6D or 5Diii for ML what is the best a continuous focus

[shhd]

Hello all
Please help me choose
I decided to buy 5Diii or 6D
I am interested in the video more than the pictures
6D has a continuous focus while 5D iii does not have a continuous focus
Does the continuous focus work on ML
I ask those with knowledge and experience to help me make a decision because I do not know how these cameras handle ML
I apologize if the place is wrong
I just wanted the owners of the camera to help me so I wrote here
In general, if you are going to buy a user and money is not an issue, what to choose for ML

[Levas]

For video I suggest the 5d III.
The 5dIII has better options for video with Magic Lantern (higher resolution options and framerate options).

Don't expect much from autofocus.
As far as I now, there is no autofocus in videomode, both on the 6d and 5dIII
On the 6d, you can half press the shutter button while in liveview(and not recording) and then it autofocus on where the focus box is, or if facedetection is on, it focuses on a face.
But it only does autofocus while doing a half shutter press on the shutter button and while not recording, during recording there is no autofocus, none.
Furthermore, the autofocus in liveview on the 6d is based on contrast autofocus, so no dual pixel autofocus, not even dedicated autofocus pixels anywhere on the sensor.
Not sure, but  expect the 5dIII has the same autofocus options in video as the 6d, that is none 

[shhd]

thank you very much 
I will look for 5D iii
Also, the speed of writing on the memory is limited in the 6D, it is only about 40MB, while the 5D iii reaches 90MB
Is that correct

[masc]

Also, the speed of writing on the memory is limited in the 6D, it is only about 40MB, while the 5D iii reaches 90MB
Is that correct

Depends on what we are talking about. 5D3 reaches ~130 MB/s when writing on CF and SD at the same time. For the 6D it also exists a SD overclocking, but for the exact numbers please read the corresponding threads.

[Levas]

Standard the 6d has a write speed of about 40Mb/s.
With latest SD_UHS module, it can write about 80Mb/s.

[shhd]

thank you very much all
Found 5D III took 20,000 photos, worth $ 1,000
I found 6D took 6000 pictures worth $ 600 and I'm confused

I own this memory



ard Typ
SDHC   
Storage Capacity
32 GB   
Bus Type
UHS-II   
Speed Class
10   
UHS Speed Class
U3   
Speed (X-rating)
1733x   
Bus Interface Speed
UHS Class II   
Data Transfer
Read Speed 300 MB/s Maximum
Write Speed 260 MB/s Maximum
Write Speed 30 MB/s Minimum   
Durability   
Shockproof, Temperature Extremes, Waterproof, X-Ray Proof   
Operating Temperature
-13 to 185°F/ -25 to 85°C   
Storage Temperature
-40 to 185°F/ -40 to 85°C   
Built-in Write-Protect Switch
Yes

And I found this clip on YouTube

Have or put a link to a topic explaining the memory speed on 6D 

[Danne]

Buy the 5diii and then test that card against the 240Mb/s sd uhs patch. That's a card we havn't tested yet.
There's really no comparison between these two cameras. It's an achievment from thebilalfarouk that the 6d can record with speeds up to 80Mb/s. But that won't reach the magnitudes of the 5diii even when used with a single decent cf card.
If all you want to do is talk and put up photos of your equipment in multiple posts please do or get a 5diii and do some good by testing the sd uhs patch for the community.

[shhd]

Thank you and I am ready to do anything for society
I just want you to explain to me what is required even when I got the camera, do the required test and provide you with the result
How do I activate
sd uhs
I am new to ML and many things I still do not know

[Danne]

Nothing to do except enabling it from the menu. When ready I can enable the extra setting. Most likely it won't work but if you're getting the camera might as well test.

[shhd]

Can you put a picture showing the place of the option in the list

5D iii will be with me soon and I will be ready for any trials and hope to clarify the risks that may occur in the event there are risks

[Danne]

No more risk than enabling the default patch. No camera at hand. Search the sd uhs thread.

[a1ex]

I'd rather make the user aware of the risks, even if they are small. Just saying

[shhd]

 

thank you all
I will read the topic on uhs

a1ex

I love you

[Danne]

For all I know. A temporary corrupted, or in worst case, a broken card? Mostly fixed by formatting the card on computer or simply restarting camera. Heard of one user breaking his card, not sure what to think of that particular tester. Once the card was corrupted he formatted in camera so maybe stay off that idea at least. If card gets slow or acting strange a restart is what to do. Select 160Mhz or simply deselect the option to patch altogether.
More unlikely damaging sd controller, at least never heard of that. With that said. If anything breaks, you get to keep and bury all the pieces .
Please feel free to share anything more on this topic a1ex, my experience derives more or less from repeated sd patch testing trying to figure out what you have coded and lately testing out refined patch numbers from the bilalfakhouri.

[a1ex]

Well, there is one report where a SD card died after attempting to overclock, and that report looks legit to me (OK, it's just one case from maybe some hundreds of testers). There's also my old CF card, which died after an incorrect software request, I think (an overflow in my testing code, but I'm not 100% sure - IIRC overclocking wasn't even active when that happened). What I want to say is that card-related experiments are not exactly risk-free (OK, nothing is, but this area in particular is a bit more problematic), and the users should be aware of that, especially if testing on expensive cards.

Thing is, we don't really know what happens electrically (whether the voltages are correct for the UHS modes, whether the timings are correct or just on the edge, i.e. happening to work with most cards, but not with others etc - one would have to probe the SD signals with an oscilloscope and check them against the specs, which I don't think any of us did). Also, we don't know exactly what the low-level registers are doing - we only know that some values work at least on some cards, with different levels of performance and reliability, and others values don't work at all. Whatever is behind those values, is still a mystery (of course, there are some educated guesses about what each register might do, but that's pretty much it).

Another anecdote, not related to ML, but from another experiment I did: a 5V sensor connected via UART to a 3.3V microcontroller. Apparently fine at first sight, so I didn't think too much about it. Worked fine for a few days of continuous use, then it stopped - the MCU pin that was receiving data from the sensor, appears to be toast. So, you can't really say there is little to no risk, even after several hours of continuous testing, when there is a pretty clear report of damaged hardware (in my opinion), even if all other tests were apparently safe. That scenario might or might not happen again - we simply can't tell.

Not saying users should not run the tests, but - in my opinion - they should be aware of the small chances of damaging the SD card hardware. OK, there is the "you get to keep both pieces" thingie, but that doesn't mean we should rely on it. OP asked specifically what are the risks for this particular test, so...

Food for thought: https://marcan.st/2011/01/safe-hacking/

[Danne]

You're the right guy to shed some light. I'll keep this post for future reference.

[70MM13]

i have a suspicion that my 5d3 was potentially damaged by playing rather aggressively with sd overclocking.  behaviour with a few cards all changed at once after a single failure.

luckily, it is working (so far, having done several hours of recording since) at 160mhz with the sandisk extreme pro 170, but each time i hit record i wonder if it will fail.

i am not complaining, and i am happy to help in my small way.  it may simply be a coincidence that 3 cards all changed their behaviour at the same time (although i was only using one of them at the time of the failure), but time will tell.

of course, as always, a1ex is right.  everyone should be well informed that this is risky!

but i am very happy to take the risk!  this has made the 5d3 a monster video camera!!

thanks to the efforts of so many passionate people!

[shhd]

There is no problem with card damage when applying the test, I can buy another card and this is very little for community service here, ML community deserves more than card damage
Really you are great people 
I wait for the arrival of 5D 3

[shhd]

Dears
Today I finalized a deal with someone who owns 5Diii, she has taken 29,000 photos and the body is in very good condition
Wait for her to take the test
And please, I bought the camera for $ 650. Do you see this as a good deal?

[Danne]

If intetested the sd_uhs ball is running here:
https://www.magiclantern.fm/forum/index.php?topic=23041.msg229630;topicseen#msg229629

[shhd]

OK. A day or two up and 5D3 are now on the way

[Danne]

Ah, yes. Thought you gotten the camera already.

Anyway. Testing like this would help:
https://www.magiclantern.fm/forum/index.php?topic=23041.msg229639#msg229639

[shhd]

Still waiting for 5D iii arrival 
The seller appears hesitant after seeing that I'm sticking to the camera 
Question please
Can write speeds of SD and CF cards be combined in 5D  ?

[Danne]

Do you think reading from here and use search function will answer your questions?
https://www.magiclantern.fm/forum/index.php?topic=23041.msg216899#msg216899

[shhd]

Hello
I tried your copy on my girlfriend's camera and we only had a CF card, we activated the raw video only and shot some stills, but always from time to time the message appears canon
The camera stopped working
Restart the battery and take it out
The night structure is quite stable
Raw Video 14 BIT
1920X1080
9:16
Writing speed 82
23.976 FPS