I was following some advice from PageSpeed Insights from Google and one of them is:
Formatting and properly compressing images can save many bytes of data.
In the meantime I've optimized them with a program - JpegMini - and even then there's still the suggestion to optimize the images (even after 30 seconds). Is there any other way to optimize them?
OptiPNG :
$ optipng /tmp/saturno.png
** Processing: /tmp/saturno.png
1041x1041 pixels, 8 bits/pixel, 195 colors in palette
Input IDAT size = 119947 bytes
Input file size = 120793 bytes
Trying:
zc = 9 zm = 8 zs = 0 f = 0 IDAT size = 118148
zc = 9 zm = 8 zs = 1 f = 5 IDAT size = 113455
zc = 9 zm = 8 zs = 3 f = 5 IDAT size = 112036
Selecting parameters:
zc = 9 zm = 8 zs = 3 f = 5 IDAT size = 112036
Output IDAT size = 112036 bytes (7911 bytes decrease)
Output file size = 112846 bytes (7947 bytes = 6.58% decrease)
jpegoptim :
$ jpegoptim /tmp/jupiter.jpg
/tmp/jupiter.jpg 1920x1080 24bit N Exif IPTC XMP ICC Adobe [OK] 175249 --> 174634 bytes (0.35%), optimized.
An alternative would be imagemin , via the terminal using imagemin-cli or if you prefer via imagein-app application, but the latter does not work as well (by the trial version). There is also the possibility via Grunt or Gulp .
In case you use via terminal imagemin-cli :
# instalar o imagemin via npm
npm install -g imagemin-cli
# optimizar todas as imagens JPG/PNG/SVG de uma pasta
imagemin pasta_com_images/* -o pasta_de_destino/
Using the reference images (see below) for comparison it is verified that:
c-1920.jpg had a 6.7% reduction pia01969-saturn-voyager1.png had a reduction of 11.8% Another alternative for bitmap images would be ImageOptim , via the terminal using Imageoptim-cli or if you prefer via imageoptim for MAC OS , or for other platforms . There is also the possibility via Grunt or Gulp .
In case you use via terminal imageoptim-cli :
# instalar via npm
npm install -g imageoptim-cli
# optimizar todas as imagens JPG/PNG de uma pasta
imageoptim -d ./pasta_com_images/
Using the reference images for comparison it is verified that:
c-1920.jpg had a reduction of 21.1% pia01969-saturn-voyager1.png had a reduction of 10.7% For this small experience, ImageOptim considerably reduced the JPEG image more than Imagenin (21.1% instead of 6.7%). The PNG image was close, obtaining the best image reduction (11.8% instead of the 10.7% of ImageOptim). On the other hand, imagemin also optimizes SVG images.
For the comparison, we used the source images JPG and PNG used in the Lacobus response .
For other examples see the following comparison
Hope you can help someone!
