diff --git a/comfy/ldm/lightricks/vae/causal_video_autoencoder.py b/comfy/ldm/lightricks/vae/causal_video_autoencoder.py
index 1e61ec755..1a15cafd0 100644
--- a/comfy/ldm/lightricks/vae/causal_video_autoencoder.py
+++ b/comfy/ldm/lightricks/vae/causal_video_autoencoder.py
@@ -293,8 +293,14 @@ class Encoder(nn.Module):
         outputs = []
         samples = [sample[:, :, :1, :, :]]
         if sample.shape[2] > 1:
-            n = max(1, max_chunk_size // (2 * frame_size))
-            samples += list(torch.split(sample[:, :, 1:, :, :], 2 * n, dim=2))
+            chunk_t = max(2, max_chunk_size // frame_size)
+            if chunk_t < 4:
+                chunk_t = 2
+            elif chunk_t < 8:
+                chunk_t = 4
+            else:
+                chunk_t = (chunk_t // 8) * 8
+            samples += list(torch.split(sample[:, :, 1:, :, :], chunk_t, dim=2))
         for chunk_idx, chunk in enumerate(samples):
             if chunk_idx == len(samples) - 1:
                 mark_conv3d_ended(self)