A few notes to add. I would suggest you consider either turning off music track in your video so your voice is more clear and easier to understand or better yet, apply auto ducking so that when you talk the music track volume ducks under your voice volume and we can hear your speech more clearly. Second note about the super scale option in resolve. Since resolve uses lot of AI type effects and assistant tools, including for motion estimation and masking etc. I was hoping they have up scaling also running AI, but according to their manual its just a more advance traditional method. Overall I find it ok, but not great. It can smudge fine details and it does not have ability to apply context sensitive sharpening, so its pretty much limited in controls. I haven't tried it with CG renders like you did, which probably could do the job well, but for organic type footage with people and lot of fine details, it is somewhat limited compared to other options. Currently Topaz Video enhance AI does the best job, and its best to go with version 2.6.4 instead of latest 3.1 since its quite unreliable and buggy at the moment. Anyway, just few notes to add. Cheers!
Topaz AI is definitely the best AI upscale toolset in the market for sure! I appreciate your willingness to share your knowledge and expertise. I was aware of the traditional methods for super scale in previous versions, but as of Davinci 18, their website and documentation includes the super scale in its newest additions to AI powered tools. I may be mistaken, but it’s in big red letters if you scroll down on their homepage for Davinci Resolve 18. With audio, was it the starting music or the music during the tutorial that bothered you? Either way I’ll stick to keeping the music lower in general.
@@owenjenkinsofficial Cool thanks. As for the super scale being AI power or not. Its a bit confusing. Originally like you I though it was AI power. Its advertised as you pointed out on their page as one of the AI features. However it could be marketing oversight. Once I went to see the official Resolve 18 manual. They mention every feature that is powered by their neural engine and mention each time a feature is using it. With super scale the AI parts is absent. So I'm not sure if I should trust the manual or their webpage. I would think manual is written by people who know the features and website is primerally for advertising. I personally am leaning to manual as authoritative source on this, but weather its AI or just more complete version of standard algorithms it does work a bit better than the previous ones.
@@owenjenkinsofficial Like I said, its hard to know what the actual algorithm does under the hood, but according to resolve manual this is what they say. Super Scale: Sets a very processor-intensive and high quality upscaling algorithm that actually creates new pixels for the resized image. The possible values are: None, 2x, 3x, 4x, and Auto. For more information on Super Scale, see Chapter 11, “Image Sizing and Resolution Independence.” The DaVinci Resolve Sizing Pipeline This section discusses the various sizing controls that are available in DaVinci Resolve, and how they work together. “Super Scale” High Quality Upscaling (Studio Version Only) For instances when you need higher-quality upscaling than the standard Resize Filters allow, you can now enable one of three “Super Scale” options in the Video panel of the Clip Attributes window for one or more selected clips. Unlike using one of the numerous scaling options in the Edit, Fusion, or Color pages, Super Scale actually increases the source resolution of the clip being processed, which means that clip will have more pixels than it did before and will be more processor-intensive to work with than before, unless you optimize the clip (which bakes in the Super Scale effect into the optimized media) or cache the clip in some a UHD timeline, or when you find it necessary to enlarge a clip past its native resolution in order to create a closeup. You may find that, depending on the source media you’re working with, setting Sharpness to Medium yields a relatively subtle result that can be hard to notice, but setting Sharpness to high should be immediately more preferable, while also sharpening grain and noise in the image to an undesirable extent at the default settings. However, while raising Noise Reduction will ameliorate this effect, it will also diminish the gains you obtained by raising Sharpness. In these cases, it’s worth experimenting with keeping Sharpness at Low or Medium so that Super Scale sharpens all aspects of a clip, but then using the Noise Reduction tools of the Color page (with their additional ability to be fine-tuned) to diminish the unwanted noise. TIP Super Scale, while incredibly useful, is a processor-intensive operation, so be aware that turning this on will likely prevent real-time playback. One way to get around this is to create Optimized Media for clips in which you’ve enabled Super Scale, since Optimized Media “bakes in” the Super Scale effect. Another way to work is to create a string-out of all of the source media you’ll need to enlarge at high-quality, turn on Super Scale for all of them, and then render that timeline as individual clips, while turning on the “Render at source resolution” and “Filename uses > Source Name” options.
@@owenjenkinsofficial When their Neural Engine is used, they usually specifically say it. Like in this section on frame interpolation for slow motion effects. That is why I'm not sure if super scale is really aI or not. Just a minor geeky observation. Motion estimation mode: When using Optical Flow to process speed change effects or clips with a different frame rate than that of the Timeline, the Motion Estimation pop-up lets you choose the best-looking rendering option for a particular clip. Each method has different artifacts, and the highest quality option isn’t always the best choice for a particular clip. The default setting is “Project Settings,” so all speed-effected clips are treated the same way. There are several options. The “Standard Faster” and “Standard Better” settings are the same options that have been available in previous versions of DaVinci Resolve. They’re more processor efficient and yield good quality that are suitable for most situations. However, “Enhanced Faster” and “Enhanced Better” should yield superior results in nearly every case where the standard options exhibit artifacts, at the expense of being more computationally intensive, and thus slower on most systems. The Speed Warp setting is available for even higher-quality slow motion effects using the DaVinci Neural Engine. Your results with this setting will vary according to the content of the clip, but in ideal circumstances this will yield higher visual quality with fewer artifacts than even the Enhanced Better setting.
Either way it’s pretty powerful and has worked well for me regardless. The optical flow is also amazing. I look forward to the future with optimism to see what innovations happen next!
A few notes to add. I would suggest you consider either turning off music track in your video so your voice is more clear and easier to understand or better yet, apply auto ducking so that when you talk the music track volume ducks under your voice volume and we can hear your speech more clearly.
Second note about the super scale option in resolve. Since resolve uses lot of AI type effects and assistant tools, including for motion estimation and masking etc. I was hoping they have up scaling also running AI, but according to their manual its just a more advance traditional method. Overall I find it ok, but not great. It can smudge fine details and it does not have ability to apply context sensitive sharpening, so its pretty much limited in controls. I haven't tried it with CG renders like you did, which probably could do the job well, but for organic type footage with people and lot of fine details, it is somewhat limited compared to other options. Currently Topaz Video enhance AI does the best job, and its best to go with version 2.6.4 instead of latest 3.1 since its quite unreliable and buggy at the moment.
Anyway, just few notes to add. Cheers!
Topaz AI is definitely the best AI upscale toolset in the market for sure! I appreciate your willingness to share your knowledge and expertise. I was aware of the traditional methods for super scale in previous versions, but as of Davinci 18, their website and documentation includes the super scale in its newest additions to AI powered tools. I may be mistaken, but it’s in big red letters if you scroll down on their homepage for Davinci Resolve 18. With audio, was it the starting music or the music during the tutorial that bothered you? Either way I’ll stick to keeping the music lower in general.
@@owenjenkinsofficial Cool thanks.
As for the super scale being AI power or not. Its a bit confusing. Originally like you I though it was AI power. Its advertised as you pointed out on their page as one of the AI features.
However it could be marketing oversight. Once I went to see the official Resolve 18 manual. They mention every feature that is powered by their neural engine and mention each time a feature is using it. With super scale the AI parts is absent.
So I'm not sure if I should trust the manual or their webpage. I would think manual is written by people who know the features and website is primerally for advertising. I personally am leaning to manual as authoritative source on this, but weather its AI or just more complete version of standard algorithms it does work a bit better than the previous ones.
@@owenjenkinsofficial Like I said, its hard to know what the actual algorithm does under the hood, but according to resolve manual this is what they say.
Super Scale: Sets a very processor-intensive and high quality upscaling algorithm that actually creates new pixels for the resized image. The possible values are: None, 2x, 3x, 4x, and Auto. For more information on Super Scale, see Chapter 11, “Image Sizing and Resolution Independence.”
The DaVinci Resolve Sizing Pipeline
This section discusses the various sizing controls that are available in DaVinci Resolve, and how they work together.
“Super Scale” High Quality Upscaling (Studio Version Only)
For instances when you need higher-quality upscaling than the standard Resize Filters allow, you can now enable one of three “Super Scale” options in the Video panel of the Clip Attributes window for one or more selected clips. Unlike using one of the numerous scaling options in the Edit, Fusion, or Color pages, Super Scale actually increases the source resolution of the clip being processed, which means that clip will have more pixels than it did before and will be more processor-intensive to work with than before, unless you optimize the clip (which bakes in the Super Scale effect into the optimized media) or cache the clip in some a UHD timeline, or when you find it necessary to enlarge a clip past its native resolution in order to create a closeup.
You may find that, depending on the source media you’re working with, setting Sharpness to Medium yields a relatively subtle result that can be hard to notice, but setting Sharpness to high should be immediately more preferable, while also sharpening grain and noise in the image to an undesirable extent at the default settings. However, while raising Noise Reduction will ameliorate this effect, it will also diminish the gains you obtained by raising Sharpness. In these cases, it’s worth experimenting with keeping Sharpness at Low or Medium so that Super Scale sharpens all aspects of a clip, but then using the Noise Reduction tools of the Color page (with their additional ability to be fine-tuned) to diminish the unwanted noise.
TIP Super Scale, while incredibly useful, is a processor-intensive operation, so be aware that turning this on will likely prevent real-time playback. One way to get around this is to create Optimized Media for clips in which you’ve enabled Super Scale, since Optimized Media “bakes in” the Super Scale effect. Another way to work is to create a string-out of all of the source media you’ll need to enlarge at high-quality, turn on Super Scale for all of them, and then render that timeline as individual clips, while turning on the “Render at source resolution” and “Filename uses > Source Name” options.
@@owenjenkinsofficial When their Neural Engine is used, they usually specifically say it. Like in this section on frame interpolation for slow motion effects. That is why I'm not sure if super scale is really aI or not. Just a minor geeky observation.
Motion estimation mode: When using Optical Flow to process speed change effects or clips with a different frame rate than that of the Timeline, the Motion Estimation pop-up lets you choose the best-looking rendering option for a particular clip. Each method has different artifacts, and the highest quality option isn’t always the best choice for a particular clip. The default setting is “Project Settings,” so all speed-effected clips are treated the same way. There are several options. The “Standard Faster” and “Standard Better” settings are the same options that have been available in previous versions of DaVinci Resolve. They’re more processor efficient and yield good quality that are suitable for most situations. However, “Enhanced Faster” and “Enhanced Better” should yield superior results in nearly every case where the standard options exhibit artifacts, at the expense of being more computationally intensive, and thus slower on most systems. The Speed Warp setting is available for even higher-quality slow motion effects using the DaVinci Neural Engine. Your results with this setting will vary according to the content of the clip, but in ideal circumstances this will yield higher visual quality with fewer artifacts than even the Enhanced Better setting.
Either way it’s pretty powerful and has worked well for me regardless. The optical flow is also amazing. I look forward to the future with optimism to see what innovations happen next!