fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1       2              4       5       6       end
  

Archive 2009 · Pre-production 1Dmk4 tests with Ai-servo

  
 
SLD
Offline
• • • • •
Upload & Sell: On
p.3 #1 · Pre-production 1Dmk4 tests with Ai-servo


rprouty wrote:
I agree with everyone here..


I agree with you

Serious, Look at the second test, the RING OF FIRE AI SERVO worked pretty good on the lady's shoulder as it's closer to the camera then her head, lol....



Nov 01, 2009 at 10:22 AM
scalesusa
Offline
• • • •
Upload & Sell: Off
p.3 #2 · Pre-production 1Dmk4 tests with Ai-servo


My first shots from my 7D were that way as well, I just set it in full auto mode with AI Servo and all focus points as a quick attempt to see what it could do. It was a disaster.

As soon as I took the time to set the AF modes properly, it was a whole different story. I found it nearly impossible to focus on what I wanted in a busy scene with all points on and AI servo.

I never attempted to show those first photos to anyone or complain, because I knew I had to learn how to use the new AF system. Once i experimented with the different modes and learned to use them, it was perfect every time



Nov 01, 2009 at 10:26 AM
digitalbug30d
Offline
• • • •
Upload & Sell: Off
p.3 #3 · Pre-production 1Dmk4 tests with Ai-servo


According to the 30d manual AI servo is used for moving subjects when the focusing distance keeps changing...so people are wrong here since AI servo works the same in all canon DSLRs even had to look in my eos3 manual yep pretty much the same...still AI servo isnt what is recommended with the above shots no matter your opinions read the manuals kids...thanks


Nov 01, 2009 at 10:28 AM
Tom_W
Offline
• • • • •
Upload & Sell: Off
p.3 #4 · Pre-production 1Dmk4 tests with Ai-servo


Series #1 worked exactly as designed - Canon cameras are programmed to keep a watchful eye on Nikon and this copy did just that.

Seriously, taking a pre-production demo camera whose settings are totally unknown and trying to use it to define a possible issue is just silly. There'll be plenty of copies out in the real world in another month or so. Then we can see how well it works. In the meantime, this thread is DPR material.



Nov 01, 2009 at 11:03 AM
garyvot
Offline
• • • • •
Upload & Sell: On
p.3 #5 · Pre-production 1Dmk4 tests with Ai-servo


scalesusa wrote:
My first shots from my 7D were that way as well, I just set it in full auto mode with AI Servo and all focus points as a quick attempt to see what it could do. It was a disaster.

As soon as I took the time to set the AF modes properly, it was a whole different story. I found it nearly impossible to focus on what I wanted in a busy scene with all points on and AI servo.

I never attempted to show those first photos to anyone or complain, because I knew I had to learn how
...Show more

Yes, I've noticed this too. But the problem here is that the camera does not seem to behave as designed or intended: when area-AF is enabled, the camera should focus on the nearest subject within the AF grid.

In the first sequence above, the camera appears to be grabbing focus at random on either a background subect, the foreground subject, or something in-between.

One should NOT have to use single-point AF exclusively in AI-Servo shooting to obtain consistent results. There are legitimate reasons to use area-AF when focus tracking, but the camera must produce consistent, predictable behavior.



Nov 01, 2009 at 11:04 AM
Tom_W
Offline
• • • • •
Upload & Sell: Off
p.3 #6 · Pre-production 1Dmk4 tests with Ai-servo


n0b0 wrote:
You probably get flamed because you worded the title like those attention grabbing sensational headlines which article usually turn out to be nothing important.


+1



Nov 01, 2009 at 11:06 AM
garyvot
Offline
• • • • •
Upload & Sell: On
p.3 #7 · Pre-production 1Dmk4 tests with Ai-servo


keithreeder wrote:
That's crap, Fred.

This isn't about "defending Canon", it's about being pig-sick of uninformed, knee-jerk allegations of problems without anything whatsoever to back the allegations up, and, as in this case, a clear lack of any objective basis for the accusation.

Look again that the thread title. It is clearly implying that there are indeed AF problems with the 1D Mk IV (adding a question mark doesn't change what is being suggested), and this thread just looks like another example of someone desperate to be the first with bad news.

I wouldn't dream of starting a thread like this on the basis of
...Show more

Keith, respectfually I have to disagree with you here. Based on my experience with EOS cameras, this camera does not appear to be working as expected, especially in the first sequence. I'd say the evidence for this is objectively and fairly presented.

I don't get the impression the OP is trying to be inflammatory or is trolling (and we know who some of the trolls are on this board).

You are right, of course, that one example taken in an uncontrolled setting does not "prove" anything about the Mark IV autofocus yet. But based on Canon's track record it's certainly legitimate for people to question the behavior of the camera.

Edited on Nov 01, 2009 at 11:27 AM · View previous versions


Nov 01, 2009 at 11:10 AM
Tom_W
Offline
• • • • •
Upload & Sell: Off
p.3 #8 · Pre-production 1Dmk4 tests with Ai-servo


rprouty wrote:
I agree with everyone here..


And I agree with nobody here...

... except you!



Nov 01, 2009 at 11:11 AM
Schlotkins
Offline
• • • •
Upload & Sell: Off
p.3 #9 · Pre-production 1Dmk4 tests with Ai-servo


Any information on shutter speed, ISO, f-stop etc? To me it looks like the camera is tending to grab onto the higher constract subjects (nikon and coke can). I tried to used my 1D MKII under poor lighting with all 45 AF points enabled one time and it exhibited the same behavior. Not saying this is how it's 'supposed' to work, but I can imagine the camera being progammed to do that when all AF points are active.

I should also note I agree it's way too early to say anything about AF performance. Canon would have to be absolutely nutty to not have the AF working properly out the door this time.

Chris



Nov 01, 2009 at 11:23 AM
garyvot
Offline
• • • • •
Upload & Sell: On
p.3 #10 · Pre-production 1Dmk4 tests with Ai-servo


Schlotkins wrote:
Any information on shutter speed, ISO, f-stop etc? To me it looks like the camera is tending to grab onto the higher constract subjects (nikon and coke can).
Chris


Agreed, but I think it should be biased to focus on the nearest subject, and once focus is acquired, it is not supposed to jump around (much) from frame to frame.

Would be interesting to know what the AI Servo tracking sensitivity and 1st/2nd image priority CFs were set to.



Nov 01, 2009 at 11:30 AM
AGeoJO
Offline
• • • • • • •
Upload & Sell: On
p.3 #11 · Pre-production 1Dmk4 tests with Ai-servo


Normally, Canon or Nikon and maybe other manufacturers would even let you shoot (using your own CF/SD card, that is) with a preproduction model at a trade show. For the very reason that some thing like this would hit the internet. Well, it is not difficult to be sneaky and do that though.... . Even a lowly 5D could pull that off, I would say. Let's face it, Canon is not that dumb to make a similar mistake twice . Get real, folks!


Nov 01, 2009 at 11:32 AM
Adam L
Offline
• •
Upload & Sell: Off
p.3 #12 · Pre-production 1Dmk4 tests with Ai-servo


Tom_W wrote:
In the meantime, this thread is DPR material.


And so is the reaction to it.



Nov 01, 2009 at 11:37 AM
vpk24_astro
Offline
• •
Upload & Sell: Off
p.3 #13 · Pre-production 1Dmk4 tests with Ai-servo


I think you guys need to be careful about talking about what the camera is supposed to do. All AF sensors use the same physics to detect focusing distance - its the software algorithm that changes from mode to mode. If I were designing a camera to shoot action sports with and I put in a mode designed specifically to shoot action and I had enough computing power (you don't need much it turns out), I'd be using some sort of Bayesian probability algorithm - i.e. what should I be expecting to focus track most of the time? Football/soccer/any kind of foot running players move at roughly the same speed at full tilt and depending on which AF points trigger, I can determine direction. Once I've done that, it is easy to predict where my focus point should be in 3-D. I can incorporate corrections for the person slowing down or coming to dead stop instantaneously - my point is this is stuff that follows predictable routines in a certain sense. Canon probably spends a lot of money fine tuning the parameters in the software which is what you pay for in a 1-series body.
Now, if I use a Bayesian algorithm designed to track fast moving subjects under the usual conditions of various sports on a man walking across the room, I'd expect errors because the processor will have a hard time believing its tracking correctly since it expects the guy to be going a lot faster.
I'm not saying this is what Canon does - I don't work for them or any other camera company, but it's probably how it works give processing power and requirements. Just a thought.



Nov 01, 2009 at 01:01 PM
digitalbug30d
Offline
• • • •
Upload & Sell: Off
p.3 #14 · Pre-production 1Dmk4 tests with Ai-servo


Not this again....


Nov 01, 2009 at 02:22 PM
Velu01
Offline
• • • •
Upload & Sell: Off
p.3 #15 · Pre-production 1Dmk4 tests with Ai-servo


I'd say, change the topic's title into "I was one of the very few who got a 1DIV in my hands but I wonder why I got these results !?".

Ai servo ... so that's not suitable to shoot stationary subjects hah !? .... ( some use it to shoot portraits while using large aperture ( f1.2) in order to increase the numbers of keepers ) ...

Rgds
Velu




Nov 01, 2009 at 02:49 PM
Sahid Limon
Offline
• • • •
Upload & Sell: Off
p.3 #16 · Pre-production 1Dmk4 tests with Ai-servo


Wow, I had no idea that this was the reaction this thread was going to create. Being called the next "Rob Galbraith"... just wow. Daryl, I'm still impressed how you manage to know everything that I was doing about the test, without ever being there or seeing it in person. I should ask you for predictions in the stock market. Oh, and just so you know, the subject can move up, down, side to side, and the Ai-servo should still be able to track the subject. Direction and orientation shouldn't be an issue here.

Anyway, I am sorry to all of those who this thread or the title has injured or cause any suffering. I was simply trying to present some data of the new camera, and compare results from anyone else who had experience with the test models, not to cause a riot among the forums. I've been on this forum for some time, and the last thing I need is to be labeled a troll on here. It's sad that people will jump on the messenger and not even look at the message, but that seems to be the case here. Just because I posted these images, does NOT mean I wasn't impressed with the camera. Remember, I am trying to buy one... so naturally, I would like to know everything about it before I drop the $5k. I did high ISO tests, and general focusing and fps tests, all which were spot on, and actually exceeded all the current 1D series. I may be a Canon fanboy such as some of the posters here, but at least I can realize that any company can have a bad apple, or make a mistake in their production from time to time. I tested ONE pre-production body from the 3 they had at the show, and did not get the results I wanted. Is that my fault? Could be? Is that the cameras fault? Could be as well. I was simply presenting the information and trying to verify with others results if it was my mistake... not blaming the entire stock of cameras. Oh, and just to clarify... the Canon representative told me to put in my own CF card to test the camera. Another person asked if Canon were giving out demo CF cards, and the representatives response was the following: "You didn't bring a CF card? You always bring a good CF cards to these things."

Anyway, last thing I want is a hoard of attack on me for simply asking a question of a pre-production camera (which I should say was shown only a month before the actual release, so it should be solid). I'm a practical guy, and rather have things go peacefully. I am sorry to all those who were offended by this thread, or lost sleep over it. I hope that at least to some, the information was helpful and served a purpose. If any of you are considering actually buying this camera, then I suggest you test it out when you get it. Take what you will from this thread. I changed the title, but I don't think that will change anyone's stand on the subject. Just to avoid any other unnecessary arguments, I am going to send the moderator a message and ask him to go over it, and either lock or delete this thread if he sees fit. I guess an abundance of information can be a bad thing as well. Thank you all for your input.



Nov 01, 2009 at 03:01 PM
Nick Nishizaka
Offline
• • •
Upload & Sell: Off
p.3 #17 · Pre-production 1Dmk4 tests with Ai-servo


darryn patch wrote:
Pilot error without a doubt 100%, The scenario that you've shot in using AI tells me you really haven't got much of an idea of how AF works or what mode of AF to use in which situations. Would that be correct? yes.

I'll give you 2 bobs worth of advice for free, Ai Servo is solely designed for moving subjects. not stationary objects.

Go and educate yourself on AF and then pick up the camera and try again.


What's ironic about your post is that the subject was indeed moving and you missed that part.

You can take that 2 bobs of advice and...well you know the rest champ.



Nov 01, 2009 at 03:43 PM
Nick Nishizaka
Offline
• • •
Upload & Sell: Off
p.3 #18 · Pre-production 1Dmk4 tests with Ai-servo


PetKal wrote:
One way or another, I am curious to find out what the 1DMk IV performance will be under really challenging Servo AF conditions, such as that shown on the image below, which IMO goes a bit beyond the difficulty of capturing the Galbraith's runner or shoppers browsing in a store.. It is not only the small target cross-section and fast head-on approach which poses a problem, the background is also a veritable AF magnet.
If Mk IV can help me with shots like those, then I'd consider buying it.
The first shot is full frame taken with 1D MkIIN +200mm
...Show more

A duck flying straight at you? Honestly, is that more challenging that a runner coming right at you?



Nov 01, 2009 at 03:47 PM
Nick Nishizaka
Offline
• • •
Upload & Sell: Off
p.3 #19 · Pre-production 1Dmk4 tests with Ai-servo


Sahid Limon wrote:
Wow, I had no idea that this was the reaction this thread was going to create. Being called the next "Rob Galbraith"... just wow. Daryl, I'm still impressed how you manage to know everything that I was doing about the test, without ever being there or seeing it in person. I should ask you for predictions in the stock market. Oh, and just so you know, the subject can move up, down, side to side, and the Ai-servo should still be able to track the subject. Direction and orientation shouldn't be an issue here.


This is FM...aka DPReview #2. Don't be so surprised.



Nov 01, 2009 at 03:48 PM
David Manning
Offline
• • •
Upload & Sell: Off
p.3 #20 · Pre-production 1Dmk4 tests with Ai-servo


Wow this camera does suck, but I'll wait for my pre-ordered copy to see for myself!


Nov 01, 2009 at 03:55 PM
1       2              4       5       6       end




FM Forums | Canon Forum | Join Upload & Sell

1       2              4       5       6       end
    
 

Welcome back
Log in to your account