hi can you make a video showing the flexebility of those passes and how much can be chanegd through the passes or is it just lighting and color not like adding more fake environment
Quick question! I rendered as multilayer exr with the file outputs in blender, but when pulled into davinci, it is only one of the 4 layers as a sequence. I had overwrite checked in the image sequence settings but i noticed you also had that selected... Would that mess up the layers? Edit: Never mind lol I just figured it out. Sorry, I am still learning davinci.
@@timenotspaceproduction What are you gonna do when multiple frames get a graphical glitch? Render it out all over again? You can just drag and drop image sequences into most video editing softwares and it'll play as a video, and it won't be compressed like an mp4.
I had to take down the last video because of a background music track. I had misunderstood the license, so I couldn't earn anything from the video. We should be all good now.
Could you please upload your tutorials with no background music at all when you are speaking? The music serves no purpose and for some people it makes the video extremely stressful to watch. Because music is such a personal thing no matter what you choose some people will hate it, and that's where the stress comes from. In addition some autistic people (like me) will also struggle to understand what you are saying even if the background music is quiet. I had to stop watching this video about half way through because the music was causing me a lot of stress.
Amazing explanation! And thanks for the little clip in the background being honest about its usefulness. As a professional CG artist I've found that its not necessary for a lone artist in 99% of cases. Studios do this because going back down the pipeline for minor changes means dozens of people have to redo certain parts (and get paid for those hours haha). I've used passes a few times to adjust my render, but I have never found a usecase for rendering out every collection in its own renderlayer yet. I think the best use-case for rendering so many separate layers and passes is to create a killer VFX breakdown of your project ;)
As a hobbyist, I actually use some of those tricks (especially the view layers) to deal with performance issues. For example, hair can take a long time to render, and using a separate view layer for it can help. This way I don't have to render the hair again when I do minor modifications to the scene. It's also a good way to have a complex background or volumes without the long render time that comes with it. It's a nice way to introduce separation of concerns into my workflow. (Of course, having a better machine can help with that haha!)
Totally agree! If the project does not specifically ask for passes in compositing, I also believe that it's unproductive to do that. But at the same time, I've had clients that specifically asked for minor tweaks in volumetrics or lighting and that was a total pain considering I didn't render in passes. The lesson for me was to always check project requirements and non-negotiables. Render passes are one of the tools we can use as cg artists to make our work easier, but it doesn't mean we have to use it all the time.
I like rendering individual layers 'cause I use Blender mostly for paintover in Photoshop, but I don't need half of the passes he showed, by just having every part separated and a ID pass to select everything more easily is enough.
Blender is SO damn amazing! I'm still confused by basic shit because it's not my primary obsession lol. When I see channels like this do these things it looks like magic even when I watch you do every step!🔥🤯
Very cool. I've been wondering about all of this for like a decade! But I stayed away because it was the big bad wolf until you came along and blew his house down. It all clicked as I watched. Thanks for the tour! Cheers
Really like how you explain things. I have something to add for this video: - ACEScg, ACEScc, ACEScct are all working colorspace. You should treat your 3D render like camera footage: capture as much info as possible. Therefore, ACES2065-1 (AP0 - linear) is the colorspace for handover and archive. While cg, cc, cct or (AP1 - linear, log, log) are the purpose of working colorspace. SO, render them in ACES2065 instead of ACEScg is better. However, you have to do another mapping from ACES2065 to ACEScg, which takes time. I think fusion on resolve 19 has that color management for fusion as a project setting, which can automatically do things for you. (I’m a colorist, I don’t work much in Fusion page so… learn it yourself, IDK fusion 😅) - For color grading purposes, output it to ACEScg and let the colorist do their job (colorist will map those out to ACEScc or cct, or whatever working colorspace they prefer). PLEASE, PLEASE,PLEASE don’t do rec709 transform or anything. Once you squeeze things down to a smaller gamut, there is less information to work with, and it hurts. You would not want to waste all the effort of doing all those render separately and then squeeze them into a sRGB or rec709 and hand it to the next step 🙂, b/c they will punch you 🤛. Hope those help some 3D + VFX people🤞.
I think this video is pure gold and is a subject that's rarely covered on YT... it would be amazing Robin if you could do a tutorial video encapsulating all these processes.. even just for one scene.... I don't think any other blender channel has done this :)
What an absolute banger of a video. I might not follow or need the complete workflow but I know I would need parts of it. But leaving the point of what I need or do not need, it was an absolute delight and a knowledge boost to watch the entire video. The way you told everything in sequence was so awesome and easy for me to understand and I don't even use resolve. Thank you for making this video. You, good sir, earned yourself a subscribe.
i dont know how much things have changed on 3Dsmax but i remember i used to just click on the Passes tab on the render window and select the ones i wanted, havent done that in a long time, other than rendering just zdepth and material ID
Wow, I've needed to see this video for a long time. I was making view layers way too hard for myself. I was also using up memory placing holdouts and shadow catchers into my other layers to represent my objects from another layer... That was brutal on my ancient PC. Also I was completely unaware of the Channel Boolean node in Resolve, I've been super confused why those options weren't in the merge node.
I've followed along now and is done with a draft render. I'm now not sure where to begin when i have all this flexibility in my layers and passes. I'm hoping you're diving into this in the next video! Love your style and focus.
I did not even consider using layers before, but recently i had to because how Grease Pencil interacts with semi transparent materials. It's not as complicated as i thought before, but there are extra steps to be used for sure.
Ornithopter, or thopter shortened. Anyways thank you for showing how to split all the passes prerender it's pretty much necessary with how much physics simulations go into creating realistic scenes from light, gravity, fluid dynamics collision, etc.
about the feedback you asked for in the first video, the song is pretty nice, helps to keep the viewer engaged but doesnt steal focus, the pacing of the tutorial is pretty good, honestly only thing i have to say is that your speech could be more consistent but honestly i do not mind and imo it does not hurt the tutorial, overall great stuff, love to have those in sights into the pro workflow, i dont intent on going that deep but i for sure will be rendering layers individually whenever my PC becomes concernealing loud, and having the option to change individual aspects without re-rendering the full image is actual gold, quick iterations are the core of computer based creative processes, really enjoyed the video, for sure can picture myself coming back to check a thing or two
Great video, thanks for the info. I've done this kind of rendering for years in different software but never in Blender. It's good to see how to do render layers correctly (in Blender), it looks like it works quite well. However from what I can tell Blender can't do one important render task. The ability to apply specific shaders to specific objects per render layer. This is the feature for me that's holding Blender back from being really flexible for rendering. I think on almost every job I've ever done I've needed this feature for at least one shot. The fact its not possible so far in Blender is really weird.
Blender should add denoise checkboxes on that node if it requires it. Or if it supports multi-selection to pull out to conjoin to a multi-input denoise filter node.
It can be a good idea to denoise the volume passes separately to everything else as they often have no relevance to the normal and albedo passes. If you can afford to wait.
This was really helpful. I've experimented in the past with trying to do a proper "Professional" pipeline for my projects, but it always just turns into a confusion mess (The Holdout feature was something I just learned about in this vid). I'll try something like this again on my next big project.
A better way of getting an accurate preview in DaVinci is to set the project colour management to have ACES as the input and rec.709/2020 or whatever your deliverable needs to be as the output. That way you see a colour correct image with no need for a LUT and no need for OCIO or CST, and your timeline has a wide gamut for grading.
Hello Robin. Thanks for the great content! I had already watched the previous video, and I came here to like and view it to help with engagement. It's a great video. I had already tried to understand this type of workflow, but your video was the first one in which I really understood how this type of composition works. Do you have any videos that teach this same workflow, however, for composing animations? I would like to adapt this workflow to compose a sequence of images for an animation, but I can't find anything as good as your latest video. Again, thanks for your great content.
Hey, I should have mentioned that, but the workflow is exactly the same. The only difference is that instead of importing still images into either Fusion or the Blender compositor, you import image sequences.
One thing that i dont understand is why are you setting up output node in compositor? Blender does set up multilayer exr with all the passes and viewlayers without it. I guess if you want each viewlayer as separate exr?
I use scenes instead of view layers as it's easier to automate file path settings. It would be nice to be able to set a single view layer to render with command line rendering. I'd need this as my render farm does all of my rendering. Also we rarely use holdouts in VFX houses, we use deep compositing to do holdouts in nuke. Such a shame blender and fusion don't support deep rendering
Doesn't rendering in a multilayer reduce the performance of reading and processing the sequences? I mean in scenarios when you render a bunch of AOV's but only end up using a few of them. Haven't really tested myself but maybe someone knows here.
I have an issue. Trying to make a simplified version of this workflow to composite in Blender as the last step as well. And I am following your nodeflow to the letter, but I have a problem with compositing volume to look correctly. In my demo scene I have a bunch of random objects and a thick smoke in the center. It is rendered as its own separate viewlayer/exr and when I composite it into the nodeflow it looks as if it's on 20% opacity and no matter what I do I can't get it to show properly. Used alpha over, mix, math nodes etc. Not sure what I am doing wrong.
Hi, Can you explain why, if you export a Multi-layer EXR and plan to open it in DaVinci or After Effects, you go through the compositing process and ensure to export each layer separately? If I'm not mistaken, and this is how I actually work, I just export a Multi-layer EXR and can access the different passes through the embedded information in the file itself.
So with working with an external compositor, (I use Fusion), would you denoise the passes in Blender and export each frame there, or would you denoise them in the external compositor of choice before moving on? I don't know how the denoising works in Davinci so I'm unsure if it's temporal like Optix, which makes animations flicker less. Great video by the way!
Thanks! You have a few options in Resolve. There's a built-in denoiser which is indeed temporal, but it's not great. If you don't want to spend money on a third-party tool like Neat Video, your best bet is to denoise them in the Blender compositor prior to the file output node. But that is not temporal I believe. Not without doing the whole complicated temporal denoising workflow. Or have they integrated it lately without me noticing?
19:26 why is the math wrong for jpeg? is it because grading 8-bit can be destructive, cause then we can just use higher depth exr with sRGB... also CST nodes won't replicate OCIO nodes, right? would've been nice if they did as it has both gamma (ACEScc & ACEScct) and color space (ACES ap0 & ap1) for aces.
Great tutorial, Is there a way to colour correct/grade each individual render layer in resolve non destructively? So instead of affecting everything at once it would only affect the one layer and that layer wouldn’t have to be re- rendered and added back in.
![ไปอยู่ไสมา - ช๊อปเปอร์ โตเกียวมิวสิค [ OFFICIAL MUSIC VIDEO ]](http://i.ytimg.com/vi/sWuVNbRTJbc/mqdefault.jpg)
Animation? Yes! The workflow is the exact same. Just replace the single image files with image sequences.
hi can you make a video showing the flexebility of those passes and how much can be chanegd through the passes or is it just lighting and color not like adding more fake environment
but i love .mp4 files i do not like image sequences ):
Quick question! I rendered as multilayer exr with the file outputs in blender, but when pulled into davinci, it is only one of the 4 layers as a sequence. I had overwrite checked in the image sequence settings but i noticed you also had that selected... Would that mess up the layers?
Edit: Never mind lol I just figured it out. Sorry, I am still learning davinci.
@@timenotspaceproduction What are you gonna do when multiple frames get a graphical glitch? Render it out all over again? You can just drag and drop image sequences into most video editing softwares and it'll play as a video, and it won't be compressed like an mp4.
@@xodiz yeah you're right
I had to take down the last video because of a background music track. I had misunderstood the license, so I couldn't earn anything from the video. We should be all good now.
Thanks for the video regardless 🙏
Thanks for the video regardless (1)
I was wondering why it was taken down. Good call btw. It is an awesome video and it will gather good views.
Could you please upload your tutorials with no background music at all when you are speaking? The music serves no purpose and for some people it makes the video extremely stressful to watch. Because music is such a personal thing no matter what you choose some people will hate it, and that's where the stress comes from. In addition some autistic people (like me) will also struggle to understand what you are saying even if the background music is quiet.
I had to stop watching this video about half way through because the music was causing me a lot of stress.
The music is just distracting anyway
it being a re-upload is a good enough excuse for me to re-watch it
Haha same
If it weren't for this re-upload, I wouldn't have seen it
was it really a reupload, i swear i watch this video. was confused why it says 1 month ago
A good excuse to rewatch the video for a 4th time!
Amazing explanation! And thanks for the little clip in the background being honest about its usefulness.
As a professional CG artist I've found that its not necessary for a lone artist in 99% of cases. Studios do this because going back down the pipeline for minor changes means dozens of people have to redo certain parts (and get paid for those hours haha). I've used passes a few times to adjust my render, but I have never found a usecase for rendering out every collection in its own renderlayer yet.
I think the best use-case for rendering so many separate layers and passes is to create a killer VFX breakdown of your project ;)
As a hobbyist, I actually use some of those tricks (especially the view layers) to deal with performance issues. For example, hair can take a long time to render, and using a separate view layer for it can help. This way I don't have to render the hair again when I do minor modifications to the scene. It's also a good way to have a complex background or volumes without the long render time that comes with it. It's a nice way to introduce separation of concerns into my workflow. (Of course, having a better machine can help with that haha!)
@@corentinsakwinski4500 exactly! I have only 8gigs of vram and I need to render things separately to fit into that.
Totally agree! If the project does not specifically ask for passes in compositing, I also believe that it's unproductive to do that. But at the same time, I've had clients that specifically asked for minor tweaks in volumetrics or lighting and that was a total pain considering I didn't render in passes. The lesson for me was to always check project requirements and non-negotiables. Render passes are one of the tools we can use as cg artists to make our work easier, but it doesn't mean we have to use it all the time.
I like rendering individual layers 'cause I use Blender mostly for paintover in Photoshop, but I don't need half of the passes he showed, by just having every part separated and a ID pass to select everything more easily is enough.
saving this video so i can come back to it in a few months once im half decent at blender
Blender is SO damn amazing! I'm still confused by basic shit because it's not my primary obsession lol. When I see channels like this do these things it looks like magic even when I watch you do every step!🔥🤯
As a beginner trying to make my works more controlled and professional, this is absolutely a godsend. Thank you!
Why do you talk to the camera like im having a meltdown and you are trying to let me know everything will be ok? Will it be ok tho? You promise?
This is single handedly the best blender tutorial video I've ever seen. Every chapter of it felt composed and to the point. Really well done man
A piece of gold! The only video you need to render and composite your works.
Very cool. I've been wondering about all of this for like a decade! But I stayed away because it was the big bad wolf until you came along and blew his house down. It all clicked as I watched. Thanks for the tour! Cheers
Really like how you explain things. I have something to add for this video:
- ACEScg, ACEScc, ACEScct are all working colorspace. You should treat your 3D render like camera footage: capture as much info as possible. Therefore, ACES2065-1 (AP0 - linear) is the colorspace for handover and archive. While cg, cc, cct or (AP1 - linear, log, log) are the purpose of working colorspace. SO, render them in ACES2065 instead of ACEScg is better. However, you have to do another mapping from ACES2065 to ACEScg, which takes time. I think fusion on resolve 19 has that color management for fusion as a project setting, which can automatically do things for you. (I’m a colorist, I don’t work much in Fusion page so… learn it yourself, IDK fusion 😅)
- For color grading purposes, output it to ACEScg and let the colorist do their job (colorist will map those out to ACEScc or cct, or whatever working colorspace they prefer). PLEASE, PLEASE,PLEASE don’t do rec709 transform or anything. Once you squeeze things down to a smaller gamut, there is less information to work with, and it hurts. You would not want to waste all the effort of doing all those render separately and then squeeze them into a sRGB or rec709 and hand it to the next step 🙂, b/c they will punch you 🤛.
Hope those help some 3D + VFX people🤞.
I saved the previous one to my playlist, I got so worried where it had gone,
thank god this one popped up in my homepage.
bro is megamind😩🔥
Okay, this video has convinced me of the value in this channel for me to sub. Youve really iutdone yourself Robin.
Im gonna try to do this myself now.
This comment is just for the algorithm to understand I loved it and show it to more people. (since you had to delete the original one)
Liking and Commenting for the Algorythm.
This was an amazing video, and deserves the views it had on it's original upload.
Great work Robin!
Wow! I knew before about render layers but never rendered it like that, and finally it seems that ocio fixes exr colors
That's a THICC render!
I think this video is pure gold and is a subject that's rarely covered on YT... it would be amazing Robin if you could do a tutorial video encapsulating all these processes.. even just for one scene.... I don't think any other blender channel has done this :)
This is one of the best videos I've seen in a long time
That blade runner 'render' looks more like one of the actual model shots they used. Would not be surprised if that was a practical shot.
I've been told by commenters that it is indeed a real model, yes. I think that's hilarious.
keep reuploading because i will KEEP WATCHING
What an absolute banger of a video. I might not follow or need the complete workflow but I know I would need parts of it. But leaving the point of what I need or do not need, it was an absolute delight and a knowledge boost to watch the entire video. The way you told everything in sequence was so awesome and easy for me to understand and I don't even use resolve. Thank you for making this video. You, good sir, earned yourself a subscribe.
Phenomenally easy explanation of a complex subject!
i dont know how much things have changed on 3Dsmax but i remember i used to just click on the Passes tab on the render window and select the ones i wanted, havent done that in a long time, other than rendering just zdepth and material ID
This is a very insightful video, thank you Robin.
Just watching your video made me feel like a big boy, this is the first time i commented in TH-cam in years, absolutely amazing video
Please make a video showing how you made the dust from the flying thingy. It looked soo good and realistic i never managed to reach that point
Wow, I've needed to see this video for a long time. I was making view layers way too hard for myself. I was also using up memory placing holdouts and shadow catchers into my other layers to represent my objects from another layer... That was brutal on my ancient PC. Also I was completely unaware of the Channel Boolean node in Resolve, I've been super confused why those options weren't in the merge node.
you really make the subjects you cover come alive!
This was a great refresher video for when I first learned this in university over a decade ago, thank you! :)
What a bunch of nicely composed information I will need but always was too much of a hassle to get started with. thank you man
This is a lovely tutorial! So high quality and well paced. This is rare on youtube! Thank you:)
I've followed along now and is done with a draft render. I'm now not sure where to begin when i have all this flexibility in my layers and passes. I'm hoping you're diving into this in the next video! Love your style and focus.
Gracias por volver a subir el video, justo el día que quería aplicarlo a mi proyecto
best video on composite so far
This is where the magic's at. Thanks!
I did not even consider using layers before, but recently i had to because how Grease Pencil interacts with semi transparent materials. It's not as complicated as i thought before, but there are extra steps to be used for sure.
This is the exact video I was looking for like a month ago. Thank you so much and great work!
Ornithopter, or thopter shortened. Anyways thank you for showing how to split all the passes prerender it's pretty much necessary with how much physics simulations go into creating realistic scenes from light, gravity, fluid dynamics collision, etc.
I would like this video twice if I could.
about the feedback you asked for in the first video, the song is pretty nice, helps to keep the viewer engaged but doesnt steal focus, the pacing of the tutorial is pretty good, honestly only thing i have to say is that your speech could be more consistent but honestly i do not mind and imo it does not hurt the tutorial, overall great stuff, love to have those in sights into the pro workflow, i dont intent on going that deep but i for sure will be rendering layers individually whenever my PC becomes concernealing loud, and having the option to change individual aspects without re-rendering the full image is actual gold, quick iterations are the core of computer based creative processes, really enjoyed the video, for sure can picture myself coming back to check a thing or two
Great video, thanks for the info. I've done this kind of rendering for years in different software but never in Blender. It's good to see how to do render layers correctly (in Blender), it looks like it works quite well.
However from what I can tell Blender can't do one important render task. The ability to apply specific shaders to specific objects per render layer. This is the feature for me that's holding Blender back from being really flexible for rendering. I think on almost every job I've ever done I've needed this feature for at least one shot. The fact its not possible so far in Blender is really weird.
I enjoyed the comedic moments. Nice!
Am I having deja vu? Guess not time to watch this new banger!
I needed this video more than I knew.
Love the video. Greatin seeing and finally understanding this workflow. Especially tips like turning on 'Indirect only' etc. Thanks!
this summed up can be said as "you trade off image quality and your pc not dying for ten times the render and post processing time"
The best explainer of these theme
I will integrate this in my workflow soon! Thanks a million man!
This is so helpfull!!🙏
wow! this video have everything i need to get better, thank you for sharing!
Nothing short of amazing
Yes! So glad this is back up! 💪🏼💪🏼
I'm so excited to watch this video!
Excellent video bro
Blender should add denoise checkboxes on that node if it requires it. Or if it supports multi-selection to pull out to conjoin to a multi-input denoise filter node.
GOATED tutorial
Ahh, so its not just me that has an 'Everything' view node, I just call it ALL, but yes it's very useful having this layer.
Top notch content!
awesome dude! thanks man:)
Time to watch it again! Thanks for the awesome resource
Good stuff Robin! I'll keep an eye out for your videos.
It can be a good idea to denoise the volume passes separately to everything else as they often have no relevance to the normal and albedo passes. If you can afford to wait.
Time ago looking for a tutorial like this, thx!
This was really helpful. I've experimented in the past with trying to do a proper "Professional" pipeline for my projects, but it always just turns into a confusion mess (The Holdout feature was something I just learned about in this vid). I'll try something like this again on my next big project.
Thanks for this piece of gold, new subscriber here !!!
A better way of getting an accurate preview in DaVinci is to set the project colour management to have ACES as the input and rec.709/2020 or whatever your deliverable needs to be as the output. That way you see a colour correct image with no need for a LUT and no need for OCIO or CST, and your timeline has a wide gamut for grading.
Yes, that is also a good workflow!
BROOOO thank you. love it!!!
Hello Robin. Thanks for the great content! I had already watched the previous video, and I came here to like and view it to help with engagement. It's a great video. I had already tried to understand this type of workflow, but your video was the first one in which I really understood how this type of composition works. Do you have any videos that teach this same workflow, however, for composing animations? I would like to adapt this workflow to compose a sequence of images for an animation, but I can't find anything as good as your latest video. Again, thanks for your great content.
If you have already created or will create a course about this, I will be the first on the waiting list haha @robinsquares
Hey, I should have mentioned that, but the workflow is exactly the same. The only difference is that instead of importing still images into either Fusion or the Blender compositor, you import image sequences.
best vid i have ever seen
your video style reminds me of vox . really nice
Ahh the ornithopter the CGI vehicle with 1000 subframes data ahaha
Great tutorial dude, make sure you set the alpha to do nothing in channel booleans.
woo Affinity go!
Nice, I wanted to watch it again
non i watch againt it wanted i Nice
well efforted video
they say at my studio, it's a 3D artist's job to get 2D (compositors) what they need.
amazing information!
great video
One thing that i dont understand is why are you setting up output node in compositor? Blender does set up multilayer exr with all the passes and viewlayers without it.
I guess if you want each viewlayer as separate exr?
THANK YOU!!🥰
yeah, the best video, again
I don’t think I’m ready for this looool
I use scenes instead of view layers as it's easier to automate file path settings. It would be nice to be able to set a single view layer to render with command line rendering. I'd need this as my render farm does all of my rendering. Also we rarely use holdouts in VFX houses, we use deep compositing to do holdouts in nuke. Such a shame blender and fusion don't support deep rendering
Thank you
Doesn't rendering in a multilayer reduce the performance of reading and processing the sequences? I mean in scenarios when you render a bunch of AOV's but only end up using a few of them.
Haven't really tested myself but maybe someone knows here.
Please also cover all auxiliary passes as well, specially uv, world position, normals
Jesus just when i thought im close to understanding all of blender some new tutorial reminds me that i know very little😭. where does it end!?
awesome
Yess thank you thank you thank you
i fuking love u bro, this is GOLD
I have an issue. Trying to make a simplified version of this workflow to composite in Blender as the last step as well. And I am following your nodeflow to the letter, but I have a problem with compositing volume to look correctly. In my demo scene I have a bunch of random objects and a thick smoke in the center. It is rendered as its own separate viewlayer/exr and when I composite it into the nodeflow it looks as if it's on 20% opacity and no matter what I do I can't get it to show properly. Used alpha over, mix, math nodes etc. Not sure what I am doing wrong.
Hi,
Can you explain why, if you export a Multi-layer EXR and plan to open it in DaVinci or After Effects, you go through the compositing process and ensure to export each layer separately? If I'm not mistaken, and this is how I actually work, I just export a Multi-layer EXR and can access the different passes through the embedded information in the file itself.
I'd be curious to know as well
So with working with an external compositor, (I use Fusion), would you denoise the passes in Blender and export each frame there, or would you denoise them in the external compositor of choice before moving on? I don't know how the denoising works in Davinci so I'm unsure if it's temporal like Optix, which makes animations flicker less.
Great video by the way!
Thanks! You have a few options in Resolve. There's a built-in denoiser which is indeed temporal, but it's not great. If you don't want to spend money on a third-party tool like Neat Video, your best bet is to denoise them in the Blender compositor prior to the file output node. But that is not temporal I believe. Not without doing the whole complicated temporal denoising workflow. Or have they integrated it lately without me noticing?
deep compositing is just better
Is it more compute intensive to use this workflow? Or does the render time equal to the beauty render time?
More compute intensive, yes.
19:26 why is the math wrong for jpeg? is it because grading 8-bit can be destructive, cause then we can just use higher depth exr with sRGB... also CST nodes won't replicate OCIO nodes, right? would've been nice if they did as it has both gamma (ACEScc & ACEScct) and color space (ACES ap0 & ap1) for aces.
Great tutorial, Is there a way to colour correct/grade each individual render layer in resolve non destructively?
So instead of affecting everything at once it would only affect the one layer and that layer wouldn’t have to be re- rendered and added back in.