Who started the WebP fairy tale?

WordPress Mobile Speed

Updated


There are no affiliate links on PagePipe.

UNBELIEVABLE. Look at how many site owners are drinking the WebP image Kool-aid. “Drinking the Kool-Aid” refer to a person who believes in a dangerous idea because of perceived potential high rewards. That’s right we’re saying WebP is *dangerous*. There is no reward. Why? Because it wastes time and money and your personal energy. Chasing ideas that don’t make a significant difference in speed or ranking is wasteful. True entrepreneurs can’t afford to waste limited resources.

WebP is a speed hoax.

Oddly, the hoax is most perpetrated by professional speed optimizers. And image plugin authors say it’s state-of-the-art “best practice.” We hate WebP with derision.

Why derision? Because WebP is so fake, we don’t know whether to laugh or puke. Or both.

REFERENCE: https://pagepipe.com/dont-use-webp-image-format/

You may say, “Prove it doesn’t help speed. We issue a challenge.”

OK. We accept your challenge. But if you spent even a little money converting your media library to WebP, you won’t appreciate how this fairy-tale story ends. You may feel angry or foolish.

The WebP Torture Test

So we took an optimized PNG image (45k). We put it through a WebP online tool – and the resultant file size is 25.5k. That is a 43 percent decrease in file size.

Whoopee! Sounds impressive, doesn’t it? 43 percent better! But better than what? Is the speed better? Let’s check.

This is the image. It’s one we formerly used to advertise our speed services.

Does that fantastic WebP reduction make the page 43 percent faster? No.

Surely at least the image loads 43 percent faster? No.

What? It makes no difference at all in speed?

Skeptical? Keep reading. We’ll explain with real numbers.

Side-by-side comparison of images.

Now we’ll upload both the PNG and a WebP conversion onto our WordPress test page. Then we’ll make separate measurements to find the speed difference of each image using a browser timer addon.

Note: What curiosity is this? The LiteSpeed dashboard claims:

Significantly improve load time by replacing images with their optimized .webp versions.

Why did they say that? To appear cool offering a fashionable magic image format? Imitating the big and popular technology doesn’t make things cool.

Who started this WebP fairy tale?

We’ll soon prove the WebP web myth is pure rubbish. Not significant at all for speed improvement.

Who claims it is significant? Google influences many ignorant blogs who assume Google speaks the truth. Who claims to have invented WebP? Genius Google does. Did they? No. They acquired the technology inadvertently in a corporate buyout. They didn’t know what good it was. It had no purpose. It was supposed to help videos. Not static PNG and JPEG images.

WebP is an image format employing both lossy and lossless compression and supports animation and alpha transparency. Developed by Google, it is designed to create files that are smaller for the same quality, or of higher quality for the same size, than JPEG, PNG, and GIF image formats.

https://en.wikipedia.org/wiki/WebP

Sorry, Wikipedia. Google did not develop this. They bought it in a business error. WebP is face-saving salvage after the flub. “What should we do with this silly thing?”

Nudging the marketing guy next to him, “Let’s tell website owners this stupidity we paid for is good for still images. Then we’ll look smart instead of dumb.”

WebP as speed goodness is now regurgitated over and over on web blogs. And in plugins read.me files. And, by image optimization plugin authors. And, on server caches. Incredible.

It’s not true. It isn’t goodness for speed. At best, it’s benign.

Didn’t anyone test this WebP thing in a real web environment?

WebP may slow your site if you create and load the WebP images using offsite remote image storage. CDNs like ShortPixel for example.

TRUTH: WebP does make images lighter.

But does WebP make a difference in real speed? You know: page load time in milliseconds?

TRUTH: Not a bit. It makes no difference in speed.

We’ll demonstrate. You can try this experiment for yourself.

Wait! Seriously? Everyone says, “WebP improves my web core vital metrics.” It may improve some lame scores. But it won’t improve your page speed. Page rank improves not one iota either.

Because various online speed tests vary so widely, for this test we use a browser timer. We prefer using a browser add-on or extension. In this case: app.telemetry page speed monitor. This handy tool checks the page load times of any web page. It shows the offset and duration of each page load step based on W3C navigation timing. And it’s free for Chrome and Firefox. Nice.

WebP Test One

So here’s what we use for the test setup:

  • Browser timer
  • Stock default WordPress Twenty Twenty One theme.
  • Homepage without images.
  • Hosting: GreenGeeks with LiteSpeed cache enabled.

Naked homepage results – no image.

initial load: 489 milliseconds.
Primed cache: 185 milliseconds.

Add 1 PNG image results.

We add the 45k PNG test image shown above top. And voilà —

Initial load: 300 milliseconds.
Primed browser cache: 185 milliseconds.

Wait! Isn’t that faster? Shouldn’t adding the image slow down the page?

The installed image is being lazy-loaded in parallel with the page favicon. The image addition made no change. Because of image lazy load and parallel loading. Lazy load is now a default in WordPress core since August 2020. Is it better than a standalone lazy load plugin? Not always. Test to make sure. Primed cache speed is identical in this case.

Replace PNG with WebP image.

Now we remove the PNG image and replace it with a optimized WebP image. Sadly, it won’t upload to the media library until version 5.8 is released on July 20, 2021. So we do a plugin workaround. We installed: https://wordpress.org/plugins/webp-converter-for-media/

WebP Converter for Media – Convert WebP & Optimize Images is a free plugin from the WordPress directory of over 60,000 open-source plugins.

The bogus promise in the plugin description:
Remember that a better-optimized website also affects your Google ranking.

No it doesn’t. Not again. Shocking how the WebP propaganda runs so deep. But why would this plugin author lie to us? Oh, no. He drank the toxic Kool-Aid, too.

We convert the JPEG image installed on the page to WebP using the conversion plugin default settings. We then purge the LiteSpeed cache.

WebP image -after results.

The image is now a 31K WebP format derived from the same original PNG. A little lighter than the other online tool WebP conversion. 36.8 percent lighter than the original PNG image. Bonus.

Now, will it make a difference in speed?

BEFORE WebP conversion (PNG)
Initial load: 300 milliseconds.
Primed browser cache: 185 milliseconds.

AFTER WebP conversion
Initial load: 320 milliseconds
Primed cache: 189 milliseconds

Seriously! Slower after conversion to WebP conversion.
Wasn’t it 36.8 percent better in goodness?

So? WebP conversion made zero difference in WordPress speed.

WebP Test Two

Let’s get super serious and torture test.

We download a huge open-source jaguar image from Pixabay.
Weight: 4.1 megabytes.
Dimensions: 5616 x 3744 pixels.
A monster size image. It would ruin page speed in the good old days of site-building. We upload it “as-is” without any preparation.

https://pixabay.com/photos/jaguar-stains-look-fiera-stalking-3370498/

Uploaded as a featured image:
WordPress does it’s magic and resizes the huge JPEG to: 1366px × 911px (scaled to 1166px × 777px). It also changes the JPEG compress quality setting to 82.
Final result: 223.78 kilobyte featured image — not the original 4.1 megabytes.

No image optimization plugin did that. WordPress core did.
4 megabytes down to 224 kilobytes.
Think about that. Wow. WordPress protects you from overloading your site with bloated images.

The WebP plugin wouldn’t convert a featured image.
It failed.
We decide to try an inset image in the page body in stead for testing.

On the page, we inset the huge JPEG image. WordPress automatically converts it to:

JPEG: 1024 × 683 pixels.
(scaled to 608px × 406px by the browser)
Weight: 114.46 kilobytes.

Initial Page Load time: 299 milliseconds.
Primed cache: 184 milliseconds.

Next, we convert the uploaded inset image with the installed WebP plugin.

Dimensions: 1024px × 683px
(scaled to 608px × 406px by the browser)
Image Size: 84k

BEFORE

Initial Load time: 299 milliseconds.
Primed cache: 184 milliseconds.

AFTER

Initial Load time: 290 milliseconds.
Primed: 189 milliseconds.

Inconsequential, insignificant speed change from WebP.

Let’s check some more numbers with speed test results instead of a timer.

WebPageTest.org:
TTFB: 329 ms
load time: 624 ms.

Now let’s uploading a properly-sized JPEG image so the browser and WordPress don’t have to do resizing or scaling. We use GIMP free image editor using save for web GIMP plugin features.

68.2 kilobytes
608×405 pixels optimized.

It’s then changed some more by WordPress during upload. The only change is a slightly smaller weight: 66.6 k. We take another measurement:

initial load: 289ms
primed: 201ms

Again inconsequential change in speed.

speed test results: WPT.org:
TTFB: 280 milliseconds
load time: 559 milliseconds

WebP image format doesn’t help speed performance. It makes images lighter — but it doesn’t impact speed measurements.

        

Test 2

no image

JPEG original

WP upload

WebP conv.

change

hand opt

change

weight k

0

4100000

114

31

83

66.6

47.4

dimension px

 

5616 x 3744

1024 × 683

1024 × 683

 

608×405

 

initial load

489

 

299

290

9

289

10

primed cache

185

 

184

189

-5

201

-17

REFERENCE: https://pagepipe.com/webp-is-a-speed-hoax/

Godspeed-

Steve Teare
performance engineer
July 2021

 

PagePipe Site Tuning Services for Speed

Instead of band-aid approaches, we drill down to the root cause of your slow site. This is origin optimization. Also known as site tuning. To do this, we analyze site components:

  • Hosting
  • Theme
  • Plugins
  • Scripts and third-party services.
  • Images and media library.
  • We minimize globally loading plugin effects.

Find out more details about Site TuningGet Speed!