feed into two DIGICs just as does the 1Dx, but there is another processor between them, so not clear about path. It may be for sorting the dual pixel AF data.
Click to expand...
The 7D Mk II noise is so fine, that they may be reading the 40 million photosite off the sensor individually and then combining them after the ADC, instead of combining them on the sensor and then reading them.
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top