0Test de nouveaux outils de compression d'image

Je me sers Optimiseur d'image EWWW pour compresser tous mes JPEG et PNG images sur ce site depuis plusieurs années. EWWW utilise OptiPNG et PNGOUT et tout ceux-ci sont des outils puissants il existe des alternatives là maintenant que je voulais tester et comparer.


I am aware of new­er image formats that may offer smal­ler sizes but until they are broadly sup­por­ted we will have to con­tin­ue to work with PNG et JPEG.

I have already seen some com­par­is­ons show­ing some impress­ive per­form­ance by some oth­er formats but I wanted to test this for myself with my real­world images. Based on the claims presen­ted I decided to start with Pingo and see how much it reduced the size of my total image lib­rary.

Il y a 3549 des images qui sont un mélange de PNGs, JPEGs, and Webp files and range in size from over 1000×1000 pixels to small thumb­nails of 60×60 pixels.

Sans perte

taille
Ori­gin­al116Mb (121,848,736 octets)
après Pingo113Mb (118,310,160 octets)

3505 images were pro­cessed in 434 seconds which is impress­ively quick com­pared to the pre­vi­ous tools I’ve used. cependant, seulement 3455kb (3.37mb) a été sauvé qui est 2.5% sav­ing.

Des­pite the lack of reduc­tion, I will still be look­ing to migrate to pingo as it is vastly faster than the tools used by EWWW and achieves effect­ively the same res­ult

lossy

taille
Ori­gin­al116Mb (121,848,736 octets)
après Pingo94Mb (98,133,668 octets)

Lossy com­pres­sion was much more impress­ive in size reduc­tion with a 22.6Mb (19%) reduc­tion but this is to be expec­ted as my pre­vi­ous tools were not using lossy com­pres­sion. I had decided in advance that a sig­ni­fic­ant reduc­tion would be needed for me to con­sider using lossy com­pres­sion. This reduc­tion is right on the bor­der­line. The ulti­mate decision comes down to how much the smal­ler dimen­sion (plus souvent servi) images saved — the major­ity of the sav­ings were with ori­gin­al images not with the smal­ler images so I’m not per­suaded that the reduc­tion is worth­while at this point, espe­cially as I already serve webp images to over 50% of vis­it­ors and these gained the least

The next step is to fig­ure out how to auto­mate the pro­cess with pingo which isn’t sup­por­ted by EWWW.

Pense que nous avons manqué quelque chose? Faites-nous savoir en commentant ci-dessous. Si vous souhaitez vous abonner s'il vous plaît utiliser le lien d'abonnement dans le menu en haut à droite. Vous pouvez également partager avec vos amis en utilisant les liens sociaux ci-dessous. À votre santé.

Laisser un commentaire