Weird, unannounced and breaking problem: all images generated by ChatGPT are now downloaded in webp format, instead of PNG that was standard even a few days ago. And that’s not only for newly generated images, but also for the ones that we originally downloaded as PNG.
I wonder if OpenAI has no slightest idea that such change may actually break people’s routine, add additional steps (like converting of unpopular webp, because many editors don’t work with it), reduce the value of the service and just annoy users? Isn’t that enough to at least announce the change beforehand?
Have been using Dalle since beta and never really known them to ask for true input from users. However, as there are many capable people here, I asked Chat if a gpt can be created to convert easily from webp to png and this is what is shared.
Anyone that is good with some basic coding, be easy and expect, highly used!
Yes, you can create a ChatGPT-like model or system that converts WebP files to PNG. This system would involve two main components: a chat interface for interacting with users and a backend service that handles the conversion of WebP files to PNG format. Here’s a broad overview of how you could approach this:
Chat Interface: This would be similar to ChatGPT, where users can send commands or requests through a chat interface. The interface would accept file uploads or URLs of WebP images that users wish to convert to PNG.
Conversion Backend: When a WebP file is received, the backend service processes the file, converting it from WebP to PNG format. This could be done using existing libraries or tools that support image format conversion, such as ImageMagick, Python’s PIL (Pillow) library, or similar.
Integration: The chat interface communicates with the conversion backend, sending it the WebP files and receiving back PNG files. The system would then provide a download link or directly send back the converted PNG file to the user through the chat interface.
User Feedback Loop: Optionally, you could implement a feedback loop where users can rate the conversion quality or report issues, allowing you to improve the service over time.
Creating such a system requires expertise in natural language processing (for the chat interface), web development (for the service that handles requests and serves files), and image processing (for the actual conversion of image formats).
If you’re looking into creating a simple script without the chat interface, you could focus solely on the conversion part using a scripting language like Python. Here’s a very basic concept of how the conversion could be handled in Python using the Pillow library:
from PIL import Image
def convert_webp_to_png(input_webp_path, output_png_path):
with Image.open(input_webp_path) as img:
This script doesn’t involve ChatGPT-like interactions but demonstrates the core functionality you’d incorporate into your system. For a ChatGPT-like interface, you would need to build or use an existing chat platform and integrate your conversion logic into it, handling file uploads and downloads accordingly.