Messages

Thanks @Danne and @Levas. It's definitely bleeding edge! That's great to hear it's coming to the crop_rec code soon though. With chroma smoothing at 3x3 the focus pixels are gone - at 2x2 they're still slightly visible.

Yes it should definitely be possible to get a few more frames with real time view mode, or global draw turned off

@Danne Have tested your 4k build and it works, see video. Able to record 4 or 5 frames at a time at 4096x2552 resolution at 9fps. That's the first time I've ever recorded anything in 4k! Liveview is scrambled, but with global draw on it works - you get the usual pixelated low fps preview.

In one of the exterior shots close to the horizon area the focus pixels make an appearance (small area only). I've looked at the various options to remove these in MLV App but they don't seem to go away.

Also, I tried to see if it was possible to select a lower resolution with this build (2.5k or 3k), but it didn't seem to work so I did no further testing. I'm assuming this 4k build is only designed to work with 4k resolution?

Understanding how these registers works because these registers controls write & read speeds .

Or someone helping me with this. 

Thanks for bringing this up @Bilal. I guess that's probably the most important next step to take in order to make higher resolution recording feasible. What sort of coding or steps would be needed to help you with this? I don't know how to help but maybe could ask around, although I imagine anyone who looks at this would need to have specific knowledge about coding for Magic Lantern?

[MLV proof:]

@theBilalFakhouri  14-bit lossless works yes, and I'm using the EOS M. With one or two of Danne's earlier builds it was possible to get liveview to work and record fine with lower bit depths, but it's true that the recorded footage had glitches.

I did try your 2520x1384 adtg_gui file yesterday and all looked fine in liveview, but unfortunately when I looked at the footage on my computer it was jumping around and was glitchy. Danne posted an example (see here below) which is exactly what I was getting with my footage:

MLV proof:
https://bitbucket.org/Dannephoto/magic-lantern/downloads/2520x1384.MLV

I'll try again to double check in case I did something wrong, but in any case Danne's newest build (2520x1320) works 100% as you can see in my video.

A higher write speed is definitely needed though. I wonder what it will take to push to 208MHz?

I got lower bit rates to work without error messages in earlier builds but the resulting footage had glitches. I might investigate that again when I have time. I can imagine it will be hard to find a way around that though - let's just hope there will be a way to achieve higher write speeds with the SD overclock hack in future, or that an SD card company manages to release a faster UHS-I card!

Thanks for changing the lua scripts file names, that will save some time!

Here's some more footage from the flamenco event a few weeks back!

@fortymileswest thanks. With youtube videos you have to write

   it's probably the same with vimeo?

This is also an issue with the EOS M. I've just been shortening the lua script file names every time a new build comes out - but the full file name remains in the magic lantern scripts menu so it's not a problem.

Thanks Danne and Bilal, that's some incredible progress and really appreciated! With your latest 2.5k build (newest 16th June version), it works well with no glitches or corrupted lines. I've made a quick video recorded at 2520x1320 and also 2336x1314 for a slightly longer record time. This is 14 bit lossless (I briefly tried 10 bit lossless but immediately got error messages).

I had tried your 16th June build yesterday @Danne before you posted the edited version and was getting glitchy footage that was jumping everywhere. Same with Bilal's adtg_gui file. But this version is perfect.

Sorry for my delayed replies. I've tested both new builds and getting the same results as @loknar with the black lower part of the image and the thin colour line. 

I mostly tested in 14 bit lossless which is fine, but also briefly tried 10 and 12 bit lossless with the original 2.5k build. Some of those clips worked but sometimes they were glitchy - I can recheck that if you like?

However with today's build when I tried 10 and 12 bit lossless on the 2.5k build, all the resulting footage was glitchy and unusable (it looks ok on liveview when you're recording it), and 14 bit lossless is fine as expected.

Thanks so much for your help @Danne and theBilalFakhouri (and Nikfreak)! I've tested both of your builds Danne using your automated lua script, and they work without crashes but there's an issue with the bottom part of the recorded frame which seems to correspond to the increased height resolution area (see images below). Is this also happening with the 100D?

2.5k




3k




The liveview preview works fine with the 2.5k build, with the 3k one it's garbled - but that doesn't affect the recorded video.

I'll take a closer look this weekend once I get the chance. Personally I really like the idea of being able to do 2520x1256 which is close to 16:9.

@theBilalFakhouri aha, that's better. Now getting 68.7mb/s write speed (72mb/s read speed). Thanks so much!

Cheers @Danne - I'm following the work you're doing with the 100D which is very interesting. Once I have the adtg_gui file I'll get on with it and post my findings here.

@theBilalFakhouri  thanks, if you're able to prepare the adtg module file for the eos m I'd really appreciate it (there's no rush though). Then I can start working on it!

I'd like to look into the adtg module for the EOS M to increase the max resolution in crop mode. There's some nice progress being made on this with the 100D.

To get started do I just put the adtg_gui file into the modules folder and enable it, or is it necessary to create a custom-made build specifically for the EOS M to make it work?

@theBilalFakhouri thanks for this - running the benchmark with fps override at less than 12fps now gives me around 62mb/s write speed. Unsure how to use benchmark testing whilst in Play mode on the EOS M though, since I can't access the magic lantern menu in that mode.

I got 71MB/s with 6D and SanDisk Extreme Pro 95MB/s 64GB.

That's very impressive - maybe it's something to do with the EOS M rather than the card. I have this same card, and also tried another one which was the same.

@OlRivrRat  That's a UHS-II card. Canon cameras use UHS-I, so there's no point trying an expensive UHS-II card - I'm guessing they'd end up being slower in a Canon than a fast UHS-I card. Although I don't know if anyone's tested this.

@Sapporo  Wondering if there's a faster alternative to Sandisk cards, in particular to their Extreme Pro 95mb/s V30 UHS-I SD cards? I've tested two of those and the max write speed I can get on an EOS M is 55mb/s. I've seen people get faster benchmark results but unsure exactly which cards they were using.

Liveview is realtime when global draw is off or if you select real-time preview from the raw menu when global draw is on. However this is in 5x magnification crop mode, so the liveview preview is zoomed in 5x from the 1.6x crop aps-c sensor (i.e. 8x crop compared to 35mm full frame). At a resolution of 2520x1080 the crop factor of the video recorded is 3.33x compared to 35mm full frame. So what you see on screen is zoomed in more than twice than what's recorded, which makes proper framing tough, but it's real time and in colour. If you half press the shutter button you get a pixelated and slow framerate preview of the correct frame which is useful for composing a shot.

Do the other eos cameras work the same way in 5x magnification crop mode? My 50D is the same, only the preview is in b&w, not colour.

Here's a new test I've done using @Danne's May 23rd build with the SD UHS overclock hack module on the EOS M.

Recording 2520x1080 resolution 12-bit lossless raw, using a Sandisk Extreme Pro 95mb/s 64gb SD card. The max write speed I'm getting in benchmarks with this card is 55mb/s. Recording time varies a lot depending on what's in the frame (high contrast and highlights shortens recording time considerably). With a higher max write speed I'm sure recording times would be continuous at 12 bit.

This was recorded during warm weather of around 26 degrees C. I didn't let my camera get above 42 degrees C.

See this thread for more info about this module: https://www.magiclantern.fm/forum/index.php?topic=12862.0

@a1ex  Thanks for your work on this and the warning about card access, that's very useful information. So maybe it's safer not to use the Lua script (to autorun the overclock test) as I guess it adds even more to the risk.

Changed name to SDOclock.lua & now it shows up ~

@Danne  I've just tried your new build on my eos m (with Sandisk extreme pro) and it works. Thanks so much for this! I initially had the same issue as @OlRivRat where I had to change the name of the lua file to get it to appear in the scripts menu, but that was easy to fix.

I'm wondering if we still need to run the benchmark test every time the camera is turned on or is this not necessary?

@Danne  I've now tested your 27th April build and it works fine on my eos m with a fast card, same good results as the previous build. Thanks for this!

You are aware that this build skips the tests and goes right into patching(160mhz) when selected? Card needs to be about 95mb. Works on 100D at least.

Good to know - I wasn't aware! I was using a 95mb/s card so should have been ok. Might give it another try when I get a chance.

@keel - for the SD hack I used files from @Danne's bitbucket page: https://bitbucket.org/Dannephoto/magic-lantern/downloads/  (I hope it's ok to share this, if not I'll delete the link)

I downloaded the "magiclantern-Nightly.2018Apr17.EOSM202_sd_uhs.zip" file from 17th April 2018 and put that in my card (not the 27th April file which wouldn't do anything when I tried to run the overclock test). Then I downloaded his sd_uhs module file from 19th April and put that into the modules folder in ML (replacing the sd_uhs module that was already there).

After that I followed @IDA_ML 's instructions on how to run the SD overclock test and the card benchmark test, ensuring it's all done in photo mode. His instructions are here: https://www.magiclantern.fm/forum/index.php?topic=16040.msg200229#msg200229

And it worked. I used a Sandisk extreme pro 95mb/s 64gb card. I'm only shooting 24fps.

Do let me know if there's a better way to do this!

@Walter Schulz, that's a very good point!

Well, my question was actually about APSC-sensor based DSLRs.  Can a speed booster turn them into full-frame cameras when operating in Life View with full-frame lenses?  Do such speed boosters exist and if yes, does anyone on this forum have any experience with them? 

Sorry I misread your question. Unfortunately that's optically impossible. The flange distance (between sensor and EF mount) on EOS DSLRs is 44mm so there's no space for a speedbooster. You could hack one and put it inside the camera housing, but you'd have to take your camera apart and remove the mirror mechanism. On the other hand, EF-M mount flange distance is only 18mm (as it's mirrorless), so you can add a speedbooster in front of that mount.