Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Dragon aliasing color shift

Mark Wilkinson

Mark Wilkinson

Well-known member
Joined
Jun 28, 2009
Messages
495
Reaction score
0
Points
0
Location
LA/ATL
Hello Redheads,

I was looking at the recent CML camera tests and comparing Alexa and Dragon - which seem essentially identical once a curve is applied to the raw files.

But if you zoom way in, there is a color shift in the aliasing.

I assumed it was a Bayer pattern issue, but the Dragon has it and not the Alexa.

I'm curious what causes it.

Any ideas?

Dragon.jpg
Alexa.jpg

dragoncu.jpg

Alexacu.jpg
 
Last edited:
And as a separate question - what color is that chip second in from the left - orange or red?
 
Michael Hastings said:
If it was a lens issue it seems like it should be on both of them?
Click to expand...

Resolution and maybe even internal image processing.

Also, I wasn't there checking focus and it could be a few other things. Hard to tell without being there in this case.
 
I believe the second to left chip is Red (255, 0, 0) and was a struggle for all the single chip cameras to reproduce.
 
Those cropped/zoomed screen grabs seem to be the same resolution. Did you upscale the Alexa footage or down-res the Dragon footage?
 
I just looked at the R3D which I think is: H001_C010_0124DC.0000000F and it looks very different from the down sampled images posted above. What is the precise source of the down sampled images?

Graeme
 
Graeme Nattress said:
I just looked at the R3D which I think is: H001_C010_0124DC.0000000F and it looks very different from the down sampled images posted above. What is the precise source of the down sampled images?

Graeme
Click to expand...

Should be H002 I believe. I think his sample is from Dragon Skintone Daylight folder.
 
Starting with the light path; the interdependence of the diffraction coefficients of the lens elements, the OLPF, the airy disc, etc have the potential to create unique artifacts - but I don't think that's it.

My best guess is that the Alexa is doing more sophisticated on board processing to create a "prettier" image at the expense of some granular precision. In other words, I believe the ARRI electronics are more aggressive in smoothing out the image per metrics of the image as a whole vs Dragon. RED has typically been more doctrinaire about giving us "what the sensor saw" then allowing greater control over the RAW development process per creative intent. With great power comes great responsibility, perhaps even to the point of fine tuning scaling metrics. I would also point out that ARRI biased their whole system to make clean 1080/2K from the full chip. RED took on the far more difficult task of making good imagery at various source and output resolutions. not to mention accommodating windowing for HFR/lens coverage.

Cheers - #19
 
I'm not seeing the color shift in question when looking at the R3D. Zooming in all the way to 400%, it's shift-free. There does seem to be some CA in the lens but again, no color shift in the "alias". Looks like the it might be a downsample issue in the creation of the screenshot/still.
 
I've tried the Dragon skin tone R3D from daylight and tungsten shots, Stop 0, H001_C010_0124DC.0000000F and F002_C010_0125TA.0000000F and zooming in I see mild CA, but nothing really glowing. Downsizing the exported dlx in photoshop to 1920x1080 they still look fairly monochrome with just a hint of the CA poking through.

Yup, what was the workflow to produce the down sampled image?

Graeme
 
You must log in or register to reply here.
Back
Top