Canon 5D Mark III

Started by a1ex, June 14, 2012, 04:45:09 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

hjfilmspeed

Well it would be awsome if this could be remedied by ML but dont get your hopes up to high and overlook the other sweet features of ML.

Andrzej

Yes. I am very greatful what Alex has been done witch Magic Lantern in 1 year. Will donate for sure.
The true about 5dm3 it that all it users suffer from low resolution. In my opinion this is Ahilles heel in video therms.
Each way is god to fix this.
Iam not an expert. All my knowledge i have, come from my experience and internet (information what i chose to believe).

Some time ago i did resolution test's using 5dm3 in 720p comparing prime to zoom lenses.
Main comparison was done between Canon 24-105 f4.0 at 85mm f4.0 and Falcon 85mm at f1.4
Conclusion is that prime lenses producing more lines (20-100 vertical pixels !) (5dm3 720p)
This also was visible during my work on source material i did.

I was asking myself why? From scientific tests done by "optyczne" Canon should produce more lines at 85mm f4 comparing to falcon 85mm f1.4. And this is true in photo mode. In video is not. It is opposite.

We know that prime lenses wide open produce strong chromatic aberration http://en.wikipedia.org/wiki/Chromatic_aberration
This is something to do witch light diffraction. Maybe chromatic aberration is a quality factor in this case? So maybe OLPF filter ?



It was not scientific test, but proper enough to see the difrience.

I will do more tests witch only Falcon 85mm comparing f1.4 to 11 and publish them in about 20-30 days when i find the time for this.

Chromatic aberration come out on strong contrast.
Canon 5dm3, 550d, Canon 24-105mm 4.0 IS, Canon 18-55mm 3.5-5.6 IS, Sigma 30mm 1.4, Sigma 20mm 1.8, Sigma 70-300mm 4.0-5.6, Samyang 14mm 2.8, Falcon 85mm 1.4, Zeiss 180mm 2.8 and more...

hjfilmspeed

"Canon should produce more lines at 85mm f4 comparing to falcon 85mm f1.4. And this is true in photo mode. In video is not. It is opposite."

Canon L lenses are suposed to super sharp but you are comparing a zoom lense with a prime. Prime has less glass then a zoom so it should be sharper depending on the prime lense of course. Im not sure why its the oposit in video mode. Also are all your in camera lense profiles turned off?

I agree the resolution issue kills me especialy after you look at true 1080 or 720. I just did a test between the s2 jpeg and all-i 1920x1080 (i will post soon) and its realy strange what happeneds to the image comparing the video to the still (looks more like sampling methods to me). the difference was huge and I realy dought the olpf could fix this. Go for it and test the lenses but i think its the cameras default video smapling method.

hjfilmspeed

even though removing the olpf did help a tiny bit
it still most likely wouldnt resolve 1000 lines like it
it should. again i think its deeper then that
unfortunetly. i also fear that the 5d3 might not
have the power to resolve to much more @ 24p


Andrzej

Canon: http://www.optyczne.pl/84.4-Test_obiektywu-Canon_EF_24-105_mm_f_4L_IS_USM_Rozdzielczość_obrazu.html
Falcon: http://www.optyczne.pl/178.4-Test_obiektywu-Falcon_85_mm_f_1.4_Aspherical_IF_Rozdzielczość_obrazu.html
Wide open, Canon is sharper. Settings were the same. Lens profiles off.
I did fast test on falcon 85mm comparing f 1.4 to 11. No difference in resolution. So is no chromatic aberration.

I found this interesting: http://www.eoshd.com/comments/topic/495-canon-5d-mark-iii-polarizer-improves-sharpness/

Canon 5dm3, 550d, Canon 24-105mm 4.0 IS, Canon 18-55mm 3.5-5.6 IS, Sigma 30mm 1.4, Sigma 20mm 1.8, Sigma 70-300mm 4.0-5.6, Samyang 14mm 2.8, Falcon 85mm 1.4, Zeiss 180mm 2.8 and more...

jplxpto

Quote from: Andrzej on October 10, 2012, 12:18:06 PM
Yes. I am very greatful what Alex has been done witch Magic Lantern in 1 year. Will donate for sure.
The true about 5dm3 it that all it users suffer from low resolution. In my opinion this is Ahilles heel in video therms.
Each way is god to fix this.
Iam not an expert. All my knowledge i have, come from my experience and internet (information what i chose to believe).

Some time ago i did resolution test's using 5dm3 in 720p comparing prime to zoom lenses.
Main comparison was done between Canon 24-105 f4.0 at 85mm f4.0 and Falcon 85mm at f1.4
Conclusion is that prime lenses producing more lines (20-100 vertical pixels !) (5dm3 720p)
This also was visible during my work on source material i did.

I was asking myself why? From scientific tests done by "optyczne" Canon should produce more lines at 85mm f4 comparing to falcon 85mm f1.4. And this is true in photo mode. In video is not. It is opposite.

We know that prime lenses wide open produce strong chromatic aberration http://en.wikipedia.org/wiki/Chromatic_aberration
This is something to do witch light diffraction. Maybe chromatic aberration is a quality factor in this case? So maybe OLPF filter ?



It was not scientific test, but proper enough to see the difrience.

I will do more tests witch only Falcon 85mm comparing f1.4 to 11 and publish them in about 20-30 days when i find the time for this.

Chromatic aberration come out on strong contrast.


"Yes. I am very greatful what Alex has been done witch Magic Lantern in 1 year."   

This guy is one of our heroes. He is very knowledgeable and always ready to help others.


Andrzej

Quote from: jplxpto on October 10, 2012, 11:12:25 PM

"Yes. I am very greatful what Alex has been done witch Magic Lantern in 1 year."   

This guy is one of our heroes. He is very knowledgeable and always ready to help others.



Pretty valuable post...
Glad to be your hero  8)
Canon 5dm3, 550d, Canon 24-105mm 4.0 IS, Canon 18-55mm 3.5-5.6 IS, Sigma 30mm 1.4, Sigma 20mm 1.8, Sigma 70-300mm 4.0-5.6, Samyang 14mm 2.8, Falcon 85mm 1.4, Zeiss 180mm 2.8 and more...

Indy

yes, Alex is doing an amazing work on Magic Lantern.
This project won't such a success without him!

Of course all other contributions are important too!

Indy

jplxpto

Quote from: Indy on October 11, 2012, 01:13:18 PM
yes, Alex is doing an amazing work on Magic Lantern.
This project won't such a success without him!

Of course all other contributions are important too!

Indy


Yes, this guy is fantastic ... next year he will win the Nobel Prize of hacking.

hehehe

bonny lester

Does any one know if the alpha release allows continuous video for 12 hours outputed on the hdmi or do i need a mk 2 and v2.3 as i know this will do it.

Will be being used to capture video onto pc.

Thanks

lourenco

I really hope mjpeg will be possible. I am sure a lexar 1000x CF card is fast enough. I guess we still do not know if 5d mark III is able to do it or not until alex tries. 

I am thinking if someone not concern about shallow dept of field or low light performance a good camcorder might have to do for now for higher resolved lines of resolution.  I might look at Panasonic TM900 for 1080p @ 60p for now.  One review http://www.camcorderinfo.com/content/Panasonic-HDC-TM900-Camcorder-Review/Sharpness-Performance.htm and other http://translate.google.com/translate?sl=auto&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fwww.ixbt.com%2Fdivideo%2Fmark-ii-vs-tm900.shtml
5D Mark III, CF Lexar 1000X 32GB, 24-105 F4L

yvan

just a little question as I'm completely addicted to ML now(I can't imagine to work again without it).
Will the next release have the possibility to have the histogram under the frame as for the mark II to have a full clean frame ?

I also have a little dream : with ML informations, have some transparency above and under the frame (you can see the 4/3 full picture, as in mark II in normal view but it disapeared in mark III) to see the boom operator before being in the frame.

hjfilmspeed

So what does everyone think? Can the 5d3 handle higher resolving video in 1920xwhat ever? Just wondering what everyone thinks. ;D

squig

Quote from: hjfilmspeed on October 16, 2012, 01:17:09 AM
So what does everyone think? Can the 5d3 handle higher resolving video in 1920xwhat ever? Just wondering what everyone thinks. ;D

If Alex can get it recording uncompressed 1080p to CF cards then it should resolve more detail. It's the compression and the noise reduction that stunts the MKIII.

hjfilmspeed

i think ur right. the rez is there. you can see it when
u digitaly zoom in to focus. its also in even
the lower quality jpegs. then on the same image
in video mode the rez is gone. =(

ilguercio

Quote from: hjfilmspeed on October 16, 2012, 03:52:15 PM
i think ur right. the rez is there. you can see it when
u digitaly zoom in to focus. its also in even
the lower quality jpegs. then on the same image
in video mode the rez is gone. =(
???
Canon EOS 6D, 60D, 50D.
Sigma 70-200 EX OS HSM, Sigma 70-200 Apo EX HSM, Samyang 14 2.8, Samyang 35 1.4, Samyang 85 1.4.
Proud supporter of Magic Lantern.

hjfilmspeed

Quote from: ilguercio on October 16, 2012, 03:57:22 PM
???
Sorry maybe im wrong but if your in movie mode and your not recording and you zoom in using canons focus assist, there seems to be decent detail. Remember that spot you focused on. Take an s2 jpeg(it will crop out at 1920x1080 if you shoot it in video mode) and take a 1080p all-i video (with zero in camera sharpening) . Bring both up in you editor. then zoom in on that same spot on both the s2 jpeg(1920x1080) and your 1080 video. Youll see a hudge difference and youll see what 1080 is suposed to look like.

You also notice that the detail you saw in the canon "focus assist" is in fact present in the s2 jpeg and not in the video.

I think what squig said is what the solution could be but man would that be a task. Dont know how these ML guys do it. I wouldnt even know where to start.

Im sorry i deffinetly dont mean to beat a dead horse just wish it was possible to see why this is.

@!ex

What are the odds of seeing more still shooting features in the next ML alpha?  I don't shoot much video, but would love to start testing out the still features.

a1ex

The feature set is already announced.

@!ex

Quote from: a1ex on October 17, 2012, 11:02:12 PM
The feature set is already announced.

Sorry, Kinda new here. I've looked all over the forum and on the mkiii alpha download page, and I can only find the feature set for the alpha1.  Am I missing something (obviously I probably am), please drop some knowledge on me.

a1ex

It was buried in the forum.

Quote
Proposed feature set for Alpha 2:

- Magic Zoom
- Focus peaking with grayscale image
- HDR video
- Gradual ISO transitions (this is new feature, and 5D3 is the only camera where it works flawless... yet)
- Display adjustments (brightness/contrast/saturation)
- Clean HDMI out with pillarboxes
- UniWB display correction (bleeding edge)
- Zoom on old button or M-Fn (not quite reliable though, I may disable this)
- Display presets

None of these things require changing any permanent camera settings, so they should be pretty safe. Backend things being tested: DIGIC controls and state object hijacking.

Besides that, it will contain quite a few fixes for bugs found by the 7D guys (which were also present on 5D3).

@!ex

Thanks for the quick reply a1ex (nice handle by the way).  So, any hints on when we can expect more of the stills features, since most of the new features slated for the alpha2 are video-centric?  I don't mean to pester, just really anxious/curious. 

Chucho

I remember when a1ex first join the google mailing list. There was already an Alex in the mailing list, so he added the 1 so he wouldn't be confuse with the other Alex.

hjfilmspeed

Clean hdmi? wait .... realy? Like external recorder clean or external monitor clean?

RenatoPhoto

Alex said 5D MIII Alpha 2 should be out this week!  I am patiently awaiting.....

Why is 5D MIII Alpha 1 not listed in downloads in Bitbucket?  Just curios to know how many downloads so far..
http://www.pululahuahostal.com  |  EF 300 f/4, EF 100-400 L, EF 180 L, EF-S 10-22, Samyang 14mm, Sigma 28mm EX DG, Sigma 8mm 1:3.5 EX DG, EF 50mm 1:1.8 II, EF 1.4X II, Kenko C-AF 2X