JPEG Compression Test: Google Photos vs. JPEGmini

May 30, 2015 - photo frame


In this essay we will take a demeanour during Google Photos’ new print application performance. I’ve been regulating a module called JPEGmini for a integrate years now to restrict my JPEG images. Its application of JPEGs is lossy, nonetheless it claims to do so withdrawal a perceptual design peculiarity probably unchanged. As distant as we can tell, a claims are flattering accurate, and it has literally helped me cut a distance of some of my design folders in half.

As I’m certain many of we are aware, Google just denounced Google Photos, and with it announced total storage space for photos and videos. The total partial comes with a caveat: Google will request lossy application to your files.

Like JPEGmini, Google claims to be means to request lossy application to images nonetheless changing a perceptual peculiarity of a image. If we investigate a uncompressed and dense images with a computer, we can see differences — nonetheless by eye, they demeanour identical.

So we a did a small exam with a splendid design to review JPEGmini’s application with Google’s compression. Take a demeanour during this image:


This shows a comparison between a uncompressed image, Google’s compression, JPEGmini’s compression, and finally JPEGmini practical to an already Google-compressed image.

You can confirm for yourself if there is any detriment in visible acuity between a original, a Google-compressed, a JPEGmini-compressed, and a double-compressed image. To make it easier to perspective a images side-by-side, here’s an Imgur album with any particular frame, that we can bucket it apart browser tabs and click behind and forth.

Below, I’ve constructed a list display a opposite application ratios.


In my personal opinion, all 4 images have a same perceptual design quality. Google’s process is remarkably good. The rebate in design peculiarity is roughly imperceptible, nonetheless it manages to scarcely cut a record distance in half. On a other hand, JPEGmini manages to eek out an additional 40% application with usually pardonable design degradation. Most conspicuous of all, though, is that by mixing a two–Google and afterwards JPEGmini–it doubled a application even of a already-compressed image!

Of course, a double-compressed design has a many critical plunge of design quality, nonetheless even so, it is so teenager that, we think, it is still roughly positively unnoticeable in vast photographs.

Finally, for a some-more severe research of a tangible differences between any of a 4 exam cases (original, Google, JPEGmini, Google+JPEGmini), we installed them adult in Photoshop and used a disproportion covering outcome to calculate a tangible changes in a strange design that a application algorithms performed. You can see that here:


What should be immediately transparent is that Google’s application algorithm was a slightest spiritless by a prolonged shot. The original, uncompressed design was usually really somewhat altered by Google’s method, and nonetheless it still was means to scarcely cut a filesize in half. JPEGmini’s application is good, too, eeking out an additional 40% application over Google’s while still causing comparatively small repairs to a strange photo.

The biggest burst in design peculiarity plunge is from requesting JPEGmini to an already Google-compressed image. No pain no gain, we guess, nonetheless even so, we cruise it exhibits remarkably small plunge cruise it is scarcely a 100% alleviation over possibly Google or JPEGmini’s algorithms alone.

A Better Test Image

Next, we redid a comparison with a somewhat improved image. The strange design was usually a print taken on an 2014 Moto X. This time around, we used a studio exam design taken by a Nikon D750 DSLR ($2000 camera) taken from DPReview. Because Google usually allows a max print distance of 16 megapixels and a strange exam design was 24 MP, we downloaded a RAW format shot and resized it down to 16 MP in Photoshop before starting a test. You can squeeze a new exam images here.

And now a updated table:


This time a formula are a small bit different, and there are some engaging results. Google’s algorithm performs a small bit better, and we see that a best altogether application is still when we perform a Google application first, and afterwards a JPEGmini application on tip of it.

Taking a demeanour during a disproportion maps, we can see that JPEGmini and Google restrict a JPEGs a small bit differently. JPEGmini’s disproportion maps are all scarcely monochromatic. This means that JPEGmini isn’t altering a colors of a strange design really much. Google, on a other hand, is altering a colors utterly a bit.


That’s not indispensably a bad thing if Google can restrict colors imperceptibly.

…But here’s a thing–it doesn’t. Take a demeanour during a tone wheels usually to a right of a bottles during a bottom of a strange image. JPEGmini compresses these well, nonetheless Google creates some nasty outlines around a edges of a circles.


Other than that, both JPEGmini and Google vanquish a sum in a gray credentials differently, and we consider that it’s adult to personal welfare as to that one is better.

Overall, we cruise Google and JPEGmini both have considerable algorithms. Google’s beats JPEGmini in application some of a time, nonetheless it looks like there are still some kinks to work out to forestall aberrations and other artifacts. Nevertheless, we consider that this is something Google will urge over time.

Given that Google has their possess web browser and their possess mobile OS, they’re some-more than able of even pushing out new design formats or new standards. Until then, if we have some critical images that we wish to save from teenager design artifacts, it competence be advantageous to keep an additional backup outward of Google Photos.

About a author: Brian Young is a photography fan who’s meddlesome in JPEG application technologies. He has also common his commentary on Reddit.

small.wp_rp_excerpt { line-height:115%; font-style:normal; } .related_post_title { } ul.related_post { line-height:120%; } ul.related_post li { list-style-type:none; clear:both; margin:0 0 0 3px; } ul.related_post li a { font-weight: bold; display:block; margin:0 0 5px 0; } ul.related_post li a:hover { text-decoration:underline; } ul.related_post li A img { width:130px; height:auto; }

More frame ...

› tags: photo frame /