Comfy-Org/ComfyUI · error · ValueError

Unsupported process_res_method: {method}

Error message

Unsupported process_res_method: {method}

What it means

compute_target_size() chooses a resolution-scaling rule by method string: 'upper_bound_resize' (scale longest side to process_res) or 'lower_bound_resize' (scale shortest side). Any other string raises ValueError. Callers normally pass the default, so the error appears only when a custom method string is forwarded.

Source

Thrown at comfy/ldm/depth_anything_3/preprocess.py:36

def _round_to_patch(x: int, patch: int = PATCH_SIZE) -> int:
    down = (x // patch) * patch
    up = down + patch
    return up if abs(up - x) <= abs(x - down) else down


def compute_target_size(orig_h: int, orig_w: int, process_res: int, method: str = "upper_bound_resize") -> Tuple[int, int]:
    """Compute (target_h, target_w) for a single image.
    upper_bound_resize: scale longest side to process_res, then round each dim to nearest multiple of 14 (default upstream method).
    lower_bound_resize: scale shortest side to process_res, then round."""

    if method == "upper_bound_resize":
        longest = max(orig_h, orig_w)
        scale = process_res / float(longest)
    elif method == "lower_bound_resize":
        shortest = min(orig_h, orig_w)
        scale = process_res / float(shortest)
    else:
        raise ValueError(f"Unsupported process_res_method: {method}")

    new_w = max(1, _round_to_patch(int(round(orig_w * scale))))
    new_h = max(1, _round_to_patch(int(round(orig_h * scale))))
    return new_h, new_w


def preprocess_image(image: torch.Tensor, process_res: int = 504, method: str = "upper_bound_resize") -> torch.Tensor:
    assert image.ndim == 4 and image.shape[-1] == 3, f"expected (B,H,W,3) IMAGE; got {tuple(image.shape)}"
    B, H, W, _ = image.shape
    target_h, target_w = compute_target_size(H, W, process_res, method)

    # (B, H, W, 3) -> (B, 3, H, W)
    x = image.movedim(-1, 1).contiguous()
    if (target_h, target_w) != (H, W):
        # Upstream uses cv2 INTER_CUBIC (upscale) / INTER_AREA (downscale).
        # Lanczos in ``common_upscale`` is anti-aliased and produces the
        # closest pixel-wise match in a sweep across {bilinear, bicubic,
        # area, lanczos, bislerp}. Used in both directions for simplicity.

View on GitHub (pinned to 1c6d8d45b3)

Solutions

1. Use "upper_bound_resize" (default) or "lower_bound_resize" exactly.
2. If exposing the option in a node, make it a fixed combo of the two supported values rather than free text.
3. Map/normalize any upstream method name to one of the two supported values at the node boundary.

Example fix

# before
h, w = compute_target_size(H, W, 504, method="fit_longest")

# after
h, w = compute_target_size(H, W, 504, method="upper_bound_resize")

assert method in ("upper_bound_resize", "lower_bound_resize"), f"unsupported method {method!r}"

def is_supported_resize_method(method: str) -> bool:
    return method in {"upper_bound_resize", "lower_bound_resize"}

When it happens

Trigger: Calling preprocess_image()/compute_target_size() with method set to a value other than the two supported strings — e.g. a custom node exposing a free-text combo, or code copied from upstream that used a different method name like 'keep_aspect'.

Common situations: Custom nodes wrapping DA3 preprocessing with user-visible method dropdowns; drifting method names between upstream repo and this port.

Related errors

• At least 4 points are required to compute a homography.
• Unknown reference view selection strategy: {strategy!r}. Mus
• Unknown DINOv2 backbone variant: {backbone_name!r}
• Invalid affine shape: {ext.shape}
• Invalid rotation matrix shape {matrix.shape}.

AI-assisted analysis of Comfy-Org/ComfyUI@1c6d8d45b3 (2026-08-14). Data as JSON: /api/errors/763ed6625dbb85f9. Report an issue: GitHub.