If you've ever tried auto-rigging and got weird shoulders, broken elbows, or twisted wrists, the starting pose is often the reason. The two most common bind poses are T-pose and A-pose. Both can work, but they behave differently for automatic skeleton fitting and skinning.
What is a T-pose?
A T-pose is the classic: arms straight out horizontally, forming a "T" shape. This pose makes limb lengths and joint directions very clear to algorithms because arms are fully extended and symmetrical.
What is an A-pose?
An A-pose keeps arms angled down (usually 30-45 degrees). Many manual rigs prefer A-pose because it can produce nicer shoulder deformation for certain characters.
Why many auto-riggers prefer T-pose
- Clear joint positions: shoulders, elbows, wrists are easier to detect
- Symmetry: both arms align cleanly for left/right matching
- Less ambiguity: fewer "almost" positions that confuse fitting
When A-pose can be better
- high shoulder realism is more important than speed
- your character has heavy clothing or armor around shoulders
- you plan to manually polish weights anyway
The practical rule
If the tool expects T-pose, use T-pose. If you feed an A-pose into a system tuned for T-pose, you often get incorrect shoulder orientation, weird arm twist, or uneven skinning.
Quick checklist for a "good T-pose"
- Arms at ~90° from torso (horizontal)
- Elbows straight (not hyper-bent)
- Hands neutral (not extreme claw)
- Feet flat, legs straight
- Model centered and scaled correctly
If you're targeting fast auto-rigging and predictable exports, start with T-pose and keep geometry clean.