EOS 5D Mark III Firmware Coming April 30, 2013

[ddashti]

What exactly is wrong with the autofocus at f/8?

[K-amps]

What exactly is wrong with the autofocus at f/8?

Does not exist where min aperture falls above F5.6 with phase AF at the moment... this change will bump it up to F8. This not you setting aperture in cam, but open wide (what the cam does when assessing AF). When capturing the image, the Cam will reduce F to whatever level you chose... the limitation still is what the value is wide open.

[jm977]

From Canon USA's site:

Product Advisories
Thank you for using Canon products.

Affected Products
EOS 5D Mark III Digital SLR Camera
EOS-1D X Digital SLR Camera

Details
Depending upon the shooting conditions, it may take slightly longer for the EOS 5D Mark III and EOS-1D X digital SLR cameras to acquire focus when using the Speedlite's AF Assist Beam, compared with that of the EOS 5D Mark II and EOS-1D Mark IV digital SLR cameras using the Speedlite's AF Assist Beam.

We are developing a solution to enhance the focusing feature for the EOS 5D Mark III and EOS-1D X and plan to release a firmware update based on following estimated schedule:

    EOS 5D Mark III: (estimated) Late April, 2013
    EOS-1D X: (estimated) Late May, 2013

So this update does not address the issue with AF-assist beam lag. Is that a separate update?

I was wondering this myself as the Canon website does not specifically mention this.  I thought that this was one of the big driving issues for a firmware update so I find it odd it's not in included.

First, it was f/8 AF then the clean HDMI out.  Canon has also said they will update the AF-Assist issue, but it came later, and will likely be rolled up in a furure update.

Absolutely.  However, the confusion comes from the Canon USA press release which does not mention the AF assist issue. 

"New Firmware Update For Canon EOS 5D Mark III Digital SLR Camera Provides Uncompressed HDMI Output Support And Improved AF Performance

LAKE SUCCESS, N.Y., October 23, 2012 – Canon U.S.A., Inc., a leader in digital imaging solutions, today announced a new firmware update for the EOS 5D Mark III Digital SLR camera that significantly improves the camera's performance and usability. In response to requests from professionals working in the fields of cinema and television production, the firmware update enables the use of uncompressed HDMI Output support, making possible more efficient video editing and monitoring procedures. Additionally, the upgrade supports the advanced needs of photographers through improved AF performance when capturing still images.

Uncompressed HDMI Output Support

When shooting video, HDMI Output makes possible the recording of high-definition uncompressed video data (YCbCr 4:2:2, 8 bit) from the EOS 5D Mark III to an external recorder via the camera's HDMI terminal. This, in turn, facilitates the editing of video data with minimal image degradation for greater on-site workflow efficiency during motion picture and video productions. Additionally, video being captured can be displayed on an external monitor, enabling real-time, on-site monitoring of high-definition video during shooting.
Improved AF Functionality

Even when the EOS 5D Mark III is equipped with an extender and lens making possible a maximum aperture of f/8, the firmware update supports AF employing the camera's central cross-type points (currently compatible with maximum apertures up to f/5.6). Accordingly, the update will allow users to take advantage of AF when shooting distant subjects, benefitting sports and nature photographers, particularly when using telephoto lenses.

The new firmware update will be available, at no charge, in April 2013 from the Canon U.S.A. website and can be downloaded by end users or through Canon Factory Service Centers"
I just find it odd that it's left out of the press release since such a big deal was made of it.

[RGomezPhotos]

Good stuff.  If the big MP camera that we've heard about isn't 90% of what rumors spec it to be, I'll probably go with the 5D MKIII...

[Barrfly]

 I'm slightly confused, maybe someone can clear this f8 focusing thing up for me.
When I use my Sigma 50-500mm f/4.5-6.3 APO DG OS HSM with a 2X converter my widest usable aperture is effectively f13 and the auto focusing works fine.

 Am I missing something ?

[East Wind Photography]

Yep, the issue is with Canon lenses and Canon teleconverters.  Some 3rd party extenders have worked all along.

I'm slightly confused, maybe someone can clear this f8 focusing thing up for me.
When I use my Sigma 50-500mm f/4.5-6.3 APO DG OS HSM with a 2X converter my widest usable aperture is effectively f13 and the auto focusing works fine.

 Am I missing something ?

[Roger Jones]

Anyone know of some good (cheaper) video recorders to use with that clean HDMI out that would be better than the compressed video on the SD card?

yes.  Atomos Ninjia2 is a bit on the expensive side; the alternative will be Blackmagic Design HyperDeck Shuttle 2 SSD Video Recorder at $327.75, which does not have a monitor, but will do the work.

http://www.bhphotovideo.com/c/product/841023-REG/Blackmagic_Design_HYPERD_PT2_HyperDeck_Shuttle_2_SSD.html

Lowest price may not lead to the lowest TCO. The shuttle only works with SSDs while the ninja is supposed to work with spinning disks which will make media costs much lower. It would be practical to use the disks to archive camera masters. Ninja has a built in monitor, if you already have one its not an issue but if you don't and you can get by with the one on the ninja thats a savings as well. Also it remains to be seen exactly what the 5dIII will output on the HDMI. Ninja can detect the cadence on a 60i signal and record the stream as 24p. Not sure if the shuttle will be able to do that. If not you will have to de-interlace as a post process which adds time to the work flow. I was tempted by the shuttle and I already have monitors but I think I'm going to opt for the ninja after the new firmware ships and the field reports come in.

[LetTheRightLensIn]

Well...theydid say, "April"....

b&

Arggh why do they always hold it every single time to the last day of the month. It would've been a cool test of external video recorders to record falling cherry/apple blossoms. April 30th will be JUST a week too late. damn can't believe they pulled that last day of the month thing again.

I guess it will be in time for the second of spring migration (f/8 AF) at least.

And it is awesome that were are getting these two upgrades (why they didn't also remove the AutoISO shutter speed crippling I have no clue though).

[Krob78]

Well...theydid say, "April"....

b&

Arggh why do they always hold it every single time to the last day of the month. It would've been a cool test of external video recorders to record falling cherry/apple blossoms. April 30th will be JUST a week too late. damn can't believe they pulled that last day of the month thing again.

I guess it will be in time for the second of spring migration (f/8 AF) at least.

And it is awesome that were are getting these two upgrades (why they didn't also remove the AutoISO shutter speed crippling I have no clue though).

I'm glad I didn't expect it until  May!  I was hoping for April 30th, as usual but secretly praying for April 15th!  Oh well, it's as I suspected... that's not too bad, other than migrations as you state...

[Rienzphotoz]

Yippy ... finally, AF at f/8 is less than a couple of weeks away ... been waiting for it patiently - better late than never.

[Krob78]

Yippy ... finally, AF at f/8 is less than a couple of weeks away ... been waiting for it patiently - better late than never.

+1 Indeed! 

[pwp]

Hope they fix the AF points so that the active point is red

Oh yes please...

-PW

[East Wind Photography]

I doubt that will happen, ever, due to the 5D3 design.  In order for it to happen they have to take away metering speed.  They already stated that the red AF points affect the metering and they dont have a way to deal with that in firmware.

Hope they fix the AF points so that the active point is red

Oh yes please...

-PW

[Krob78]

They have stated that, true enough.  Doesn't it seem a bit hard to believe though?  Seems like there should be some type of fix or code that can be written into the firmware.  Seemed odd to me when they made that statement, obviously I'm not a software, firmware, hardware engineer though, so my inability to understand the why's or why not's regarding this issue are somewhat limited I'm sure! lol!

I doubt that will happen, ever, due to the 5D3 design.  In order for it to happen they have to take away metering speed.  They already stated that the red AF points affect the metering and they dont have a way to deal with that in firmware.

Hope they fix the AF points so that the active point is red

Oh yes please...

-PW

[JonAustin]

I doubt that will happen, ever, due to the 5D3 design.  In order for it to happen they have to take away metering speed.  They already stated that the red AF points affect the metering and they dont have a way to deal with that in firmware.

Hope they fix the AF points so that the active point is red

Maybe I'm missing the point completely (wouldn't be the first time by a long shot), but illumination of the active AF points in red is already provided (Canon 5DIII instruction manual, page 103: VF display illumination).

It's true that when this function is set to either Enable (always) or Auto (only in low ambient light), the duration of the illumination is quite brief (1/4 second or so), but -- according to my VF display -- the metering is achieved (virtually) simultaneously with the illumination of the active AF points, so I don't get the "affects metering" argument.

It seems to me that a firmware update could be easily implemented to select the duration of the AF point illumination; i.e.: the brief duration currently available; 1, 2, 4, 8 seconds (just throwing out possible values here); or as long as the shutter button is depressed halfway.

Again, if I'm missing the point entirely here, I would love to be schooled. I do appreciate that metering can remain active as long as the shutter button is depressed halfway, so as to modify meter readings in changing ambient light levels during composition (i.e., passing clouds or focus/recompose). But it seems to me that an algorithm could be developed to take AF point illumination into consideration in the metering calculation, if / when the light spill from the illuminated AF points does impact metering; something like: adjust metering by x% when (count of AF points illuminated x light emitted per illuminated AF point) > y% of ambient light level of metered scene.