Pages: 1 [2] 3   Go Down

Author Topic: Notes on "The Jargon of Video"  (Read 73502 times)

D Fuller

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
    • AirStream Pictures
Re: Notes on "The Jargon of Video"
« Reply #20 on: July 16, 2015, 06:57:21 pm »

The ProRes codec is called ProRes4444 because it has the ability to contain an alpha channel as well as 4:4:4 color. The Panasonic records an 4:4:4 signal using that codec, but does not produce an aplha.
Logged
business website: www.airstream.pictures
blog: thirtynineframes.com/blog

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: Notes on "The Jargon of Video"
« Reply #21 on: July 16, 2015, 08:37:25 pm »

That's my understanding too, thus
My surprise to see an alpha channel in capture.

The question is that if this not recorded channel
Acts as a zero data one (bypassed) or if it adds size to the file?

What's the point? The only utility I see is that a fx artist
Could use the files as it. But fx artists don't like very much
Those codecs don't they. Here they all bark if you
Don't deliver EXR.

Really I don't see the benefit of this prores flavour.
There must be one but I can't see it. And is it RGB?
Logged

D Fuller

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
    • AirStream Pictures
Re: Notes on "The Jargon of Video"
« Reply #22 on: July 16, 2015, 09:34:52 pm »

ProRes 4444 is the name of the codec. It's named that because it is capable of holdinng an alpha channel in addition to 4:4:4 color. (I believe it is the only one of the ProRes codecs that can carry an alpha channel.) But that doesn't mean it has to carry an alpha channel.

So the Panasonic uses a Codec named ProRes 4444, but the video data it puts into that file is 4:4:4 color. No Alpha.

Similarly, ProRes 4444 is capable of holding 16-bit color, but you can put 8-bit color into it if you like. The file will be smaller.
« Last Edit: July 16, 2015, 09:40:48 pm by D Fuller »
Logged
business website: www.airstream.pictures
blog: thirtynineframes.com/blog

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: Notes on "The Jargon of Video"
« Reply #23 on: July 16, 2015, 10:32:22 pm »

Thanks Daf.
Logged

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: Notes on "The Jargon of Video"
« Reply #24 on: July 17, 2015, 07:15:13 pm »

Aja made a camera?

On micro 4/3, I think it's a dead-end road.

On Lightworks:

Yeah, this bloody editor has won with Avid, more oscars
Than all the competion together by a huge difference.
Serious editors, at the very high-end are cuttin with it.

BUT...but but but...

It's a specialist's tool. Versatile? No. It's IMO the very
Best NLE for cuttin stories, and specially feature. In this,
It has no rivals and the way you work on it is completly
Different than any other NLE currently available. It gives
You a freedom and flexibility that you don't find anywhere
Else. And it's designed FOR the editor. Period.
The 2000 euros console is worth every penny.
Damn good in multicam too.
Everything in its implementation helps the story-teller.

And this is where the film ends and the director says "cuuut"...
LW needs a "old fashion" colaborative workflow.
This why it's almost inexistant in small houses because
They have better options out-there.
The specialists are not common, and LW is not for everyone.

Now, it's that good at what it does that once you start to get
Used cutting on it, you can't touch a normal NLE anymore.
All LW editors say the same story. But the "elite" that
Work on it has full editshare support  (monthly expensive)
And they don't do anything else. It's cine stuff. i love it,
But I have to say that for the ones who need to be multitask,
It's definatly not a good choice.

On Avid:

Being discrete,
Avid is given the battle for the mass market lost. They know
BM is unstoppable.
They actually don't target anymore the market share in wich
BM is winning. Resolve will become the Photoshop of motion.
No one can stop this. avid targets the big productoon houses,
Broadcast, live...and they will be less and less involved
Into the segment we are in.
May they try to attract revenues and "kid" generation with a free version,  it's not going to work. Realisticaly, probably will center
In the future on their broadcast marketplace. They are already strong in it where resolve is not going to put the mess.

IMO, the best options would be or Premiere Pro, or Resolve. (yeah...after effects...layers...)
Doing everything from A to Z in one of those platforms.
PP ingests almost everything, included ArriRaw native. I hate the package but it works very well to be fair.
If you don't mind roundtrippin to color-conform, Lightworks.
But it's very different and only makes sense if you put the editing tasks at the top of your priorities.
Now, yes...Lightworks IS intuitive. In a few days of practise
You're done.


Avid is unintuitive. Completly unintuitive. And as I pointed, they may not be suitable for the segment
We are in for very long time.

And something that pissed me off, when they sacrified DS,
The vendors were still pushing for newcommers to get into
Until the last minute. "buy this car that the company is
Going to stop producing in 3 months". Very stock-market
Kind of attitude.


Ps: LW has Fusion integration, and that is really powerfull
When you need to compo.

Ps2: a funny thing that is happening: the colorists, are now
Stepping away more and more from Resolve. Because they
Don't like the fact that it's becomming an all-in-one.
The competition, Baselight, Nucoda, Mistika, and Scratch
Are pushing the developpement to their color apps while Resolve
Is investing a lot of ressources to be the swiss army knife.
We will see more and more colorists using other platforms.
But BM knows that. They don't want the color artist on
An expensive standalone Baselight or Nucoda restauration stuff or the oscar winners editor market, they want to rule the WW multitask. The shooter
Who edit, color and deliver.
Prices will jump-down because of that. The more revenues
For BM, the less revenues for the small studios. Mathematic.
I hope everybody is aware and preapared for that.

Cheers.

D.F,sorry, this has gone way out of topic.
« Last Edit: July 17, 2015, 08:49:24 pm by fredjeang2 »
Logged

John Brawley

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 77
Re: Notes on "The Jargon of Video"
« Reply #25 on: July 18, 2015, 04:11:08 am »



So I imagine that DNG files from BMC are actually
444 in 8bits like the tiffs? And not 10bits.



Blackmagic DNG's from any of their cameras are 12 bit LOG encoded files that generally unpack as 16 bit linear files in an application like Resolve.

They are far far better than 8 or even 10 bit tiffs. DNG is a tiff file basically by the way, with some extra metadata.  cDNG is the motion version of a DNG which is also a DNG with timecode and audio.

They are unencoded and therefore do not have a colour sub sampling applied to them.

Colour sampling or chroma subsampling is the various ratios that get applied to how many times the luminance channel is "sampled" Vs the two colour difference channels.  

That is what YUV is. (sometimes also called component video)  It's fundamentally different to RGB which is what most sills photographers would be used to working with and what the BM camera stores DNG files as.

When we talk about colour sampling, we are talking about what's known as "encoded" video.   It's in a video form ready to edit, do post work, watch etc.  DNG's are basically like stills RAW photos.  

So in YUV form, the Y channel is the brightness or luminance channel and is ALSO the green channel.  It gets summed or subtracted from the two colour difference channels (the U and the V part of YUV) so therefore we're creating a video signal by taking the BRIGHTNESS signal and then multiplying or subtracting it from the brightness signal of the Red and Blue signals to generate the other colours.  

This was a kind of way of having compression before data compression was around in the analog days, the logic being that green was the colour we're most sensitive too, so all the resolution was stored there and the colour could then be stored in two smaller less frequently sampled channels.

RBG has the brightness encoded within each individual colour channel.  YUV does not.

https://en.wikipedia.org/wiki/Chroma_subsampling

RAW video cameras like those from Blackmagic and Arriflex have un-mosaiced DNG files, so it's not really the correct terminology to describe them using video encoding terms because they arguably haven't been encoded yet.  There is no 4:2:2 or 4:4:4 until they get turned into YUV video.

Many many many people also confuse that ratio with the CMOS bayer sensor ratio of green / blue and red pixels and assume they mean the same thing.  There are many heated discussions on video oriented forums about CMOS sensors needing to oversample to be able to generate 4:4:4 video because they don't have enough pixels to generate "full bandwidth colour".  Because the ratio of RGB photosites tends to be twice as many greens to blues and reds you get a lot of people sating CMOS sensors only have 4:2:2 colour which is also the wrong terminology completely (and ignores what happens during the mosaic de-mosaic process.

It's a RATIO for the ENCODED video

Sony, back when they introduced this novel new HD format called HDCAM that got used to shoot this new Star Wars film called the Phantom Menace was originally a 22:11:11 format ! Which compared to the original SD it was.  

So that nomenclature now is very very basterdised and misunderstood and has pretty much lost it's original meaning :-)

Other RAW cameras like RED and SONY do other secret things that make it harder to work out what's going on. RED have an SDK that does some secret stuff and until recently, you couldn't get anyone from RED to actually commit to what the bit depth of the cameras was. (it's apparently at least 16 bit internally now, but was originally probably 12-14 bit)

RED also famously sued Sony a little while ago when Sony introduced their RAW format for the F55 / F65.  From what I could tell reading the papers lodged, RED have a patent on the compression they do because they convert to the RBG sensor data to YUV FIRST and then apply different encoding and compression techniques to the Y signal from the UV signals.  Sony tried the same technique of compressing the half encoded YUV signal and got sued.  They later settled out of court.

By the way, someone was mentioned the Metadata being carried by RED ?  DNG's have metadata embedded in the file itself as well, and It's up to the application handling the file to then interpret it.

JB

EDIT

here's a link to some DNG's from the little POCKET cinema camera.  It's "only" a 1920 sensor size so tiny, but i think you'll find a camera that can shoot 24 of these every second is pretty compelling.

https://copy.com/s36D39T6q7oa

(these are quite old and from a prototype pocket camera, but you get the idea.




« Last Edit: July 18, 2015, 04:21:18 am by John Brawley »
Logged

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: Notes on "The Jargon of Video"
« Reply #26 on: July 18, 2015, 05:34:30 am »

Even out of topic for awhile now,
This thread is tremendously interesting,
Thanks to the apportations of all you guys.

Clearly explained, it helps to clarify points in those
Technical areas.

I have a Blackmagic camera in order. Will receive it
From Germany very soon.
Never had the oportunity before to work with cDNG
And know nothing on the workflow in post.
It will be a complete discovery from scratch (no pun intended)

So in terms of looking for a workflow with BM,
I guess it should't be very different from SDK.
« Last Edit: July 19, 2015, 01:41:39 pm by fredjeang2 »
Logged

John Brawley

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 77
Re: Notes on "The Jargon of Video"
« Reply #27 on: July 18, 2015, 08:45:33 am »


So in terms of looking for a workflow with BM, I don't
Really trust internet examples of hobbyists and I need
To find in the middle of the mess, where are the
Reliable examples of workflow pipelines.
I guess it should't be very different from SDK.


What do you want to know ? :-)

In all honestly, most of the time people who aren't making a living from a camera will find shooting RAW to be onerous.  It's very taxing,  It takes extra time and resources to manage and get through.

What is really nice though is that the BM cameras have RAW as well as several iterations of ProRes (and sometimes DNx) so you can switch between the codec you want, even on the same card !  So you can just choose what you need for each shot basically.

I'm about to start shooting an AMERICAN Science Fiction TV show with lot's of VFX...guess what format we're shooting ?  Not RAW, not even 4K, but 1920 ProRes.  And this is a big show.

Very very few have the patience and resources to deal with RAW for a whole shoot.  Most don't want to deal, nor do they get enough of an adavantge to outweigh the baggage that comes with RAW shooting.

So, if you have the TIME and resources in terms of horsepower with your IT side of things then knock yourself out.  But most be find things get a teeny bit better than ProRes for a lot more overhead.

JB



Logged

Kevin Gallagher

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 962
Re: Notes on "The Jargon of Video"
« Reply #28 on: July 18, 2015, 09:28:45 am »

 I just wanted to pop in here and say how refreshing it is to read a thread where there is a civil and mature exchange of ideas rather than the usual genital waving (mine is bigger and I'm smarter because I've known him longer and I've written some silly article) that goes on in some quarters. Not to mention how much this neophyte to video is learning from you guys!!


Kevin in CT
Logged
Kevin In CT
All Animals Are Equal But Some Are More Equal
 George Orwell

John Brawley

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 77
Re: Notes on "The Jargon of Video"
« Reply #29 on: July 18, 2015, 09:19:40 pm »


Compared to shooting a c-log and adding a lut, I find shooting raw about the same process, maybe even easier.


Totally.  I think if you're shooting LOG then in terms of post, RAW is just the same steps with a different starting point and larger files.  Some cameras even load their RAW files to a "log" looking start position cause that's what people are used to :-) 

In Resolve for example, with BM DNG's, there's a RAW camera tab and a little check box where you can choose "BMD FILM" which is their take on LOG.  But you don't have to start there, and Resolve (for example) still maintains the same precision, just working from a different curve.


I can see proRes capture if it's episodic and their doing the base edit on stage or in a trailer on set as they're shooting.


Yes.  I typically shoot at least 200 mins per day of rushes and typically more like 300 mins on bigger days, across several cameras.  And that's day in day out, over weeks.  It quickly becomes terabytes of information.

Not saying either way is better, but I see people all the time complain about raw, then shoot some flavor of codec in a flat log, then the editor has to apply a lut and either let the nle transcode in the background or transcode prior to editorial, which has to be done anyway if they produce dailies.


Yes so in episodic drama all the editors use AVID because they have the shared storage so much better set up and organised.  Typically we transcode all rushes to DNx for editorial and 264's for the various producers and TV network peeps to look at as well as dailies for us  on set.

At the end of the day it drops on the colorists to try to get it all matched and sometimes that's difficult.

I'm not defending RED or saying shooting a proRes is a wrong way, I just think raw doesn't really slow us down, it actually allows us to speed up the workflow in the longterm as we usually come out with a pretty close proRes from either CineX or Resolve prior to editorial and depending on the production, whether you edit from a proxy, a flat log file with a lut, or attempt to edit directly from the raw, the time involved always seems to be the same.

Maybe it's because I oversee the process much like a still photographer.



Indeed.  RED has always suited owner operators better.  And those that shoot / direct.

I don't have any skin in the game and like to be camera agnostic and use the best camera for the job, but I have used RED on many jobs.  These days though I find them pretty difficult to deal with as a company.  I had some major issues which I'm not allowed to talk about because I had to sign an NDA before they would fix them.  That kind of attitude isn't really the way I like to work.


I have noticed that Resolve has a much richer look to my R3d files, though Cine-x seems to hold more detail and a little more range, (but that's just an opinion, nothing based on fact).

Welcome to the club.  Users of Sony RAW and Arri RAW have noticed the same thing.  Like the differences between C1 and ACR, Resolve's de-bayering algorithm is very very very good and preferred by many and many in HOLLYWOOD like Resolve for this reason.

I have coloured using Resolve on Red ROCKET based systems for real time performance, but on output we switched off the RED ROCKET and use the software based RESOLVE de-bayer and it's much much nicer (though non realtime off course).

So even with the camera manufactures themselves OWN software, you can often get better results using other applications, and that's what I like about open formats like DNG. 
 
My take on motion/video/film whatever we call it now, is there is no free lunch.

Where matching color and look is an issue for us is this "real" lifestyle look where we go inside and out, changing countries and cities and then trying to match.  

Usually on a brutal schedule.


But I'd also argue that when you're on a punishing schedule is when you WANT the salvation of RAW....It lets you hide the sins of the shoot... :-)

JB






Logged

D Fuller

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
    • AirStream Pictures
Re: Notes on "The Jargon of Video"
« Reply #30 on: July 19, 2015, 12:18:08 pm »

I think it might be time for a new thread. :-)
Logged
business website: www.airstream.pictures
blog: thirtynineframes.com/blog

Christopher Sanderson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2692
    • photopxl.com
Re: Notes on "The Jargon of Video"
« Reply #31 on: July 19, 2015, 10:11:10 pm »

I think it might be time for a new thread. :-)

or not... this thread is very valuable and my personal thanks to all those who have contributed. It has made for great reading.

Perhaps we just shorten the thread title to "On video..."

D Fuller

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
    • AirStream Pictures
Re: Notes on "The Jargon of Video"
« Reply #32 on: July 20, 2015, 03:04:42 pm »

OK, then. If it's proving valuable to people, let it continue.

When I started the thread, it was really with the hope that it might persuade Michael Reichmann to correct his "Jargon of Video" article. There is so much wrong information about video out there on the web, and LuLa is such a respected source that it really ought to be a dource of accurate information about photography...

But since then, the thread has taken on a whole different direction, and if people are getting good stuff from it, then rock on.

DAF

or not... this thread is very valuable and my personal thanks to all those who have contributed. It has made for great reading.

Perhaps we just shorten the thread title to "On video..."
Logged
business website: www.airstream.pictures
blog: thirtynineframes.com/blog

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: Notes on "The Jargon of Video"
« Reply #33 on: July 20, 2015, 03:44:49 pm »

OK, then. If it's proving valuable to people, let it continue.

When I started the thread, it was really with the hope that it might persuade Michael Reichmann to correct his "Jargon of Video" article. There is so much wrong information about video out there on the web, and LuLa is such a respected source that it really ought to be a dource of accurate information about photography...

But since then, the thread has taken on a whole different direction, and if people are getting good stuff from it, then rock on.

DAF


Yeah, the thread as started to have its own dynamic and truth that it was not anymore really the topic for wich you first wrote.
I apologize to DAF for been part of hijacking it and in the end when I realized that I was completly out of topic for so long I deleted my
latest posts because I didn't want to be unpolite to the thread's creator.

But I think it is indeed very interesting for many people, as I found in it very valuable informations-corrections and I'm sure not being the only one.

What I found particularly interesting in the all thread is in fact what I would like to see more often in internet.
Long time experienced cine pros, answering younguer and-or not so experienced people, all in a very educated way.
But more important: everything was very clear.
The most experienced helping the less experienced, clarifying points. It's that way IMO
that real knowledge is passed to the next generation.
I'm the "next" generation, and I feel thankfull when more experienced people take their time to answer and
pass the knowledge.

Yes, probably for very experienced people, this sort of threads may be too long and sort of boring,
but I beleive that many many people have extrated really good informations all along.
And this is priceless.

Thank to all you guys.
 

  
« Last Edit: July 20, 2015, 04:16:05 pm by fredjeang2 »
Logged

D Fuller

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
    • AirStream Pictures
Re: Notes on "The Jargon of Video"
« Reply #34 on: July 20, 2015, 10:15:31 pm »


Does anyone work a non punishing schedule anymore?
LOL! No one I know. Right now I'm trying to work 8 scenes in 8 (apparently) different locations into a budget that is scaled for one day. It's just the way things are going.

What I like about raw is I know I've got most of the data available off the sensor and I hate shooting these flat logs on cooked files, regardless of camera, regardless of lighting and on set control.  Even if you can apply a look to a monitor for a better view.   You just never know what's going to happen in post until you get there.
Totally agreed. IMO it's as close as you can get to the ability to recover a shot that we had with film. The way you can, in Resolve, push the curves around to adjust the pivot points for contrast in a raw file gets me to where I want to be much better than anything I've done with baked footage.

My most common working method is to shoot using RedGamma of some variety (depending on what we agree looks nice) then in post, switch it all to RedLogFilm with a RLF to RG3 LUT as a starting point in Resolve. I find it gives me a lot of flexibility in getting where i want. Also, turning the LUT off and looking at the LOG gamma lets me see all the dynamic range the footage has to offer if I need to.

Like BC, I'm pretty tenacious in my hold on the coloring of the footage. I prefer to work with a colorist who is better than i am, but if the budget doesn't allow that, I will do it anyway, even if there's no money allocated for it. That's where the look is finalized, and it's my reputation.

Those camera maker videos where they show a well known DP shooting a short with a few ___fill in the brand____ cameras and saying how easy it is to hold the detail out the window are so misleading, because nobody knows the time element and budget they had for post, or how many watts they're pulling on the main subjects to balance.

They give the impression that all you need is a camera and some sticks and you too can shoot a movie.  
Maybe the worst offender is Apple. Those "shot on iPad" spots are insane. "You can do this. It's shot on an iPad. On a crane. With an 18K. But you can do it 'cause you've got an iPad..."

Moving on...

P.S.   I have two of the first red rocket cards, keep one in a mac silver desktop, one in a portable enclosure (which only allows for 2k). Now I don't bother because the new macs, even the new Imacs can crunch data like crazy and though not real time, can blow through dailies and proxy's almost into real time from the RED files and though I like Cinex as I think it holds a little more data, Resolve is faster, at least for me.

I agree, and it's not just faster, it has so many ways to approach and think about color that you can find a path that works no matter how you think about color. And I think it has an amazing color engine under the hood. It seems easier to get where I want, even if I don't need secondaries.

DAF
Logged
business website: www.airstream.pictures
blog: thirtynineframes.com/blog

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: Notes on "The Jargon of Video"
« Reply #35 on: July 20, 2015, 11:32:17 pm »


My most common working method is to shoot using RedGamma of some variety (depending on what we agree looks nice) then in post, switch it all to RedLogFilm with a RLF to RG3 LUT as a starting point...//... Also, turning the LUT off and looking at the LOG gamma lets me see all the dynamic range the footage has to offer if I need to.


I certainly agree 100% on this approach. And I smile because got the same "tic" of turning off and on the LUT to check the room available.
« Last Edit: July 20, 2015, 11:34:01 pm by fredjeang2 »
Logged

D Fuller

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
    • AirStream Pictures
Re: Notes on "The Jargon of Video"
« Reply #36 on: July 20, 2015, 11:36:29 pm »

Totally.  I think if you're shooting LOG then in terms of post, RAW is just the same steps with a different starting point and larger files.  Some cameras even load their RAW files to a "log" looking start position cause that's what people are used to :-) 
It's not necessarily that the files are RAW that makes them larger, That might true for Arri, where RAW files are uncompressed, and I don't know a lot about BM, but RED .r3d files are quite a lot smaller than equivalent resolution RGB files. How much smaller depends on what compression you are using. An hour of RedCode UHD (3840x2160) 24fps at 5:1 compression is about 198 GB of data. An hour of ProRes4444 (5.4:1 compression) is almost 3X as heavy--about 536 GB, and ProRes(HQ) (also 5.4:1, but only 10 bits) is 357 GB. You have to go all the way to ProRes (LT) (12:1) at 166 GB to get a smaller file than an .r3d at 5:1.

It's really shooting 4K or 5K or 6K that makes the files heavy, not shooting RAW.

JB, I'm sure you know this, but for those who arre wondering how this can be, it's because a RAW file contains only one value per photo site, but an RGB file contains three (one each for R, G & B).

RED has always suited owner operators better.  And those that shoot / direct.
I'm sure this is true. But I suspect that outside of episodic TV, where sheer volume makes transcoding a barrier, it's more about training people than anything inherent in the workflow. Owner-operators and very small companies can adapt very quickly if they find something that suits them. For a larger company, with a lot of people and stuff, it's harder--more expensive and more complicated. And Red, at least, makes it more challenging by releasing more firmware updates than anyone in history, so the training is never-ending.

Like the differences between C1 and ACR, Resolve's de-bayering algorithm is very very very good and preferred by many and many in HOLLYWOOD like Resolve for this reason.

I have coloured using Resolve on Red ROCKET based systems for real time performance, but on output we switched off the RED ROCKET and use the software based RESOLVE de-bayer and it's much much nicer (though non realtime off course).
Are you sure it's the de-bayer that's better? I was under the impression that Red and Arri control the de-Bayer pretty tightly through their SDKs. In Red's case, they think it is necessary to keep control of the color science. They even shut down Scratch's early efforts at GPU de-Bayer to keep it within the SDK methods.

I absolutely agree that Resolve makes richer files more easliy than anything else I've used, but I always thought it was some magic Resolve was working once the video was in RGB. It is, however, well documented that software deBayer is different from Red Rocket de-Bayer, so doing software-only output from Resolve seems to really make the best of both the RED SDK and Reslolve's magic.

Logged
business website: www.airstream.pictures
blog: thirtynineframes.com/blog

John Brawley

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 77
Re: Notes on "The Jargon of Video"
« Reply #37 on: July 21, 2015, 12:03:20 am »


It's not necessarily that the files are RAW that makes them larger, That might true for Arri, where RAW files are uncompressed, and I don't know a lot about BM, but RED .r3d files are quite a lot smaller than equivalent resolution RGB files.

ARRI RAW is always totally uncompressed and so they do tend to be big.  BM have had uncompressed, but recently changed to a lossless (ZIP style) compression which is 1.5:1 but still exactly the same as uncompressed and they also are now offering 3:1 compression as an option too, which for them is a lower file size that the new ProRes XQ codec.  I know in the future they will probably do higher rates of compression.


And I think that was a large reason that RED was so successful and the true reason they were so revolutionary...They made RAW possible   To go back to 2008 when they first started really shipping to have a way of shooting RAW (Arri weren't really there at that point)





How much smaller depends on what compression you are using. An hour of RedCode UHD (3840x2160) 24fps at 5:1 compression is about 198 GB of data. An hour of ProRes4444 (5.4:1 compression) is almost 3X as heavy--about 536 GB, and ProRes(HQ) (also 5.4:1, but only 10 bits) is 357 GB. You have to go all the way to ProRes (LT) (12:1) at 166 GB to get a smaller file than an .r3d at 5:1.

It's really shooting 4K or 5K or 6K that makes the files heavy, not shooting RAW.


Indeed, and that's what I was saying above, with BM's 3:1 compression and a similar rate in REDCODE, you can get RAW "for the same price" as baked in LOG ProRes files...



 But I suspect that outside of episodic TV, where sheer volume makes transcoding a barrier, it's more about training people than anything inherent in the workflow. Owner-operators and very small companies can adapt very quickly if they find something that suits them.


It's more the case that contrary to popular opinion, 90% of what we see on TV is still shot 1920, and shooting RAW generally forces you to shoot "sensor native" so your origination resolution is higher than your delivery resolution.  Usually it's a great thing, but not when you have hundreds of hours of rushes that need to be accessed across multiple workstations.  It is an onerous amount of IT infrastructure to do that.


Are you sure it's the de-bayer that's better? I was under the impression that Red and Arri control the de-Bayer pretty tightly through their SDKs. In Red's case, they think it is necessary to keep control of the color science. They even shut down Scratch's early efforts at GPU de-Bayer to keep it within the SDK methods.


Yes, it's their de-bayer algorithm and their "precision", because they are 32 bit "float" they can dynamically adjust as needed.  From what I understand (and I'm a DP first not a colourist / engineer) the SDK does the basic matrix conversions aka colour science, but the WAY those conversion are rendered is down to the Debayer / Demosaic.  So, like C1 and ACR working with Canon files, they each produce a subtle but noticeable difference.  It's alos one of those things that end up becoming highly subjective too.

And I think someone else was asking about Using C1 to do processing of motion DNG's earlier ?  Yes it can be done and it CAN look good but it's rather tedious and not worth the effort for most.  Also C1 has no tools for motion based colour correction.  Many of the tools in resolve for example can be animated over time.  So you can draw a window and grade inside that window differently, but you can also then track and move that window over time.  C1 has none of those basic motion tools which you really miss....

I absolutely agree that Resolve makes richer files more easliy than anything else I've used, but I always thought it was some magic Resolve was working once the video was in RGB. It is, however, well documented that software deBayer is different from Red Rocket de-Bayer, so doing software-only output from Resolve seems to really make the best of both the RED SDK and Reslolve's magic.


Resolve is also a very very mature toolset that's been around a long time. They have been the number one choice when money was no object at the top end of town for many years.  When Blackmagic bought them only a few years ago, a base system cost upwards of 500K.  Since version 8 they've been making it easier and easier to use.  V12 is totally amazing now and when I think back to how far they've come.

I started using Resolve in V8 and it was incredibly complex in terms of the UI.  I'm also form the school that really prefers a colourist to be doing the work too, and I'd never needed to lear how to colour correct, but Resolve has now given me a better sense of what's possible and how to TALK to a colourist about what I want.

I urge anyone who hasn't used Resolve to download the LITE version now.  Unlike a lot of other crippled demos, it's basically go not limitations to use it, other than the fact that it's output is limited to 1920 (but you can still work with 4K+ files) and some tools like 3D and noise reduction are disabled.  But it's just as functional as a colour corrector and has amazing tool set for a free version.

JB
Logged

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: Notes on "The Jargon of Video"
« Reply #38 on: July 21, 2015, 12:30:17 am »

On Resolve 12,
Do we know how the roundtripp with Fusion is?
This point is important, as BM also owns Fusion,
Hope they made it as straightforward as with LW.

I have a question on DNG. The metadatas aren't separate, no?

John, you seem, correct me if I'm wrong, to prefer the DNG
Approach than proprietary files.
I wonder why. I really don't see the advantage.
Remember Phase one, Hassy...only Leica went DNG
But it's often true that brands tend to have their own
"magic recepy" like it's the case for Red.
True that Red color science is a bit of a strange animal
But it works very well.
What is the real advantage of DNG?
« Last Edit: July 21, 2015, 12:43:52 am by fredjeang2 »
Logged

John Brawley

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 77
Re: Notes on "The Jargon of Video"
« Reply #39 on: July 21, 2015, 12:59:13 am »

On Resolve 12,
Do we know how the roundtripp with Fusion is?
This point is important, as BM also owns Fusion,
Hope they made it as straightforward as with LW.

I don't use Fusion :-)  I don't know how it is.  

I have a question on DNG. The metadatas aren't separate, no?

No, as mentioned earlier it's embedded in the file itself.

John, you seem, correct me if I'm wrong, to prefer the DNG
Approach than proprietary files.
I wonder why. I really don't see the advantage.
Remember Phase one, Hassy...only Leica went DNG
But it's often true that brands tend to have their own
"magic recepy" like it's the case for Red.
True that Red color science is a bit of a strange animal
But it works very well.
What is the real advantage of DNG?


DNG is an open format.

It's non proprietary.  It means that ANYONE can make an application to open and use DNG files. Right now if i store a DNG on google drive or drop box and open that window in a WEB BROWSER I can see a preview of the image ! You don't need to license or access or work through another SDK from the host camera manufacturer.

That means you're able to start working with an image in its more native form.  You don't really know what kind of special work Sony or RED are doing to their files before they get to you(despite the fact they are "raw".  

Philosophically, I like the idea of being transparent in the way the image is "processed".  With an SDK, it's like you have to put the "film" through a special pre-wash and you don't really know what that pre-wash is doing....

JB




« Last Edit: July 21, 2015, 01:00:59 am by John Brawley »
Logged
Pages: 1 [2] 3   Go Up