Still have 1,000 or so shots to go through before I have a better idea on how it worked.
But in general the focus for sure seems a lot better. I have to get used to some of the controls as they are so much different. Nothing major, just will take some getting used to.
The shutter sounds really weird. Way different than any camera I have ever used. Almost sounds broken or loose. The few images I have looked at seem nice up to 50% then kinda fall apart after that so Im not liking that at all. Not sure what is going on yet, but like I said, I have a lot of images to go through so I hope I can figure it out?
The few images I have looked at seem nice up to 50% then kinda fall apart after that so Im not liking that at all. Not sure what is going on yet, but like I said, I have a lot of images to go through so I hope I can figure it out?
Maybe your AF needs calibration, either by micro focus adjustment, or by Canon Service. I'm finding my 5D3 to be just as enjoyable to pixel peep as my 5D2, and more often due to the better AF.
My 7D was a different story. Compared to the 5D2 (and 3), which isn't really a fair fight, viewing at 100% just wasn't usually satisfying. I had to learn to judge 7D images at 50%. My 5D3 is not like that at all.
-Jack
An "accurate" reproduction of a scene and a good photograph are often two different things.
Good looking MX shot! What do you mean when you say you don't like the 5D3 images at 50% or greater?
Its weird. It looks raserized? Not sure thats the right term for the look but its not smooth tones and the edges look very digital. Not a focus issue and its the same look at 100 to 400 asa.
Also got lots of errors. No error #, just an Err and it stopped working.
Turned if off and on then it was fine till it did it again. Only did it in auto focus mode, but not sure if that means anything as I was in AF mode most of the day testing all the new settings.
Its weird. It looks raserized? Not sure thats the right term for the look but its not smooth tones and the edges look very digital. Not a focus issue and its the same look at 100 to 400 asa
What raw converter are you using? Try shooting some L JPEG's and see how those look.
Its weird. It looks raserized? Not sure thats the right term for the look but its not smooth tones and the edges look very digital. Not a focus issue and its the same look at 100 to 400 asa.
Are you using DPP? If not, try it. If so, change the RAW sharpening method from "Unsharp Mask" to "Sharpness", I like that better.
Also got lots of errors. No error #, just an Err and it stopped working.
Turned if off and on then it was fine till it did it again. Only did it in auto focus mode, but not sure if that means anything as I was in AF mode most of the day testing all the new settings.
Sounds very fishy, I'd take/send the camera back to where you bought it. No errors with my 5D3, and the images are stunning at 100% in DPP.
-Jack
An "accurate" reproduction of a scene and a good photograph are often two different things.
Ha.
I was at the store doing just that (returning it) a little while ago.
Home now with the new one. Will do some testing later and let ya all know how it goes.
0
Matthew SavilleRegistered Users, Retired ModPosts: 3,352Major grins
edited April 16, 2012
Studio shot it's 2nd wedding with the 5D mk3, and for a second time the lead shooter decided there was a SIGNIFICANT issue with "focus confirmation hesitance". In fact this lead shooter even switched back to the mk2 for a little while just to be sure they weren't just imagining the difference.
Hopefully we just have a lemon? We'll approach Canon about the issue, since the studio is right down the street from CPS...
(Although, once again I'd like to emphasize, there is a SIGNIFICANT improvement in the keeper, in-focus rate. It just hesitates a little bit before confirming focus, which is noticeable enough to cause worries about missing candid moments...
Studio shot it's 2nd wedding with the 5D mk3, and for a second time the lead shooter decided there was a SIGNIFICANT issue with "focus confirmation hesitance"....
=Matt=
I can replicate the same issue, and here's the fix:
Use one of the TWO Single Point AF modes
Do NOT use any mode that requires the camera to select the AF point
Consider whether or not your comparing to situations in which the 5D2 would say "focused!", when it was really "close enough!"
Same goes for AI-Servo usage. I have not found the AF Point tracking to be all that impressive. I get better results in AI servo mode my using one of the Single Point or "cluster" AF Point modes. But I am hard to please. The D700 AF Point tracking didn't even make my cut. The D3s or 1DMK4 are the only bodies with AF point tracking that is really snappy enough to rely on...for me. The good news is the 5D3 has the processing power for this to work well, I would expect firmware updates to help out in this regard. Better written algorithms for the AF point tracking/selection would make things more responsive.
Has anyone tried the different shot priority and AI focusing mode combinations yet? All of the different AI focus modes? They do significantly change the behavior of it and most people seem to be going on the standard mode and not going over the other AF scenarios....
Better written algorithms for the AF point tracking/selection would make things more responsive.
The AI servo behavior settings are the algorithms changeable at our disposal... did you change those yet? It sounds like you've only tried the AF point selection modes and left out the AI servo behavior settings. Standard mode is a little slow and changing the settings dramatically changes the behavior.
A lot of these problems are in the dark so if it keeps up I imagine something indeed will have to be changed by Canon in order to reduce that delay in the dark. I've still yet to do a low light shoot, so hopefully soon I'll be able to do this first person. So far no problems though.
The AI/servo tracking modes are the algorithms changeable at our disposal... did you change those yet?
I am actually referring to One Shot usage, but the same applies for AI-Servo mode. Anytime I let the camera pick the focus point there is a perceptible lag in initial focus lock. The AF tracking modes all come into play AFTER the initial focus lock is gained. You can play with the Focus/Release priority options, but the real problem is just the decision making algorithm of the auto point selection system itself.
These are observations from in the field use. I think every camera has it's quirks, and the 5D3 has more than it's fair share cropping up these days. Early release, bugs, who knows, I am just stoked to be using and producing with my favorite Canon ever.
The Canon 7D also has a perceptible lag in AF lock in low light, and the 7D and 5D MKIII share a very similar, if not the same, metering module. The reason that this may be significant is that both the 7D and 5D MKIII tie the metering section with the focusing section.
The AF lag in low-light might be symptomatic to the metering module integration.
On the 7D, the metering system can slow down the frame rate to favor AF accuracy. There is even a vague reference in the 7D User Manual. (Bottom of page 93.)
"In low-light areas or indoors, the continuous shooting speed may become slower even if a fast shuttter speed is set"
Is there a similar mention in the 5D MKIII owners manual? (Like maybe page number 361?)
In low light I highly recommend the use of a flash with an AF Assist light, or at least an ST-E2. (The ST-E3-RT does not appear to have an AF Assist light.)
Studio shot it's 2nd wedding with the 5D mk3, and for a second time the lead shooter decided there was a SIGNIFICANT issue with "focus confirmation hesitance". In fact this lead shooter even switched back to the mk2 for a little while just to be sure they weren't just imagining the difference.
Were they shooting with all focus points active? Letting the camera decide where to focus? If so, that's more Artificial Intelligence that you have to wait for. This was a known characteristic with the 1D line as well. Shooting action in "Ring of Fire" mode was not advised.
-Jack
An "accurate" reproduction of a scene and a good photograph are often two different things.
0
Matthew SavilleRegistered Users, Retired ModPosts: 3,352Major grins
Were they shooting with all focus points active? Letting the camera decide where to focus? If so, that's more Artificial Intelligence that you have to wait for. This was a known characteristic with the 1D line as well. Shooting action in "Ring of Fire" mode was not advised.
I can replicate the same issue, and here's the fix:
Use one of the TWO Single Point AF modes
Do NOT use any mode that requires the camera to select the AF point
Consider whether or not your comparing to situations in which the 5D2 would say "focused!", when it was really "close enough!"
Same goes for AI-Servo usage. I have not found the AF Point tracking to be all that impressive. I get better results in AI servo mode my using one of the Single Point or "cluster" AF Point modes. But I am hard to please. The D700 AF Point tracking didn't even make my cut. The D3s or 1DMK4 are the only bodies with AF point tracking that is really snappy enough to rely on...for me. The good news is the 5D3 has the processing power for this to work well, I would expect firmware updates to help out in this regard. Better written algorithms for the AF point tracking/selection would make things more responsive.
Yep, we're testing in One-Shot, Single / Spot AF, and all the different types of metering modes just in case that has an effect as well. Still no dice.
I'm not sure exactly what you're talking about when you say "consider whether or not you're comparing to situations in which the 5D mk2 said "close enough", ...because that is indeed exactly how I'm testing this. The 5D mk2 locks nice and snappy, while the 5D mk3 takes it's time to confirm. Yeah, the 5D mk2 is just plain missing focus half the time in low light, HOWEVER, I'm also testing against a D700 and 24-70, which CAN nail shots every single time at the same zero-hesitation type of speeds that I'm used to. So for me, it is un-acceptable to just call it "the price you pay for greater accuracy". That is why I hope the camera is defective, and why I'm asking if any others who have both a mk2 and a mk3 can test and confirm the same issue.
I'd agree that the D700 might not be as good at tracking as the D3s or a 1-series Canon, but I spend most of my time just tracking brides as they walk up the aisle, not cheetas or NFL wide receivers, , ...and the rest of the time I'm in One-Shot / Single Focus.
I'm not sure exactly what you're talking about when you say "consider whether or not you're comparing to situations in which the 5D mk2 said "close enough", ...because that is indeed exactly how I'm testing this. The 5D mk2 locks nice and snappy, while the 5D mk3 takes it's time to confirm. Yeah, the 5D mk2 is just plain missing focus half the time in low light, HOWEVER, I'm also testing against a D700 and 24-70, which CAN nail shots every single time at the same zero-hesitation type of speeds that I'm used to. So for me, it is un-acceptable to just call it "the price you pay for greater accuracy". That is why I hope the camera is defective, and why I'm asking if any others who have both a mk2 and a mk3 can test and confirm the same issue.
=Matt=
It may very well be defective. I am very happy with the AF performance when using Spot AF in single shot or AI servo mode.
I agree, Matt. Nikon has set the bar for AF performance in DSLRs. If I set all my infatuation for the 5D3 aside, I'd say the D700 is probably still the better camera when it comes to simple, easy, dead reliable AF.
But, we don't shoot Canon because of the AF capability. The 5D2 is probably the most popular wedding pro DSLR in history, and it's AF was/is atrocious. We shoot Canon because we love the dreamy skin tones of their FF sensors, the gorgeous bokeh of the 1.2 glass, and in my case, the ability to get OOC perfect images. No post! I love the D700, but as a compete image makin' machine, the 5D3 takes the cake handily.
Remember to view at X3 size Only 4 photos at the moment and I got at least 30 more keepers... I'm tired so will edit later.
Love this camera. Yeah I could do this on the MKII, but the MKIII really is a whole world apart with AF selection/behavior and lower noise, even at ISO 400. The quiet shutter helps a ton, and the 8 second HDR makes workflow so smooth when you can skip the 10 minute edit session with 3rd party software! Things open up so much more with on-the-fly photographs.
Remember to view at X3 size Only 4 photos at the moment and I got at least 30 more keepers... I'm tired so will edit later.
Love this camera. Yeah I could do this on the MKII, but the MKIII really is a whole world apart with AF selection/behavior and lower noise, even at ISO 400. The quiet shutter helps a ton, and the 8 second HDR makes workflow so smooth when you can skip the 10 minute edit session with 3rd party software! Things open up so much more with on-the-fly photographs.
The dynamic range of the 5D3 sensor continues to impress me. Here's another sunset shot of the Cook Inlet in Anchorage, Alaska. Shot with the 24-70, S1 JPEG. This is out of camera, y'all.
Great shot. Standard picture style? Have you played around with the Landscape picture style? It also boosts blue sky.
Any reason you're shooting in S1 other than to conserve memory cards?
Emerald picture style. I am a huge fan of this profile right now. I am shooting S1 JPEG's because they are perfect for everything I need from my snapshots. Web sharing, prints up to 16x20. For serious work, I shoot RAW, but I am finding that I create more and share more by shooting nice looking in-camera JPEGs. No editing, converting or resizing, just drag and drop to share.
Look at those alley scenes, straight off the card!!
The video is absolutely stellar, and I don't even do video! It is really inspiring. Just from an objective point of view, for video, this makes the D800 look like garbage. Make sure you watch it in HD and its like, nothing is missed. Flying birds, text on shadowed concrete, even the flecks of minerals in it. Just wow.
Filmed on location in the remote village of Nondalton, Alaska. These video interviews are part of a growing body of work that I am contributing to the book project: "Where Water is Gold: Bristol Bay and the Pebble Mine" by Carl Johnson.
Visit the Facebook page for this project here: facebook.com/BristolBayPebble
Matthew SavilleRegistered Users, Retired ModPosts: 3,352Major grins
edited April 27, 2012
...Still waiting for someone to confirm or deny that in a side-by-side comparison, their mk3 hesitates to confirm focus in low light, One-Shot, Spot / Point AF, compared to the mk2... ANYBODY?????
Comments
Link to my Smugmug site
"You miss 100% of the shots you don't take" - Wayne Gretzky
Still have 1,000 or so shots to go through before I have a better idea on how it worked.
But in general the focus for sure seems a lot better. I have to get used to some of the controls as they are so much different. Nothing major, just will take some getting used to.
The shutter sounds really weird. Way different than any camera I have ever used. Almost sounds broken or loose. The few images I have looked at seem nice up to 50% then kinda fall apart after that so Im not liking that at all. Not sure what is going on yet, but like I said, I have a lot of images to go through so I hope I can figure it out?
Good looking MX shot! What do you mean when you say you don't like the 5D3 images at 50% or greater?
Maybe your AF needs calibration, either by micro focus adjustment, or by Canon Service. I'm finding my 5D3 to be just as enjoyable to pixel peep as my 5D2, and more often due to the better AF.
See this thread for a couple of my action shots with 100% crops:
http://www.dgrin.com/showthread.php?t=217762
My 7D was a different story. Compared to the 5D2 (and 3), which isn't really a fair fight, viewing at 100% just wasn't usually satisfying. I had to learn to judge 7D images at 50%. My 5D3 is not like that at all.
An "accurate" reproduction of a scene and a good photograph are often two different things.
Its weird. It looks raserized? Not sure thats the right term for the look but its not smooth tones and the edges look very digital. Not a focus issue and its the same look at 100 to 400 asa.
Also got lots of errors. No error #, just an Err and it stopped working.
Turned if off and on then it was fine till it did it again. Only did it in auto focus mode, but not sure if that means anything as I was in AF mode most of the day testing all the new settings.
What raw converter are you using? Try shooting some L JPEG's and see how those look.
Are you using DPP? If not, try it. If so, change the RAW sharpening method from "Unsharp Mask" to "Sharpness", I like that better.
Sounds very fishy, I'd take/send the camera back to where you bought it. No errors with my 5D3, and the images are stunning at 100% in DPP.
An "accurate" reproduction of a scene and a good photograph are often two different things.
this
I was at the store doing just that (returning it) a little while ago.
Home now with the new one. Will do some testing later and let ya all know how it goes.
Hopefully we just have a lemon? We'll approach Canon about the issue, since the studio is right down the street from CPS...
(Although, once again I'd like to emphasize, there is a SIGNIFICANT improvement in the keeper, in-focus rate. It just hesitates a little bit before confirming focus, which is noticeable enough to cause worries about missing candid moments...
=Matt=
My SmugMug Portfolio • My Astro-Landscape Photo Blog • Dgrin Weddings Forum
I can replicate the same issue, and here's the fix:
Same goes for AI-Servo usage. I have not found the AF Point tracking to be all that impressive. I get better results in AI servo mode my using one of the Single Point or "cluster" AF Point modes. But I am hard to please. The D700 AF Point tracking didn't even make my cut. The D3s or 1DMK4 are the only bodies with AF point tracking that is really snappy enough to rely on...for me. The good news is the 5D3 has the processing power for this to work well, I would expect firmware updates to help out in this regard. Better written algorithms for the AF point tracking/selection would make things more responsive.
The AI servo behavior settings are the algorithms changeable at our disposal... did you change those yet? It sounds like you've only tried the AF point selection modes and left out the AI servo behavior settings. Standard mode is a little slow and changing the settings dramatically changes the behavior.
A lot of these problems are in the dark so if it keeps up I imagine something indeed will have to be changed by Canon in order to reduce that delay in the dark. I've still yet to do a low light shoot, so hopefully soon I'll be able to do this first person. So far no problems though.
I am actually referring to One Shot usage, but the same applies for AI-Servo mode. Anytime I let the camera pick the focus point there is a perceptible lag in initial focus lock. The AF tracking modes all come into play AFTER the initial focus lock is gained. You can play with the Focus/Release priority options, but the real problem is just the decision making algorithm of the auto point selection system itself.
These are observations from in the field use. I think every camera has it's quirks, and the 5D3 has more than it's fair share cropping up these days. Early release, bugs, who knows, I am just stoked to be using and producing with my favorite Canon ever.
The AF lag in low-light might be symptomatic to the metering module integration.
On the 7D, the metering system can slow down the frame rate to favor AF accuracy. There is even a vague reference in the 7D User Manual. (Bottom of page 93.)
"In low-light areas or indoors, the continuous shooting speed may become slower even if a fast shuttter speed is set"
Is there a similar mention in the 5D MKIII owners manual? (Like maybe page number 361?)
In low light I highly recommend the use of a flash with an AF Assist light, or at least an ST-E2. (The ST-E3-RT does not appear to have an AF Assist light.)
Moderator of the Cameras and Accessories forums
Were they shooting with all focus points active? Letting the camera decide where to focus? If so, that's more Artificial Intelligence that you have to wait for. This was a known characteristic with the 1D line as well. Shooting action in "Ring of Fire" mode was not advised.
An "accurate" reproduction of a scene and a good photograph are often two different things.
Yep, we're testing in One-Shot, Single / Spot AF, and all the different types of metering modes just in case that has an effect as well. Still no dice.
I'm not sure exactly what you're talking about when you say "consider whether or not you're comparing to situations in which the 5D mk2 said "close enough", ...because that is indeed exactly how I'm testing this. The 5D mk2 locks nice and snappy, while the 5D mk3 takes it's time to confirm. Yeah, the 5D mk2 is just plain missing focus half the time in low light, HOWEVER, I'm also testing against a D700 and 24-70, which CAN nail shots every single time at the same zero-hesitation type of speeds that I'm used to. So for me, it is un-acceptable to just call it "the price you pay for greater accuracy". That is why I hope the camera is defective, and why I'm asking if any others who have both a mk2 and a mk3 can test and confirm the same issue.
I'd agree that the D700 might not be as good at tracking as the D3s or a 1-series Canon, but I spend most of my time just tracking brides as they walk up the aisle, not cheetas or NFL wide receivers,
=Matt=
My SmugMug Portfolio • My Astro-Landscape Photo Blog • Dgrin Weddings Forum
It may very well be defective. I am very happy with the AF performance when using Spot AF in single shot or AI servo mode.
I agree, Matt. Nikon has set the bar for AF performance in DSLRs. If I set all my infatuation for the 5D3 aside, I'd say the D700 is probably still the better camera when it comes to simple, easy, dead reliable AF.
But, we don't shoot Canon because of the AF capability. The 5D2 is probably the most popular wedding pro DSLR in history, and it's AF was/is atrocious. We shoot Canon because we love the dreamy skin tones of their FF sensors, the gorgeous bokeh of the 1.2 glass, and in my case, the ability to get OOC perfect images. No post! I love the D700, but as a compete image makin' machine, the 5D3 takes the cake handily.
http://overfocused.smugmug.com/Bugs/Butterfly-Fields/22502197_FW4mVz#!i=1799865843&k=KnK8mRt
Remember to view at X3 size
Love this camera. Yeah I could do this on the MKII, but the MKIII really is a whole world apart with AF selection/behavior and lower noise, even at ISO 400. The quiet shutter helps a ton, and the 8 second HDR makes workflow so smooth when you can skip the 10 minute edit session with 3rd party software! Things open up so much more with on-the-fly photographs.
Congrats!
Who is wise? He who learns from everyone.
My SmugMug Site
Notice the rainbow on the left side of the image?
WHAT DOES IT MEAN??
Great shot. Standard picture style? Have you played around with the Landscape picture style? It also boosts blue sky.
Any reason you're shooting in S1 other than to conserve memory cards?
An "accurate" reproduction of a scene and a good photograph are often two different things.
Emerald picture style. I am a huge fan of this profile right now. I am shooting S1 JPEG's because they are perfect for everything I need from my snapshots. Web sharing, prints up to 16x20. For serious work, I shoot RAW, but I am finding that I create more and share more by shooting nice looking in-camera JPEGs. No editing, converting or resizing, just drag and drop to share.
:jawdrop
Look at those alley scenes, straight off the card!!
The video is absolutely stellar, and I don't even do video! It is really inspiring. Just from an objective point of view, for video, this makes the D800 look like garbage. Make sure you watch it in HD and its like, nothing is missed. Flying birds, text on shadowed concrete, even the flecks of minerals in it. Just wow.
This was my first video shoot with the 5D3. Monopod, with 70-200 II. Shotgun mic on boom pole into Zoom H4N.
<!-- This version of the embed code is no longer supported. Learn more: https://vimeo.com/help/faq/embedding --> <object width="800" height="450"><param name="allowfullscreen" value="true" /><param name="allowscriptaccess" value="always" /><param name="movie" value="http://vimeo.com/moogaloop.swf?clip_id=41117285&force_embed=1&server=vimeo.com&show_title=1&show_byline=1&show_portrait=1&color=00adef&fullscreen=1&autoplay=0&loop=0" /><embed src="http://vimeo.com/moogaloop.swf?clip_id=41117285&force_embed=1&server=vimeo.com&show_title=1&show_byline=1&show_portrait=1&color=00adef&fullscreen=1&autoplay=0&loop=0" type="application/x-shockwave-flash" allowfullscreen="true" allowscriptaccess="always" width="800" height="450"></embed></object>
Filmed on location in the remote village of Nondalton, Alaska. These video interviews are part of a growing body of work that I am contributing to the book project: "Where Water is Gold: Bristol Bay and the Pebble Mine" by Carl Johnson.
Visit the Facebook page for this project here: facebook.com/BristolBayPebble
=Matt=
My SmugMug Portfolio • My Astro-Landscape Photo Blog • Dgrin Weddings Forum