SSD run times

[Jannard]

Quick question, can you record to ssd to 100% capacity or do they function like a spinning drive where it is recommended you leave some "breathing room"?

Record 100%.

Jim

 

[David Battistella]

I have not seen REDCODE artifacts.

I am wondering if there is an extraction problem if people claim to need uncompressed because of artifacts.

I am also reading that uncompressed is a thing to have just to say it's there to please some big time idea that film is somehow uncompressed.

I think 5K and a 5:1 redcode probably out resolves film now. Films compression can change frame to frame.

Like what is the compression of say, "a vynil record or a 1/4inch tape" professional standards for a very long time.

David


David

 

[Gunleik Groven]

Analogue media has allways had the capacity to compress dynamics.

Thing is that it happens in a different way than digital compression.

bear in mind that we're talking about two different types of compression here.

1. (the definitive good one) to be able to capture and represent "more" dynamics within "a less dynamic" range.

Think
Film, Vinyl, analogue tape.

Actually, as to Dynamics some of the "problem" with digital is the by nature "uncompressed" representation of whatever is recorded.

And for images, that the precission is "different" or "upside down" as to how we hear and see. Thus the precission is higher at higher intensities, while most of our precission is at lower.

That is partly solveable through adding bit-depth and neglecting the lower precission bits.


So on the one hand we're talking about dynamic compression, which is all good.

on the other hand we're talking about data-compression, which can have its not-so-good sides when done badly.

Me too have come to realize that a few of the "issues" I thought of as data-compression artefacts in the Redcode RAW files, actually have been decoding artefacts, which through the altered processing philosophy post B30 has changed my view on compressed RAW from "it's super cool, I can do 4k at home" into:
"Gee, I really don't want to work totally uncompressed again for a while... except for VFX and online...)

 

[Jannard]

Funny... Weta has never asked for uncompressed.

Jim

 

[Gunleik Groven]

Funny... Weta has never asked for uncompressed.

Jim

LOL

I'll take that.... -

We have had better results with developed EXR (and earlier DPX) from the r3d's than going directly from the RAW files in Nuke/Shake/AE.

But maybe we're doing something wrong.


I was NOT saying "capture" uncompressed, but "work" uncompressed in some stages of the production.

If memory serves me, that was what Lightiron described for the 2k online of Social Network, too...
Sure they work without transcoding?

(Just really curious here...)

 

[David Battistella]

Still.

Are we talking put uncompressed in there to shut people up?

or

what??

People have not even seen the footage from EPIC's yet, but still it seems they have to satisfy an internal need to say that something has no compression.


Properly extracted r3d is like a good film 2K film scan. There is no reason anyone should expect or see a different quality than what they get out of 2K scan and actually they are starting with a larger frame size.

Anyhow. Just sayin'

 

[Jannard]

Still.

Are we talking put uncompressed in there to shut people up?

or

what??

People have not even seen the footage from EPIC's yet, but still it seems they have to satisfy an internal need to say that something has no compression.


Properly extracted r3d is like a good film 2K film scan. There is no reason anyone should expect or see a different quality than what they get out of 2K scan and actually they are starting with a larger frame size.

Anyhow. Just sayin'

Sorry David... many that come by here say that a properly extracted EPIC R3D file looks more like VistaVision, not a 2K film scan. Peter Jackson for one.

Jim

 

[Hrvoje Simic]

It is not about pixel peeping for step to 2.5:1. It's not about how good DP's think 5:1 r3d is.

The fact is that more data is coming out of that incredible sensor than what is coming out of the camera. It's also a philosophical and psychological matter. While 5:1 presents the most reasonable choice, 2.5:1 presents the ultimate peak. No bit lost.

When they can, people will choose the lossless. Because they are humans.

Also, history has shown many times that "What one sees" at the time being and what next generations or solutions see or require tends to differentiate.

If there are technical limitations from preventing 2.5:1, no problem whatsoever.
This is fantastic as is.

If it can be done, do it.

 

[Jannard]

Just to recap... the camera is capable of 2:5:1 REDCODE RAW (mathematically lossless). The SSD's currently are not fast enough to record 24fps 2.5:1. You can get about 18fps. As soon as SSDs or other recording options can keep up, we will let you know.

Jim

 

[Cüneyt Kaya]

hoenstly i love the fact that a macrpo with rocket can work with r3d files and that i dont need a 100 TB raid array.
if i wanted umcompressed, i can take a Dalsa.
imho that reds RAW is compressed is the key to success.
hey, i can through 4x3TB drives into a macpro and am still able to edit amd finish it today.
thats future for me at least.
storm/fcp/avid/davinci in 1 box with a red rocket inside, thats another revolution happening right now.
i dont want to be able to own an epic amd not be able to work with it.
2 cents from berlin

 

[Hrvoje Simic]

Just to recap... the camera is capable of 2:5:1 REDCODE RAW (mathematically lossless). The SSD's currently are not fast enough to record 24fps 2.5:1.
Jim

I don't see any problems then.

I was under the impression that SSD limitations might be the case but didn't catch the post mentioning that.

 

[Timothy Kotthoff]

Jim

I understand REDCODE to capture RAW data from the sensor that is compressed via a wavelet transform and entropy coder - the exact details of how that all works are, understandably, proprietary. It does not use Inter-frame - but rather intra-frame compression, where each frame gets compressed on it's own. The data is kept in RAW mosaic form etc.

This is is all sensical to me but I am trying to understand how bit depth (as it relates to the Epic) is determined. Does the codec decide based on the compression ratio being used whether it is 12, 14, or 16 bit? Or do we the user, define bit depth in post after de-bayering?

The reason that it seems so important to me is that there is quite a difference between the levels of luminance or shades of grey in a 12 to 14 to 16 bit sampling.

Thanks and best!

Tim

 

[Rocco Schult]

It should influence contrast ratio, which technically is just another term of what you are saying.
Given it is a constant on the Epic, it should use the same bit depth over all compression ratios, unless other options are explicitly talked out.
Proving will tell if this is true or if some compressions ratios compromise e.g. for the sake of speed.