scowl wrote:
If that is the case then that means the months and months of Canon's Mark IV testing failed since they were not able to find and solve all the tracking problems until weeks after the camera had been released, despite the horrific Mark III fiasco which began years ago.
No it doesn't. Maybe it means that the camera performed at some level and someone believed it could perform better. The previous "good" level does not mean that it failed. Moreover, as thorough as beta testing is, there are always unforeseen circumstances where the camera yielded performance that someone wasn't happy about. That doesn't mean it was bad performance, but maybe not at the level they wanted.
chez wrote:
It costs an organization a lot of money releasing a new version of firmware. Canon would not do this so close to the original release of the camera unless it was to fix an issue.
Unless the software engineers are contract employees, the cost is probably built into the cost of employment. It's not like the software engineers stop working just because the camera was released.
To take this fascinating and elevated discussion in a slightly different direction, does anybody know why Canon, after having jumped through so many hoops with RG on the Mark III meeting with him, flying in technicians and bodies for additional testing by and with him, etc. etc. etc. didn't involve him in the pre-release testing of the Mark IV?
kakomu wrote:
No it doesn't. Maybe it means that the camera performed at some level and someone believed it could perform better. The previous "good" level does not mean that it failed. Moreover, as thorough as beta testing is, there are always unforeseen circumstances where the camera yielded performance that someone wasn't happy about. That doesn't mean it was bad performance, but maybe not at the level they wanted.
Unless the software engineers are contract employees, the cost is probably built into the cost of employment. It's not like the software engineers stop working just because the camera was released.
I manage a line of products and I know how much it costs the company to do a firmware release. It is not trivial. If a company does not need to do a release, they won't. Canon releasing new firmware right on the heels of the product release only means they found an issue that the firmware corrected. The cost of just releasing firmware without any added functionality is just way too costly unless it fixes a problem.
Nill Toulme wrote:
To take this fascinating and elevated discussion in a slightly different direction, does anybody know why Canon, after having jumped through so many hoops with RG on the Mark III meeting with him, flying in technicians and bodies for additional testing by and with him, etc. etc. etc. didn't involve him in the pre-release testing of the Mark IV?
Nill
Good question. Seems like he would be the ultimate pre-release tester to have on your team.
alundeb wrote:
What do we know about the situation?
The player could well have started moving or accelerating after the first frame.
No. If you look closely at the first frame her left leg is behind her, her right leg is in front of her, and she is already dribbling the ball -- she was already moving towards the camera.
With risk of embarrasing myself here: Is it expected that an AF system will track random motion between frames 100%?
I would call this particular sequence the absolute opposite of "random motion". This is a basketball player dribbling to the basket during practice before a game with nothing intruding on the focus area.
Looking through all the basketball images, I found 13 of 169 "useless". If that is the worst situation and the worst results you can find, I need to redefine my standards.
I would not turn in 56 of the 169 shots. Perhaps your boss is less picky about focus. I have to assume that any shot I turn in might end up on the front page of a media guide. I would hate to have a blurry shot with my name on it.
kakomu wrote:
No it doesn't. Maybe it means that the camera performed at some level and someone believed it could perform better.
Unless Canon firmware development is complete anarchy, someone couldn't just "believe" it could "perform better". They had to identify the exact problem (or whatever word you want to call it) and prove that it performed better after the change, then they had to justify why the change demands a new firmware release.
It's possible that Canon's firmware development is anarchy and random engineers are coming up with half-ass reckless improvements to imagined deficiencies. If that's the case then I'll be switching to Nikon.
The previous "good" level does not mean that it failed.
If there was no failure than how can they know for certain there was any improvement? These things have to be quantifiable and provable in software development otherwise a company will have no control over the quality of their product. I speak from two decades of experience, some of it very unpleasant.
Moreover, as thorough as beta testing is, there are always unforeseen circumstances where the camera yielded performance that someone wasn't happy about. That doesn't mean it was bad performance, but maybe not at the level they wanted.
Unless Canon has absolutely no consistent development process, this means someone had to identify and qualify what they weren't "happy about", investigate it, come up with an "improvement", and prove that the change made them "happy".
Software development does not succeed with vague descriptions like "level" and "happy". That leads to tweaking, false assumptions, reckless changes and customers fleeing to the competition.
All I can still see is a comparsion of the two lastest "Pro" Cameras on a site that's sponsered by Nikon, and it's saying the Mark IV isn't all that, but, the D3 is.....sorry I'm gullible, but, not that gullible that I will take RG's assesment as gosple. I've had one session with the Mark IV in horrible midday side lit light and the camera gave me my best images for those conditions related to surfing than any other camera I've used...That's 10 years of shooting surf with a DLSR......but, feel free to run around yelling "the Sky is falling, the Sky is falling....." based on Galbraith
Again, something doesn't need to fail for an upgrade to be offered. I cannot quantify the metrics Canon uses to denote quality of the AF system. All I can say is that firmware is frequently in a constant state of development before and after a product is released.
Why, because I sometimes choose to disagree with people who do nothing but bitch, whine and complain about every last thing that they take a personal dislike to and then do their damnedest to get us to feel the same way about?
On that basis, being called a fanboy (even though I'm not - I've changed brands before and would again if it was right for me to do so) is a compliment.
kakomu wrote:
Again, something doesn't need to fail for an upgrade to be offered. I cannot quantify the metrics Canon uses to denote quality of the AF system. All I can say is that firmware is frequently in a constant state of development before and after a product is released.
kakomu wrote:
Again, something doesn't need to fail for an upgrade to be offered.
Saying it over and over does not make it true. You clearly do not work in software development.
To make a software or firmware change in any reputable company, a need must be identified, qualified, justified and testable. The need may be a new feature or to fix a problem. To prove that 1.0.6 "enhances" AF tracking performance as Canon carefully worded it, they had to identify a situation where it repeatedly and consistently failed and prove that it does not fail in that situation with the new firmware. If they do not do this, obviously they can't be sure 1.0.6 "enhanced" anything.
If they are not following this industry-standard procedure then they are only pulling ideas out of their butts, throwing them at customers and letting them do their work. This is not what reputable companies do.
I cannot quantify the metrics Canon uses to denote quality of the AF system. All I can say is that firmware is frequently in a constant state of development before and after a product is released.
Yes, in the industry we call those bug fixes. We identify problems, find solutions, test them and send out fixes.
Yes, our managers might call them "enhancements" for political reasons. I once worked on a fix that management ended up saying "enhanced system reliability" because our customers were tired of their machines crashing. Everyone knew the description was a joke because machines aren't supposed to crash.
In this industry you have to be naive to think that a firmware fix that adds no features is an "enhancement". These are the things we laugh about and send to Scott Adams to put in his Dilbert cartoons.
ftemoto wrote:
I think it's worth keeping it in perspective and deciding for yourself whether the rig does your subject well. I didn't shoot the MkIII with subjects that caused RG the issues he experienced, and rarely droned long bursts, and was pretty satisfied with the results.
So far, I've felt more comfortable sooner with the handling and results going from the III to the IV that I did going from the II to the III. That's no doubt because the jump from the II to III included a big interface change, but it's also because I had a hefty amount of paranoia and confirmation bias with all the internet hysteria over the III's performance. I'm a little more comfortable this time relying on my own experience without second guessing based on RG or folks on the boards if their reports are at odds with my experience.
So, since RG says soccer is the Waterloo for this new body, and I shoot a fair bit of it, we'll see how it goes this weekend....Show more →
Bingo! Getting input from others shooting soccer is exactly what I'm looking for. I'll very interested in what you discover. And BTW: here's hoping you come back being impressed
"When some memory cards were inserted in the camera, CHA was displayed in the top control panel and images could not be captured. This issue has been resolved."
Your Nikon camera couldn't take pictures with certain memory cards. Now it can! Being able to take pictures is one heck of a "feature enhancement" for a camera!
Be aware that we cannot judge which AF points that the camera used to track to subject, if you download and check the files.
This is what GB writes:
With some multi-point configurations, including C.Fn III-8-2 (Surrounding AF Points), the EOS-1D Mark IV's metadata doesn't show which specific point(s) in a group were used for the AF calculation for that frame.
The downloadable photos show where the points were positioned, then, but there's no way of knowing which the camera actually chose.
scowl wrote:
...In this industry you have to be naive to think that a firmware fix that adds no features is an "enhancement". These are the things we laugh about and send to Scott Adams to put in his Dilbert cartoons.
Totally true. I mean, come on, does anyone really buy "enhancement"?
Why doesn't Canon offer a firmware enhancement to upgrade the 500D's video capabilities to match those of the 550D's? Now that would be an "enhancement".
