<div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/7msepxHGqq8?rel=1" start="0"></iframe></div>This workflow is designed for LTX 2.3 IC-LoRA video colorization and cinematic recoloring. Its main purpose is to take an existing video or frame-guided video input, preserve the original motion structure, and use LTX 2.3 with IC-LoRA Colorizer to rebuild the clip with more natural color, stronger lighting consistency, and a more polished film-style visual result.Unlike a simple color filter, LUT, or saturation adjustment workflow, this setup uses a generative video pipeline. The workflow does not merely apply color correction on top of the original frames. It uses LTX 2.3 video generation, image-to-video conditioning, IC-LoRA guide injection, text-guided color direction, audio-video latent handling, tiled decoding, and final video reconstruction. This makes it more suitable for AI video remastering, black-and-white video colorization, old-video restoration, stylized recoloring, and cinematic color enhancement.The workflow uses ltx-2.3-22b-dev as the main model route, together with ltx-2.3-22b-distilled-lora-384 and LTX-2.3-22b-IC-LoRA-Colorizer-0.9. The core pipeline includes LTXVPreprocess, EmptyLTXVLatentVideo, LTXVImgToVideoConditionOnly, LTXAddVideoICLoRAGuide, LTXVConditioning, LTXVConcatAVLatent, LTXVSeparateAVLatent, LTXVCropGuides, SamplerCustomAdvanced, VAEDecodeTiled, LTXVAudioVAEDecode, CreateVideo, and SaveVideo. This gives the workflow a complete video-to-video style colorization structure rather than a single-frame image edit.The key node is LTXAddVideoICLoRAGuide. It injects the IC-LoRA colorization guide into the video latent process, helping the model follow the source structure while adding new color information. This is useful when the creator wants to keep the original framing, pose, action, and temporal rhythm, but improve the visual palette. The result can be used for natural skin tones, warm sunlight, vintage film color, cinematic contrast, clothing color recovery, environmental recoloring, and more stable color across frames.The workflow also includes a prompt-controlled color direction. The example prompt asks for realistic film colorization, natural skin tone, warm light, a blue shirt, brown leather shoes, a dark gray suit, green trees, golden sunlight, a vintage cinematic palette, and stable colors across frames. This kind of prompt is important because colorization is not only about “adding color.” It needs clear material logic: skin, clothes, shoes, trees, sunlight, shadows, and the overall color grade should all remain coherent.The graph also keeps the audio-video structure through LTXVEmptyLatentAudio, LTXVConcatAVLatent, LTXVSeparateAVLatent, and LTXVAudioVAEDecode. This allows the final output to remain a usable video instead of becoming a silent frame sequence. Tiled VAE decoding is used to make decoding more stable for larger video frames and reduce memory pressure during output.This workflow is ideal for creators who want to test LTX 2.3 IC-LoRA colorization, grayscale video recoloring, old-video enhancement, cinematic remastering, AI short-video restoration, character video color repair, RunningHub demos, and Civitai workflow previews. If you want to see how LTX 2.3, IC-LoRA Colorizer, prompt-guided color control, audio-video latent routing, and tiled decoding work together, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: <a target="_blank" rel="ugc" href="https://www.runninghub.ai/post/2035010437032448002?inviteCode=rh-v1111">https://www.runninghub.ai/post/2035010437032448002?inviteCode=rh-v1111</a>Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China &amp; Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: <a target="_blank" rel="ugc" href="https://www.bilibili.com/video/BV11sAGzvE1M/">https://www.bilibili.com/video/BV11sAGzvE1M/</a>I will continue updating model resources on Quark Drive:👉 <a target="_blank" rel="ugc" href="https://pan.quark.cn/s/20c6f6f8d87b">https://pan.quark.cn/s/20c6f6f8d87b</a>These resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： <a target="_blank" rel="ugc" href="https://www.runninghub.ai/post/2035010437032448002?inviteCode=rh-v1111">https://www.runninghub.ai/post/2035010437032448002?inviteCode=rh-v1111</a>打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： <a target="_blank" rel="ugc" href="https://www.bilibili.com/video/BV11sAGzvE1M/">https://www.bilibili.com/video/BV11sAGzvE1M/</a>我会在 夸克网盘 持续更新模型资源：👉 <a target="_blank" rel="ugc" href="https://pan.quark.cn/s/20c6f6f8d87b">https://pan.quark.cn/s/20c6f6f8d87b</a>这些资源主要面向本地用户，方便进行创作与学习。

Files with SHA256: 11aa4420878031e10ec1ab11c7ca830b9d640c76f5000bb806926387f55e345a