Epic Compression Test 2

[Gunleik Groven]

Thanks for posting this test. I came across several challenging complex nature scenes that are unusable at 8:1 compression on Epic. Strangely I don't remember such scenes to be challenge for RED One (mysterium sensor).

Well, they were...
Early on, the camera actually gave up and died under these conditions...

Hahahaha

 

[Patrick Tresch]

I think for complex scenes like these, lower rezz and no HDRx may be of advantage, in addition to lower compression...

Just a hunch...

Hello Gunleik,

I don't understand your logic here. Lower Rez wouldn't help as 3:1, 5:1, 8:1,... are a compression RATIO. It's true that in lower resolution the file would be smaller, but the compression & artefacts would be the same in my understanding. We are not speaking anymore about a target compression like RC36 wich was referring to MB/sec.

Patrick

PS: And larger (5K) files using wavelet compression should hold better compression results as smaller (2K)...

 

[Gunleik Groven]

Hello Gunleik,

I don't understand your logic here. Lower Rez wouldn't help as 3:1, 5:1, 8:1,... are a compression RATIO. It's true that in lower resolution the file would be smaller, but the compression & artefacts would be the same in my understanding. We are not speaking anymore about a target compression like RC36 wich was referring to MB/sec.

Patrick

PS: And larger files using wavelet compression should hold better compression results as smaller...

I maybe off... Just looking through a bit of forest material.

Less pixels = less material to process. To me it looks a bit like the more complex the image (with high contrast + lots of green detail + lots of motion = complex) enjoys the lesser strain.

Just a practical observation that with a combo like this, one can seem to reach some kind of computational limit, as Adam has demonstrated.

I may be off, but it seems less compression = less hard computing tasks, additionally less size = less data to process, thus more processing power left to process what data is in there well.

But I haven't tested for this. Just looked at a few images with and without HDRx and at 4khd/5k under these conditions. But NOT a controlled test.

So you're right, I may be off.
That's why I call it a "hunch".

Cheers

G

 

[Patrick Tresch]

Well I understand your arguments.

Either Graeme chime in... or we will have to do some testing.

;-)

See you Gunleik

Patrick

 

[Adam Rosenbloom]

Thanks for doing this!
Oddly though, I shot something yesterday at 120fps and 12-1 compression and 5k (I know you're shooting 4k HD), with a similar high contrast forest scene, and it looked like the results from your 5-1... MUCH better than what your 10-1 looked like. Not sure why this is. Maybe the 5k?

EDIT: Taking another look at my footage, there was also a bit less contrast, but not a whole lot less than yours.

Very cool! What was your aperture? I'm suspecting that my f/16 and wide depth of field to be a big factor here. Even though in the original test this was intentional, I may have gone too far!

 

[Gunleik Groven]

Very cool! What was your aperture? I'm suspecting that my f/16 and wide depth of field to be a big factor here. Even though in the original test this was intentional, I may have gone too far!

You will need the deep DOF to trigger it. It is the many details that can be a bit much to chew on...

 

[Will Keir]

hahahah, that's not funny.

Well, they were...
Early on, the camera actually gave up and died under these conditions...

Hahahaha