This topic is in the 2009 archive. You can read it and like its posts here. It is read only.
Archive 2009Canon EOS 7D Master thread
p.74 #1Canon EOS 7D Master thread
Great news from Canon Australia! Can't wait to get hold of the 7D next week & try the new AF on the 100-400L & Sigmonster!
Sep 24, 2009 at 07:57 PM
p.74 #2Canon EOS 7D Master thread
cjtaylo wrote:
I know some Nikon fellas with the D300 doing this using an SB-800.Wow, I hadn't thought of that, thanks! One less nuisance when using my Kirk macro flash bracket.
MarkYou can do the same with an ST-E2 of course. However, in either case you loose rear-curtain sync with Canon (an unfortunate limitation of wireless E-TTL). Often I'm dragging the shutter a bit in situations where I'm using a bracket, so the cord is still useful.I was wondering about that. My flash manual indicates that the slave flash has to be in front of the master and within an 80 degree arc centered on the flash axis, assuming that both are 580EX IIs. Is that not the case when using the ST-E2 as the master?Has anyone read or know what the range on the 7D flash transmitter is? Just curious.Yes, the technical white paper says 10m indoors and 15 meters outdoors. I'm looking forward to getting rid of my cheap ebay triggers :-)
http://cpn.canon-europe.com/content/technical/eos7d.do#container
That's pretty good range
BUT are those numbers reversed? I think the ST-E2 has more range indoors with near 50 ft and 33 ft outdoors.
Added:
They are reversed I checked the link and it is the same as the ST-E2.
Gil_W wrote
WillWeb wrote
garyvot wrote
msalvetti wrote
rsg_1 wrote
I don't know if this has been brought up, but with the flash commander in the 7D, when using a flash bracket with a Speedlite, you would not need the OC-E3 off-shoe flash cord.I know some Nikon fellas with the D300 doing this using an SB-800.
Mark
http://cpn.canon-europe.com/content/technical/eos7d.do#container
That's pretty good range
BUT are those numbers reversed? I think the ST-E2 has more range indoors with near 50 ft and 33 ft outdoors. Added:
They are reversed I checked the link and it is the same as the ST-E2.
Sep 24, 2009 at 08:27 PM
p.74 #3Canon EOS 7D Master thread
cameron12x wrote:
Although it's not the final word, the title of Michael Reichman's preliminary review of the 7D says it all:
A New Body for the Next Decade
We need not be worried.
See previous page
Although it's not the final word, the title of Michael Reichman's preliminary review of the 7D says it all:
A New Body for the Next Decade
We need not be worried.
See previous page
Sep 25, 2009 at 12:30 AM
p.74 #4Canon EOS 7D Master thread
alundeb wrote:
He should have just punted... his further comments do nothing more than add confusion and possilbly raise doubts where there may be none.
"I am going to punt, and will not be describing in any detail the image quality that I have seen from the Canon 7D.
Frankly, the reason is that I am seeing things which I am not sure are a consequence of the camera's inherent characteristics or are specific to pre-production cameras or even this specific camera. This includes overall image softness and some digital artifacting. Nothing terrible mind you; just enough though that I am aware of it, and therefore don't want to make any undue assumptions."Sounds like effects of NR combined with sharpening. My 500D can do that sort of things by the default settings in DPP. Relaxing NR and sharpening helps.
This is just a general comment, and should not be taken as defence of the 7D, or denial that Reichmann may have seen something.
if you look at the RAW samples from IR you can see lots of sensor artifacting from imbalanced channels, hopefully only a few pre-pro units have that issue....
it is a little bad that MR seems to hint that maybe he has that too....
then again maybe for the test units they didnt have time to calibrate the sensors....
John McLean wrote
scott f wrote
The image quality comments are quite concerning to me."I am going to punt, and will not be describing in any detail the image quality that I have seen from the Canon 7D.
Frankly, the reason is that I am seeing things which I am not sure are a consequence of the camera's inherent characteristics or are specific to pre-production cameras or even this specific camera. This includes overall image softness and some digital artifacting. Nothing terrible mind you; just enough though that I am aware of it, and therefore don't want to make any undue assumptions."
This is just a general comment, and should not be taken as defence of the 7D, or denial that Reichmann may have seen something.
if you look at the RAW samples from IR you can see lots of sensor artifacting from imbalanced channels, hopefully only a few pre-pro units have that issue....
it is a little bad that MR seems to hint that maybe he has that too....
then again maybe for the test units they didnt have time to calibrate the sensors....
Sep 25, 2009 at 01:04 AM
p.74 #5Canon EOS 7D Master thread
skibum5 wrote:
if you look at the RAW samples from IR you can see lots of sensor artifacting from imbalanced channels, hopefully only a few pre-pro units have that issue....
it is a little bad that MR seems to hint that maybe he has that too....
then again maybe for the test units they didnt have time to calibrate the sensors....
Are you saying the 7D sensor is not that great after all? Hmmm...
I have yet to come across a single report that says the production copy behaves differently from a pre-production copy.
if you look at the RAW samples from IR you can see lots of sensor artifacting from imbalanced channels, hopefully only a few pre-pro units have that issue....
it is a little bad that MR seems to hint that maybe he has that too....
then again maybe for the test units they didnt have time to calibrate the sensors....
Are you saying the 7D sensor is not that great after all? Hmmm...
I have yet to come across a single report that says the production copy behaves differently from a pre-production copy.
Sep 25, 2009 at 02:18 AM
p.74 #6Canon EOS 7D Master thread
thw2 wrote:
it is a little bad that MR seems to hint that maybe he has that too....
then again maybe for the test units they didnt have time to calibrate the sensors....Are you saying the 7D sensor is not that great after all? Hmmm...
I have yet to come across a single report that says the production copy behaves differently from a pre-production copy.
I guess skibum is talking about the imbalance between the two green channels. That is certainly something that should be fixed by adjustment or firmware or correct parameters for the RAW conversion software.
skibum5 wrote
if you look at the RAW samples from IR you can see lots of sensor artifacting from imbalanced channels, hopefully only a few pre-pro units have that issue....it is a little bad that MR seems to hint that maybe he has that too....
then again maybe for the test units they didnt have time to calibrate the sensors....
I have yet to come across a single report that says the production copy behaves differently from a pre-production copy.
I guess skibum is talking about the imbalance between the two green channels. That is certainly something that should be fixed by adjustment or firmware or correct parameters for the RAW conversion software.
Sep 25, 2009 at 02:24 AM
p.74 #7Canon EOS 7D Master thread
alundeb wrote:
I guess skibum is talking about the imbalance between the two green channels. That is certainly something that should be fixed by adjustment or firmware or correct parameters for the RAW conversion software.
How exactly does such an artifact look like?
I guess skibum is talking about the imbalance between the two green channels. That is certainly something that should be fixed by adjustment or firmware or correct parameters for the RAW conversion software.
How exactly does such an artifact look like?
Sep 25, 2009 at 02:40 AM
p.74 #8Canon EOS 7D Master thread
alundeb wrote:
http://www.robgalbraith.com/bins/multi_page.asp?cid=7-10042-10239
In the review, he pretty much summed up my greatest fear about the 7D IQ. Quote:
What the pictures look like is perhaps easier to digest. So far, the photos we've taken with a beta 7D look a lot like they've come from a 50D, except with 2.9 million additional pixels of resolution. Canon appears to have done a masterful job of wringing out every ounce of quality from the 7D's little pixels (smaller than any Canon before), resulting in photos that are crisp, reasonably clean and eminently usable up to about ISO 1600.
If I sell my very low click 50D to pickup a 7D only to find that the images are the same or essentially similar, I am going to be really disappointed. Looking at his sample images, I can see how he came to his conclusions. I would just rather keep the 50D and put my money into something else.
He has some sample images with the new 15-85 lens.I don't know why RG uses the words he does.
"A lot like they've come from a 50D" can mean anything from worse to better, and no one knows which direction, and by how much.
Why not check comparisons rather than uncomparable test shots in arbitrary conditions?
One example, the presentation of samples from IR, by thedigitalbean on page 41 in this thread.
http://akimagery.zenfolio.com/img/v1/p1063258743.jpg
http://akimagery.zenfolio.com/img/v2/p917742329.jpg
Does the improvement from the 50D to 7D seem significant to you or not?
The Picture in the Link above
What for is this good? Its Pixel Leveld based and has nothing to do the real Image sizes....
UCSB wrote
I was reading the Rob Galbraith review of the 7D. It can be found here:http://www.robgalbraith.com/bins/multi_page.asp?cid=7-10042-10239
In the review, he pretty much summed up my greatest fear about the 7D IQ. Quote:
What the pictures look like is perhaps easier to digest. So far, the photos we've taken with a beta 7D look a lot like they've come from a 50D, except with 2.9 million additional pixels of resolution. Canon appears to have done a masterful job of wringing out every ounce of quality from the 7D's little pixels (smaller than any Canon before), resulting in photos that are crisp, reasonably clean and eminently usable up to about ISO 1600.
If I sell my very low click 50D to pickup a 7D only to find that the images are the same or essentially similar, I am going to be really disappointed. Looking at his sample images, I can see how he came to his conclusions. I would just rather keep the 50D and put my money into something else.
He has some sample images with the new 15-85 lens.
"A lot like they've come from a 50D" can mean anything from worse to better, and no one knows which direction, and by how much.
Why not check comparisons rather than uncomparable test shots in arbitrary conditions?
One example, the presentation of samples from IR, by thedigitalbean on page 41 in this thread.
http://akimagery.zenfolio.com/img/v1/p1063258743.jpg
http://akimagery.zenfolio.com/img/v2/p917742329.jpg
Does the improvement from the 50D to 7D seem significant to you or not?
The Picture in the Link above
What for is this good? Its Pixel Leveld based and has nothing to do the real Image sizes....
Sep 25, 2009 at 02:58 AM
p.74 #9Canon EOS 7D Master thread
headroom wrote:
http://www.robgalbraith.com/bins/multi_page.asp?cid=7-10042-10239
In the review, he pretty much summed up my greatest fear about the 7D IQ. Quote:
What the pictures look like is perhaps easier to digest. So far, the photos we've taken with a beta 7D look a lot like they've come from a 50D, except with 2.9 million additional pixels of resolution. Canon appears to have done a masterful job of wringing out every ounce of quality from the 7D's little pixels (smaller than any Canon before), resulting in photos that are crisp, reasonably clean and eminently usable up to about ISO 1600.
If I sell my very low click 50D to pickup a 7D only to find that the images are the same or essentially similar, I am going to be really disappointed. Looking at his sample images, I can see how he came to his conclusions. I would just rather keep the 50D and put my money into something else.
He has some sample images with the new 15-85 lens.I don't know why RG uses the words he does.
"A lot like they've come from a 50D" can mean anything from worse to better, and no one knows which direction, and by how much.
Why not check comparisons rather than uncomparable test shots in arbitrary conditions?
One example, the presentation of samples from IR, by thedigitalbean on page 41 in this thread.
http://akimagery.zenfolio.com/img/v1/p1063258743.jpg
http://akimagery.zenfolio.com/img/v2/p917742329.jpg
Does the improvement from the 50D to 7D seem significant to you or not?The Picture in the Link above
What for is this good? Its Pixel Leveld based and has nothing to do the real Image sizes....
Wrong, the sizes are equalized, according to thedigitalbean who posted them:
"I have put together a camera/ISO matrix based off the Imaging Resource RAW files and ACR 5.5/LR 2.5. Click to link view actual pixels. All other cameras were resized (up or down) to the 7D's size to keep the comparison fair."
alundeb wrote
UCSB wrote
I was reading the Rob Galbraith review of the 7D. It can be found here:http://www.robgalbraith.com/bins/multi_page.asp?cid=7-10042-10239
In the review, he pretty much summed up my greatest fear about the 7D IQ. Quote:
What the pictures look like is perhaps easier to digest. So far, the photos we've taken with a beta 7D look a lot like they've come from a 50D, except with 2.9 million additional pixels of resolution. Canon appears to have done a masterful job of wringing out every ounce of quality from the 7D's little pixels (smaller than any Canon before), resulting in photos that are crisp, reasonably clean and eminently usable up to about ISO 1600.
If I sell my very low click 50D to pickup a 7D only to find that the images are the same or essentially similar, I am going to be really disappointed. Looking at his sample images, I can see how he came to his conclusions. I would just rather keep the 50D and put my money into something else.
He has some sample images with the new 15-85 lens.
"A lot like they've come from a 50D" can mean anything from worse to better, and no one knows which direction, and by how much.
Why not check comparisons rather than uncomparable test shots in arbitrary conditions?
One example, the presentation of samples from IR, by thedigitalbean on page 41 in this thread.
http://akimagery.zenfolio.com/img/v1/p1063258743.jpg
http://akimagery.zenfolio.com/img/v2/p917742329.jpg
Does the improvement from the 50D to 7D seem significant to you or not?
What for is this good? Its Pixel Leveld based and has nothing to do the real Image sizes....
Wrong, the sizes are equalized, according to thedigitalbean who posted them:
"I have put together a camera/ISO matrix based off the Imaging Resource RAW files and ACR 5.5/LR 2.5. Click to link view actual pixels. All other cameras were resized (up or down) to the 7D's size to keep the comparison fair."
Sep 25, 2009 at 03:13 AM
p.74 #10Canon EOS 7D Master thread
skibum5 wrote:
yeah they should've at least tossed us the f/4 high-precision for the center since it does make all the f/4 dodgier for AF.
Well duh, they tossed us high precision for the center top, center middle and center bottom sensors working up to f/5.6 by incorporating interleaved horizontal line sensors at those positions into the cross type sensor.
Citing from CPN:
"f/5.6 sensors are especially good at predictive tracking focus, however, the designers wanted to find a solution that also offers high focus accuracy as well. Increasing focus accuracy can be achieved by using a higher resolution focus sensor however, simply increasing the pitch, and therefore resolution, of a single line too much leads to a lower signal to noise ratio in terms of light levels recorded. To combat this, the centre top, centre middle and centre bottom sensors (when the camera is held in a landscape orientation) have a dual zigzag horizontal line sensor.
The dual zigzag arrangement solves this problem by using two focus sensors that are slightly offset from each other. By performing AF calculations based on the two lines sampling slightly different areas, the focusing resolution is improved leading to a more precise focus result while still retaining the responsive AF tracking."
Except for the single horizontal line sensor which works up to f/8 the 7D autofocus module is wiping the floor with the 1D series module...
yeah they should've at least tossed us the f/4 high-precision for the center since it does make all the f/4 dodgier for AF.
Well duh, they tossed us high precision for the center top, center middle and center bottom sensors working up to f/5.6 by incorporating interleaved horizontal line sensors at those positions into the cross type sensor.
Citing from CPN:
"f/5.6 sensors are especially good at predictive tracking focus, however, the designers wanted to find a solution that also offers high focus accuracy as well. Increasing focus accuracy can be achieved by using a higher resolution focus sensor however, simply increasing the pitch, and therefore resolution, of a single line too much leads to a lower signal to noise ratio in terms of light levels recorded. To combat this, the centre top, centre middle and centre bottom sensors (when the camera is held in a landscape orientation) have a dual zigzag horizontal line sensor.
The dual zigzag arrangement solves this problem by using two focus sensors that are slightly offset from each other. By performing AF calculations based on the two lines sampling slightly different areas, the focusing resolution is improved leading to a more precise focus result while still retaining the responsive AF tracking."
Except for the single horizontal line sensor which works up to f/8 the 7D autofocus module is wiping the floor with the 1D series module...
Sep 25, 2009 at 03:15 AM
p.74 #11Canon EOS 7D Master thread
thw2 wrote:
How exactly does such an artifact look like?
I saw an example once, when viewed at 400%, one could see that each 4'th pixel was a tiny bit darker, or something like that. Don't remember where I found it, but skibum may help us here.
Depending on the demosaic algorithm, it may also show up as a maze pattern.
alundeb wrote
I guess skibum is talking about the imbalance between the two green channels. That is certainly something that should be fixed by adjustment or firmware or correct parameters for the RAW conversion software.I saw an example once, when viewed at 400%, one could see that each 4'th pixel was a tiny bit darker, or something like that. Don't remember where I found it, but skibum may help us here.
Depending on the demosaic algorithm, it may also show up as a maze pattern.
Sep 25, 2009 at 03:18 AM
p.74 #12Canon EOS 7D Master thread
I think I can live with that level of "damage" to the IQ...
Sep 25, 2009 at 04:27 AM
p.74 #13Canon EOS 7D Master thread
keithreeder wrote:
I think I can live with that level of "damage" to the IQ...
Yes, if that's all
I'm buying it. Preordered at the release date, not knowing for certain, with a free cancel option. After consuming all the information available so far, I am not afraid that any hidden problems are going to break the deal for me.
Most important to me, are resolution, pixel density, noise and dynamic range. I see a distinct improvement over my 500D, but everyone will have to see for themselves, as said here before :-)
I think I can live with that level of "damage" to the IQ...
Yes, if that's all
I'm buying it. Preordered at the release date, not knowing for certain, with a free cancel option. After consuming all the information available so far, I am not afraid that any hidden problems are going to break the deal for me.
Most important to me, are resolution, pixel density, noise and dynamic range. I see a distinct improvement over my 500D, but everyone will have to see for themselves, as said here before :-)
Sep 25, 2009 at 05:00 AM
p.74 #14Canon EOS 7D Master thread
alundeb wrote:
I'm buying it. Preordered at the release date, not knowing for certain, with a free cancel option. After consuming all the information available so far, I am not afraid that any hidden problems are going to break the deal for me.
Most important to me, are resolution, pixel density, noise and dynamic range. I see a distinct improvement over my 500D, but everyone will have to see for themselves, as said here before :-)
Ditto. Having checked the raw-files myself, I'm not concerned about the image comment at Luninous Landscape since it most probably stems from poorly tuned raw-converter software at time point (Adobe themselves says support is beta and NOT optimized). I'm also having a feeling that the built-in noise reduction on jpegs isn't that great, but I shoot raw and can always turn it off if jpegs are needed for action burst-shoots.
I'm buying it. Preordered at the release date, not knowing for certain, with a free cancel option. After consuming all the information available so far, I am not afraid that any hidden problems are going to break the deal for me.
Most important to me, are resolution, pixel density, noise and dynamic range. I see a distinct improvement over my 500D, but everyone will have to see for themselves, as said here before :-)
Ditto. Having checked the raw-files myself, I'm not concerned about the image comment at Luninous Landscape since it most probably stems from poorly tuned raw-converter software at time point (Adobe themselves says support is beta and NOT optimized). I'm also having a feeling that the built-in noise reduction on jpegs isn't that great, but I shoot raw and can always turn it off if jpegs are needed for action burst-shoots.
Sep 25, 2009 at 05:41 AM
p.74 #15Canon EOS 7D Master thread
charlyw wrote:
Well duh, they tossed us high precision for the center top, center middle and center bottom sensors working up to f/5.6 by incorporating interleaved horizontal line sensors at those positions into the cross type sensor.
Citing from CPN:
"f/5.6 sensors are especially good at predictive tracking focus, however, the designers wanted to find a solution that also offers high focus accuracy as well. Increasing focus accuracy can be achieved by using a higher resolution focus sensor however, simply increasing the pitch, and therefore resolution, of a single line too much leads to a lower signal to noise ratio in terms of light...Show more →
The only focus point which has the extra sensitivity at f2.8 is the centre one as per a direct quote from that canon webpage. I didn't see it said anywhere about it working at f8 (think telephoto with 2X converter), have you got a link to that, I did look for it wondering whether is was something else that canon was keeping for the 1 series (understandably actually).
skibum5 wrote
yeah they should've at least tossed us the f/4 high-precision for the center since it does make all the f/4 dodgier for AF.Citing from CPN:
"f/5.6 sensors are especially good at predictive tracking focus, however, the designers wanted to find a solution that also offers high focus accuracy as well. Increasing focus accuracy can be achieved by using a higher resolution focus sensor however, simply increasing the pitch, and therefore resolution, of a single line too much leads to a lower signal to noise ratio in terms of light...Show more →
The only focus point which has the extra sensitivity at f2.8 is the centre one as per a direct quote from that canon webpage. I didn't see it said anywhere about it working at f8 (think telephoto with 2X converter), have you got a link to that, I did look for it wondering whether is was something else that canon was keeping for the 1 series (understandably actually).
Sep 25, 2009 at 05:54 AM
p.74 #16Canon EOS 7D Master thread
Beni wrote:
The only focus point which has the extra sensitivity at f2.8 is the centre one as per a direct quote from that canon webpage. I didn't see it said anywhere about it working at f8 (think telephoto with 2X converter), have you got a link to that, I did look for it wondering whether is was something else that canon was keeping for the 1 series (understandably actually).
I wrote that f/8 focusing is the only omission. And I quoted directly from the Canon Professional Network site about the high precision focusing points located at the top, middle and bottom of the center row - which are high precision focusing line sensors all the way to f/5.6.
The only focus point which has the extra sensitivity at f2.8 is the centre one as per a direct quote from that canon webpage. I didn't see it said anywhere about it working at f8 (think telephoto with 2X converter), have you got a link to that, I did look for it wondering whether is was something else that canon was keeping for the 1 series (understandably actually).
I wrote that f/8 focusing is the only omission. And I quoted directly from the Canon Professional Network site about the high precision focusing points located at the top, middle and bottom of the center row - which are high precision focusing line sensors all the way to f/5.6.
Sep 25, 2009 at 06:12 AM
p.74 #17Canon EOS 7D Master thread
charlyw wrote:
Except for the single horizontal line sensor which works up to f/8 the 7D autofocus module is wiping the floor with the 1D series module...
I doubt that will likely be the case with the next 1D iteration; Canon will likely continue to use AF as part of their market segmentation strategy.
IF the 7D AF "approaches" the last working 1D focusing performance I'll be a very happy camper. Time will tell. We should know in the next few weeks, as people start using the 7D (it's already available on some streets in the UK).
skibum5 wrote
yeah they should've at least tossed us the f/4 high-precision for the center since it does make all the f/4 dodgier for AF.I doubt that will likely be the case with the next 1D iteration; Canon will likely continue to use AF as part of their market segmentation strategy.
IF the 7D AF "approaches" the last working 1D focusing performance I'll be a very happy camper. Time will tell. We should know in the next few weeks, as people start using the 7D (it's already available on some streets in the UK).
Sep 25, 2009 at 06:15 AM
p.74 #18Canon EOS 7D Master thread
charlyw wrote:
I wrote that f/8 focusing is the only omission. And I quoted directly from the Canon Professional Network site about the high precision focusing points located at the top, middle and bottom of the center row - which are high precision focusing line sensors all the way to f/5.6.
Read it again, it's only the center focusing point which has extra sensitivity at f2.8.
"For enhanced precision, the central AF point features extra sensitivity when used with lenses having a maximum aperture of f/2.8 or faster thanks to the diagonal arrangement of the cross-type sensor"
Beni wrote
The only focus point which has the extra sensitivity at f2.8 is the centre one as per a direct quote from that canon webpage. I didn't see it said anywhere about it working at f8 (think telephoto with 2X converter), have you got a link to that, I did look for it wondering whether is was something else that canon was keeping for the 1 series (understandably actually).Read it again, it's only the center focusing point which has extra sensitivity at f2.8.
"For enhanced precision, the central AF point features extra sensitivity when used with lenses having a maximum aperture of f/2.8 or faster thanks to the diagonal arrangement of the cross-type sensor"
Sep 25, 2009 at 06:57 AM
p.74 #19Canon EOS 7D Master thread
Beni wrote:
Read it again, it's only the center focusing point which has extra sensitivity at f2.8.
You should reread the quote I delivered in my original response. The f/2.8 X aligned cross type sensor is an additional with extra precision for those lenses. Then there are the normally precise 19 cross type AF sensors of which there are three that have horizontal (thus sensitive against vertical contrast) line type detectors which are using an interleaved double AF sensor cell array and thus are delivering higher precision information than the rest.
Read it again, it's only the center focusing point which has extra sensitivity at f2.8.
You should reread the quote I delivered in my original response. The f/2.8 X aligned cross type sensor is an additional with extra precision for those lenses. Then there are the normally precise 19 cross type AF sensors of which there are three that have horizontal (thus sensitive against vertical contrast) line type detectors which are using an interleaved double AF sensor cell array and thus are delivering higher precision information than the rest.
Sep 25, 2009 at 07:06 AM
p.74 #20Canon EOS 7D Master thread
From CPN:
For enhanced precision, the central AF point features extra sensitivity when used with lenses having a maximum aperture of f/2.8 or faster thanks to the diagonal arrangement of the cross-type sensor - a feature first seen on the EOS 50D. This diagonal placement with the sensors forming an X-arrangement means that the f/2.8 sensors do not overlap with the f/5.6 sensors at the same point as they have a horizontal/vertical arrangement. This avoids any interference between the two sensors and offers both fast tracking performance from the f/5.6 sensor and extra precision from the f/2.8 sensor.
f/5.6 sensors are especially good at predictive tracking focus, however, the designers wanted to find a solution that also offers high focus accuracy as well. Increasing focus accuracy can be achieved by using a higher resolution focus sensor however, simply increasing the pitch, and therefore resolution, of a single line too much leads to a lower signal to noise ratio in terms of light levels recorded. To combat this, the centre top, centre middle and centre bottom sensors (when the camera is held in a landscape orientation) have a dual zigzag horizontal line sensor.
The dual zigzag arrangement solves this problem by using two focus sensors that are slightly offset from each other. By performing AF calculations based on the two lines sampling slightly different areas, the focusing resolution is improved leading to a more precise focus result while still retaining the responsive AF tracking.
Iguess you are both correct, sort of.
The centre point is f2.8 (enhanced precision), but there are three f5.6 points with 'dual zigzag' (more precise focus).
For enhanced precision, the central AF point features extra sensitivity when used with lenses having a maximum aperture of f/2.8 or faster thanks to the diagonal arrangement of the cross-type sensor - a feature first seen on the EOS 50D. This diagonal placement with the sensors forming an X-arrangement means that the f/2.8 sensors do not overlap with the f/5.6 sensors at the same point as they have a horizontal/vertical arrangement. This avoids any interference between the two sensors and offers both fast tracking performance from the f/5.6 sensor and extra precision from the f/2.8 sensor.
f/5.6 sensors are especially good at predictive tracking focus, however, the designers wanted to find a solution that also offers high focus accuracy as well. Increasing focus accuracy can be achieved by using a higher resolution focus sensor however, simply increasing the pitch, and therefore resolution, of a single line too much leads to a lower signal to noise ratio in terms of light levels recorded. To combat this, the centre top, centre middle and centre bottom sensors (when the camera is held in a landscape orientation) have a dual zigzag horizontal line sensor.
The dual zigzag arrangement solves this problem by using two focus sensors that are slightly offset from each other. By performing AF calculations based on the two lines sampling slightly different areas, the focusing resolution is improved leading to a more precise focus result while still retaining the responsive AF tracking.
Iguess you are both correct, sort of.
The centre point is f2.8 (enhanced precision), but there are three f5.6 points with 'dual zigzag' (more precise focus).
Sep 25, 2009 at 07:10 AM