Messages

Hi Andy,
I did some recordings this week after upgrading to latest TL and some times the camera freezed, never seen before.
Next week I'll have to do some shoots on a Piper flying just once, so can you tell me which is the most stable version?
I'm really afraid to miss the shots!

What's the purpose of framing liveview if recording is shifted?
This bug makes crop view almost unusable for anything else than simple tests.
I'm using the 50D as B-cam of a BMCC but I'm forced to use it in full view, losing resolution and increasing aliasing, just because I never know what will be recorded in crop mode
Shots with short DOF are ok, but wide shots clearly suffer from the limitations of full view, compared to BMCC.

Adapter from a ebay store from Singapore.
http://www.ebay.co.uk/itm/181227040910

Obviously works but i'm having a issue with centering the crop, known to work well on the HD crop mode on the 600d.

I don't have the 600d so I have no idea if the crop factor is the same of the 50D raw crop mode. If it's the same and 2/3" lenses cover it without vignetting that's great!
Unfortunatly the seller you pointed doesn't ship to italy, and other canon/B4 adapters start from 200 euros!

@ Jackeatley
Hi, I'm curious to know how you mounted the B4 lens, 'cause I have a fuji zoom lens and would like to work with it!

That's a great idea, the porting to ML.
I checked your topic and I was wondering why it's not in the RAW subforum.
Also this current topic should change its title from 50D - 40D to only 50D.

thanks for the video levisdavis, very interesting.
So definition and dynamic range are on BMPCC while colors are on 50D side right?
I would be curious to see a couple of frame grabs from the shot that frames the area under the stairs, it seems full of noise but the compression of youtube makes it look 264-ish.
Can you share two jpegs from it?

Great effort SebaVuye.
But I would notice some points: I don't understand why, even using RAw, you have some overexposed areas, making the whole thing look digital.
Notice the bed on the right while the telephone is ringing, or the hand on the left while watching the photo.
Also the head in the first shots has a light problem, making it look a bit flat and not enough dramatic.
Last, but not about the quality of filming, is that original actor gives much more tension.
Anyway, an excellent exercise!

great idea, recycling footage wisely!
Noise is not so much disturbing, I would have appreciated a better white balance, skin tone was too yellow/green, even for city night lights.
Didn't understand the Kate Upton shot but appreciated it

All APS-C cover 90-100% Alexa FOV, being Super35mm which is cinema standard, which is same as APS-C.

added!
and re-modified

Basically crop factor battle looks like this:

5DIII                      1:1
50D no crop:          1:1.6
50D crop                1:3.96
BMPCC                   1:2.88
BMCC                     1:2.4
BMCPC 4k              1:1.7
GH2                       1:88
GH3                       1:2
ARRI ALEXA           1:1.5
RED                       1:1.4 (depends on format)
Canon APS-C        1:1.6
Sony, APS-C          1:1.5
Nikon APS-C          1:1.5

hi a1ex, I shot a video with the lens cap.
same lenght (more or less) of the other, than put it in subtraction mode on it.
so you suggest a single frame made with the average of the dark video?

I finally decided to make a test for the Dark Frame Subtraction technique.
It's used in photography to reduce noise and remove hot pixels, and somebody claimed it could be used in video too.
The answer is no.
Well, it removes hot pixels, but noise remains the same or even worse.
You can check by yourself in the video below

or read the whole story on my blog and download the high quality video (utube made a awful compression, even in fullhd!)

www.riccardocovino.it

I'm stuck with this workflow, simplest and most straightforward for me:

Conversion directly from flash card with RAW2CDNG with 12bit option enabled (saves 15% space) and import in Resolve 10 for CC and editing.

It has its flaws, but it's damn quick!

Yes, it's a variant of JEPG2000.
It has wavelet compression, which leads to virtually zero artifacts and no blocking.
And stores info at 16 bit, allowing for RAW-like grading.

Compression ranges from 3:1 to 15:1 or 18:1, I don't remember, But usually it's used around best quality settings.

Even the long-time hoped 10bit RAW conversion while writing data would be a huge benfit on file size without affecting quality in a visible way, but all efforts went wrong..

Just to clarify for everybody, because there's a general misunderstanding here:

File size comparison between CinemaDNG and REDCINE at 1920x1080 24P

CinemaDNG: 4GB / minute for 14bit RAW
REDCINE: from 0.4 to 1.5 GB / minute (depending on compression option) for 16bit RAW

So a compression factor would really really be helpful, now we have a "mini"RED that records 4x heavier files (at less quality).

I don't know how you got 200GB for 5 mins 2K on RED, it's more like 1GB per minute at 24p, so 120 should be around 5 GB, so way better than CinemaDNG.
And it couldn't be otherwise, or REDCINE raw compression wouldn't be so famous!

Friday night. Shot with 50D ML:

I think your video is..
oh, thank you Rommex, thank you

@KahL:
I didn't expect 50D to be involved in such kind of work, great!
Ususally covering sport events requires continuous recording, how did you manage the storage problem?
I guess You had to stop and change cards/batteries so frequently!

I shot someone in front of a green screen for a few tests and used the 3-point light system usually used for chroma keying.  The aperture was open all the way and the subject was a little dark, so I bumped the ISO to about 500 and it was just a little overexposed (by 1 slash whatever that is).  After recording, I dropped it in AE to remove the green screen and it was noisy and tough to get a clean cut.

Any tips on chroma keying with Raw?

Chroma keying and 3 point lighting aren't necessarily related.
To achieve a good chroma key the most important thing is to have uniform lighting on the green background.
So, no shadows, no gradients in lighting, no digital noise due to insufficient light.
Never put the character close to the green screen and if you need him to be dark leave enough space behind to allow you to lighten the greenscreen enough.

this is camera shake that reproducing old world war 2 times
the song says "oh bird dont sing let the soldiers sleep".
and cat goes to sleep at that moment.

i shot it in cyprus

and the idea behind this music clip is
cats were brought in cyprus to fight snakes in past times by humans
and cats were killing snakes and dying for humans
this is historical fact

so my music clip is like putting the real war song from 1945 from russia
to the cyprus cats as if they are the ones who fights the nazi (snakes) =)

You're crazy, but I appreciate your crazyness!

Hey people
chill out

my first professional looking music video made by canon 50d and latest tragic lantern from Andy600

watch and learn

finally a music video and not the usual cats and flowers..
nice grading, but the Parkinsonian shake of the camera is really disturbing!

@Andy600:
I shot at 800 ISO, intentionally under-exposing.
I wanted to saved information on the lamp, and indeed I had them.
You can see here, in this attempt of extreme recovery, that you can see almost every wrinkle of the lamp paper!

This is a test I made to see the difference in quality from Raw to H.264.
With Raw I can get much more detail instead of H.264 compression artifacts (macro-blocking), and also the dynamic range is much wider allowing me to see the wrinkles in the chinese papaer lamp and also the details in the shadow areas.
Thanks to magic lantern for this awesome Raw feature!

You can follow my work on my website:

www.riccardocovino.it

Was that after trying Dual ISO?

no, I didn't try dual ISO.
Just put the build in the CF and first video was all green, darker than liveview, with insane contrast.