Blackmagic RAW 12:1 Video

People might not even notice on 4k Bluray on pause (and apparently ProRes produces similar effects).

I'd be completely shocked if any human eye could parse the difference between cDNG and BRAW in real world (non-zoomed) footage mastered to DCP and projected on a theater screen. DCPs are compressed too. Not as heavily as BluRay, but compressed nonetheless.

No way you could ever tell the difference.

People might not even notice on 4k Bluray on pause (and apparently ProRes produces similar effects).

I'd be completely shocked if any human eye could parse the difference between cDNG and BRAW in real world (non-zoomed) footage mastered to DCP and projected on a theater screen. DCPs are compressed too. Not as heavily as BluRay, but compressed nonetheless.

Two very different sets of requirements are being conflated in this thread: Detailed VFX on the one hand, and normal editing and color grading on the other. There's a tremendous difference in whether a codec is suitable for VFX work where artists need to be able to manipulate an image in ways that require fine grainless details unseen by the naked eye, compared to whether an image in a narrative drama or documentary is suitable for editing and color grading (even extreme color grading). I can't comment on VFX work, as it's not what I do, but I can say that based on what I've seen so far out of BRAW, it is more than sufficient for editing, color grading and mastering for even cinema deliverables.

Yes, I am planning on submitting this film to film festivals without any special affect editing so I am very happy that BRAW can help me achieve this

Detailed VFX on the one hand, and normal editing and color grading on the other. There's a tremendous difference in whether a codec is suitable for VFX work where artists need to be able to manipulate an image in ways that require fine grainless details unseen by the naked eye

For Redcode, there is basically one parameter you can control that determines the compression quality, which is the target ratio. The VFX pros actually don't recommend the lossless 2:1 setting, but rather the 4:1 setting (that might be different for Helium and Monstro footage, because the compression is more efficient with higher resolutions).

Now my guess is that the while the braw settings will be different because there are more options available, the concept will be the same. Some settings will be fine for most situations, some will only be optimal if you're downsampling for delivery, and you'll have to bump up the quality settings for VFX work.

Same idea... just a new set of parameters to experiment with.

Lol. Guys, if there was a one thing I know, is that a lot of people don't see too well, and you shouldn't lean on their opinion too much. I think the deliverable to cinema is too heavily compressed too (frinding on what your using, and resolution) and problems you see in the samples preferably filtered out). Now, that macroblocking is so large as to show up on 4k as a coursness. So, at FullHD you reduce the number of pixels involved and smooth out a bit. If it was FullHD raw Bayer, it would be more stark.

So 4k to 2k at cinema compression might smooth things out a bit too.

Now, back to looking at the branches on the tops of far off mountains, the lines between the pixels at the local VMAX, and the FullHD pixels on my TV, since my treatments have immensely cleared up my vision.

Lol. Guys, if there was a one thing I know, is that a lot of people don't see too well, and you shouldn't lean on their opinion too much. I think the deliverable to cinema is too heavily compressed too (frinding on what your using, and resolution) and problems you see in the samples preferably filtered out). Now, that macroblocking is so large as to show up on 4k as a coursness. So, at FullHD you reduce the number of pixels involved and smooth out a bit. If it was FullHD raw Bayer, it would be more stark.

So 4k to 2k at cinema compression might smooth things out a bit too.

From my BRAW tests (lossles/3:1 raw vs BRAW Q0/Q5):

Yes there is a visible difference (loosing some image detail, with massive compression is not exactly a new concept) - but you have to zoom in 800%-1000% on a still to actually see that difference.

In a running clip, on a TV or screen - no way you can spot the difference.
After exporting it with a delivery codec, they all look the same, even zoomed in.

After exporting it with a delivery codec, they all look the same, even zoomed in.

In the end, it all looks horrible being viewed on a 720p smartphone screen with night light filter active.

Pixel peeper test - straight out of the camera, no correction - 800% crop

As I said before, some image information loss as expected at that level of compression, but nothing you see at 100%, and sure not in a running clip.

BRAW-COMPARSION.jpg (748.31 KiB) Viewed 2234 times

Amazing. There are differences when you pixel peep, but when you’re watching the video, I can see you’re not likely going to be bothered by the lower data rates as you are listening and watching a story unfold.


Sent from my iPhone using Tapatalk

I agree Rick, once distributed, the viewer is not going to see any difference, unless you screen at an IMAX!
Cheers

So what everybody is saying, on the various threads.. is that the differences only matter to VFX artists that need to work with the cleanest possible image... for all other cases, BRAW is going to be plenty good, including for large budget films because in the end, everyone but a few of the people working on the film are going to see the highly compressed deliverable which basically does away with any of the differences.

I assume that the select few that need the ultimate clean image would be those doing green screen at a fine detail (and even then I am guessing that is questionable), or maybe working with rendered/animated images like cartoons or sci-fi space ships light sabers, etc.. is that about right?

I agree Rick, once distributed, the viewer is not going to see any difference..

Q0 is at about the same level as Prores4444 (plus the benefits of raw), but no chance to see the differences at 100%.

Just for kicks, I encoded those clips to x264 - one of the codecs all our material has to go through at one point, if people should watch it. After that, all the differences are gone, and all the clips look basically the same. I hope you guys can now sleep at night

Frank, I'm talking about CDNG compared to various BRaw. If you want to preserve more, you might want to start off with more stark detail versus more muddied detail in BRaw. The macroblocking is spread across enough pixels it may matter, but to an encoder into a consumer format, that might not compress too well, reducing quality elsewhere.

As I said, no matter if you come from CDNG, Prores or BRAW, after delivery encoding, they look all the same.
There is a threshold of image detail in the original material that matters, everything above that doesn't. Just the nature of long GOP massive compression beasts.

But again non of that matters at 100% - unless you crop in on a still to 800% or more, you not able to see it anyway.

Button line:

I always loved CDNGs (never cared about file size), but I often ended up using some sort of diffusion, to get rid of the harshness.
A lot of that stuff, folks think is resolution or image detail, is actually just aliasing. With BRAW, I get a much more pleasing image right out of the camera, no need for a third party OLPF or diffusion filters.

So what everybody is saying, on the various threads.. is that the differences only matter to VFX artists that need to work with the cleanest possible image... for all other cases, BRAW is going to be plenty good, including for large budget films because in the end, everyone but a few of the people working on the film are going to see the highly compressed deliverable which basically does away with any of the differences.

I assume that the select few that need the ultimate clean image would be those doing green screen at a fine detail (and even then I am guessing that is questionable), or maybe working with rendered/animated images like cartoons or sci-fi space ships light sabers, etc.. is that about right?

But if you’re working in vfx you’re going to downsample to 4k at best. And that’s if you’ve got the budget to render 4k. Marvel don’t have that budget. At which point you’ve probably got the money to shoot on whatever you want.

Star Wars episode 2 was shot 8 bit, 3:1:1 Color at 1440 x 1080 and the person on the street doesn’t mention the matte lines. Plenty of other stuff about that film but not the matte lines but I digress

Everyone who has shot it raves about it. So just use it if you want it’ll be great

I’ve got a day off so I’m finally going to shoot some braw myself!

Thanks for telling me that Gavin. Now I know, that's why it looked so bad. The issue people fail to realise, is that the subconscious (visual and recognition system included) is more intense. Even if you casually glance by weird stuff, your subconscious can still be disturbed by it.

Frank. Yes, aliasing etc, but I'm talking about real detail. People are not complaining BRaw is producing better detail, but less.

Now, codecs are designed to compress base line movement and areas plus detail, introduce alaising, false detail (macro blocks), it replaces real data, remove detail just gives it less to work with to produce the best result. So, it is all about packing the best representation in the stream, dumbing down the data is not going do it, unless the dumbed down image happens to map to the codec functions really well. So, yes, it is likely to affect a rubbish consumer image to be a bit more rubbish. However, what is being described as happening to the BRaw image, is maybe looking like some sort of heavy anti-noise diffusion.

Frank, that's a pretty bad example.
Try some contrasting detail, with colours as another example, like static frabrics. The monotone pastel red/orange roof tiles shot that way are just hiding issues. Even with your diffusion, just glancing at it there is obvious difference. What I'm going say proves the point. You saved that jpeg at reasonably low quality (like 72-150mb/s h264 in equivalent bitrate, but I don't know what quality efficiency per version of each codec) and frankly, there is not much to compress. But look at the vertical tiles beneath the window, you go down the list and they look less and less real (except the two bottom ones of course). I just looked at the tiles to the left, and saw the adjustment in them first. However, what resolution is this really. The minute miriad of pixel details has the same effect, their preservation on consumer video helps the visual appeal. It is all about how people "feel" about the image, how it subconsciously and consciously strikes the viewers mind. I remember the police interviews in The Bill when they switched over to digital betamax. They used a cinemagraphic lighting style which really brought out the facial details and facial hair, on our lousy free to air broadcast channels. The art of low bit stream encoding, is what to keep, as I pointed out to the industry 14 or more years ago. I recommend ambarella for a reason, before that low bit rate h264 was rubbish from what I saw.

I've sized up the samples at distance to precisely emulate a reasonable feild of view (measured it out to just at the lower end of the optimal feild of view for cinema seating). I can see the differences, but if you don't look for them, the levels are more pleasing at the top sample. You can see the mark in the 3rd or more tile beneath the window in, stands out, while is just gets less noticeable below it. This contrast of detail stands out. You could empty out much of the texture on that mark and it would still look like something realise there (the low contrast between the pixels in there help make it less noticeable) but if you reduced the contrast of the market itself compared to its surroundings, it would look less strikingly real. Thus the levels of the whole pattern of tiles benefit from this in the uncompressed sample. Years ago, when David Newman invented cineform raw, he showed us samples of cineform compared to uncompressed sample. The scene was very much the same, but the contrasts went fuzzy, shallower, much less pleasing in parts. Like cineform at high compression ratios, BRaw seems to suffer somewhat, except what I'm seeing at best quality is sort of what I would expect at 12:1. So, I would hope BRaw will have a CDGN beating mode one day. Something is happening there, we don't know, but maybe it is some sort of sacrifice in debayering to remove noise that some are pointing at.

Now, another issue is the debayer mechanism. It's going to effect the accuracy and level of detail available fur a transcode.

So what everybody is saying, on the various threads.. is that the differences only matter to VFX artists that need to work with the cleanest possible image... for all other cases, BRAW is going to be plenty good, including for large budget films because in the end, everyone but a few of the people working on the film are going to see the highly compressed deliverable which basically does away with any of the differences.

Yes. So let the complainers complain and just go shoot stuff... not that it matters anyway, since once you finish the film, no one will be able to tell what you shot it with anyway if you and your team did a good job.

I assume that the select few that need the ultimate clean image would be those doing green screen at a fine detail (and even then I am guessing that is questionable), or maybe working with rendered/animated images like cartoons or sci-fi space ships light sabers, etc.. is that about right?

Even most of them won't need cDNG. If seeing Frank G's examples isn't enough, then odds are nothing will be enough.

I always loved CDNGs (never cared about file size), but I often ended up using some sort of diffusion, to get rid of the harshness.
A lot of that stuff, folks think is resolution or image detail, is actually just aliasing. With BRAW, I get a much more pleasing image right out of the camera, no need for a third party OLPF or diffusion filters.

Well said!
I agree with this 100%

+1

We went through the same thing in the audiophile world, where a lot of critically acclaimed very expensive systems introduced aliasing that created “presence”. As an ex low end studio engineer this drove me crazy after fighting to eliminate this in the recording,

Don't worry about what they say, the truth is, when your dealing out top quality work to air, the consumer codec is a paper bag you got to be careful handeling. If it matters it feels better to watch the image derived from a better source, then look at that. If it is a fast full in documentary about school tuck shop wars, it don't matter as much.

Let's look at weddings. One day those people are likely to have options to have 8k high frame rate auto 3D wide vision screen, and may well have. So if it's going show up a better impression of it on a wider feild of view on their, it matters. But Bluray only goes so far. However, FullHD Upscaling might look worse. I don't think 8k is strictly needed, but 4k is a nice compromise. That little bit more pleasing levels and detail you get out of it on that feild of view will make it feel better. What we see as a one or two bits worth of difference in the original matters if you can squeeze it through. I'm sure if people viewed it objectively they would agree (charge them more for the premium treatment, and get a good projector in a black out room to show them. If they are not going to pay you for extra days of work then don't worry as much.
But knowing weddings, you can't even grarantee everything will turn out premium, so, an option on the proviso filming turns out). I'm still waiting to see what the 8k camera turns out like next month (blitz this) or a micro 4k. I've missed out on a trip to Spain from I'll health, so I'll be watching for the reviews to compare it against the pocket like the rest of you guys. Good 4k is worth more than 8k that's not so good, so, until I can see which one is the best buy, I'm hanging out.

Actually, I'm going to be your good classic example. DVD used the mizerly mpeg2 codec, yet produced surprisingly good results. But the truth is they put a lot of work putting major motion pictures to DVD (and cleaned up noise). The result was a much better picture on the little datarate it had. If you look at footage on TV, you see vast differences in quality that comes through between major productions and something slspiedthrough, like the news. The differences dissapear, yeah sure?