Re: HDR and camera evaluations
Geoff Boyle
Yes I know Steve but it does give us a representation of what we need to bear in mind when shooting HDR.
Apart from not putting bright windows behind the talent or putting bright lights in a different part of the picture to the talent.
Yes, I have been watching HDT shows on my HDR TV, not the monitor, the TV in my living room.
Cheers
Geoff Boyle NSC FBKS Cinematographer Netherlands www.gboyle.nl
From: cml-raw-log-hdr@... <cml-raw-log-hdr@...> On Behalf Of Steve Shaw
As the display is not a HDR display, that really shows that using a gamma that is not the default Rec709 will produce the better images.
|
||||
|
||||
Re: HDR and camera evaluations
As the display is not a HDR display, that really shows that using a gamma that is not the default Rec709 will produce the better images.
That is what a colourist does naturally. Steve
|
||||
|
||||
Re: New test uploads
Geoff Boyle
I’ve now added Eva, Pure, Venice, Terra, Movi, Gemini, Monstro & X-H1
All there ins raw from, some missing from h264
Cheers
Geoff Boyle NSC FBKS Cinematographer Netherlands www.gboyle.nl
From: cml-raw-log-hdr@... <cml-raw-log-hdr@...> On Behalf Of Geoff Boyle
Sent: 19 July 2018 17:55 To: cml-raw-log-hdr@... Subject: Re: [cml-raw-log-hdr] New test uploads
I’ve added some more, it’s now Alexa Mini, BMD UMP, Canon C200 and Canon C700 6K to 4K.
That’s it until tomorrow.
|
||||
|
||||
HDR and camera evaluations
I’ve only had a quick look at this but I thought that you’d like to see the preliminary results…
I’ve only looked at the Alexa so far, I’m using Resolve in ACEScct and the only grade I’m doing is a timeline colour balance to make the T16 Kodak Grey neutral.
Viewing is an Eizo CG318 which is 31” 4K and HDR 350, however it can simulate PQ1000.
With the ODT set to 709 and the monitor to 709/1886
Best overall look is at T16
Highlight retention is T11
Shadow detail is T32
With the ODT set to Rec 2020 ST2084 1000 P3 gamut clip and the monitor set to PQ1000
Best overall look is T11
Highlight retention is T2.8/T4 split
Shadow is still T32
With the ODT set to Rec 2020 and the monitor set to HLG
Highlight Retention is T8
Shadows still T32
The simple answer is that, bearing in mind the best overall image on screen, HDR has one stop more useable shadow detail and two and a half stops more highlight detail.
Or, three and a half more stops which kinda agrees with the numbers, rec 709 100nits, HDR 1000nits
Visually the difference is stunning.
Cheers
Geoff Boyle NSC FBKS Cinematographer Netherlands www.gboyle.nl
|
||||
|
||||
Re: New test uploads
Geoff Boyle
I’ve added some more, it’s now Alexa Mini, BMD UMP, Canon C200 and Canon C700 6K to 4K.
That’s it until tomorrow.
Cheers
Geoff Boyle NSC FBKS Cinematographer Netherlands www.gboyle.nl
|
||||
|
||||
New test uploads
I’ve started the uploads of the new tests but I’m, pausing to give you a chance to rip the shit out of me.
Please comment before I spend any more time.
Only the Alexa and BMD-UMP are finished.
Please also read the info page.
Cheers
Geoff Boyle NSC FBKS Cinematographer Netherlands www.gboyle.nl
|
||||
|
||||
Re: RAW evaluations and graded versions
Geoff Boyle
I did have a plugin for Resolve that produced wonderful ProRes files. Unfortunately when I went for a complete reinstall on a new machine I couldn’t reinstall the plugin. It needed to be registered to work and Apple had had a word with them and they no longer offer the plugin.
Scratch have kindly sent me a 30 day license for Scratch but unfortunately my machine doesn’t work with this license. I will chase this up once I am further down the road.
All the raw files are ready for upload, I’ve trimmed every shot to 1 second except for faces where they’re 5 seconds.
I’m waiting until Kodak get me the scans of the neg 😊
After they are all up and properly linked I’ll get back to making compressed versions, slightly compressed and very compressed.
I’ve done several of these already but I’m having second thoughts about the way I’m doing them.
Basically, decide which exposure is nearest to “correct” and mark it, then match every other take to it using offset only and only matching the grey patch on the Kodak chart.
I have been doing a timeline colour correction based on the selected frame, no exposure change but neutralising any colour shift. If anyone wants to see what it all really looks like then they can get the raw files.
The complication is that I’m doing this on the sequences lit with “real” tungsten, with the “substitute” ones I’m in two minds, I’m obviously not altering colour as we want to see the shift but do I match exposure to the one from the tungsten set or do I match all the substitute shots to the same stop as I used for the real light even though it may well be the wrong exposure?
I think that we need to see the difference.
I’ll also post a chart of each camera with exposure variation from measured and also what rating gives the most exposure for the lowest noise, i.e. the point where the waveform stops being fuzzy 😊 or, do I post the amount of NR at “correct” exposure that is needed to give a clean waveform?
Cheers
Geoff Boyle NSC FBKS Cinematographer Netherlands www.gboyle.nl _._,_._,_
|
||||
|
||||
Re: RAW evaluations and graded versions
Paul Curtis
I think they've output since v9, we're on v11 at the moment. I think all the product versions write prores (base Nuke upwards) They output all flavours properly, including alphas and 4444 and even 4444XQ and they're Apple sanctioned. I did mention to them at one point that they really ought to make a bigger deal of this because i stumbled on it by accident not marketing. The only negative is that QT is quite slow under windows (in general, not Nuke specifically) cheers Paul Paul Curtis, VFX & Post | Canterbury, UK
|
||||
|
||||
Re: RAW evaluations and graded versions
Virgil Kastrup
Yes. Exactly the reason I recommended Scratch as opposed to the reverse engineered ffmpeg ProRes.
On Tue, Jul 17, 2018 at 4:16 AM Clark Graff <clark@...> wrote:
--
------------------------------------------------------------------------- Virgil Kastrup post production supervisor Denmark +45 28181579
|
||||
|
||||
Re: RAW evaluations and graded versions
David Rasberry
"The Video Output Formats section listing, including ProRes, has a
footnote saying "video codecs may require installing additional
libraries or drivers." What libraries/drivers (if any) have you found that it needs to use the various ProRes output flavors it supports? The product looks good on the site, lots of features, is inexpensive as these things go. -Bob Bob Kertesz BlueScreen LLC
Hollywood, California" The only codec I needed to load externally was Cineform. Used to be I just installed the GoPro app and it included the libraries, but the most recent version didn't. I found the last versions on Cnet. I'm definitely not in a class with you guys, so I'm mostly just a fly on the wall observer here. Audio and Video systems engineer by trade. Serious hobby filmmaker mostly shooting simple docs. I am currently shooting with a Digital Bolex and my workflow is a combination of Davinci Resolve for raw processing, Cineform DI out of Resolve, edit with Lightworks from "best light" corrected Cineform DI's. I use FS4k for transcoding compressed video formats to Cineform, Mostly casual footage shot with an iPad and Iographer rig. Recently been experimenting with Slimraw "lossless" 10 bit log compressed raw. It is a Cineon log curve and works well with the color managed transforms in Resolve.
|
||||
|
||||
Re: RAW evaluations and graded versions
Mark Kenfield
Until (literally) anyone comes out with a simple drag-and-drop (and folder structure free) alternative. I highly doubt that Prores is going anywhere. I'm still at a loss as to why not a single codec manufacturer has picked up on the significance of Prores's drag and drop simplicity. It's the central pillar upon which its success stands. Mark Kenfield Cinematographer Melbourne 0400 044 500
|
||||
|
||||
Re: RAW evaluations and graded versions
I also am a fan of Cineform RAW. It’s now open-source and well supported in Resolve.
toggle quoted messageShow quoted text
On Jul 17, 2018, at 3:13 AM, Paul Curtis <paul@...> wrote:
|
||||
|
||||
Re: RAW evaluations and graded versions
Paul Curtis
But reverse engineering a proprietary 'standard' is probably not ideal in this case - there's no guarantee of broad compatibility or future robustness IMHO. Personally i'd still prefer to see Cineform take over because i do believe it's better quality. However it doesn't have a fruit behind it and also now that Adobe integrated it into their applications their version of the codec appears slightly different to the previous one and caused me some issues with past footage. Also i don't know if the codec itself is standalone anymore. cheers Paul Paul Curtis, VFX & Post | Canterbury, UK
|
||||
|
||||
Re: RAW evaluations and graded versions
Geoff Boyle
Not drinking.
https://cinematography.net/subs.html
It takes one off donations as well 😊
Cheers
Geoff Boyle NSC FBKS Cinematographer Netherlands www.gboyle.nl
From: cml-raw-log-hdr@... <cml-raw-log-hdr@...> On Behalf Of Leonard Levy
Sent: 17 July 2018 05:29 To: cml-raw-log-hdr@... Subject: Re: [cml-raw-log-hdr] RAW evaluations and graded versions
Who’s with me ?
Lenny Levy , DP San Rafael , CA
|
||||
|
||||
Re: RAW evaluations and graded versions
Bob Kertesz
The Video Output Formats section listing, including ProRes, has a
footnote saying "video codecs may require installing additional
libraries or drivers."
toggle quoted messageShow quoted text
What libraries/drivers (if any) have you found that it needs to use the various ProRes output flavors it supports? The product looks good on the site, lots of features, is inexpensive as these things go. -Bob Bob Kertesz BlueScreen LLC Hollywood, California DIT, Video Controller, and live compositor extraordinaire. High quality images for more than four decades - whether you've wanted them or not.© * * * * * * * * * *
On 7/16/2018 7:37 PM, David Rasberry
via Cml.News wrote:
|
||||
|
||||
Re: RAW evaluations and graded versions
Leonard Levy
Let’s get real guys . Instead of overloading beleaguered Geoff with more & more ideas he won’t use, let’s just chip in for his favorite booze . He needs a serious break. Just $10 bucks from each of us should yield something better than I’ve ever even tasted . I’ll start in with $10 if Geoff can provide a PayPal address . Jesus it’s the least we can do for putting up with all these insane suggestions . Geoff are you still drinking ? If so Post a PayPal address ?Dont be shy . If not alcohol then go out for an incredibly fine dinner ! We all owe you . Thank god I didn’t have to do these tests ! Who’s with me ? Lenny Levy , DP San Rafael , CA
|
||||
|
||||
Re: RAW evaluations and graded versions
David Rasberry
Personally I prefer Cineform.
|
||||
|
||||
Re: RAW evaluations and graded versions
David Rasberry
Footage Studio 4k is a practical inexpensive easy to use transcoding utility that I've found very useful for a couple of years now. Transcodes pretty much any common encoded video format (not raw) . Writes all flavors of Prores, DNXHD/HR. and Cineform DI's from proxy to 4:4:4:4 16 bit. Acrovid - FootageStudio - Format conversions, Video Standard conversions, Frame rate conversions, SD-HD conversions, Video Denoise, Overcrank.
|
||||
|
||||
Re: RAW evaluations and graded versions
simon@...
On 2018-07-16 20:55, Daniel Rozsnyó wrote:
I too would have mentioned ffmpeg as the tool to use to avoid a Mac-Windows round trip, had I not been busy when I first saw Geoff's post.
But respectfully, I completely disagree about "approved" or "licensed" products automatically having better encoding quality. Take, for instance, commercial h264 encoders vs open source x264. x264 beats everything else I've ever tested hands down for quality at similar bitrates and can be an order of magnitude faster.
Of the three Prores encoders in ffmpeg, one stands head and shoulders above the others and that is prores_ks, as shown in the example Zach posted.
All lossy encoders generate artifacts, that's a given. ffmpeg's prores_ks encoder just generates different artifacts to the Apple Prores encoder. In my experience, neither stand out as being worse than the others, just different.
Some time ago, I spent some time working on some improvements for the DCP encoder we developed in-house. We started with an open source JPEG2000 encoder which consistently made better pictures than a couple of (incredibly expensive) hardware DCP encoders and our work improved things further. So it's definitely not a case of you get what you pay for :-)
Simon
--
Simon Burley Direct: 01342 395 003
|
||||
|
||||
Re: RAW evaluations and graded versions
Clark Graff
>>>As has been touched on, there are a few windows apps that do encode
I believe that Scratch can OUTPUT ProRES correctly. I have done it on a few jobs without issue.
~clark
Clark Graff Film Maker- Production Designer – VFX Sup – DIT – Workflow – Editor – Propellerhead SoCal ~ Toronto ~ Vancouver 805-253-5496 cel www.4nr.com
|
||||
|