CMMAP, so well I loaded the movie the Dark Knight and started tinkering.
Guess I’ll start a new thread for it. : )
CMMAP, so well I loaded the movie the Dark Knight and started tinkering.
Guess I’ll start a new thread for it. : )
44rh1n thanks, very good advice. I’ll plan on keeping the source in HDR then.
44rh1n thank you for the tutorial about the Dolby Vision trim analysis. I’ve just begun using Resolve, and your helpful tips are much appreciated. This analysis also seems to be exactly what winoni71 was looking for as well. I will certainly be trying this on an upcoming edit. And based on your comment: “Once it’s done, you’ll notice that the image no longer looks flat and ugly because you’re now seeing the SDR version created by the Dolby Vision analysis.” it looks like I can further color grade after the analysis and before the rendering. Sweet!
EDIT: or perhaps a best practice for users of Resolve would be be to import the file container with HEVC HDR into Premiere, apply the Dolby vision, and export that to a DNxHD or DNxHR to be used as the new source to edit?
CMMAP thank you for watching the edit [AGAIN] and taking notes! I will take a break from the Dark Knight series with no definitive plans to continue onto the Dark Knight yet. For two reasons…
One, the changes I want to make which would make it worth it to me require VFX beyond my skill level. The desired changes for The Dark Knight would include the civilian hostages exploding (I would have them push the detonator button to kill the prisoners on the other ferry, but guess what the jokes on them and it explodes their own ferry instead) and also changing the look of the new skinny Batsuit to something better.
Two, I would like have an edit published on Fanedit.org, so I will make my next effort on something without much VFX where I can focus much more on the sound engineering and cuts to the score being absolutely clean. My next effort will also be done in Resolve, so I’m learning the basic functionality of that program now. It’s really amazing with its color capabilities.
Watching that video you can definitely tell that Williams hated the idea but hey the music editor had a gun to his head, what can you do.
Vader also really hated those force chokes, it was the emperor’s dumb idea.
V2.1 released - color correction added to fix washed out HDR > SDR conversion.
A NLE is a non-linear editor, such as Adobe Premiere.
The easiest and fast way would be to do what you did again just choosing a LUT you like better.
Below is a decent and fast option to apply HDR > SDR color correction to the entire video:
In Premiere put three adjustment layers above your video:
Top layer - add a LUT which improves the picture, such as SL GOLD RUSH (or the LUT you already tried), and reduce intensity, perhaps 50%.
Mid layer - add the SDR conform effect here and adjust the brightness and contrast appropriately. Leave soft knee typically at 100%.
Lower Layer - add saturation here of perhaps 115-120% if the colors are washed out.
If you like the SDR conform effect by itself just use it alone.
I am trying to convert a bunch of HDR movies into SDR using “TMPGEnc Video Mastering Works 7” which allows the use of LUT files and I was wondering if there’s someone here willing to help me out.
I have successfully imported the “Rec2020ToRec709_CSTDefault” LUT file, which I suppose is the one to use, and the results are good, but I’d like to make them better, but I’m not sure how.
The conversion LUTS (HDR > SDR) generally are fixing the overall overexposed and washed out look you would get without correction. The issue is that one LUT will not adequately correct every scene in a film. Each scene has different light and color, so to achieve good results you would have to color correct scene by scene, ideally using an NLE.
Adobe Premiere has a built in “SDR conform” tool, and while you can apply it to the entire video, you can also create an adjustment layer and tweak the SDR conform settings scene by scene.
Krausfadr, you´ve totally nailed it now for the continuity regarding batmans voice! Just awesome. From my end i´ve nothing further and deem it perfect.
Whoops, sorry. Sent the line to Hal in a PM at first. Here’s the link: https://drive.google.com/file/d/1jOfypfXSDVkf7Z0p2xVB2ktg0RfwdxRW/view?usp=sharing
So, the extra layer was kind of weird. One of the problems I noticed was that there was a harsh cutoff between “become” and “A DYad”. I tried just lowering the volume of “A DY” a bit, but it still sounded uneasy to me. So I isolated the “mm” sound from “become” and layered it in a bit sooner. Then I blended it in with the “A” sound more. The end result (hopefully) was that it would go phonetically from “become. a Dyad.” to “become [ma] Dyad.”
Rather odd all around. I’m glad that it seems to have worked!
Here’s how you can quickly make an LFE channel:
The reason I asked is because I heard about this tool which is supposed to work wonders for converting from 2.0 to 5.1: https://nugenaudio.com/haloupmix/
and was wondering if anyone here had tried it or something similar.
Sounds like you are a wizard if you did this manually in Audition!
I’ve used Hybrid, and its an excellent program, but I prefer dmMediaConverter and Staxrip for ease of use. As for converting a blu-ray rip to Prores, I just use makemkv, switch the mkv container to mp4 using dmMediaConverter WITHOUT ANY RE-ENCODING, and export to Prores using Premiere.
The 4k is now available. 12.2GB.
If you use Staxrip and want proper de-banding, you have to use the VS filters under the restoration category:
Restoration > RCR > Colorbanding > f3kdb
Thanks for letting me know. Handbrake is no stranger to me, but didn´t know that either. But haven´t used it since i realized that it only has the 8-bit pipeline.
Didn´t expect it to have just 12 GB but nice 😃
That’s why I stopped using Handbrake, because it reverts to 8bit during processing. The final filesize can vary wildly with crf since its all about content. So for example, on one 2 hour film 4K crf 20 could result in 20GB and in another film-- half that.
Hal 9000 said:
sherlockpotter will solve this or it’ll just remain poop.
Oh god, Hal himself just hinged the success of whole project on me. Fuuuuuu…
How does this sound for a v2 of the line? If it’s not perfect yet, let me know; but I’ve been listening to the same two seconds of audio for the past few hours, and I’m starting to lose objectivity. I’ve heard it so many times - and I’m hyper aware of where the different seams are - that I’m not sure if it sounds fluid or not. I need an fresh set of ears!
I’ve added an extra layer to smooth the transition a little more, re-pitched a couple of syllables, and I’ve been shuffling the different pieces around by a few thousandths of a second to make it flow better. Thoughts? Be brutal.
Sherlock, could you please post your final clip for everyone in a WAV file? Excellent work.
Also when you say you “added an extra layer” what does that mean precisely?
As much as I appreciate the work done on the “become a Dyad” line I just don’t like that idea narratively. For me, the fact that they are a Dyad from the start explains so much about their connection even before they met. I don’t want to beleaguer the point so I guess I’ll just leave it at that…
Saying “we’ve become a dyad” ends the possibility of always having been a dyad. Saying “we are a dyad” leaves it open to interpretation in either direction… they could have always been one or recently become one. For me, becoming a dyad in the force is more interesting because it means that perhaps Snoke inadvertently made Rey and Ren uber powerful by facilitating the bridging of their minds. I like the change and it’s done seamlessly here. What starts off as a strong connection can become a dyad given the right environment.
Definitley appreciate that you encode it for me, if you didn´t started yet crf 20 is fine, no need for crf 18.
CMMP, I have a 4K encode almost finished going at crf20. Also I realized that Staxrip (the program I’m using to encode) does not automatically apply a de-banding filter to smooth out unnatural gradients (different than deblocking)-- Handbrake, a different program but without true 10 bit end to end, seems to always have de-banding on by default. So in addition to the better crf of 20 now there is proper de-banding, and the picture is noticeably better (perhaps even close to perceptually lossless in most scenes now). Final filesize is estimated at just over 12GB and should be ready later today.
Kylo already defines dyad in the movie. Two that are one. And when Palpatine is healed it instantly and clearly expands that definition.
Jar Jar Bricks said:
I would like an opinion - does poppa’s added in line feel natural to the rest of you? It’s the “you were right” line.
I only ask because I know what line was changed, and as a result I always am anticipating the subtle differences. Maybe it sounds just fine and I’m imagining it.
I feel like this is an important thing to discuss if we’re going to try the same thing again with another line.
I agree, the you were right line stands out yes but it was still done so well by poppasketti. It’s very difficult to blend two completely different voice actors as the same character in the same scene with a different sfx chain. He cleverly spliced in pieces of Driver’s voice along with his own. So it’s a matter of how important is the change versus it not matching 100%. My preference would be no V.O. from a different voice actor because for me it’s always very noticeable. This is noticeable but it’s still very acceptable.
From what I remember, the line is very punchy maybe with a lot of brickwall limiting and in your face with a little extra sfx applied to help cover up the different voice.
Amazing job on the ship intro text. Great example of superb VFX work.
Also, what software or plugins did you use to implement your 2.0 to 5.1 conversion?
So now just sitting an thinking about it, “Why have the clones of him as an old guy?” Kinda like, you are going to go on a road trip, and instead of filling up your tank, you fill only fill it up a couple of gallons.
Intellectually this makes sense, and I like it. In implementation though with the very brief scene, the casual viewers would have no idea wtf was going on unless you subsequently sat them down and explained what they were seeing.
Many thanks to CMMAP for the critical eye and comments. I encourage anyone who watches the edit to do the same.
Below is what I plan so far for updating to v1.3. If anyone else has additional comments on how to improve in the next day or so, I will gladly take them into account.
If it makes a bigger difference in encoding time, crf 20 would be probably enough quality i can live with for sure. I know that i´m kind of odd in this regard 😄
- 1:11:59 (when the guys head got smashed into the mirror) --> looks like a or 2 or 3 frame drops. Checked with the original and there isn´t any stutter
- 1:29.05 - 1:29.27 --> seems to be off again like 1:08.33 but less distorted
I can encode at crf 18. It won’t take any longer to encode than crf 20. I have a fast PC so a full 4K encode using HEVC (H265) usually takes about 10-14 hours depending on the content. I just set it at night and the next morning its done. Trying the same thing on my macbook would take about 60 or more hours to encode. lol
When the guy gets smashed into the mirror, yes I did cut his head turning which was necessary to remove the peeing the way the scene and audio was done. So I’m ok with this. At this point I won’t be making any more changes to the video portion unless I find a serious glitch like optical flow errors.
The scene at about 1hr29m (Batman and Rachael in the Batcave). Yes I definitely see your point again with the Bat vocals. I realize also that this and the scene at Gordon’s trashcan both had some noise reduction I applied. I am removing it which will also help the issue quite a bit.
- 1:08.33 to 1:09.06 --> his voice seems way different than in the rest of the film, more hollow and “siri”-like. Is it possible to get it like in 1:20.44-1:20.54?
Good point here. The voice in this scene is just not great. In fact I went back to the original scene to listen and it seems like a joke.
The original scene suffers from two problems. First Bale used an especially bad voice, and second the EQ was not great. So I have played around with adding some more low end to this and also to modulate his voice down just a slight bit more. I changed around the effects used as well and while I will never be able to make it sound like the better voices he used, I can make it much much less bad. Maybe even slightly good.
Thanks for the help. And I’ll respond to the rest of your post as well.