In the ongoing battle for image format supremacy, a new web tool called OnlyJPG has emerged promising to convert various image formats into optimized JPEG files using Google's advanced jpeglib with perceptual XYB modeling. While the tool aims to solve the universal compatibility problem that still plagues newer formats, early user testing has revealed significant conversion issues with WebP and AVIF files that produce unexpectedly poor results.
Technical Hiccups in Modern Format Conversion
Users quickly discovered that while OnlyJPG works excellently with existing JPEG files—often achieving substantial size reductions with minimal quality loss—the tool struggles significantly when converting from WebP and AVIF formats. One tester reported that WebP images processed through the service emerged incredibly pixelated and sometimes even larger than the original files, despite using the tool's default quality setting of 90. The same user found that AVIF files suffered from similar conversion problems, suggesting a systematic issue with how the XYB color model handles these modern formats.
If I run WebP images from reddit through your site they all end up looking terrible. If I run them through Squoosh.app I get the expected minimal degradation in quality and a 30%ish reduction in size.
The workaround discovered by users involves switching to legacy mode using YCbCr color encoding instead of XYB, which successfully processes WebP files without the quality degradation. This suggests the issue lies specifically with the interaction between XYB compression and the color profiles of WebP/AVIF source images, though the exact technical cause remains unclear.
OnlyJPG Conversion Issues:
- WebP to JPEG with XYB: Severe quality degradation, sometimes larger file size
- AVIF to JPEG with XYB: Similar quality problems as WebP
- WebP to JPEG with YCbCr (legacy mode): Works correctly
- Existing JPEG to JPEG with XYB: Works well with significant size reduction
The Right Tool for the Right Job
Beyond the conversion issues, the discussion highlighted important considerations about when JPEG compression is actually appropriate. Several commenters pointed out that JPEG's lossy compression, while excellent for photographs, performs poorly with certain types of images. Screenshots, digital art, and images containing text or sharp edges often suffer from visible artifacts when converted to JPEG, making PNG a better choice for these use cases.
The community also noted alternative tools like Squoosh.app, which offers similar conversion capabilities while handling WebP files more reliably. Some users expressed frustration with the tool's documentation, particularly around handling transparency—a feature JPEG doesn't natively support, requiring users to understand when to stick with PNG format instead.
Image Format Comparison for Different Use Cases:
- JPEG: Best for photographs, universal compatibility
- PNG: Ideal for screenshots, digital art, images with text
- WebP: Good modern alternative with better compression than JPEG
- AVIF: Advanced compression but limited software support
- GIF: Suitable for memes and simple animations only
Privacy and Practical Concerns
Privacy-conscious users raised questions about web-based image conversion tools generally, wondering how to verify that images aren't being uploaded to external servers. The suggestion to test conversion while offline highlights the trust issues surrounding web applications that process user files. For those preferring local processing, traditional tools like ImageMagick remain popular alternatives, though they require more technical expertise than the simple web interface offered by OnlyJPG.
Some users also criticized the tool's own web page for being unnecessarily heavy, noting that a 7.5 MB download seems counterproductive for a service focused on reducing file sizes and digital waste. This irony wasn't lost on the community discussing the tool's environmental claims.
As of UTC+0 2025-10-17T19:24:16Z, the conversion issues with WebP and AVIF files remain unresolved, presenting both a technical challenge for the developers and a practical limitation for users hoping to standardize their diverse image collections into universally compatible JPEG format. The situation demonstrates that even with advanced compression algorithms, format conversion remains a complex problem with unexpected edge cases that only emerge through real-world testing.
Reference: Only JPG