ComfyUI_ASCIIArtNode Introduction
ComfyUI_ASCIIArtNode is a collection of ComfyUI nodes that turns images and video frames into artwork made from characters. You can create a single ASCII art image, reveal an image as though it is being typed, or transform a video into moving, color-changing character art.
The extension gives you several ways to control the result. You can use one color for a classic text-based look, borrow colors from the source image, add a blinking cursor, or animate a second pass that colors in the characters. The nodes produce images or batches of image frames, so you can connect their output to the rest of your ComfyUI workflow.
How ComfyUI_ASCIIArtNode Works
Imagine dividing an image into a grid and choosing a character for each part of that grid. A visually light character can represent one brightness level, while a denser character can represent another. Together, the characters form a recognizable version of the original image.
The ascii_charset setting supplies the characters available for this effect. Their order matters: it normally runs from sparse or light-looking characters to dense or dark-looking ones. The invert_brightness_mapping option reverses which end of that sequence is used for lighter and darker areas.
The nodes then render those characters into images using your chosen font and background. Depending on the node, the characters can all share one text_color or take colors from corresponding areas of the source image. Animation nodes create a batch of frames that reveals characters over time. Video nodes instead work from an input batch of video frames: one keeps the character choices fixed while updating their colors, and the other updates both characters and colors on every frame.
ComfyUI_ASCIIArtNode Features
Controls shared by most nodes
char_widthsets the artwork’s width in characters; its height is generally calculated to preserve the image’s proportions. A larger width can show more detail, but animations may take longer to generate. The author suggests starting around 80–120 and increasing it as needed.font_pathpoints to a.ttfor.otffont file. A monospaced font, such as Courier New, Consolas, or DejaVu Sans Mono, gives each character consistent spacing and usually produces the clearest grid.font_sizecontrols how large the rendered characters appear.ascii_charsetcontrols the visual texture. Changing the characters or their order can make the same image look noticeably different.background_colorsets the output background using a hex color such as#000000for black.invert_brightness_mappingreverses the usual relationship between image brightness and character density.
Static image
ASCII Art Generator (Static) converts one input image into one ascii_image. It renders the result at the input image’s dimensions, scaling the ASCII artwork to fit. Set text_color to give every character the same color, such as #FFFFFF for white. Choose this node when you want a finished still image without an animation.
Typing animations from one image
These nodes reveal the ASCII version of a single image over a sequence of frames:
| Node | What you see | Main controls |
|---|---|---|
| ASCII Typing Animation Generator | Characters appear in one uniform text_color. | chars_per_frame |
| Color ASCII Typing Animation | Each new character takes a color from the corresponding part of the source image. | chars_per_frame |
| Realistic Color ASCII Typing Animation | Image-colored characters appear with a flashing cursor at the current typing position. | chars_per_frame, cursor_char, cursor_blink_frames, cursor_color |
| Two-Pass Color Typing (Concurrent) | New characters appear in initial_text_color while earlier characters transition to their image-sampled colors. A flashing cursor is included. | initial_text_color, chars_per_frame, cursor controls |
| Sequential Two-Pass Color Typing | The whole image is typed in initial_text_color, pauses, then receives its image-sampled colors in a second pass. Both passes include a cursor effect. | initial_text_color, chars_per_frame, cursor controls, pause_frames_between_passes |
chars_per_frame determines how many new characters appear in each frame. A higher value reveals the artwork in fewer frames; a lower value gives the typing effect more time to unfold. For cursor-enabled nodes, cursor_char chooses a symbol such as _ or |, cursor_color sets its color, and cursor_blink_frames controls how many frames each visible or hidden blink state lasts. In the sequential node, pause_frames_between_passes sets the length of the hold before coloring begins.
Each typing node outputs an IMAGE batch, not a ready-to-play video or GIF. Connect that batch to a compatible animation or video-saving node, such as a Video Combine node from VideoHelperSuite or a Save Animated WebP/GIF node, when you want a playable file.
Video conversions
Both video nodes accept video_frames as an IMAGE batch and return an ASCII-art frame batch:
- Video to Color Static ASCII Art: Chooses the characters from the first input frame and keeps them in place. Their colors change with subsequent frames. Use it when you want the text pattern to stay steady while color carries the motion.
- Video to Dynamic Color ASCII Art: Recalculates both the characters and their colors for every frame. Use it when you want changes in the video’s shapes and brightness to appear in the character pattern itself.
Troubleshooting ComfyUI_ASCIIArtNode
The ASCII image looks uneven or hard to recognize.
- Try a monospaced font so characters line up consistently.
- Adjust
char_width. Increasing it can reveal more detail. - Try a different
ascii_charset, paying attention to the order from sparse to dense characters. - Toggle
invert_brightness_mappingif light and dark areas seem to have the opposite emphasis from what you intended.
An animation takes too long to generate.
- Lower
char_widthto reduce the number of characters. - For a typing animation, increase
chars_per_frameso the reveal uses fewer frames. - Test with a shorter input frame batch when working with video.
The typing node runs, but I do not have a playable animation.
The node outputs individual images as an IMAGE batch. Connect that output to a compatible ComfyUI node that saves image batches as a video, GIF, or animated WebP.
The font does not render or produces an error.
Check that font_path points to an existing .ttf or .otf file. If a font name alone does not work, provide its full file path.
The node does not appear in ComfyUI.
Restart ComfyUI, then check its startup console for an error associated with the custom nodes. If the message mentions PIL or Pillow, the dependency may be missing from the Python environment used by ComfyUI.
Why do the characters stay the same throughout my video?
That is the intended behavior of Video to Color Static ASCII Art: it selects characters from the first frame and changes only their colors afterward. Choose Video to Dynamic Color ASCII Art if you want the characters to change with each frame.
What is the difference between the two two-pass typing effects?
Concurrent typing begins changing earlier characters to their final colors while new ones are still appearing. Sequential typing finishes the initial-color reveal first, optionally pauses, and only then begins coloring the artwork.
