Holographic Matrix of all things_SDXL(Nature section)_Matrix Hologram

Also not too important instructions:

Based on feedback from many fellow Taoists, there is a desire to enhance the "dot matrix" holographic effect, which would be more suitable for presenting objects with softer or more intricate contours. Therefore, this time, not only did we add support for more types, but also added the dot matrix artistic effect.

Compared to 3D holographic art SDXL (vehicle episode), the main updates for this LoRA iteration are as follows:

1. ) Added scenes of characters (ancient and modern, from home and abroad), individual buildings, building clusters, while also considering LoRA's compatibility and practicality, it also supports types such as animals/plants and marine creatures. Take a look at the example images or the return image area!

2. ) You can try starting with "Dot Matrix Hologram" to evoke it. Correspondingly, for the wireframe hologram effect, you can write "Wireframe Hologram" or not write anything, letting the model express itself.

3. ) This time, I also listened to suggestions from a few AI players on Sweet Potato🍠: they felt that a purely holographic effect was somewhat monotonous and could lead to aesthetic fatigue. After my own testing, the data also provided intuitive feedback that suggested a combination of virtual and real, creating contrast for more artistic and playable effects.

4. ) Therefore, in this training set, I added a few images combining real scenes and holographic images. If you want more holographic semi-transparency, you can try writing "glowing 3D objects."

5. ) 🙏 Thank you all for your valuable suggestions. I am dedicated to taking advice, so I welcome more comments and discussions. The mixed reality version of LoRA is already on its way and will be released this month. The visual effects will be further enhanced (in my personal opinion).

————————————
Furthermore, the following optimizations have been made this time:

1. ) The background is relatively cleaner now, without any more unnecessary, meaningless lines and points;

2. ) To enrich the image, when drawing aerial or overhead perspectives, there will occasionally be more pronounced depth of field, similar to a miniature landscape effect; there are random particle effects;

3. ) All watermarks have been thoroughly cleared. If they appear again, it definitely won't be LoRA's fault.

Recommended Parameters:

Base model recommendation: sd_xl_base_1.0_0.9vae (no guarantee for other base film effects)

Sampling method: DPM++2M Karras (up to you)

Iteration steps: 30~40

CFG: 7~8

Weight: Performance is good between 0.6 and 0.8

Size: 768*1024 (most images in the training set are 3:4 vertical compositions)