First off...Best way to see differences is with a "difference"effect (BCC has one). Then add a contrast color correct after, to see it more clearly. This will show you stuff you can't see normally on a monitor.
Second off. As of 2018.10 we have an uncompressed higher than HD resolution. In the current release it is 32 bit RGB and YUV float, which is crazy big. Subsequent releases will have 10 bit RGB and YUV versions, so Media Composer is now capable of approved Netflix workflows. They are currently cumbersome since 32bit float is so huge, but will become more reasonable, promised by the end of this year.
Per the statements of AVID from the Ask Avid event on Wednesday, all the colorspace math is being done in current Media Composer with 32 bit float math. The rest of the processing in Media Composer is 16 bit integer, but that is supposed to change sometime in 2019, with a re-write that is supposed to include eliminating the current issues with 3rd party codecs and an increase in render quality in other areas.
You can hear about all of this on the live stream replay of the event, which is on the Avid Editors of Facebook page, or possibly this link:
| |  | Avid Watch our exclusive live interview with Tom Cross, ACE, editor of First Man. And get ready for "The Latest,... | | | |
It starts off with a panel of the edit team from First Man, but then opens up to a Q&A event like last year, with the CEO and key officers of the Media Composer team. Also, during the live demo after Q&A there is a very interesting detail from the Chief Color Scientist at AVID, about how Media Composer is handling luts, where the various settings are, and letting us know that all that math is being done 32 bit float, while the app is "currently" only doing 16 bit integer for other rendering.
Dave Hogan
Burbank, CA
On Thursday, November 8, 2018 6:19 PM, "John Moore bigfish@pacbell.net [Avid-L2]" <Avid-L2@yahoogroups.com> wrote:
I know when I use to dub VHS to 1 inch etc... the image didn't get any better but subsequent use did not deteriorate much. Obviously dubbing to an inferior format like VHS to VHS would add much more noticeable noise etc..
Now in the 4K world I have been using Avid with success in a number of shows. Now there is a new directive to finish completely in Resolve. I totally understand Resolve handles media better and has 32 bit floating point etc. Our source material is Varicam 35 AVC-Intra YUV 10bit Rec 709 shot in Panasonic Vlog. My traditional understanding is Avid processes at 10 bit, whether timeline quality is a factor in the processing still seems not clearly defined, but I always run in 10 bit timeline for renders etc...
So the basic argument is they want us to work from the AVC-Intra master files in Resolve and not transcode to DNxHRHQX. My understanding of HQX is it is a 12 bit codec so in my mind the transcode shouldn't add any noticeable image degradation. In practice I see no detectable difference on my monitors when linking to ama'd AVC-Intra master files or the HQX transcodes.
I'm trying to understand how anyone would see a difference in image quality that matters using 10bit source material in Avid vs Resolve. I understand Resolve is much more powerful and has more robust color correction tools but I am working on well lit stage shows that don't call for a super sophisticated grade. In the days of tape we would just take the truck video in more or less unity when we got to HD cam and just add a slight overall tweak to assure no scope alarms would sound. Hence while I understand on paper the Resolve flow is cleaner I just don't understand how anyone would ever see a difference in the final product.
I'm not trying to ignore or refute the network concern I just want to learn what can look for in the image that makes Avid noticeably inferior in this particular work flow.
John Moore Barking Trout Productions Studio City, CA bigfish@pacbell.net
