> ## Documentation Index
> Fetch the complete documentation index at: https://lemonslice.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Avatar Image Tips

## How to get the best quality avatars?

The reference image is the foundation of how your LemonSlice avatar looks and talks. The LemonSlice model picks up on every detail of the image and uses it to infer micro-expressions, poses, and even background motion. Follow the tips on this page for best avatar quality.

<Info>
  If your image is bad, your avatar will be bad. The LemonSlice web app allows you to iterate on your images rapidly using our built-in AI image editor.
</Info>

**Don't zoom out too much.** Our model renders a fixed number of pixels. Use as many of those pixels for the character, not the background.

<Frame>
  <img src="https://mintcdn.com/lemonslice/3Q4QxR_2hl9HcIP7/images/zoom_level.png?fit=max&auto=format&n=3Q4QxR_2hl9HcIP7&q=85&s=ec087845cd1792650850c51b15c8d36d" alt="Zoom Level" width="1600" height="723" data-path="images/zoom_level.png" />
</Frame>

**Larger faces have best facial details.** Make the head as large as possible. If you want your avatar to use hand gestures, use a half-body portrait while maintaining a relatively large face.

<Frame>
  <img src="https://mintcdn.com/lemonslice/3Q4QxR_2hl9HcIP7/images/large_face.png?fit=max&auto=format&n=3Q4QxR_2hl9HcIP7&q=85&s=42144947d912b76ef745f6fe039baa1c" alt="Large Face" width="1600" height="624" data-path="images/large_face.png" />
</Frame>

**Use a neutral expression and look directly into the camera.** Your reference image controls the expression and pose of your avatar throughout the call. For example, a reference image of a person looking up may cause the avatar to always be looking up.

<Frame>
  <img src="https://mintcdn.com/lemonslice/3Q4QxR_2hl9HcIP7/images/ExpressionandPose.png?fit=max&auto=format&n=3Q4QxR_2hl9HcIP7&q=85&s=d5f23befa2aec01e79b673b9ff9778de" alt="Expressionand Pose" width="1600" height="723" data-path="images/ExpressionandPose.png" />
</Frame>

**For non-human or cartoon characters, use an image with the mouth slightly open.** This will show the model how you want the mouth to animate.

<Frame>
  <img src="https://mintcdn.com/lemonslice/3Q4QxR_2hl9HcIP7/images/Mouth.png?fit=max&auto=format&n=3Q4QxR_2hl9HcIP7&q=85&s=7ed61b19953f50226fdf6223905c3533" alt="Mouth" width="1600" height="684" data-path="images/Mouth.png" />
</Frame>

**For humans:** Use this approach if you're noticing degradation on the quality of the teeth.

<Info>
  We've found that running your image through this image edit prompt does this the best:

  *The exact same image, but they are talking (with their mouth open).*
</Info>

<Frame>
  <img src="https://mintcdn.com/lemonslice/TXIgeVGZHhykbxU4/images/HumanMouth.png?fit=max&auto=format&n=TXIgeVGZHhykbxU4&q=85&s=ed0a89dbc724083e835c05451fb37e82" alt="Human Mouth" width="1600" height="684" data-path="images/HumanMouth.png" />
</Frame>

**Drive simple animations from a single image.** Our model intelligently animates simple scenes using the reference image, the agent prompt, and the idle prompt. Keep your animations scoped to relatively simple, looping animations like waves in water and atmospheric effects. The model can also animate people in the background, but it works best when they're out of focus.

<Columns cols={2}>
  <Column>
    <video src="https://mintcdn.com/lemonslice/In5NRFQHKTq5DO3B/videos/splashingwaterinwaterfountain.webm?fit=max&auto=format&n=In5NRFQHKTq5DO3B&q=85&s=b0ab72a637263a583773eff4b8a9e4bb" controls={true} autoPlay={true} muted={true} loop={true} playsInline={true} data-path="videos/splashingwaterinwaterfountain.webm" />
  </Column>

  <Column>
    <video src="https://mintcdn.com/lemonslice/In5NRFQHKTq5DO3B/videos/simple_people.webm?fit=max&auto=format&n=In5NRFQHKTq5DO3B&q=85&s=2e95cb0b76f795f380b6f76772b68fcc" controls={true} autoPlay={true} muted={true} loop={true} playsInline={true} data-path="videos/simple_people.webm" />
  </Column>
</Columns>

**When holding objects, use one hand in a natural pose.** This keeps the avatar stable and expressive, maintains clean motion, and avoids awkward or unnatural positioning that can make the interaction feel off.

<Frame>
  <img src="https://mintcdn.com/lemonslice/ivegaDvmTC3bBwQD/images/Holding.png?fit=max&auto=format&n=ivegaDvmTC3bBwQD&q=85&s=c5c21d3fa079421e1928da62a4e4c9be" alt="Holding" width="1600" height="684" data-path="images/Holding.png" />
</Frame>

## Choosing a voice

**Voice intonation impacts facial expression.** For example, choosing a voice with an angry intonation will yield an angry expression. Likewise, a monotonous voice will produce a neutral facial expression.

**If your avatar's mouth looks off, choose another voice.** Some voices work better than others.
