Is BRAW 444, 422...?

The place for questions about shooting with Blackmagic Cameras.
  • Author
  • Message
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostMon Jun 08, 2020 3:44 pm

Andrew Kolakowski wrote:You can't compare it as RAW is not for viewing. RAW needs to be debayered to be a meaningful image (otherwise it's just grey pixels). Answered been already given that good debayering will have eg. 80% efficiency, so you will get (badly saying) 80% of 4:4:4 (if sensor resolution=output resolution). Add a bit of oversampling and you will get full 4:4:4 worth of data image. This applies to any RAW, as BRAW is nothing specific compared to other RAW formats at the end.
People are too much exited about BRAW- its nothing that special at all compared to the RAW formats. Its based on same base principle. Rest are just details.



If you were addressing me, my point was if they had originally not made it fully exotic custom format, but a hopefully normally usable and viewable format, you could convert to Bayer if you wish. Then hopefully no special software or processing is needed, except when you convert to Bayer, you could then use the footage either way in the extra quality recording mode.

I saw Braw, and asked for an extra quality mode, and let them to braw. Because it's the shiny new toy syndrome, people get upset about, including those two in that English site, so you can't do too much with marketing effect in full swing, even though people see the difference without looking at the pixels, which is just a tool which described what you are seeing. You see it too. But all this insulting 'pixel peeping' and a litany of insults to logic just continued even about questioning absolute performance. This is what needs to be realised, it is the circling hord, not the defenders of what is real. Often it is several people circling one or two people, and a few thousand standing around ignoring allowing them. The fall of society. I think people have realised things are not all absolutely perfect and not too get into such intense debates about this one. A few options to kick Braw to the next, minor, level is really not that much to ask. That next minor level is going be bigger, but it's Braw they want, do don't worry about it. But people go on about how big of a system they need, or how much extra storage they need to store 6k footage, that is just the price you pay to play with the extra quality. No use complaining about it, the only solution is too use the latest compression standards, or beyond those, it's no use putting down 6k becsuse of it, or compromising things, lower options are still there. The only point of the situation is that that there is better compression technique and chips options which may help, but otherwise it is what you get.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

Måns Winberg

  • Posts: 70
  • Joined: Tue Aug 13, 2013 1:32 pm
  • Location: Lund, Sweden

Re: Is BRAW 444, 422...?

PostMon Jun 08, 2020 4:02 pm

Andrew Kolakowski wrote:Yes, this is a "positive" effect of mild compression.


As far as I can tell from the patent application, actually, no. It seems to be a method to reduce the effect of deviant single pixels, such as dead pixels and random noise pixels, by not using only the original Bayer mosaic value. Those deviant pixels are high frequency, since they are single pixels that differ significantly from their surrounding pixels. By calculating an average from a number of surrounding pixels these deviations are averaged out. The original Bayer value of the pixel to be calculated is given more weight than the surrounding pixels. This is not compression, just a different way to decide the value. The averaging has to be done anyway for the chrominance values, when you start from a Bayer mosaic, but here it's done for the luminance values as well.

Since the red and blue pixels in a Bayer mosaic are quarter resolution, just like the two color signals in 420 chroma subsampling, the RGB debayer stage can be skipped. The raw Bayer data can be stored as a YCbCr420 component video signal without loss of information, and unpacked in the computer as RGB values (or perhaps as a Bayer mosaic, but I don't know if that would have any advantages). Wavelet and discreet cosine transform, JPEG2000 and JPEG, are used for compression of the YCbCr420 signal.
Last edited by Måns Winberg on Mon Jun 08, 2020 4:19 pm, edited 1 time in total.
Offline

Andrew Kolakowski

  • Posts: 9207
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Is BRAW 444, 422...?

PostMon Jun 08, 2020 4:11 pm

I'm saying that mild compression (ProRes etc.) always has "smoothing" effect. Not directly related to BRAW.

BRAW doesn't touch wavelet. It's plain simple DCT which been confirmed by those who tried to reverse it:
https://cml.news/g/cml-raw-log-hdr/topi ... 0,25749037
Offline

Måns Winberg

  • Posts: 70
  • Joined: Tue Aug 13, 2013 1:32 pm
  • Location: Lund, Sweden

Re: Is BRAW 444, 422...?

PostMon Jun 08, 2020 4:32 pm

Andrew Kolakowski wrote:I'm saying that mild compression (ProRes etc.) always has "smoothing" effect. Not directly related to BRAW.

BRAW doesn't touch wavelet. It's plain simple DCT which been confirmed by those who tried to reverse it:
https://cml.news/g/cml-raw-log-hdr/topi ... 0,25749037


Im just saying that the smoothing noise reduction calculations aren't a form of compression.

I take your/their word for it. In the patent application it just says that it can be used: "Preferably compression uses transform coding, for example using wavelet transform or discrete cosine transform or the like." "Compression can be performed using any suitable of compression techniques (e.g., including techniques based on discrete cosine transforms (e.g., JPEG) and wavelet transforms (e.g., JPEG 2000)."
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostMon Jun 08, 2020 6:50 pm

So, if they are leaving it that open, why don't they just use the cineform component compression technique, or better still, base it on open free cineform. You then might get 20:1+ Braw at high speed on post. Nvidia had started offering more wavelet assistance.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

Andrew Kolakowski

  • Posts: 9207
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Is BRAW 444, 422...?

PostMon Jun 08, 2020 8:17 pm

Cineform RAW is not free, although it's quite cheap and most important- it's probably free of RED patent problem. There is 1 problem with Cineform- it has very VBR nature in current code which is not live recording friendly.
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 7:47 am

Wayne Steven wrote:cineform component compression technique
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

Hendrik Proosa

  • Posts: 3015
  • Joined: Wed Aug 22, 2012 6:53 am
  • Location: Estonia

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 12:24 pm

Andrew Kolakowski wrote:Cineform RAW is not free, although it's quite cheap and most important- it's probably free of RED patent problem. There is 1 problem with Cineform- it has very VBR nature in current code which is not live recording friendly.

Since Cineform raw compresses the bayer pattern formed component planes directly it vioates red patent afaik. Wasn't this the main thing that killed SI-2K camera which recorded cineform raw when red was still in its infancy? VBR or CBR makes no difference in writing a file, braw has also VBR modes.
I do stuff.
Offline

Andrew Kolakowski

  • Posts: 9207
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 12:46 pm

No, because Cineform was first with RAW compression and apparently there is a special deal between RED and Cineform. I don't know how much truth is there.
SI-2K camera was a PC inside which was far from optimal.
VBR does make a huge difference to recording, specially if Cineform is very "sensitive" codec. If a clean source may end up with eg. Y peak then noise one may push it to eg. 2xY, which can be a problem during live recording in devices with fixed/limited storage speed. I spoken with David many times about it. He said that people were asking about more restricted mode for use in devices like cameras.
What would be nice to have is hard peak limit (something like ProRes has). Cineform was developed as codec for film scanning or high-end mastering, not a "recording" format. Very different principles where taken into account during core design. Biggest problem is/was also lack of reference FPGA/ASIC implementation.
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 1:09 pm

Notice that the Red patent starts just above the Si's basic specs, avoiding patent conflict with them.

Anyway, I was not talking about the raw version, but something comparable with Braw to use instead of the jpeg internals.

Variable is why buffering is used, but open ended size is not good, you need an implementation with maxinum reduced size that the system can handle.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

Hendrik Proosa

  • Posts: 3015
  • Joined: Wed Aug 22, 2012 6:53 am
  • Location: Estonia

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 1:21 pm

Andrew Kolakowski wrote:No, because Cineform was first with RAW compression and apparently there is a special deal between RED and Cineform. I don't know how much truth is there.

Interesting. I know they were first and my impression was that RED simply pulled the carpet from under them, with help from devoted believers like PJ.
Andrew Kolakowski wrote:VBR does make a huge difference to recording, specially if Cineform is very "sensitive" codec. If a clean source may end up with eg. Y peak then noise one may push it to eg. 2xY, which can be a problem during live recording in devices with fixed/limited storage speed. I spoken with David many times about it. He said that people were asking about more restricted mode for use in devices like cameras.

Sounds like implementation limitation, not something ineherent to vbr. Wavelet and DCT should be pretty similar in how bitrate limitation would work, but maybe macroblock based compression is more "resilient" to hard limits since you can screw up just one macroblock...
I do stuff.
Offline

Andrew Kolakowski

  • Posts: 9207
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 1:44 pm

Wayne Steven wrote:Variable is why buffering is used, but open ended size is not good, you need an implementation with maxinum reduced size that the system can handle.


Except BM cameras seems to have very little buffer (RAM is expensive?), so they seems to be quite prone to dropped frames.
Offline

Andrew Kolakowski

  • Posts: 9207
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 1:45 pm

Hendrik Proosa wrote:
Andrew Kolakowski wrote:No, because Cineform was first with RAW compression and apparently there is a special deal between RED and Cineform. I don't know how much truth is there.

Interesting. I know they were first and my impression was that RED simply pulled the carpet from under them, with help from devoted believers like PJ.
Andrew Kolakowski wrote:VBR does make a huge difference to recording, specially if Cineform is very "sensitive" codec. If a clean source may end up with eg. Y peak then noise one may push it to eg. 2xY, which can be a problem during live recording in devices with fixed/limited storage speed. I spoken with David many times about it. He said that people were asking about more restricted mode for use in devices like cameras.

Sounds like implementation limitation, not something ineherent to vbr. Wavelet and DCT should be pretty similar in how bitrate limitation would work, but maybe macroblock based compression is more "resilient" to hard limits since you can screw up just one macroblock...


It all can be done, but it would require some development on top of current Cineform code.
Offline

Andrew Kolakowski

  • Posts: 9207
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 1:46 pm

Wayne Steven wrote:Notice that the Red patent starts just above the Si's basic specs, avoiding patent conflict with them.


Yep, maybe this is the thing :) It's also another reason why RED patent should never be granted as it was nothing new or unique.
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 5:20 pm

Andrew Kolakowski wrote:
Wayne Steven wrote:Variable is why buffering is used, but open ended size is not good, you need an implementation with maxinum reduced size that the system can handle.


Except BM cameras seems to have very little buffer (RAM is expensive?), so they seems to be quite prone to dropped frames.



Thanks for the big warning Andrew. It should never really drop a frame itself unless something is failing, or storage is the problem.

Anyway, not good that RAM is limited (I guess you mean on FPGA). That's not great, why haven't they just moved into an ASIC solution, like ARM+GPU or with video capture compression circuit? I mean the pocket series is great but big. They could probably license Braw to ambarella. Ambarella offers 6:1+ Braw on license, and sells under 6:1 chips cheap to BM, and BM pocket drop to quarter size maybe with longer battery life. Maybe I should just try to contact my contact there about doing a better jpeg mode?

I should have said, that that was why I said to set a reduced fall back maximum datarate, to protect the buffer from overflowing . Problem might be solved then.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

John Brawley

  • Posts: 4267
  • Joined: Tue Aug 21, 2012 7:57 am
  • Location: Los Angeles California

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 8:05 pm

Andrew Kolakowski wrote:Except BM cameras seems to have very little buffer (RAM is expensive?), so they seems to be quite prone to dropped frames.


It doesn't happen if you use supported media / matched codecs and the correct formatting protocols.

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles
Offline
User avatar

Ulysses Paiva

  • Posts: 994
  • Joined: Sun Sep 01, 2013 8:32 pm
  • Location: Pernambuco, Brasil

Re: Is BRAW 444, 422...?

PostTue Jun 09, 2020 10:18 pm

John Brawley wrote:
Andrew Kolakowski wrote:Except BM cameras seems to have very little buffer (RAM is expensive?), so they seems to be quite prone to dropped frames.


It doesn't happen if you use supported media / matched codecs and the correct formatting protocols.

JB


Exactly. Never had a single dropped frame since the old BMCC 2.5. Although I had lost almost 2 cameras due to failures along time. But no dropped frames also on the P4K.
Ulysses Paiva
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostWed Jun 10, 2020 7:59 am

A certain manufacturerbid rumoured to have three 40mp+ cameras coming out in one series, maybe starting below $2k, with 8k and raw recording (internally by the looks of it, though I'll believe it when so see it confirmed) in at least one of them. Sept+.

Makes everything a bit irrelevant if actually true, time for next level of Braw.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline
User avatar

antoine

  • Posts: 497
  • Joined: Sat Nov 24, 2018 2:20 am
  • Real Name: Antoine Dornstetter

Re: Is BRAW 444, 422...?

PostThu Jun 11, 2020 7:12 pm

BRAW is most probably 4:2:2
BRAW Studio FREE and Premium for Adobe Premiere Pro and After Effects, with a brand new Desktop .BRAW Player and Color Grader, as well as an automatic White Balance Color Picker tool
Offline

Howard Roll

  • Posts: 2506
  • Joined: Fri Jun 03, 2016 7:50 am

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 11:49 am

From the patent doc it’s YCbCr420. Doesn’t matter though it’s an apples and toast comparison. 4:0:0 is all you need for uncompressed bayer. 4:2:0 is literally make data because they are legally obligated to do so.

Good Luck
Offline

Hendrik Proosa

  • Posts: 3015
  • Joined: Wed Aug 22, 2012 6:53 am
  • Location: Estonia

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 12:39 pm

It isn't apples and toast to compare braw with chroma subsampled video because it literally is chroma subsampled video, not bayer data.
I do stuff.
Offline

Howard Roll

  • Posts: 2506
  • Joined: Fri Jun 03, 2016 7:50 am

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 1:38 pm

Yes, if the 4:2:0 stricly references the demosaiced data which I do not believe it does. From the doc it appears as if the R and B channels still reference raw data. Here is an exerpt where it discusses the formulation of the R and B channels from only real R and B pixels not virtual/demosaic pixels as it does earlier with the luma channel.

"In this case, for pixels where a blue pixel is captured, a chrominance image pixel is generated to generate a blue channel chrominance image. As such, the blue chrominance image includes only pixel values corresponding to the blue pixels of the CFA, this applies mutatis mutandis to the Red chrominance image. Thus the chrominance image is at lower resolution than the luminance image."

At some point there's got to be some raw in Braw, or not.

Good Luck
Offline

John Brawley

  • Posts: 4267
  • Joined: Tue Aug 21, 2012 7:57 am
  • Location: Los Angeles California

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 1:42 pm

Do the pictures look good ?

Am I able to manipulate the white point while grading ?

Does the image stand up to heavy image manipulation in the grade ?

How does it key ?

I just shot a 100 million dollar series using a mix of Arriraw and Braw. About 30% of the show is BRAW.

Anyone want to watch it and tell me which shots are “only” 422 or the chromasubsampled ones ?

It’s alright I’ll wait....

Actually you won’t be able to tell. Which is what matters.

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles
Offline

Gavin_c_clark

  • Posts: 299
  • Joined: Thu Apr 16, 2015 3:51 pm

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 2:34 pm

What channel/streaming service can we see that on in England John?

I don’t care so much for chroma subsampling pixel peeping or generally 96% of this thread as such I just would be interested to see it!
Offline

John Brawley

  • Posts: 4267
  • Joined: Tue Aug 21, 2012 7:57 am
  • Location: Los Angeles California

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 2:48 pm

It’s been on Hulu for the past couple of weeks.



It will be on Starz very soon if not already and channel 4 later in the year.
https://www.starz.com/gb/en/series/5797 ... s?season=1

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 3:51 pm

Howard Roll wrote:From the patent doc it’s YCbCr420. Doesn’t matter though it’s an apples and toast comparison. 4:0:0 is all you need for uncompressed bayer. 4:2:0 is literally make data because they are legally obligated to do so.

Good Luck


Yep pretty much all I was saying it would be. One does, another does design.

I took a sleeper hours ago and the difference between the clips in the video is obvious. I may not be able to tell much which camera shot which anymore with the improvements and grading in my weakened state, but not always.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

Oyvind Fiksdal

  • Posts: 390
  • Joined: Sun Jul 02, 2017 7:12 pm

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 4:46 pm

John Brawley wrote:It’s been on Hulu for the past couple of weeks.



It will be on Starz very soon if not already and channel 4 later in the year.
https://www.starz.com/gb/en/series/5797 ... s?season=1

JB


Yes. I can most defiantly see the 30% you have ruined this show with... Its outrageous and a damnation in the lights of motion picture. Why would you EVER, in your right mind, meddle awesomeness with waste?

But seriously. This show looks really good. It’s a good testament that the tools we have today is just that. Tools that can make really good content even though they differ some. I’m a strong believer in Steve Yedlin’s testament that most RAW cameras, these days, are basically good enough to achieve any look or feel. Its more about framing, lenses, lighting and post work.

One thing that sometimes is forgotten. File size. How to handle files when data becomes big. ARRIRAW is awful in this regard. That’s where BRAW truly shines. There is a notion that big companies don’t care about file size. That’s wrong, IMO. RED code has proven that to me. And did we actually notice any difference in the output quality? No, we are not able to tell.

It’s like convincing kids that pure lemonade taste much better without mixing water into it. "It’s pure and tasteful! 100% RAW. 444" When the fact of mater is... sour at the end. I don’t think most “kids” like that... but they have to taste it before believing it! BRAW is defiantly NOT equal to 444, but the output quality tastes good regardless. So why should we care? What do we actually gain by knowing that it's more like 422? Is that a dealbraker?
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 5:28 pm

Howard Roll wrote:patent doc it’s YCbCr420.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 5:53 pm

I have again asked myself recently, why not just put primary sensors in an square array, but with one white, red green blue, and 3D calculating what each point in the image looks like in each primary from the overlap. You could use complementaries, you could use five colour+ hybrid array. You get maybe 20-30 stops, fantastic noise removal for a clean image from.all the se sensors, and 3D mapping, to produce a pure 4:4:4 that out does foveon and low light with the monochrome sensors help. If you want to be professionals you have to give up on these Bayer single chip consumer chips. High bit depth HDR with high frame rates 4:4:4 low noise compresses well as a bonus. On 8k or 4k combined single recorded frame footage from the camera, it won't be three times more than Bayer normally. Each time you add one of these more compressible features the compression rate for the same quality goes up. 16k+ chips will be coming, which helps in pixel value alignment, but at 8k, the averaging between sensor blue alignment is more acceptable, because at 8k you are falling below the visual colour resolution, unless you sit up close to the screen, even at 5k you can expect reasonable results. I originally speculated if Bayer without demosaicing, would be useful at 8k.

Is 4:4:4 Braw more realistic?

Thanks.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

John Brawley

  • Posts: 4267
  • Joined: Tue Aug 21, 2012 7:57 am
  • Location: Los Angeles California

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 5:59 pm

Oyvind Fiksdal wrote:

But seriously. This show looks really good. It’s a good testament that the tools we have today is just that. Tools that can make really good content even though they differ some. I’m a strong believer in Steve Yedlin’s testament that most RAW cameras, these days, are basically good enough to achieve any look or feel. Its more about framing, lenses, lighting and post work.


Thank you for the kind words. That is the point really. We get a bit lost in the minutiae of the tech and forget to step back and ask...how does it LOOK and how does it WORK (in terms of workflow on set and in post.

Oyvind Fiksdal wrote:One thing that sometimes is forgotten. File size. How to handle files when data becomes big. ARRIRAW is awful in this regard. That’s where BRAW truly shines. There is a notion that big companies don’t care about file size. That’s wrong, IMO. RED code has proven that to me. And did we actually notice any difference in the output quality? No, we are not able to tell.

It’s like convincing kids that pure lemonade taste much better without mixing water into it. "It’s pure and tasteful! 100% RAW. 444" When the fact of mater is... sour at the end. I don’t think most “kids” like that... but they have to taste it before believing it! BRAW is defiantly NOT equal to 444, but the output quality tastes good regardless. So why should we care? What do we actually gain by knowing that it's more like 422? Is that a dealbraker?


Yeah workflow is very important. This was my first BRAW show and it worked remarkably well considering how "new" the codec is. We had no trouble at all working with the files, including VFX handoffs.

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles
Offline
User avatar

rick.lang

  • Posts: 17173
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 7:03 pm

Was that BRAW Q0, John?
Rick Lang
Offline

John Brawley

  • Posts: 4267
  • Joined: Tue Aug 21, 2012 7:57 am
  • Location: Los Angeles California

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 7:13 pm

rick.lang wrote:Was that BRAW Q0, John?


Yeah that's what I went for because I could. But I have been using Q5 a lot.

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles
Offline
User avatar

rick.lang

  • Posts: 17173
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 7:35 pm

I used ProRes 444 on the URSA Mini 4.6k and BRAW Q0 on my BMPCC4K for my spec movie pilot. As you say, if I can manage it, it’s my preference.
Rick Lang
Offline

Oyvind Fiksdal

  • Posts: 390
  • Joined: Sun Jul 02, 2017 7:12 pm

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 9:14 pm

Ok. I haven’t been reading everything in this thread. So, hope I don’t step over. But. IMO.

No doubt that q0 have the edge over q5. It’s foolish saying otherwise. But in most cases q5 hold up even for FX work. There are other problems more pronounced than q0vsq5, like lens distortion/artifacts and bad lightning.

Nevertheless. I agree with Rick, regarding compression. Simply... shot with the best option available when possible. Like Q0 instead of Q5. That seem true to me. I’m a nerd. so, I can’t stop myself agreeing with this. Many of us get obsessed with information. But in the end, with today’s technology. It’s a curse. Truth is. That extra data is seldom needed. It’s a force pulling us toward that sharp RAW “lemonade”. Even if we feel the bitter taste afterwards realizing somethings off. You want to be safe, you want a guard, a backdoor, the best of the best. And that’s the nature of the OP question. Do I have my back covered with BRAW?… Short answer: YES. What you lose with BRAW compered to uncompressed RAW (ARRIRAW) is not a dealbreaker. Wait and see… ARRI will come with compressed RAW. And we “all”, ironically, know why.

And for those who are thinking “wait…but the medium becomes faster, better and cheaper. So eventually we can all shoot uncompressed RAW!”. STOP. That’s a FALSE argument. Because its not proportional to the rest of the equation(capitalism). Steve is right when he says that 20k-30K resolution is coming. It’s all about capitalism at that point. Bigger and better…BUY! Meaning. You need even faster and bigger discs etc. It’s the same with any business. Same with cameras. BRAW is almost a cheat in this regard. Like RED code. It cheats the system and give us an opportunity to be even more efficient when “we are not supposed to”. Use BRAW and you will not think about this 444, 422, 420 nonsense. You will be thrilled with the speed and IQ you get and soon forget what you thought matter…. I feel like a Nemo I Matrix now.
Offline
User avatar

rick.lang

  • Posts: 17173
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Is BRAW 444, 422...?

PostFri Jun 12, 2020 11:50 pm

Thanks, Oyvind. I don’t plan on moving to 8K (famous last words), but I’m not at all concerned shooting UHD and DCI 4K now. I think BMD has been forward thinking all along introducing the BMPCC6K to deliver 4K. But for now I’ve many options on the BMPCC4K to deliver my usual 2K and HD.

BRAW is another example of that forward thinking and I think the film industry will give it a fair chance now that they’ve seen it used on The Great. It’s nice to no longer worry about optimizing media for that matter as the Mac Pro digests everything I’ve thrown at it up to this point. Sadly I haven’t been able to do any new client work yet but that’s another story
Rick Lang
Offline
User avatar

Uli Plank

  • Posts: 21278
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Is BRAW 444, 422...?

PostSat Jun 13, 2020 1:13 am

Q5 and Q0 are actually even overlapping a bit. A very low detail, low contrast scene in Q0 will not have more data-rate than a very high detail and contrast shot in Q5.
That's why there are no choices like Q1 to Q4. Not needed.

BRAW is a damned clever codec, IMHO. And not more load on my machine than non-RAW ProRes.
No, an iGPU is not enough, and you can't use HEVC 10 bit 4:2:2 in the free version.

Studio 18.6.5, MacOS 13.6.5
MacBook M1 Pro, 16 GPU cores, 32 GB RAM and iPhone 15 Pro
Speed Editor, UltraStudio Monitor 3G, iMac 2017, eGPU
Offline
User avatar

rick.lang

  • Posts: 17173
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Is BRAW 444, 422...?

PostSat Jun 13, 2020 5:19 am

Correct, Uli. Q0 wants to see contrast and edges and if a scene is very flat, data rates will be lower.

Maybe that’s why BRAW Q0 is a good fit for a cinema camera where the images that are most important often have many cinematic indicators to heighten their importance to the viewer including some areas of key contrast with very important details to be revealed.
Rick Lang
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostSat Jun 13, 2020 5:42 am

rick.lang wrote:Thanks, Oyvind. I don’t plan on moving to 8K (famous last words), but I’m not at all concerned shooting UHD and DCI 4K now. I think BMD has been forward thinking all along introducing the BMPCC6K to deliver 4K. But for now I’ve many options on the BMPCC4K to deliver my usual 2K and HD.

BRAW is another example of that forward thinking and I think the film industry will give it a fair chance now that they’ve seen it used on The Great. It’s nice to no longer worry about optimizing media for that matter as the Mac Pro digests everything I’ve thrown at it up to this point. Sadly I haven’t been able to do any new client work yet but that’s another story


Thay are just doing what Red and Arri did, overres and down scale, it up svdme in the case of fuzzy Alexa, very flexible. But that and reframing is what 8k-32k is about. I could use 120 degree+ lens and not have track on 32k camera. We are coming to a point that a group of actors could do it all using camera stands and post. The studios are going have a feild day, they control distribution in cinema "Crew, what crew! a they chuckle, thinking of the extra dollars profit on $30 tickets, or something. They can just pick actors they know can compare a scene, and acting, and let them go. Hi, make us a movie, here $200k and 1% on sales. See you to 2 weeks or months, and send them out to make some cheap ass hit Drama with quality acting and quality story and script, chuckling at the suckers, praising each other for the scheme to get the studio back into profit, while they try to pay the "staff" as little as they can. I've been working out how to get to do it with 80-90%+ less production staff, even rspid automated set construction and resetable sets. Of course, it may never come to that, but it is in itself a quality story subject for a movie. :)
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostSat Jun 13, 2020 6:28 am

So, I can shoot feature films 8k with 43 mbit/s or so, on a phone. But, it's about selling extra quality. It's really about how a dramatic production makes one feel. So, post 8k resolutions areil also about preserving cinema lens characteristics that give the picture it's charm, according to corner light horse CTO (boy that guy really goes out there and finds it :) ). I personally think, once you hit 8k, you have enough to do a lot. You can go more, if you want too. 32k is going make nice 8k pixels, as 8k should make nice 2k pixels, but, a foveon 8k setup is going be less data than either a 16k or 32k, why bother if you are going frame 8k or post frame 2k? Sure, low light might not be as extreme, and you might need Kubrick's trick, but that's little price to pay for 3/4 the data of 16k and 3/16 the data of 32k.

Uncompressed data at the extreme end is a bit of a blessing. Image compression takes up significant hardware resources, but it's simply not very easy to do storage for. But future storage could put an uncompressed feature in a microdot maybe. Maybe, one day. But I look forwards to designing something that gets near that, but I've been nice and supportive of people in the real world, as some of you know, but the real world is exploitative and rude, for no reason but themselves, and the authorities give tacid? approval to their behaviour. So heart braking, and don't matter what you do, you can't break through to them to improve the world and make it a happier place for people.

Transmission of data has just taken a huge leap. In Australia they are testing out a new device over local fibre optic lines. What's that 40 Terabits. Just on an 8k extreme image processing file I think I calculated a few Terabits or terabytes per second. Of course extreme 8k holographic is like 96k then overes and 180 degree is 576k or so at over 16 bits, at 200 or more fps. Even in delivery at 96k, Ivfint see present optics being of much use without compression. Of course I would construct an optic with millions of channels, which I realised how, but have another scheme I am interested in.

The industry is far behind where it should be.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline
User avatar

rick.lang

  • Posts: 17173
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Is BRAW 444, 422...?

PostSat Jun 13, 2020 3:39 pm

Wayne, having a very wide-field image in 32K with fixed position cameras without any crew once the system is installed sounds perfect... for surveillance. But for most creative purposes, especially narrative films, of almost no value. There are those who rely on very precise framing to 6K downscaled for 4K deliverables which is going to get us through the next five years at a minimum I think. For those who like to reframe in post, 8K makes a lot of sense for 4K deliverables

Capturing in 16K/12 follows those use cases if delivering for 8K distribution. With the onset of 8K consumer television, there’s a niche market that may grow but as you know for most consumers, entertainment room sizes mean it’s unlikely the majority of consumers will ever go beyond 4K television screens in the next ten years if ever.

When you grow up, Wayne, you should consider a career in cosmology string theory as you’re one of a very few individuals in the world that can ‘visualize’ 11 dimensional space-time and I can see right now before me your draft 11,592 page PhD dissertation on proving the existence of 22 dimension strings!
Rick Lang
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostSat Jun 13, 2020 5:40 pm

Why should I use the report buttons the first time, on my who cares change, on you?

Seriously. You are not seeing the big picture, and yes the setup maps 3D space within it for most camera movements, plus there is another twist that goes to the next levels. So, while it may appear limited, I can assure you that maturity of design sees what is happening.

Now, strings are rather old hat, seems to be taking many attempts to prove. When people don't know what they are doing they keep adding more complexcity than the minimum needed to delay the inevitable failure of proof of efficiency. Rather like a banged up 350 inch 8 with roots super charger nitro the kitchen fridge, whatever on the old jalopy, rather than one of the simpler setups, or a Tesla that blows it, and maybe get similar milage. :)

I was thinking today the string itself might be an issue, a wrong notion that holds the theory base back. As you know, I have long worked on my own theory, which oddly at times sounds similar to the others. I mean the cause of gravity they are pushing works in with causing the mechanism I'm looking at, which is more implicit without needing gravitons, is that what they mean, but both have a similar effect, where you would maybe experiment to see which it is. Good to have 10%+ of my capability back.
Last edited by Wayne Steven on Sat Jun 13, 2020 5:54 pm, edited 1 time in total.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostSat Jun 13, 2020 5:50 pm

The industry is immature, they keep trying to reinvent the V8, when the future is vastly different. As long as they keep talking about the V8 they did when they were kids, they are not advancing to hybrids, flying cars, anti gravity space ships, whatever. We should be 100-1000 years ahead by now, we could have been here thousands of years ago, but all the party boys have been in charge wrecking things for their own ambition. Unfortunate.

The practical convenience, and bottom dollar is what is going to lead to things. Marketing is a bit hard to waste lots of money on in the end. Look at Arri and Red. You not got anything until you can deliver better results like dual gain and reliability.
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline
User avatar

rick.lang

  • Posts: 17173
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Is BRAW 444, 422...?

PostSat Jun 13, 2020 8:33 pm

Wayne Steven wrote:The industry is immature, they keep trying to reinvent the V8, when the future is vastly different. As long as they keep talking about the V8 they did when they were kids, they are not advancing...


Yes, I think my view of Ks are within current context rather than a vastly different context which I haven’t foreseen that can transform the game.

Look at Arri and Red. You not got anything until you can deliver better results like dual gain and reliability.


Agree. One has a start but one longs to be at an elite performance level.

Strings: I used to attend grad student physics lectures at UofT in the late 60s and that’s when I began to learn about strings. Trouble was I thought it was all theory and math at that time without a basis in reality or a way of being proven. More than fifty years later the complexity grows but my original misgivings remain.
Rick Lang
Offline

Oyvind Fiksdal

  • Posts: 390
  • Joined: Sun Jul 02, 2017 7:12 pm

Re: Is BRAW 444, 422...?

PostSun Jun 14, 2020 11:38 am

rick.lang wrote:
Wayne Steven wrote:The industry is immature, they keep trying to reinvent the V8, when the future is vastly different. As long as they keep talking about the V8 they did when they were kids, they are not advancing...


Yes, I think my view of Ks are within current context rather than a vastly different context which I haven’t foreseen that can transform the game.


Oh.... I agree..the irony. The K’s are undoubtedly more about capitalism than artistry. But that is also another irony on its own term. The wheel of capitalism is build so that we can earn and spend money. The K’s are necessary in this case, it’s part of the wheel at the moment. Whoever change the game, becomes either winner or loser. It’s a risk. And the “boys” seldom want or even need to take that risk.

That’s what holding it back in most cases and that’s why we should vote for something more inventful.

That is also a bit of the problem with this topic. Some are stucked in 420vs422 etc. The bayer sensor strategy is something else and make it more or less irrelevant. It’s like asking where to full gasoline on an electric car.
Offline

Wayne Steven

  • Posts: 3362
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Is BRAW 444, 422...?

PostSun Jun 14, 2020 11:55 am

But my gripe with it, is you have to store extra drive data over Bayer, reducing the compression efficiency. Forgetting detail issues for the moment, though you need more data for that, some want less drive space (that's life).

BM should do a data store and recovery package for people. Store and database footage, and migrate to newer backups, or archive to someong term format, all fire and event proof cabinets. Using drives these days going to result in losses, so a solution is needed, and BM is basically a production services equipment company, who could model the software/services side in resolve team and hardware side in house. Imagine two $1000 stores (in-house and backup) which you populate with media, for pretty high data protection, with encryption locked hardware to make it useless read and backup only if stolen and put on a new machine, to deter theft. Various people would go nuts for it. It could be a turn key auto sort these content product. Actually, every YouTube channel needs at least one. :)
aIf you are not truthfully progressive, maybe you shouldn't say anything
bTruthful side topics in-line with or related to, the discussion accepted
cOften people deceive themselves so much they do not understand, even when the truth is explained to them
Offline

cj-adams

  • Posts: 35
  • Joined: Fri Jun 12, 2020 12:43 pm
  • Real Name: CJ Adams

Re: Is BRAW 444, 422...?

PostSun Jul 12, 2020 4:11 pm

Ok..One thing I would love answered.. is.. back in the day we used to want 4:4:4 or 4:2:2 for doing things like green screen keying.. etc. we WANT that chroma info. So when using Braw.. is this almost a 4:2:0 situation.. just like whats coming out of my canon dslr?

1 - Is Braw actually going to give me something i can work with?

2 - Do i need to do Q0 or would 3:1 @6k be fine for 4k finish?

3 - Any advantage of using Prores on the BMPCC6k instead?

Thanks
CJ
Offline

John Brawley

  • Posts: 4267
  • Joined: Tue Aug 21, 2012 7:57 am
  • Location: Los Angeles California

Re: Is BRAW 444, 422...?

PostSun Jul 12, 2020 4:49 pm

cj-adams wrote:1 - Is Braw actually going to give me something i can work with?


Yeah. Functionally the same as RAW....

cj-adams wrote:2 - Do i need to do Q0 or would 3:1 @6k be fine for 4k finish?


I have been using lessor settings with great results. I would only go to lower compression ratios when it was for super critical resolution work...like compositing or for plate shots, or for very large screen / cinema.

I'm a fan of the Q (variable) settings, because it usually gives you a bit more space on your cards for similar imaging outcomes. Q5 is amazing.

It's really a smart codec....I haven't been burned by it in any way I can say after months of daily use against an Arri RAW main camera....

cj-adams wrote:3 - Any advantage of using Prores on the BMPCC6k instead?
CJ


Depends WHICH proRes. If you have a UMP for example that can shoot 12 bit 444 ProRes, then it's very good, nearly as good as most BRAW.

If you have "only" a P6K or P4K that only offer 10 Bit 422 ProRes then BRAW is better in every way image wise.

Workflow though could mean there's an advantage to ProRes in post, but it really depends on the post workflow. For example If I'm on a show and the data wrangler is using something to transcode dailies that doesn't read BRAW. Luckily that's not happened to me, and BRAW is becoming very widely accepted.

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles
Offline

cj-adams

  • Posts: 35
  • Joined: Fri Jun 12, 2020 12:43 pm
  • Real Name: CJ Adams

Re: Is BRAW 444, 422...?

PostSun Jul 12, 2020 4:51 pm

John - thanks for the explanation. in my case green screen work and fur detail. Does this change the recondition?
Offline
User avatar

antoine

  • Posts: 497
  • Joined: Sat Nov 24, 2018 2:20 am
  • Real Name: Antoine Dornstetter

Re: Is BRAW 444, 422...?

PostSun Jul 12, 2020 5:53 pm

I haven't made any test but theoretically ProRes 4444 will contain more chroma information so better for green screen keying yeah
BRAW Studio FREE and Premium for Adobe Premiere Pro and After Effects, with a brand new Desktop .BRAW Player and Color Grader, as well as an automatic White Balance Color Picker tool
Offline
User avatar

antoine

  • Posts: 497
  • Joined: Sat Nov 24, 2018 2:20 am
  • Real Name: Antoine Dornstetter

Re: Is BRAW 444, 422...?

PostSun Jul 12, 2020 5:54 pm

John Brawley wrote:
cj-adams wrote:
It's really a smart codec....I haven't been burned by it in any way I can say after months of daily use against an Arri RAW main camera....

Have you done chroma keying ? That's the question being asked :)
BRAW Studio FREE and Premium for Adobe Premiere Pro and After Effects, with a brand new Desktop .BRAW Player and Color Grader, as well as an automatic White Balance Color Picker tool
PreviousNext

Return to Cinematography

Who is online

Users browsing this forum: No registered users and 105 guests