Messages

@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.

That 4k Video looks like upscaled 1080p. I've seen crisper 4k content, but maybe the small depth of field makes it look softer than usual..

Well, my 5D4 has arrived but I won't get time to play with it until this weekend.  I could try to shoot some comparison footage - would you like to see any formats in particular?

I would very much like a night scene of a street or something similar, mostly dark with a few bright lights (streetlamps), high ISO, maybe the highest the 5DIII does with auto iso and then 2 raws with the same image and lens from the 5DIII and 5DIV. No longtime exposures, just fast enough to be handheld. To see if there is any improvement for my lowlight theater photography (besides the now actually and finally visible focus points).

Because I would have to downgrade my camera and don't like that much, after all it's a tool for daily use that needs to work.

And if this sort of test is allready been done there is no need to do my own.

Is it correct that 113 can do higher resolutions in raw recording than 123 without framedrops?

The 664 looks good too, 365g in weight. It might have a little more modern display with better viewing angles I suppose.

Can you really focus better with that thing? What about the focus peaking? As good as the internal one of ML?

And how does it affect ML if an external HDMI monitor is connected? Does it save some resources if I turn off the internal monitor? Probably doesn't matter because it still has to compute the image for the external one. But it could keep the temperature down I guess.

I dont see the use of this, that number is always changing. So that wouldn't be very accurate.

Also no significant difference between MLV and Raw for me. Both do 7 seconds of 1920x672 50fps.

Well the internal Benchmark with everything (GD, Liveview) off shows this:

Benchmark complete.
ML Nightly.2014Rug07.5D3123, ffdld533b1dc (5D3-123) tip
Mode: playback, Global Draw: OFF
CF MI Corporation SILICONMOTION SM2236RB
Write speed (buffer=16384k): 118.8 MB/s
Read speed (buffer=16384k):   152.8 MB/s
Write speed (buffer=16384k): 114.9 MB/s
Read speed (buffer=16384k):   152.8 MB/s
Write speed (buffer=15625k): 119.5 MB/s
Read speed (buffer=15625k):   152.9 MB/s
Write speed (buffer=4096k): 112.4 MB/s
Read speed (buffer=4096k):   141.1 MB/s
Write speed (buffer=3906k): 181.8 MB/s
Read speed (buffer=3906k):   139.2 MB/s
Write speed (buffer=2048k): 183.8 MB/s
Read speed (buffer=2048k):   128.8 MB/s
Write speed (buffer=1953k): 87.6 MB/s
Read speed (buffer=1953k):   123.9 MB/s
Write speed (buffer=128k): 32.4 MB/s
Read speed (buffer=128k):   34.9 MB/s

Yes 5D3 with 123.

Its not faster, the first 1066 card I got was the same speed as the 1050 I got before that.

I benchmark with CrystalDiskMark on the PC. Actually that newest 1066 card is 4 Mbit faster than the first. But the first 1066 was as fast as the 1050 (84 MBit seq writing)..also 1050 and 1066 isn't that big a difference. Still these cards are advertised with a 100Mbit writing speed.

In the camera I only check if get to record continuously in a higher resolution (thats all that matters) in comparison to the other card.

Just tried my second Komputerbay 128GB 1066x. Its also not faster than my 1050 128GB Komputerbay.

It also doesnt go beyond 1920x1080 25fps  and 1600x560 with 50fps in continuous recording.

If you have the time, what seq. write speed do you get with CrystalDiskMark on your 128gb card?

This is so annoying, the 64GB cards are almost as expensive as the 128GB. And if they actually are faster I would only use them for high frame rate stuff since my 128GB card works fine for normal 25 frames 1080p.

But the high frame rate stuff of course needs more space. So I wouldn't get much onto the smaller card.

I'm looking for an external monitor for my steadicam rig. And yes I am cheap, so I have my eye on those Lilliput monitors. But there are just so many types.

Whats the newest most useful model?

I really like the focus peaking option, that could save some CPU resources if ML doesnt have to do it. Unless the focus peaking of the monitor is inferior, than my 5D has to keep doing it.

Do I need 1080p (do they even come with 1080p, they are cheap after all?). Of course the more pixels the more accurate the focus peaking. But maybe less is already enough and saves some battery time?

Those monitors seem to come with different battery adapters. So what battery type keeps it running the longest. I probably won't give it the batteries from my camera because, well they go into the camera. But if I decide on different battery of course I would need an additional charger.

Some help to bring light into the confusing Lilliput model lineup would be great.

http://www.manfrotto.com/clamp-on-electronic-remote-control-for-canon-hdslrs-mvr911eccn

That doesnt control the joystick but you control focus directly with this via USB cable. Works fine during ML usage.

Or if you have more than one hand free, this seems nice too: http://www.aputure.com/blog/2012/08/15/introducing-v-control-usb-follow-focus/

But the 128 ist the best value for the price both 64 and 256 cost more per GB.

I guess no one has seen a 5DM3 writing more than 100MBs but it would be great to max out my cameras capabilities with the memory card.

I got two 128GB cards, one 1055 and one 1066, both are at 85MBs in writing. Thats 1920x1080 with 25fps and sound stable so far.

But its not the advertised speed of the card. For slow motion shots it would be great to get a faster one, haven't decided yet if i should order some more for testing.

I am on 123 and get a lot of bad frames as soon as I exceed the frame rate of 35 in 25 frames mode, resolution doesn't seem to matter. But all below 35 is clear so far. For 50 frames mode my limit is 60.

Thanks.

Hello,

is there a fast way to batch convert dngs into a preview movie? I accidently used the old raw module and already deleted the source files. Only have the dngs and don't want to do the lengthy process with Photoshop to just preview the files.

Is there a fast solution for Windows like "RawFlow" that works with the dng files as a starting point?

Does the 4GB limits also affect raw video recording?

How is raw video on the battery, I would imagine it could save some compared to normal encoded video because the CPU doesn't have to compress it..but then its more data. Will it empty my batteries faster?

That sounds good. Could I put ML on a small SD card and just remove it to turn it off?

No option to check the recorded raw video on set is not so great though.

Hello there,

is the 2 second delay you never get rid of again still and issue? Or can I get my camera back as it was after using ML?

Also if watching a raw video stream is not working very well on the camera about about a connected android smartphone? Would this work to check the recording?