From Below / From Above (ILXL)
This slider shifts the camera perspective toward either a low-angle “from below” look or a high-angle “from above” look, giving you stronger control over how characters are framed. It’s built for ILXL and tuned to keep faces, bodies, and proportions stable while still pushing the angle noticeably.
The slider blends smoothly both ways, making it ideal for dynamic posing, stylised realism, or scenes that need stronger directional emphasis.
To intensify the effect:
"from below"or"from above"Both will push the angle harder and exaggerate the perspective shift.
Model Support:
Fully compatible with ILXL, PONY, NAI, SDXL
Description
FAQ
Comments (22)
Great work! should come in handy.
@Benpharten thanks ❤️ please leave any feedback and suggestions once you play around. This helps me fix the issues for future version.
My man! Appreciate all the awesome stuff you've been putting out.
❤️ @mrwizard9001 thanks. More to come
Now we're talking. The kind of sliders that were surprisingly missing. Thanks!
❤️
Thanks for making this. Are you thinking about making a Left-Right angle slider?
@Rosettasees you can't really make left or right, chdckpoint does not know this. Although saying that, I can make slider to do from side - this will give you side profile and everything in between. Can also do from front and back.
@klaabu That would be very helpful for a little more camera control (this one for above, below is pretty useful already).
@klaabu That would be brilliant, thank you!
@einar_rainhart @Rosettasees no worries. I'll do from front - from behind and I'll see what I can do about the side control one.
@klaabu Looking forward to it!
Definitely useful. The slider alone effect is sometimes muted or enhanced based on certain LoRAs you are using, but this is to be expected. Thanks for taking the time to create this!
I had high hopes for this Lora, but unfortunately it's a disappointment. Yes, there are images "from below" for negative strengths and "from above" for positive ones. But this Lora doesn't provide a smooth transition between the two perspectives. Different strengths only offer variations from the same perspective; only a change in sign alters the perspective.
@Samtiger_ger thanks for your feedback.
This LoRA does what it says it does. Not sure if you expected something else from it.
@klaabu
What advantage does your Lora offer if I can just as easily add 'from above' or 'from below' to the prompt?
If, on the other hand, I could dynamically adjust the viewing angle using a slider, that would be truly useful and a huge added value.
@Samtiger_ger it seems like it offers nothing to you. To others, it offers exactly what it does.
Take care (:
Incredible work! It can be used to easily shift the camera angle and perspective, which seems like it may not be useful in many situations, but it can be used to keep consistent angles and camera height/position. Keep it up you've been putting great stuff out lately!
Thank you for your support @dashfitness . Im about to drop zoom in/out slider, this will either elevate the perspective:)
It seems to vary a lot in terms of effectiveness depending of the picture that you re trying to generate. That being said this is still a must have if you don't want to rely on forcing a high score on tags such as "from below".
Honestly I've found it's better not to increase the weight of individual tokens anymore, but to just repeat them. "from below, from below" almost universally provides better results than (from below:1.2). At least with the way comfyui handles weighting. I think A111/forge style does weighting differently so I can't speak for that.
Thank you for the feedback :)