Canon 5D Mark IV

[Simonwb]

@a1ex here's my first batch of test images:

To check white levels: https://www.dropbox.com/sh/mv9ukwyueeh5cq7/AAB7hCeEqFzvqNZNAofza8XAa?dl=0

To check crop area: https://www.dropbox.com/sh/l4h7xiue8vc2mda/AADQnaxi14kRglNFmKHmba7Va?dl=0

For colour profiling: https://www.dropbox.com/sh/bo9gc02eyde7yjm/AADmjsp-v6OKXEJfL8l3HCP3a?dl=0

I'm a real newbie at this, so I hope these files are useful.  Unfortunately, the darktable colour profiling page completely mystified me.

On the dynamic range & SNR curve: could you explain what you mean by "a bracketed set of images....5 x 1EV"; I understand bracketing, but not the 5 x 1EV part.

On the dual pixel refocusing: what do you mean by front focus and back focus?  The only terminology I'm familiar with is back button focussing...

[Simonwb]

@rawvideo

Here's my attempt at your requested night scene shots: https://www.dropbox.com/sh/q1fpp5bg37mxkob/AADfvKFyGV2p4-Kuw4Z-sHvra?dl=0

Hope they work for you.

[a1ex]

Thanks, will check them later.

5 x 1 EV is the notation used by ML (5 pictures spaced 1 EV apart).

For your last question: http://photo.stackexchange.com/questions/14437/what-do-frontfocus-and-backfocus-mean

[rawvideo]

@Simonwb

Nice, thanks for the effort.

Seems I wont upgrade for the better image quality but just for the better AF in the dark, WLAN and better AutoISO.

[Simonwb]

@a1ex here's my dual pixel raw focus bracketing shots:

https://www.dropbox.com/sh/1yj3ne852z2aupb/AAD-CuZds0RzIk5OkPEQwJTAa?dl=0

Thanks for the link explaining front/back focus; I don't know if/how to force this in-camera using autofocus, so I did it using manual focus via EOS Utility remote shooting on a Macbook, hopefully finely enough and in a systematic way for you - I'm assuming you can pick up the apertures used etc used on each shot via the exif information?

[chris_overseas]

The first firmware update is available for 5D4:

http://www.canon.de/support/consumer_products/products/cameras/digital_slr/eos-5d-mark-iv.aspx?type=firmware&firmwaredetailid=tcm:83-1475064

[a1ex]

Tried to run it in QEMU, with some missing bits from 80D:

Code Select Expand


BootLoaderEND
K349 READY
K349 ICU Firmware Version 1.0.2 ( 7.4.6 )
ICU Release DateTime 2016.08.08 13:27:10


Code Select Expand


[      init:fe2b5f0d ] task_create(PowerMgr, prio=20, stack=400, entry=fe2b5e35, arg=0)
[      init:000022bf ] task_create(DbgMgr, prio=1f, stack=0, entry=223b, arg=816454)
[      init:fe0e77d7 ] task_create(Startup, prio=19, stack=2800, entry=fe0e76c9, arg=816c78)
[      init:fe0de4a5 ] task_create(TaskMain, prio=1d, stack=0, entry=fe0dd9a1, arg=0)
[      init:fe0e14b7 ] task_create(RomRead, prio=11, stack=400, entry=fe0e0ef7, arg=600006)
[      init:fe0e14c9 ] task_create(SFRead, prio=11, stack=400, entry=fe0e0ef7, arg=620008)


Code Select Expand


Setting BOOTDISK flag to FFFFFFFF
BootLoad
CF Detect High
SLOT_A LOAD OK.
Open file for read : AUTOEXEC.BIN
File size : 0xD120
Now jump to AUTOEXEC.BIN!!
Open file for write : ROM1.BIN
File close.
Write done.
Open file for write : ROM1.MD5
File close.
Write done.




Any volunteers to try the following?
- LED blinking (already tested)
- display test (please PM me for a FIR)
- ROM dumper (please PM me for a FIR)
- jumping to main firmware

And, after getting it to run on 80D:
- running user code alongside main firmware

[jrumans]

Hey A1ex, I exceeded 5 PMs per hour and can't send right now. The Firmware Jump didn't appear to do anything, which I imagine is intentional.

[a1ex]

FYI, the display test went like this:



Second picture is my attempt to understand what happened and reproduce the error in QEMU.

After the fix:

[nikfreak]

Great progress. Is this your 5D IV? 

[a1ex]

Artist: Jeff Rumans

Some low-level details:

At first I've noticed the display uses different registers, and that resolution changed to 900x600 (from 720x480) but I've assumed the rest of the code is similar to 80D (and all other D6 cameras).

Code Select Expand


[DIGIC6]   at 0x00104B68     [0xD2018228] <- 0x310000  : BMP VRAM
[DIGIC6]   at 0x00104B68     [0xD2018200] <- 0x2580384 : Display resolution
[DIGIC6]   at 0x00104B68     [0xD2018230] <- 0x2580384 : Display resolution


The 80D requires the display buffer address to be shifted by 8, so I've assumed the 5D4 does the same. With this assumption, everything looked fine in QEMU; I couldn't get Canon's error screens, but I assumed they were not present in the firmware update.
I've also assumed the image had 8 bits per pixel and used a color palette (just like 80D).

Now, why does the first screenshot look like a broken error message?

First hint: that error message (a hardcoded image in ROM, copied to RAM) must be nearby.
Second hint: from the number of garbage lines and the horizontal shift, we can find out the position of the error message, and cross-check it with the values written by Canon code to BMP VRAM registers.

ML was setting the buffer to  0x44000000. Shifted by 8, it ended up as 0x440000.

Canon code could set the display buffer to either 0x310000 (Loading), 0x3B0000 (Not found) or 0x450000 (Error). The error screen from the first image must be at 0x450000. The camera started to display a few lines of garbage (from 0x440000 to 0x450000), then continued with the error image.

How many lines of garbage? floor(0x10000/900) = 72. Since the division is not exact, the image was also shifted by 0x10000 - 900*72 = 736 pixels.

Counting the pixels on the first screenshot gives about 73 lines of garbage and about 500 pixels of horizontal shift. The difference is pretty large in the horizontal direction.

Looking in QEMU log, near the display resolution there is a register that looks close to horizontal resolution:

Code Select Expand


[DIGIC6]   at 0x00104B68     [0xD2018228] <- 0x310000  : BMP VRAM
[DIGIC6]   at 0x00104B68     [0xD201822C] <- 0x3A0     : ???
[DIGIC6]   at 0x00104B68     [0xD2018230] <- 0x2580384 : Display resolution


0x3A0 = 928, so that might be the BMP pitch. That is, the image in memory is a bit larger than the one actually displayed. Let's check the math: floor(0x10000/928) = 70, 0x10000 - 928*70 = 576. A little closer, right?

Turns out, this value was right

By doing this math, I had only one failed attempt, and that attempt was actually showing exactly what I had to change.

(I now wonder why GregoryOfManhattan said the display test did not work at all a few weeks ago - from the current results, it should have at least turned on the backlight, and maybe also displayed some garbage)

The Firmware Jump didn't appear to do anything, which I imagine is intentional.

This one is not exactly good news, but I'll keep trying.

[g3gg0]

cool progress

[a1ex]

Please find the ROM dumper: DUMP_5D4.FIR (confirmed by jrumans)

You will need a small SD card.

If it locks up or if you get incorrect MD5, try formatting the first 256MB from the card as FAT12, and leave the rest as unpartitioned space.

[wicq]

Does this mean ML Team is working on the 5D mark iv?

[Walter Schulz]

No, posts are generated by an AI developed by linguistic and cs departments at Universiteit van Amsterdam.

[Greg]

What when AI switch to Sony?

[Walter Schulz]

One of the premises to make the AI more human-like is making illogical decisions part of the scheme. It will therefore not switch to Sony or RED.

[wicq]

Does this mean ML Team is working on the 5D mark iv?

[Oswald]

Does this mean ML Team is working on the 5D mark iv?

It means just only that Microsoft Tay AI is shitposting to the forum.

[axelcine]

@wicq: The 5D4 is being investigated - the same as 760D, 80D and 7DII, so a lot of people are working in their spare time testing and reverse engineering the cams. It is tedious work, like collecting stamps. If you really want to understand the job, these magnificent guys are doing, then read the 5D3-story here. Post by post. Victories, failures, retests, corrections... it is a tale of courage, stubbornness, willingness to go far beyond the call of duty - as if there ever was such one. You could base a Pulitzer-prizewinning novel on these pages:
June 14, 2012:
http://www.magiclantern.fm/forum/index.php?topic=1006.0
September 7, 2012:
http://www.magiclantern.fm/forum/index.php?topic=2602.msg10621#msg10621
March 10, 2015:
http://www.magiclantern.fm/forum/index.php?topic=14704.0
Read "The Machine" by Tracy Kidder. That one also won the Pulitzer Prize. That book will give you some idea of what A1ex and his peers are really doing.

[wicq]

Thanks axelcine 

[ddelreal]

@wicq: The 5D4 is being investigated - the same as 760D, 80D and 7DII, so a lot of people are working in their spare time testing and reverse engineering the cams. It is tedious work, like collecting stamps. If you really want to understand the job, these magnificent guys are doing, then read the 5D3-story here. Post by post. Victories, failures, retests, corrections... it is a tale of courage, stubbornness, willingness to go far beyond the call of duty - as if there ever was such one. You could base a Pulitzer-prizewinning novel on these pages:
June 14, 2012:
http://www.magiclantern.fm/forum/index.php?topic=1006.0
September 7, 2012:
http://www.magiclantern.fm/forum/index.php?topic=2602.msg10621#msg10621
March 10, 2015:
http://www.magiclantern.fm/forum/index.php?topic=14704.0
Read "The Machine" by Tracy Kidder. That one also won the Pulitzer Prize. That book will give you some idea of what A1ex and his peers are really doing.

Well said axelcine, well said.

[olofen]

I am a DEVOTED user of Magic Lantern on my mkII....
Also being an artist this is my way to be able to get all what I need for my work not being able to pay a lot for my equipment
If it is possible to make the mkIV shine as well in its context I will be the first to honor this fantastic bunch of guys working in the dark for people like me
THANKS AGAIN ALL OF YOU 
olof thiel
photographer
stockholm
sweden

[D_Odell]

After the fix:

Awesomely cool!

[DeafEyeJedi]

Be sure to catch how much better Canon colors look and better yet the 5D4's Shadows bodes well against A7R II but not so much when it comes to Highlights...