Welcome to our community

Be a part of something great, join today!

  • 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 :)

What exactly is the quality difference between 1/4 Debayer vs. Full ?

David McDonald

Active member
Joined
Nov 2, 2011
Messages
37
Reaction score
0
Points
0
I'm working on a show that was shot over a year ago, and it's finally being finished. Everything was transcoded a long time ago to 1/4 debayer for editing offline, and now that the online editing is being done, we've discovered that a couple of the R3D files are missing (it appears the footage may have been transcoded off of the SSD Card and the R3Ds may not have transferred to the hard drives properly).

So this means our only option may be to live with one of these 1/4 debayer ProRes shots in the final film (it's an indie feature). I think it's only one shot. How bad is it to keep one of these files in the finished film? I was under the impression that the resolution isn't necessarily worse (it's 1920x1080), it's just lacking color information.

I've had similar situations with music videos for YouTube, where I've had really rushed deadlines and didn't have time to re-transcode and re-color correct a shot and ended up keeping a 1/4 debayer shot in the cut and no one could tell....so I'm of the opinion that you can get away with it for one or two shots.

Does anyone have some more insight into this? What exactly is 1/4 debayer lacking vs. Full quality?
 
I always thought it was reduced resolution. I know in RedCine-X, when you click 1/4, you're getting 1/4 the width and 1/4 the height. You can, of course, then take that and up-res it on export to 1080p, and it'll look crappy when played back at 1080p.
 
What is 1/4 lacking? Picture information. It's akin to an uncompressed audio file to an mp3, there just is not the same amount of data there so something has to give.

1/2 res is 2k about 3/4 of your image information is thrown away QT
1/4 is about 1k about 7/8 of your image is thrown away when making the QT

if you are finishing 1080, then you might be ok. I won't go into why its important to save r3ds.

David
 
I would find it rather surprising, if line skipping was employed. My guess is rather, 1/4 means each block of 2x2 pixels are averaged to form a new virtual pixel. After this the debayering algorithm is run some way with the virtual pixels. But as said, all this is just speculation until somebody from RED chimes in.
 
Actually, line skipping *is* used. Well, sort of. Not quite. Er, here: http://www.red.com/learn/red-101/redcode-file-format

In short, one of the magical things about the Red R3D file format (and there are MANY magical things about it) is that you don't have to read out the entire file in order to get a smaller preview. The data is written in such a way that the first 1/16th of data you would need to produce a 1/16 size image is written first, and thus accessed first when read back. Following that is the data needed to produce a 1/8th size. And so on. This way, whether your CPU is too slow, or your drive IO speed, or both, the data is written in such a way that it gives you the maximum playback speed possible.

Brilliant.

If you render out at 1/4 debayer, then, I'm pretty sure, you're skipping every other line vertically and horizontally, and then doing that AGAIN.
 
interesting, i just out put a clip shot in 4k, in 1080p in full, 1/2 premium/good and 1/4 debayer... no visible difference. no time saved in converting to pro res in red cine x.

i repeated the same thing but outputing to 4k with the same results.

what am i missing?
 
interesting, i just out put a clip shot in 4k, in 1080p in full, 1/2 premium/good and 1/4 debayer... no visible difference. no time saved in converting to pro res in red cine x.

i repeated the same thing but outputing to 4k with the same results.

what am i missing?

With red rocket card or without?
 
Without. I would assume RR would be faster. Not too concerned with speed tho its nice. Was really looking at quality.

Faster and it always debayers full quality. As for your conclusions, the quality differences will be obvious when comparing them at 4k. Also the lower resolutions will exhibit stair casing on angled lines. But to your point, downscaling to 1080 will hide these artifacts somewhat.
 
Stair casing with lower capture resolution or output?

when looking at footage in cine x full/ half/ quarter etc radically changed it. I see that details. But not when I out put. Are you saying that with RR the quality I'd debayer is even higher? I always use full when exporting. It's just tonight I started playing. Thanks
 
Just to be sure, you are aware that the full/half/quarter etc. setting of the viewer has no effect on output deBayer quality? That is a separate setting in your output preset. The fact that your render time did not change would suggest you were always outputting at the same quality.
 
If your shot is not focused to 4K spec then you won't see any difference.
 
Back
Top