Workflow Overview: Lightspeed Wan Series
These workflows are optimized for the Lightspeed Wan 2.1 and 2.2 Image-to-Video (I2V) models by darksidewalker. They are categorized into the streamlined SVI-V2-Pro series and the feature-rich Standard (v12 and below) series.
1. SVI-V2-Pro-I2V (Barebone Series)
The SVI series is designed for users who want a clean workspace without unnecessary overhead. It focuses on the core I2V logic with two critical utility modules.
A. Resolution Management (Resolution Master)
To ensure compatibility with Wan 2.2’s specific latent requirements, the source image must be precisely resized.
Step 1: Load the source image into the Load Image node.
Step 2: Use the dropdown (next to the TV icon) to select a target resolution (e.g., 480p, 720p).
Step 3: Click the TV Icon to activate the selection.
Step 4: Under 'Actions', set the slider to 16 and click the Snap button.
Note: This "snapping" ensures the dimensions are multiples of 16, preventing tiling artifacts or model errors.
B. Temporal Smoothing (Interpolation)
Located in a dedicated group, this module handles frame rate conversion.
Primary Use: Converting low-FPS raw generations (e.g., 16 FPS) into smooth, high-fidelity output (e.g., 60 FPS).
2. Standard Wan Workflows (v12 and Below)
The legacy/standard workflows are "all-in-one" solutions, containing a broader suite of post-processing and utility tools:
Image to Video
The core generative engine for Wan 2.1/2.2.
FLF Loop
Facilitates seamless looping by utilizing First-Last Frame logic.
Downscale Image
Safely shrinks oversized source images to prevent VRAM OOM (Out of Memory) errors.
Extract Last Frame
Pulls the final frame of a generation to be used as a starting point for the next clip (Storyboarding).
Combine Videos
Merges multiple generated clips into a single sequence.
Interpolation
Increases frame rate for smoother motion.
Upscale Video
Increases the spatial resolution of the final video output.
Description
Rearranged some of the nodes to make it easier(-ish) in case anyone disconnects something by accident.
Added manual switches for CLIP and VAE models.
Now you can toggle between 1 (Checkpoint) or 2 (Loader).
Removed the 2 KSamplers and replaced them with a single WanMoeKSampler
This is connected to WanMoEScheduler
Sampler and Scheduler will have to be set on these nodes.
You probably should use the same scheduler setting on both, but you do you.
Check their project pages if you want explanations on how they work.
I'm just throwing spaghetti at the wall with these based on the info on their project pages.
I don't claim to be an expert on Wan or how it works. You're all either learning with me or just enjoying the success and failures of my experimenting!
Tested on the High+Low versions of this model.
A 480p 5-second video takes 2 minutes on my 5060ti 16GB + 64GB RAM (using the recommended 4 steps). This is plenty fast enough for me considering my hardware.
FAQ
Comments (10)
I like the workflow; I was able to work with it, but I noticed something. I’ve been working with Daxamur’s workflow https://civitai.com/models/1853617?modelVersionId=2165705 which uses three KSamplers,He told me that it helps improve motion, and it really does. With his workflow, I’m able to create 9-second videos at resolutions between 820p and 1080p in less than 10 minutes. However, that workflow isn’t very friendly with fluids — it tends to make them disappear.
Now, in his setup, the upscale and interpolate processes run at the end, once the x-second clip has already been generated. In your workflow, however, the upscale happens while the clip is still being generated, which makes the generation process take much longer.
I’m not an expert, but is there a way for the interpolate and upscale steps to be done at the end, after the video has already been generated?
I don't have an upscale in my flow. So I'm not really sure what u mean?
If you're talking about the upscale at the start, that's a downscale for taking larger images and shrinking them before they go into the KSampler.
Regarding 3 KSamplers: I already tried the 3 KSampler thing and it wasn't for me.
Interpolate is at the end on mine. Right before it's combined and saved. On my latest version, I explain why it is where it is.
In the end, this workflow is designed for my use and I definitely don't claim that it's better than anyone elses.
@Gladas I’m not an expert, but is there a way for the upscale steps to be done at the end, after the video has already been generated?I’m not making any comparisons, and I’m not saying one is better than the other. Both are very good, both have their flaws, and both have their good things. I was only asking if, in the workflow you made, the resize and image part can be done at the end of the video. I shared the link to the other workflow, I don’t know if you’d like to take a look. It was just a suggestion. I feel like you took it the wrong way. But anyway, thanks for replying.
I took a look at their workflow and yeah I'll pass. Definitely not for me. Their workflow saves 3 versions of the video and for me that's 100% not necessary. I just want the interpolated version.
"Now, in his setup, the upscale and interpolate processes run at the end, once the x-second clip has already been generated. In your workflow, however, the upscale happens while the clip is still being generated, which makes the generation process take much longer."
As stated earlier, my interpolate runs after the KSamplers, so again not sure what workflow you're looking at or using, cuz it's not one of mine. Every version of this workflow has it at the end right before the Combine Video node which is not during the generation process.
There is no upscale on any of these workflows unless you are using it to upscale the Image being converted to video (e.g. taking some small image and bumping it up in resolution). The "upscale" that is happening before the KSamplers is the initial image (for me) being downscaled and unless you're using some crazy big upscale model, it shouldn't be much of a hit. There's even a note explaining the why it is where it is starting from the Wan 2.2 v2 version.
Did you try bypassing the upscale? That's kind of the reason why I have them in those group nodes and have the handy groups bypasser on the left side of the flow.
Plus workflows can easily be modified, so why not modify their's or mine yourself?
Do you want the video to be upscaled at the end? Did I understand that correctly?
@g1263495582 if your asking me: then no I have no interest in upscaling. The guy who started this thread sounded like he does. IDK.
@Gladas Sorry, I forgot to @ him.
You are the best, thank you for the work you have done.
Thanks!