Writing prompts like "view from above" or "45-degree angle" is a pain. What if you could just drag a 3D sphere to control your AI camera?
In this video, I introduce the new "Multiple Angles" LoRA for Qwen Edit 2511, combined with Terry Jia's amazing visual control node. I'll show you two powerful workflows: one for precise Single Image Angle Editing and another for generating 5-View Character Sheets in one click.
- - - - - - - - - - - - - - - - - - - - - - - - - -
🔽 GET THE FREE WORKFLOWS & 1000 BONUS COINS 🔽
✨ Register on RunningHub through my link to get 1000 FREE RH Coins!
💰 Plus, claim 100 extra coins every single day just by logging in.
1️⃣Workflow: Qwen2511 Multi-angle editing:
https://www.runninghub.ai/post/2010055896453357569/?inviteCode=rh-v1291
2️⃣Workflow: Qwen2511 Multi-angle editing(Batch Version):
https://www.runninghub.ai/post/2009903047484383233/?inviteCode=rh-v1291
🔽Free download models/Loras 🔽
Model links:
1️⃣Download link for Qwen Edit 2511 (huggingface):
https://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI/tree/main/split_files/diffusion_models
2️⃣Download link for Qwen Edit 2511 (Online storage):
https://pan.quark.cn/s/56442da3e9c5
plug-in:
1️⃣Download link for ComfyUI-qwenmultiangle (GitHub):
https://github.com/jtydhr88/ComfyUI-qwenmultiangle.git
LoRA:
1️⃣Qwen-Image-Edit-2511-Lightning(huggingface):
https://huggingface.co/lightx2v/Qwen-Image-Edit-2511-Lightning/tree/main
2️⃣Qwen-Image-Edit-2511-Multiple-Angles-LoRA (huggingface):
https://huggingface.co/fal/Qwen-Image-Edit-2511-Multiple-Angles-LoRA/tree/main
3️⃣Qwen-Image-Edit-2511-Multiple-Angles-LoRA (Online storage):
https://pan.quark.cn/s/17218f5bd3de
4️⃣Download link for Qwen Edit 2511 FP32 (Online storage):
https://pan.quark.cn/s/56442da3e9c5
- - - - - - - - - - - - - - - - - - - - - - - - - -
💡 Why is this LoRA special?
It was trained on over 3000 sets of 3D Gaussian Splatting data, giving Qwen a near-native understanding of 96 different camera positions. Combined with the "Qwen Multiangle" node, you get a visual UI to control Azimuth, Elevation, and Distance perfectly.
If this tool helps your character design workflow, please SMASH that LIKE button and SUBSCRIBE for more advanced ComfyUI tutorials!
#Qwen2511 #ComfyUI #CameraControl #AICharacterDesign #StableDiffusion #RunningHub #LoRA #Workflow #GenerativeAI
Description
FAQ
Comments (13)
C站的工作流有问题啊 R上的就能用 你看我的图
不会吧.我刚刚重新下载回来打开,没发现问题啊.没看到你的图片,你的报错是什么
I have 2 questions:
- Where can I get the skin texture lora (KOOK) if I cannot get Chinese apps?
- The sliders in he 3D view interface do not seem to interact with the 3D preview. I am not sure if the image will take instructions from the sliders above it or the 3D graphic below.
1.https://www.modelscope.cn/models/KookYan/KOOK_Qwen_2512_zshx_V2 Available for download here
2. The slider and 3D view are typically interactive; you may observe in my video demonstration that both can control the final angle. Additionally, should you be uncertain whether it has succeeded, you may examine the 'Show Any' node, where you can view the node's output content.
Hey, all my images are full black. do you have a solution ?
Does the issue of pure black images only occur when using Qwen series models?
You can message me privately, and I'll reply as soon as I see it.
The solution for those who might encounter this issue: disable Sage Attention at Comfy startup, or start your Run.bat file with very few parameters:
@echo off
where python
python main.py
pause
@Celebigoat This is an old issue from when the Qwen model was first released.
Encountered the same problem many months ago, didn't remember what caused and what's the solution. Maybe it was the lightning lora version.
@rocren3 If you encounter this issue, you need to change Cross attention from Sage attention to Xformers.
兩個問題請教
1.生成會變成4張預設角度，只有1張是自訂
2.要怎麼讓人物穿原本的服裝就好，不要變成韓式棒球服
1. 如果你只需要单次输出1个角度,那就选择同页面的另一个工作流,如果你是要4个角度,那你就必须每一个角度节点都进行调整.
2. 服装是由提示词驱动的,所以你可以在提示词中写上"保持服装细节不变"等内容