Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

mac-slap-happy

macrumors member
Original poster
Jul 25, 2007
96
1
Hey everyone. I'd like to announce our new app that is currently on the MAS.

littlePNG is a PNG optimizer that can drastically reduce the file sizes of your PNG images. Using either lossless or lossy compression, it is possible to reduce file sizes up to 90% in some cases! Just drag your PNG files into littlePNG, and click one button to optimize and compress the entire batch.

• Optimize your website graphics for faster loading times.
• Compress your PNG graphics for your iPhone or Mac bundles.
• Choose lossless or lossy based compression algorithms.
• No need to learn command line tools!
• Built using the latest quantization tools.

littlePNG is perfect for web developers, or iOS/Mac application developers. Reduce the bandwidth requirements of your website, or reduce the overall size of your app bundle.

Download on the MAS: https://itunes.apple.com/us/app/littlepng/id977861919?ls=1&mt=12
 

Attachments

  • 1.jpg
    1.jpg
    66.2 KB · Views: 234
  • 2.jpg
    2.jpg
    75.5 KB · Views: 182
  • 3.jpg
    3.jpg
    61.3 KB · Views: 194
Last edited:
I've attached two images for comparison. I had to actually convert these to JPG and reduce the pixel size for this forum, but you should be able to see how well the optimization works. This image started out at 1.45 MB. After optimization, the image was down to 319 KB (a 78% reduction in file size). The difference in quality between the two is minuscule even with the dramatic file size reduction.

I'd be happy to answer any questions and welcome all comments. I would also be happy to optimize your PNG file and repost it here if you'd like to see the results on your own images. Just post your PNG file and request optimization.

Thanks!
Brian
 

Attachments

  • mario1.jpg
    mario1.jpg
    229.2 KB · Views: 190
  • mario2.jpg
    mario2.jpg
    244.7 KB · Views: 189
I can see banding in places like the button and shadow under the hat, and the eye becomes a bit brighter, so... it just reduces the number of colours in the PNG?
 
Last edited:
What would this offer me over ImageOptim other than the option to choose lossy compression? Can you still retain meta data, or offer any automator hooks?
 
I can see banding in places like the button and shadow under the hat, and the eye becomes a bit brighter, so... it just reduces the number of colours in the PNG?

Hi Pakaku. It does reduce the color palette, but in an intelligent way. Colors are chosen in a way to minimize the impact of the optimization process. As you noted, there is a small impact in quality for the more color rich PNG files (shading, gradients, etc). littlePNG uses adaptive dithering to mitigate those effects. For many common uses of PNG files such as buttons, banners, etc. there is often no difference at all.

Note that the above only applies if you're choose the "Lossy Compression" option. If that option is turned off, the total space savings is reduced, but quality is exactly as the original (i.e. lossless).
 
What would this offer me over ImageOptim other than the option to choose lossy compression? Can you still retain meta data, or offer any automator hooks?

Hi blackboxideas. What littlePNG offers is the capability to combine both lossy and lossless compression for incredible space savings in an easy to use batch UI. In a normal workflow with lossy compression turned on, the image is first ran through the quantization routines to reduce the color palette and apply the dithering to minimize the impact to the image. Once this is done, the image is then ran through the lossless process to gain an additional 1% - 20% compression in a lossless manner. All of this is transparent to the end user ... it operates as one single process during the workflow.

Note that if you turn off the lossy compression option, you'll still get the lossless compression. Space savings aren't as dramatic, but I have seen compression savings of upwards of 35% with it alone. It entirely depends on the image, and not all images will show savings this high with strictly lossless compression. Combine the two methods however, and the savings are dramatic. This is the strong point of littlePNG.

Meta data should be maintained, as the software isn't doing anything to specifically strip it out. I will most likely expose more options down the road, so at some point soon I will add an "advanced" configuration panel which will allow users to strip metadata if they so choose, along with many other configurable options. For version 1.0.0 however, I chose to keep the workflow simple, and preset the software with the options that generate the highest quality image with the greatest space savings.
 
Last edited:
I've processed the Mario image again, and uploaded the variations for download. This should allow for a better demonstration of the results. Click the URL's below to view the images.

Original Image: http://emediadesigns.com/littlePNG/Mario-Original.png
Original Size: 1.45 MB

Lossless Optimized Image: http://emediadesigns.com/littlePNG/Mario-Lossless.png
New Size: 973.14 KB
Savings Over Original: 34.38%

Lossy Optimized Image: http://emediadesigns.com/littlePNG/Mario-Lossy.png
New Size: 319.35 KB
Savings Over Original: 78.47%

As you can see, there are gains to be had with most images using only lossless compression. Add in the lossy + the lossless, and the savings are much more dramatic.

If anyone has any questions, please let me know. Thanks!
 
Oops! Got my library names mixed up earlier. Lossless compression is powered by the OptiPNG library, not ImageOptim. Sorry for any confusion.
 
Version 1.1.0 of littlePNG has been released on the MAS. Based on the comments here, I've exposed more options in the advanced settings.

http://apple.co/195bVVZ

• Added ability to control dithering for lossy compression.
• Added ability to adjust quality for lossy compression.
• Added option to control number of trials for lossless compression.
• Added ability to preserve file attributes.
• Added option to choose interlace type.
• Added additional logging statistics for optimized files.
• Added improved reporting mechanisms.
• Fixed bug: destination browse cancel error.

http://apple.co/195bVVZ
 

Attachments

  • 1.png
    1.png
    258.6 KB · Views: 170
  • 2.png
    2.png
    279.9 KB · Views: 162
  • 3.png
    3.png
    246.1 KB · Views: 151
Last edited:
To be honest, it looks exactly like ImageOptim. Also, the ”difference” you talk about, correct me if i'm wrong, but you could do exactly the same things in ImageOptim.

I may be wrong, it seems that you using the same libraries and almost the same GUI, but with less settings for the user to tweak the different libraries, or choose what libraries or options used. Also, ImageOptim is integrated with ImageAlpha. So not to be a party crasher, but i really don't see how this one is beating ImageAlpha.

Also, the libraries used is OpenSource (various licenses), and i don't think they are compatible with the terms & conditions at MAS.

I may sound like a grumpy old man, but MAS is full of apps that just is repackaged/rebuilt versions of OpenSource apps. Of course there is nothing wrong with this, as long as they follow the license (not very common at MAS. Specially in the Graphics and Video player department.
 
While the two programs are similar, they are not the same. I ran a sample image (the Mario image above) through littlePNG, ImageOptim, and ImageAlpha. This image started out at 1.5MB.

littlePNG: 78.47% reduction (319K)
ImageOptim: 38.6% reduction (932K)
ImageAlpha: 78% reduction (320K)

ImageAlpha does perform lossy compression, and achieved a very similar result as littlePNG. However, it does not do batch processing as far as I can tell. littlePNG has batch capability (like ImageOptim), and the ability to perform lossy compression (like ImageAlpha).

Both ImageAlpha and ImageOptim are wonderful apps, and I would never discourage anyone from using them. If you need batched lossy PNG compression however, littlePNG is more suited to the task.

OptiPNG is licensed under the zlib license. pngquant is licensed under the BSD license. Both of these licenses meet the terms and conditions of the MAS. I wanted to add additional compression libraries when I started development, but nearly all of them are under the GPL, which cannot be used under Apple's current terms and conditions. For this reason, I assume you'll never see ImageOptim in the MAS. I applaud the developer for that ... many folks would have tried to sneak it through regardless.

Thank you for your feedback.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.