I know about some 'advanced' converters, thats can optimize jpeg files piecemeal, strongly squeezing "unimportant" areas and practically without touching real content of image. In addition, there are some rumors that different converters use different optimization methods for the same JPEG quality value. So, they generate various in quality and optimization results. I need the most accurate tool to convert PNGs to JPGs practically without losses (visible to the naked eye).

The task involves choosing to convert only those images that do not contain transparent pixels (in any opacity level except of %100 for each pixel). I think, it's possible to use imagemagick, but it inconvenient to use under windows, so, i'd like to find GUI tool (Yes, it is quite tedious to write scripts under Windows and "mine" the documentation imagemagick too).

2 Answers 2


If you have experience using any Windows programming language, such as C# or VB.NET, you could use an advanced imaging library such as LEADTOOLS to easily perform the conversion in your own program (Disclaimer: I'm an employee of its vendor).

If you'd like to try, there's a free evaluation edition on www.leadtools.com. Even if you're not a hard-core programmer, the toolkit is fairly easy to use with many demo projects that you can build your program based on. There's also free email and chat support service if you need help with any toolkit function.

The code can be as simple as this:

RasterImage image = _codecs.Load("image.png");
   return; //if the image has a transparent color, don't convert it.
_codecs.Options.Jpeg.Save.QualityFactor = 2; //maximum quality with no visual loss
_codecs.Save(image, "output.jpg", RasterImageFormat.Jpeg, 24);

I use a cross platform app called Adapter. You can drop a collection of files on the app and it will convert them all to what you selected. Adapter can be found here.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.