Adds support for the following phrases in the prompt:
the camera tilts up
the camera tilts down
the camera pans to the left
the camera pans to the right
the camera zooms in on <object>
the camera zooms outCalifornia AB 2013 Training Data Disclosure
This LoRA was fine-tuned using visual data consisting of synthetic clips. The training data may include copyrighted material owned by third parties. No training data was licensed or purchased. This LoRA is provided for non-commercial use only under the terms of its distribution.
The dataset consists of 188 clips. This is the total used for training after low quality sets were culled. Dataset was created in 2026.
Image data was processed through standard resizing, cropping, normalization, and labeling steps. Synthetic images were included as part of the training dataset.
This model is intended for non-commercial, experimental, and educational use. Generated outputs may reflect copyrighted visual styles or themes associated with the underlying training data. Users are responsible for ensuring compliance with applicable copyright law, other intellectual property laws, and all other applicable laws.
Description
FAQ
Comments (4)
Thank you, Lora is following the instructions well. However, when I set Lora's weight to 1, it distorts the sound. I set it to 0.5 and there's no problem.
Thank you for letting me know! That is very strange, since I didn't do any audio training on this model. I may need to investigate further; if I can identify the cause, then I will try to make a V2 without this issue!
you might use kijai lora loaders and set the strength of audio and audio-to-video keys to 0 . and video and video-to-audio keys to 1.
Hi, first of all, thanks to ReltivlyObjectv for your work!!! Tested on WANGP / LTX2.3 DISTILLED V1.1 i2v SEED0 720P, I noticed exactly the same thing: distorted audio, less understanding of the prompt (e.g., the sea retreats and the shadow of the waves becomes dark rock?), and also incorrect gait during rapid movements (running on the beach). At (0.4), the audio distortion is less problematic and the steps are correct. I noticed that the audio distortion occurs in the second phase. Setting (1.0;0) doesn't distort the audio, but there are still slight issues with walking. I need to test further.
With the old LORA LTx2.0 (also on WANGP / LTX2.3 DISTILLED V1.1 i2v), set to (1;1), some cameras don't work (zoom out), but this didn't cause any audio problems (in fact, the voice was better!) or movement problems (which actually helped!), and it made the subject sound much more ALIVE! Link: https://civitai.red/models/2618345/camera-controls-ltx-2
To creator ReltivlyObjectv: You may not have noticed, but I've tested dozens of different types of LORAs. Your old LORA "camera-controls-ltx-2" (the new one is also available, but it's less noticeable) is the one that makes my creations more "alive," creating micro-movements of the body and face, especially when they're "doing nothing," preventing the subjects from freezing and/or going into standby mode when they finish their shot! Try a 9-16 seed 0 i2v "woman smiling at us" closeup. Without the LORA, both she and the background remain "frozen" or in unsolicited "ultra slow motion." With your LORA, she "sways realistically" in anticipation and looks around, without being told, and the sea behind her moves correctly! Have you ever thought about creating a lora focused on micro-movements of the body, mimicry, breathing and increasing the realism of acting in general?