robotsound
Well-known member
Hi a1ex,
Love the new Magic Lantern update. Thanks again for all your hard work!
Love the new Magic Lantern update. Thanks again for all your hard work!
T2i Magic Lantern for 550D in progress!
- Thread starter ...
- Start date
Moses Chisel
Active member
With the T3i/600D firmware update now available, I wonder if it can be used to figure out the 3x record facility for the T2i/550D?
Samuel H
Major Contributor
I may be wrong, but I think that old-camera focusing trick uses phase detection, which requires stuff other than software:
http://www.lensrentals.com/blog/2010/07/how-autofocus-often-works
http://www.lensrentals.com/blog/2010/07/how-autofocus-often-works
I see the option in the menu but I'm not sure as to how to activate it (surely I missed something)
I dropped autoexec & diff in the root. tried messing with focus graph.... all sort of fucus modes (although I suggested this option as means to help MANUAL focusing...
)Am I doing something wrong ?
I did. it functioned exactly as the last version
I see. so, it should work also in manual mode (not AF) ?
I'll try later today as I have things to do and maybe shoot some stuff.
But just to make things clear, lemme see if I got it right :
1) I drop the two files on a SD root - overwriting my current autoexec and creating a new "splitprism.diff" file within the root.
2) On camera menu - I engage split prism (switch it to "ON")
3) I make sure that Magic Zoom is enabled (doesb't matter what setting. or does it ?)
4) Upon pressing the magic zoom, a split prism will appear instead off magic zoom
Could someone post a screenshot of the resulting look (Live View with Split Prism)
Nope
For the time being I think I'll stop trying.
I erased the config file and set the setting just as you said. still cant see the prism, only the regular magic zoom window.
No :embarasse
It must be something with MY setup OUTSIDE Magic Lantern as it doesn't help deleting the config file.
I want to be absolutely sure, again :
I put BOTH autoexec (obviously...) AND .diff file at ROOT LEVEL, yes ?
I then engage Split Prism and upon recording, activate the Magic Zoom, yes ?
It looks like ML can't "find" the .diff file as it engages "regular" Magic Zoom (rectangular) upon
pressing the zoom.
Focus graph appear at ALL TIMES, btw.
Even as an overlay upon LiveView :smile:
No hurry. I'll wait patiently.
Thanks for your efforts !! :thumbsup:
It must be something with MY setup OUTSIDE Magic Lantern as it doesn't help deleting the config file.
I want to be absolutely sure, again :
I put BOTH autoexec (obviously...) AND .diff file at ROOT LEVEL, yes ?
I then engage Split Prism and upon recording, activate the Magic Zoom, yes ?
It looks like ML can't "find" the .diff file as it engages "regular" Magic Zoom (rectangular) upon
pressing the zoom.
Focus graph appear at ALL TIMES, btw.
Even as an overlay upon LiveView :smile:
No hurry. I'll wait patiently.
Thanks for your efforts !! :thumbsup:
I just tried the split prism (first build) and my experience is that it is a very promising feature and a good idea, but it seems a bit erratic on a longer lens.
I tried it on a fast 135mm manual lens, and found that with a shallow dof (f/1.8 wide open) it doesn't stabilise when focus is reached. It tends to re-split once you've found focus even if you don't touch the focus ring.
On a 28mm f/1.9 it is much more stable. I understand that this is a proof of concept, but it seems some damping might be needed or something like that.
On a side note, I am still seeing a minor issue that was introduced in a recent build, which is that when I turn the camera on, it starts in false colour mode.
As ever, thanks for the continuing release of new features. I keep wondering if there is anything else that could possibly improve ML on the 550d, and yet the improvements come thick and fast. Bravo Alex!
I tried it on a fast 135mm manual lens, and found that with a shallow dof (f/1.8 wide open) it doesn't stabilise when focus is reached. It tends to re-split once you've found focus even if you don't touch the focus ring.
On a 28mm f/1.9 it is much more stable. I understand that this is a proof of concept, but it seems some damping might be needed or something like that.
On a side note, I am still seeing a minor issue that was introduced in a recent build, which is that when I turn the camera on, it starts in false colour mode.
As ever, thanks for the continuing release of new features. I keep wondering if there is anything else that could possibly improve ML on the 550d, and yet the improvements come thick and fast. Bravo Alex!
ML does not look at the diff file; that's for developers who may want to try on other cameras. You don't need it.
I did all the tests in photo mode, but works when recordig too. I have no idea what the problem may be. Can you make a screenshot showing the focus graph?
I did all the tests in photo mode, but works when recordig too. I have no idea what the problem may be. Can you make a screenshot showing the focus graph?
Focus measurement is not absolute; if you know better math than Canon engineers implemented, you are welcome to improve it.
To get an absolute measurement, I'm using a scaling trick, i.e. I assume the max value of Canon's focus measurement over the last few seconds means perfect focus. I'm using some damping for the scaling factor (look at the focus graph when it rescales itself). The problem: this trick only works when you rotate the focus ring and cross the perfect focus point.
To get an absolute measurement, I'm using a scaling trick, i.e. I assume the max value of Canon's focus measurement over the last few seconds means perfect focus. I'm using some damping for the scaling factor (look at the focus graph when it rescales itself). The problem: this trick only works when you rotate the focus ring and cross the perfect focus point.
Last edited:
Hi Alex
Thank you for the updates, the split prism is great ! Though it seemed to be better before your last update on the mailing list (The heuristic for reversing the direction), at least for my case, with a manual Pentacon 28mm f/2.8, which pretty soft at 2.8.
I also have a problem like AriG with the focus graph, it is always displayed on top, even in the ML menu. I attached 3 pictures taken with my cell phone... really sorry for the quality.
It doesn't appear in the Canon menu.
I noted that there is a small rectangle "hole" below the focus graph slightly overlapping the word "false", when the menu is displayed you can see the liveview through it. Close up detail in the attached menu_close.jpg file.
Thank you for the updates, the split prism is great ! Though it seemed to be better before your last update on the mailing list (The heuristic for reversing the direction), at least for my case, with a manual Pentacon 28mm f/2.8, which pretty soft at 2.8.
I also have a problem like AriG with the focus graph, it is always displayed on top, even in the ML menu. I attached 3 pictures taken with my cell phone... really sorry for the quality.
It doesn't appear in the Canon menu.
I noted that there is a small rectangle "hole" below the focus graph slightly overlapping the word "false", when the menu is displayed you can see the liveview through it. Close up detail in the attached menu_close.jpg file.
