Messages

Yeah, by no means would I like to see the mirror gone. I rarely use LV. It just seems to me like Canon would have the ability to leave the mirror up when taking shots in LV rather than actuating it each time. So I was wondering if there was any specific reason for it. Given that Mirror Lock Up exists it seems that the only thing that it's needed for is closing the shutter at the end of the exposure, the exposure can start any time after the shutter has opened. So I wonder if ending the exposure with the shutter rather than electronically somehow gives better results.

1%, I think you're still misunderstanding me. I realize that the mirror must be up. My question is simply why does the mirror ever need to come down in LiveView mode. And why Canon would have it move when there is never any purpose for the mirror to be down while in LV mode.

Anyway, I'm only curious... It's not particularly relevant to anything.

Oh no, I'm certainly not that unfavorable of the mirror... I'm just curious about it.

Ah, I see. Do we know if there are any physical limitations to taking a full resolution image without actuating the shutter? Or is it currently just a matter of figuring it out?

Right, LiveView has the mirror up. When a photo is exposing, the mirror is up. So why is the mirror doing anything when taking a photo in LiveView?

edit: obviously, the mirror has to move when entering LiveView, I mean while in LV, maybe that's the confusion.

So my understanding is that the ML team cannot simply do whatever they want and the best we have is "screenshots" of the liveview when we want to take a picture without actuating the mirror. But is there a reason that this isn't a thing that canon just puts in? When they invented liveview, why did they make it actuate the mirror? I can at least understand the shutter, maybe shutter speed it more easily controlled mechanically than electronically. But why on Earth is the mirror moving? There must be a reason. Is it just because they didn't bother separating those two actions, mirror and shutter?

I'm getting 20/10 RW with crystal disk. Looks like it came from a Marketplace seller on newegg, I'm gonna try to RMA it.

It says "up to 30MB/s"... The thing is, it seems much faster in some cases. For instance formatting the card, even with low level seems way faster. And it seems faster on my computer. But the benchmarks and bitrate both perform worse on the class 10. Which is a bummer, it was on sale for only 10 bucks but I only got it to get higher bitrate.

I just got a Class 10 16GB SanDisk card from NewEgg, upgraded from the same but Class 6. The Class 6 got up to 12.1 MB/s write but this Class 10 only gets up to 9.4 MB/s write. And it's consistently lower anyway. I'm testing with the benchmark on a 550D. I understand that the 550D should be able to manage around 20MB/s.

So what's the deal? Why on Earth might the class 10 card be slower?

Woah! Why am I only just now finding out about this? That's awesome.

t3i is also called the 600D.

With magic lantern you can lower the bit rate of the video so more can fit inside 4GB. You can also set it to 720P and set the frame rate to 30FPS to get more time in there. This will cause weird playback until you fix it in a video editor though.

There might also be a way to get over 4GB using exfat formatting but I don't know if that actually works, I haven't looked in to it. There is also a setting to make it start recording again automatically if it is stopped but you will lose a second or two.

(I don't think there is an email notify on this forum, unfortunately.)

Edit: when I set my bitrate to .7x and use 720P mode with FPS changed to 30 the camera estimates more than 35 minutes before reaching 4GB. Just under 30 minutes with just the FPS changed. I do believe that there is still a 30 minute limit though.

Thank you very much

I've seen it mentioned that the 550D has some limitations in write speed. I am thinking about getting a faster card but there's no use in getting something that is faster than the camera. Does anyone have any benchmarks or experience with the maximum write speed of this camera so I can get an idea of what to look for?

That solution completely ignores my main point though. I only use ML about 2% of the time so now I've got a card (a good one, because it's for video) that I can't use 98% of the time.


The thing is, this feature already exists. I am just hoping for an option to swap a single parameter. Turn on ML instead of turn off.

Can someone who actually develops for ML explain how much coding this would require? I am tempted to just look at it myself, I just don't have linux set up at the moment.

I like ML for video but I only do that very rarely. I only need ML about 2% of the time. My video SD cards are my good cards so I still want to use them for photos the rest of the time, so I can't just switch cards.

I don't want to have to switch around a bunch of options every time I switch between them. Also there are other concerns, power saving and power usage. The extra power usage of ML is low but it is still a waste for 98% of the time. Plus I am using nightly builds for video and I don't want to take the risk of bugs messing up anything (a risk I am willing to take with the extra effort and caution I put into video, my video is always scripted, my photography is not). And the functions of the camera are different and inherently less intuitive(canon simply didn't design the cameras for ML).

I know it's not a huge deal but it simply boils down to this: as much as I love ML I rarely need it so all else being equal, if I could default out of it that would be really helpful to me. My hopes are that it would be a very simple tweak, I could even do it myself if turning off ML was included in the picoC API. So that despite most people here probably agreeing with you, the few who feel the way I do could benefit with hopefully little sacrifice.

I would really really like if the "hold set to bypass ML" could be reversed so that ML normally doesn't run but will start up if you are holding the set button (or something like that).

A similar feature has been requested but in a much weirder way (ML still runs but all of it's features are bypassed). I was thinking that having the option to simply reverse this already existing feature would be easy to implement.

In fact this would be easy to script if they made a "turn_off_ML" function in the TCC API. (By the way would it be considered a double post if I request this feature in the scripting subforum? Technically it could be used for other scripts)

I don't mean to pretend that I know how much work it would be, I'm not too knowledgeable beyond arduino coding and simple .net coding. But I am hoping that this is a much less demanding request than usual and that I could see it implemented if it doesn't take too much time or work.