Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

HDR Grading

Luke Neumann

Well-known member
Joined
Apr 16, 2012
Messages
819
Reaction score
0
Points
16
Location
Oregon, USA
Website
www.neumannfilms.net
Hey all,

I am trying to put together an HDR grade for a client on a really limited setup and have run into a few roadblocks. I'm wondering if anyone could lend any assistance on the following issues:

1. The Monitor: I don't have access to a really high end display so I thought about purchasing an HDR TV (Sony maybe) and using that to grade up to 1,000 nits. The client will be showing at 1,000 nits and on the same display, so I figure...grade to look good on that display and I should be good to go. This leads me to my second problem.

2. The Grading: I rendered a .DPX out of RedCineX in the REC2020 space and brought that into Premiere Pro. I turned on the HDR Scopes and everything is still squished down in the 0-100 range. The only way to bring it up is by raising exposure that just doesn't seem right. Where in the process does the transformation happen? I did the transcode to REC2020 on a non HDR display, is that the main issue here? Does the RedCineX transcode adhere to the nit range of the monitor you're using? If I completed this process on a 1,000 nit HDR TV, would the DPX sequence come in and be spread out throughout the range of 0-1,000 instead of 0-100?

It's all pretty confusing and I can't seem to find any info.

Thankful for anything anyone might be able to share.
 
50 IRE in HDR world is 100 nits. For a 1000 nits it's around 70-75 IRE. 100 IRE corresponds to 10,000 nits. It's a LOG scale.
As far as the rest goes, honestly if I were you, I wouldn't do it. You are really missing a huge chunk of necessary information and as a result, the outcome will be nothing resembling what is required to deliver a proper HDR-10 master.
But if you really, really, really want to proceed, make sure you look up the PQ 2084 gamma at the required level of brightness. PQ 2084 at 1000 nits is already in Resolve, so there is no need for RCX.
 
Thanks man,

I don't actually own Resolve (I know...I know). In the latest version of RedCineX there is an "HDR-2084" gamma space. Do you have any experience utilizing that gamma? At this point I pretty much have to proceed (even though I really really don't want to) so I'm just trying to rough out a DIY grade as best I can by myself.
 
I think it's a very, very bad idea to try to grade anything in HDR when you don't have the equipment or experience to do so.

It really starts with the monitor and knowing how to calibrate it correctly. Consumer sets are not going to get you there. There will be quite a few HDR-capable pro monitors that will be introduced at NAB in just a couple of weeks, but be warned they're not inexpensive.

My advice for anybody worrying about HDR/Rec2020 now is to just archive the original elements and come back in a year or two when you're able to buy or rent the equipment necessary, and then apply the previous grades and do an HDR/wide gamut trim pass on a calibrated display. It doesn't necessarily take that long to do.
 
That's the thing that gets me going down this path though. I mean, if I know the exact display that it will be shown on (it's for Sony and it's for a certain TV), couldn't I just do some trial and error and get it looking awesome on that one display? Might not be compatible with any other displays or really even be "future proof" but the client wants it for a certain display and so that's all I (should) really be caring about. It's tough, they love the footage, they want the footage, but they want it in HDR. So essentially, putting it on the shelf means I don't get the gig. Tough to put a bunch of money into it without a contract/payment at the same time.

Rock and a hard place, ya know?
 
apply the previous grades and do an HDR/wide gamut trim pass on a calibrated display. It doesn't necessarily take that long to do.

Do you or other colorists do just that part of the process? I have no contract lined up but I would be more interested in hiring this phase of it out if the price wasn't too steep.
 
At this point the only monitor beside Dolby Pulsar that is used for HDR grading is Sony X300, which cost a cool $30k. No one, who is not a big facility can afford such expensive equipment without reasonable return. HDR deliverables are in their infancy now and I doubt very much you'll be able to find anyone who does what you're asking for. It just doesn't make any sense as the format war just started. Spending $30k on the wrong hardware can be a costly mistake, unless you're Tecnicolor or Delux.
And yes, price will be very steep. Dolby Pulsar is $250,000 and it's a rental only.
But in a couple of years it will be very different story as monitors hit $5k price point.
 
At this point the only monitor beside Dolby Pulsar that is used for HDR grading is Sony X300, which cost a cool $30k. No one, who is not a big facility can afford such expensive equipment without reasonable return. HDR deliverables are in their infancy now and I doubt very much you'll be able to find anyone who does what you're asking for. It just doesn't make any sense as the format war just started. Spending $30k on the wrong hardware can be a costly mistake, unless you're Tecnicolor or Delux.
And yes, price will be very steep. Dolby Pulsar is $250,000 and it's a rental only.
But in a couple of years it will be very different story as monitors hit $5k price point.

My question remains though. Is there any reason why using a Sony Bravia at 1,000 nits to grade is a bad idea? Considering the alternatives, even if it is less than ideal...it's kinda my only option at this point (it seems).
 
Part of HDR is knowing the mastering display info. That info needs to be provided with the content. Then after you need to convert from P3, assuming the mastering display was P3, to 2020 by way of XYZ. You also need to scan the film and calculate the MaxCLL and MaxFALL to go with it. In the case of Dolby, you need that information per frame.

A number of titles are being converted to 2020 and calculating MaxCLL and MaxFALL via www.colorfront.com.

I would not consider HDR grading w/o a Pulsar or X300. Consumer displays will surpass the X-300s capability (1000 nits) in 2017.

At a minimum, I would read our white paper on HDR to get an overview on the various technologies. This paper will be updated as time goes buy since things are changing so fast.
 
Part of HDR is knowing the mastering display info. That info needs to be provided with the content. Then after you need to convert from P3, assuming the mastering display was P3, to 2020 by way of XYZ. You also need to scan the film and calculate the MaxCLL and MaxFALL to go with it. In the case of Dolby, you need that information per frame.

A number of titles are being converted to 2020 and calculating MaxCLL and MaxFALL via www.colorfront.com.

I would not consider HDR grading w/o a Pulsar or X300. Consumer displays will surpass the X-300s capability (1000 nits) in 2017.

At a minimum, I would read our white paper on HDR to get an overview on the various technologies. This paper will be updated as time goes buy since things are changing so fast.

Thanks Stacey, will definitely read. The only reservation I have is that Sony wants it for a specific display, so having it be future proof for displays in 2017 isn't as big of an issue as it would normally be (in my case at least).
 
My question remains though. Is there any reason why using a Sony Bravia at 1,000 nits to grade is a bad idea? Considering the alternatives, even if it is less than ideal...it's kinda my only option at this point (it seems).
I wouldn't even trust a Sony Bravia doing normal Rec709. Calibrating these monitors is harder than it looks, and they were never designed to hit normal industry specs. They're designed to look good on a showroom floor, like Costco or Best Buy.

I agree with Stacy elsewhere that things are changing very rapidly on HDR and Rec2020, and it's a moving target in terms of what's out now, vs. what will be out in six months, vs. what will be out in a year. Note that even the $35,000 Sony BVM-X300 will only do most of Rec2020. I don't know of a flat screen display that does 100% of it, but there are laser projectors like the Christie 6P that will do it.

Again, there will be a lot of action on HDR at NAB in two weeks, so I suggest you go there, talk to the venders, read the tech specs, read what SMPTE has to say about it, and then make some choices based on what you find out. I think the reality is that there's a lot of "future proofing" that can be done today, but a lot of that boils down to good archival methods and being prepared to come back and revisit HDR in a year or two. There are HDR titles being done by Technicolor, CO3, Deluxe, and so on, but I also know of experimental HD transfers being done in 1995-1999. ALL of those wound up being redone because the standards changed and because the picture quality just wasn't there. Sony and Universal each built multi-million dollar facilities to do this work, and none of it survived. I believe HDR is in a similar place right now: very early, very experimental, with very preliminary standards. It'll eventually be a very big thing, just as HD eventually became entrenched and doable.
 
By your post Luke you are selling demo footage for this display right? I think Stacey and Marc's feedback is absolutely correct in terms of it's a bad idea to do this without the proper tools to meet the current standard - but a pragmatic alternative response to that is that Sony is going to want footage that will show of it's display today no matter what - and if it's not your footage, it will be someone elses, and since show room models are never calibrated for anything other than attracting an audience anyway it likely doesn't matter that your footage is consistent or even technically accurate - just that it wows the marketing people and looks more dynamic.

Given that - I think if you were to go through this yourself using the monitor in question the one thing you will need to know is the exact settings that the demo monitor will be using. If marketing can provide you that, your grade may still be way off due to differences between the respective panels (and viewing conditions) but you can tell them that you are supplying a grade targeting the exact model and settings you were given.

Now, if the sale of the footage is entirely dependent on their QC - you are taking all the risk to get the sale obviously, so the question isn't so much whether or not it's a terrible idea to grade to a standard with a consumer display (yes it's a terrible idea) but rather whether the risk of failing QC and not getting the sale (and any associated reputational damage) plus the cost of purchasing said Sony TV is worth the value of getting the sale.

Because we all know that marketing people will put imagery up that may be technically terrible if they think it will sell the idea of the technology - so in a very mercenary sense you don't have to care about objective quality if you can satisfy the subjective needs of this particular client.
 
By your post Luke you are selling demo footage for this display right? I think Stacey and Marc's feedback is absolutely correct in terms of it's a bad idea to do this without the proper tools to meet the current standard - but a pragmatic alternative response to that is that Sony is going to want footage that will show of it's display today no matter what - and if it's not your footage, it will be someone elses, and since show room models are never calibrated for anything other than attracting an audience anyway it likely doesn't matter that your footage is consistent or even technically accurate - just that it wows the marketing people and looks more dynamic.

Given that - I think if you were to go through this yourself using the monitor in question the one thing you will need to know is the exact settings that the demo monitor will be using. If marketing can provide you that, your grade may still be way off due to differences between the respective panels (and viewing conditions) but you can tell them that you are supplying a grade targeting the exact model and settings you were given.

Now, if the sale of the footage is entirely dependent on their QC - you are taking all the risk to get the sale obviously, so the question isn't so much whether or not it's a terrible idea to grade to a standard with a consumer display (yes it's a terrible idea) but rather whether the risk of failing QC and not getting the sale (and any associated reputational damage) plus the cost of purchasing said Sony TV is worth the value of getting the sale.

Because we all know that marketing people will put imagery up that may be technically terrible if they think it will sell the idea of the technology - so in a very mercenary sense you don't have to care about objective quality if you can satisfy the subjective needs of this particular client.

Craig, this is absolutely the logic behind my question. It only needs to look good on this one display as it will only be used as showroom material. So logically...I only need it to look right on that display. I know their preferred settings ("Vivid" on the TV) and I already sent them a non HDR version that they loved.

Lol, I went into Best Buy and hooked my laptop up to a bunch of different Sony displays to see if this would work. I had exported a TIFF from RedCine in REC2020 and HDR2084. The issue I ran into was...the TIFF only displayed normal 8bit ranges and the only way to get up to 2000 nits was to raise exposure like crazy.

The HDR tools in Premiere Pro are garbage right now though. I'm not even sure they know what to do with it as they seem to only have half of the workflow implemented while the other half (like values on the scopes....am I reading nits or something else?) is missing or not complete.

Basically, I took (what I thought) was an HDR image file into premiere while hooked up to a Sony Bravia and it didn't really work out like I hoped. Scopes changed to HDR while footage stayed in 8 bit. Only way to bring it up was to overexpose. The poor best buy employee wasted so much time on me...
 
Thanks Stacey.

Do you have any thoughts on the BBCs HDR minimum brightness levels being worse than Rec709?

AJ

... I would not consider HDR grading w/o a Pulsar or X300. Consumer displays will surpass the X-300s capability (1000 nits) in 2017.
At a minimum, I would read our white paper on HDR to get an overview on the various technologies. This paper will be updated as time goes buy since things are changing so fast.
 
I'm surprised there is so much negative input from many graders here..

In my opinion, the understanding of a good CMS and a good probe is the important part here..
If you get a HDR domestic display with the right capabilities.. (Ie 10bit passthrough to the native colour capability of the display) You can set it up as a 1000nit P3 capable grading monitor without too much trouble..
You only need to take care when understanding the displays weak points.. And a good CMS should make that clear what they are...

Based on that making a 1000nits ST2084 master using a display like this and Resolve, as an example, should give "workable results" in the short term.
And why the hell not..

I expect domestic display technology to improve a lot quickly considering MD and OLED becoming more mainstream. P3 shouldn;t be that hard for manufacturers and domestic displays.

But yes, let people test all these displays coming out and I am sure one will float to the top as a good compromise.


I'm more concerned that, our editing preview screens above the timeline, can only give a very poor representation of the picture as data displays that are HDR compatible is only a problem starting to take shape and likely a very long way of in terms of the OS, manufacturers / graphics card makers etc. all getting it working together so a HDR image can be displayed at the same time as control interface. (On the same screen)

Ie, HDR is very much going to be a in a suite process. In some ways a good thing, bringing the art of colour back to the operators who take it more seriously..

James
 
James Gardiner said:
Based on that making a 1000nits ST2084 master using a display like this and Resolve, as an example, should give "workable results" in the short term.
And why the hell not..

The key is that you need to know what that mastering displays capability is and include it with the content. One area people might not record correctly is the minimum luminance. They might just assume it is black and should be 0. Let's say the display can do 1500 nits, but its black level is 0.5 nits (FYI UHDA Premium requires a mastering display to be at least 0.03 nits. The Pulsar is 0.005) then it would be wrong on playback and not a good wrong. :)

James Gardiner said:
I'm more concerned that, our editing preview screens above the timeline


Do you need to edit in HDR? I don't think many will want to. At HPA more than one person complained about limited time grading HDR before needing a break due to eye strain.
 
Do you need to edit in HDR? I don't think many will want to. At HPA more than one person complained about limited time grading HDR before needing a break due to eye strain.
I agree, Stacey. I have heard anecdotal stories from three different people at three different facilities that eyestrain and monitor heat are still considerable problems. I think it's not as much of an issue if you have a 100-nit HDR projection display on a 20' screen, which is not the same thing as a 1000-nit or 2000-nit flat-screen display that's maybe 30" or 40" and is just sitting a few feet away from you.

I expect there'll be a lot of HDR demos and new pro displays at NAB in a week.

In my opinion, the understanding of a good CMS and a good probe is the important part here... If you get a HDR domestic display with the right capabilities.. (Ie 10bit passthrough to the native colour capability of the display) You can set it up as a 1000nit P3 capable grading monitor without too much trouble..
I'd love to see a list of consumer displays that will calibrate to 1000 nits in P3. So far, I don't think there's even a lot of professional HDR displays that can reproduce 100% of P3. Projection, they can do, but those prices start at around $150K, the last time I checked. And I have yet to see any consumer displays that can handle 10-bit at all.

I think a lot of this stuff is pie-in-the-sky BS, because it's just a little too early right now. Dolby has still not shipped production versions of the Pulsar yet, and as far as I know it's just a rental item (since the estimated list price on this flat screen display is $100K). True, you can grade in HDR on a Christie 6P and archive those files, but I can tell you from experience color corrections on a projector don't translate 100% to flat screen monitors (consumer or otherwise). I would also bet that if you lined up Sharp, Panasonic, Sony, and Samsung consumer displays that are all certified for UHD Premium, none of them will reproduce the same pictures out of the box. (LG has refused to join the UHD Premium group so far, saying their sets already meet the standard.) I'm not sure if they could even do it if they were calibrated, or even if they could be calibrated at all.
 
Thanks man,

I don't actually own Resolve (I know...I know). In the latest version of RedCineX there is an "HDR-2084" gamma space. Do you have any experience utilizing that gamma? At this point I pretty much have to proceed (even though I really really don't want to) so I'm just trying to rough out a DIY grade as best I can by myself.

If you can hook up the given display and get the picture to look nice on it in the right colorspace, then you should be good to go. I think you monitor display lut on you mac or what computer you are using is not right. So even if you set RCXP to HDR-2084 your display is set to somthing else like rec709 or linear rub or such. I assume you need a HDR-2084 profile in your screen seeing in preferences on you mac/pc, no?
 
Back
Top