Howeverwhat should be easy in Premiere and Resolve never is. The tutorials I've seen all deal with adding motion blur in the Transform panel to moving layers. Not what we want. There is an echo tool, but that doesn't look any good at all - tons of ghosting and takes ages to render, won't play in real-time.
Yep we posted in another thread about adding motion back using Resolve..... I did experiment more so it depends a lot on the scene and background as just a tad of blur does not add too many artifacts in case of complex bg like a metallic fence can cause bad artifacts.
My test was a show jumping horse in a parkour the goal is to pull picture at 1/250 or 1/500 and have a usable video.
So to an untrained eye you can fool them but for real pro work not sure I would use it. But in case where photo has priority I will definitely use it.
This is a frame pulled from the video at 1/200 (for a photo a bit on the limit as you can see on the horse legs that are a bit too blurry), I have the video with added mention blur if I have the time I can pull the frame with the blur to compare.
Forget the horrible location etc. as here most of the things are still closed.
The one in resolve is close enough that I'm not going to use ND's for shots on my Z6 anymore.
Also for VFX work results are far better when you pull a key off a high shutter speed and then add the motion blur in post. Makes cleaning up hair work far easier.
It is GPU accelerated, I get Realtime playback in 4K on my GTX 2080. Just did this quick test to show how effective it is.
It's not perfect, but it's so close that 99.9% of people wouldn't be able to tell the difference. Titles go faster than I expected but you get the gist.
I'd be interested to see some stress-testing of it, like Lok did on digital rev when he did star-jumps and it really emphasised the effect. Ultimately it will come down to the motion detection and estimation so it knows what to blur.
In Resolve it would be pretty easy to bake the effect in while converting - just drag all clips to the timeline and then do an export and set it so that all the clips are individually exported. This workflow would also work if converting from RAW, and would handily be a place to apply LUTs any other things that always get added to your grades.
I am excited to see what you find Andrew. I have always wished there was a good (and fairly quick) way of doing it in either Premiere or FCPX, but the ones I have seen are not great and/or take too long to be worth it when I could just throw an ND on. But it would be great if there was.....
Of course, this depends on the order of operations - if the motion blur is applied after the speed change then the input to the effect might only be the frames that survived the speed change, so extra frame rate wouldn't help the accuracy of the effect.
I think the point that @Deadcode was making was that the extra frames can help the motion estimation better, which makes sense, although I'm not sure that it would actually work that way in how they would have likely implemented the video processing pipeline.
The problem im having is that while the motion data from the vector pass is blurring correctly inside of the character pass, it wont blur past the alpha ( that is it wont blur the edges of my characters )
I have been surfing the web for a little now and can find plenty of comparisons between Kronos and Twixtor, but none between MotionBlur (which is the motion blurring plug-in that comes with Kronos) and Reelsmart Motion Blur.
RSMB (which is the one I chose) can simply use a motion vector pass straight from my 3D app, avoiding any motion estimation to try to deduce the vectors (and therefore avoiding blobbing or the need for tweaks).
use the matte layer with inverted alpha to remove artifacting and isolate multiple passes. plus use vectors 100%, smoothing 28. filter extreme. you can try RSMB as a trial and see if you like it better. nothings perfect, after all:)
use the matte layer with inverted alpha to remove artifacting and isolate multiple passes. plus use vectors 100%, smoothing 28. filter extreme. you can try RSMB as a trial and see if you like it better. nothings perfect, after all:)
The current Chronos AI model already produces higher quality intermediate frames than anything Adobe can do. Currently the best way to add artificial motion blur is with After Effects with the Pixel Motion Blur effect. I think it would be great to see what VEAI could do on this front. High shutter speed 24/25 fps smartphone video would look much more professional.
Edit: Technically you can already do this with some extra work. For example if you would interpolate a clip with the Chronos model at 2000% you would have 19 extra sample frames. If you import and interpret that clip in a program like Premiere Pro at its intended framerate and then speed it up by 2000% with Frame Blending you could get a similar result.
I totally agree, there are some tools on the marked but often they lack from distinguishing static background from moving objects and the object as well as the background is blurred (e.g. RSMB). Chronos does a great job in separating differently moving object layers. Hence, the AI approach should work perfectly here.
I would like to see a feature that adds realistic motion blur to footage recorded at a high shutter speed. I strongly dislike having to increase the shutter speed on my camera to compensate for overexposure. However, it is the most convenient way to shoot with a wide aperture without having to purchase expensive ND filters for each lens. The downside of this approach, though, is that it eliminates natural motion blur, which is what I would like to preserve. Hopefully, someday this feature can make its way to the app.
RSMB looks at the frames either side of each frame to determine the amount of blur to apply. It analyses the actual motion in your clip and only applies blur to pixels that move, and parts of the image that move more - get more blur. This is a very computationally heavy processing. There is also a setting within RSMB to determine how many frames it looks forward and back. The higher this number the (much) greater the render time.
The composition being rendered is 11 seconds long with a few instances of the CC Pixel Polly effect and of course the two different motion blur effects. The machine has a quad core i5 processor with 16GB of RAM. Nothing too speedy, but it should give you a general idea of the test setup.
So as you can see, RSMB performed a lot faster than CC Force Motion Blur, finishing nearly in half the time! This plugin could save you tons of time if you were rendering something a bit more intensive or longer like a 30 second 3D animation composited in After Effects, etc.
Why not just use a timewarp node. I did this on a sequence that was shot with a narrow shutter angle, which looked strobey. I used TW, with motion selected and motion blur, it worked well for just adding that little bit of mo blur to take the curse off it.
We have to do this quite a bit, when stuff has been shot off-speed (high fps) but client wants to pretend it was shot at native fps. I was never very happy with the native options in Flame or even Sapphire S_BlurMotion.
The 100% TW trick can create really dodgy results when compared to simple Motion Analysis/Motion Blur. If you use motion analysis, make sure you select local rather than global. It seems to give much better results.
Work by Comen VFX. RSMB was used to give a more filmic look to footage shot on video, providing shutter speed correction. The footage looked choppy when played back at film rates. So, the client settled on RSMB to simulate the normal shutter speed you would have with film footage.
Pro license works with After Effects, Baselight, Final Cut Pro, Flame, Smoke, Fusion Studio, HitFilm, Media Composer, Motion, Natron, Nucoda, Nuke, Premiere Pro, Scratch, and Silhouette. Pro license also works with the regular product of any supported host.
It is important to note that our GPU accelerated modes of our plug-ins can provide you with different results than on the CPU. As such, it is not wise to mix CPU-generated images with GPU-generated images.
If a GPU is not supported, or we run out of GPU memory to run on, our plug-ins will return a green frame to let you know. So if you switch the Use GPU menu to ON and you see a green frame immediately, you know that you do not have a supported GPU. We do not fallback to CPU rendering in the case where there is an error on the GPU during a rendering of a sequence, because then you might get CPU and GPU calculated images intermingled and may not understand why your sequences are not looking proper from frame to frame.
First, Premiere Pro does not need to support the GPU, even though our plug-ins use OpenCL for acceleration. Nor do you need to have the Mercury engine running in OpenCL or CUDA mode to use our plug-ins in GPU mode. It is important to note that our GPU accelerated modes of our plug-ins can provide you with different results than on the CPU. As such, it is not wise to mix CPU-generated images with GPU-generated images.
OFF This option causes the plug-in to run on the CPU. This mode should be used if you do not have a GPU that our plug-ins support, or see a green frame in the middle of a rendered sequence, because this means that our GPU rendering fails somewhere in the middle of a sequence.
ON This option forces the plug-in to run on the GPU.
I wasn't planning on using this one as part of the bundle, but after finally taking the time to become familiar with it, I saw what the buzz about it was. Motion blur (or the removal thereof) is a large part of the kind of action videos I edit. Glad I took the time, you will be, too.
3a8082e126