Open pschonmann opened 1 day ago
Hi @pschonmann check if the LCP element is the correct element. If so, you can report Chrome bugs direct in https://issues.chromium.org/issues . If it seems to be the wrong element for LCP, let me know.
Im not sure if its only chromium thing. Im getting these weird results in Firefox and Edge -b option too. Anyway LCP seems to be identified okay. In chromium ctrl+shift+i and performace tab - LCP > img - same match as LCP on html report in first run.
root@docker1:~# docker run --rm -v "$(pwd)":/sitespeed.io sitespeedio/sitespeed.io:latest https://www.argul.cz -b firefox -n 1
Google Chrome 130.0.6723.58
Mozilla Firefox 131.0
Microsoft Edge 129.0.2792.89
[2024-11-07 06:34:59] INFO: Versions OS: linux 5.10.0-28-amd64 nodejs: v20.17.0 sitespeed.io: 35.3.1 browsertime: 23.2.0 coach: 8.0.2
[2024-11-07 06:34:59] INFO: Running tests using Firefox - 1 iteration(s)
[2024-11-07 06:35:01] INFO: Testing url https://www.argul.cz iteration 1
[2024-11-07 06:35:11] INFO: Take after page complete check screenshot
[2024-11-07 06:35:12] INFO: Take largest contentful paint screenshot
[2024-11-07 06:35:16] INFO: Use the visual metrics portable script
[2024-11-07 06:35:16] INFO: Get visual metrics from the video
[2024-11-07 06:35:20] INFO: https://www.argul.cz 86 requests, TTFB: 257ms, firstPaint: 825ms, firstVisualChange: 867ms, FCP: 813ms, DOMContentLoaded: 784ms, LCP: 684ms, CPUBenchmark: 60ms, Load: 1.12s, speedIndex: 1.73s, visualComplete85: 1.63s, lastVisualChange: 6.70s
[2024-11-07 06:35:20] INFO: The server responded with a 404 status code for https://popup-server.azurewebsites.net/zbozi/company/review/...
[2024-11-07 06:35:22] INFO: HTML stored in /sitespeed.io/sitespeed-result/www.argul.cz/2024-11-07-06-34-59
root@docker1:~# docker run --rm -v "$(pwd)":/sitespeed.io sitespeedio/sitespeed.io:latest https://www.argul.cz -b edge -n 1
Google Chrome 130.0.6723.58
Mozilla Firefox 131.0
Microsoft Edge 129.0.2792.89
[2024-11-07 06:35:33] INFO: Versions OS: linux 5.10.0-28-amd64 nodejs: v20.17.0 sitespeed.io: 35.3.1 browsertime: 23.2.0 coach: 8.0.2
[2024-11-07 06:35:33] INFO: Running tests using Edge - 1 iteration(s)
[2024-11-07 06:35:33] INFO: Using Edge is experimental at the moment and use the same configuration as for Chrome
[2024-11-07 06:35:34] INFO: Testing url https://www.argul.cz iteration 1
[2024-11-07 06:35:45] INFO: Take after page complete check screenshot
[2024-11-07 06:35:45] INFO: Take cumulative layout shift screenshot
[2024-11-07 06:35:46] INFO: Take largest contentful paint screenshot
[2024-11-07 06:35:49] INFO: Use the visual metrics portable script
[2024-11-07 06:35:49] INFO: Get visual metrics from the video
[2024-11-07 06:35:53] INFO: https://www.argul.cz 76 requests, TTFB: 155ms, firstPaint: 540ms, firstVisualChange: 533ms, FCP: 540ms, DOMContentLoaded: 768ms, LCP: 427ms, CLS: 0.0021, TBT: 0ms, CPUBenchmark: 41ms, Load: 928ms, speedIndex: 1.39s, visualComplete85: 733ms, lastVisualChange: 6.57s
[2024-11-07 06:35:54] INFO: The server responded with a 404 status code for https://www.argul.cz/favicon-16x16.png
[2024-11-07 06:35:55] INFO: HTML stored in /sitespeed.io/sitespeed-result/www.argul.cz/2024-11-07-06-35-33
Edge and Chrome use same codebase. You can report bugs for Firefox in https://bugzilla.mozilla.org. So the FCP and LCP is picked up from the browsers API:s, checkout https://w3c.github.io/paint-timing/#first-contentful-paint and https://w3c.github.io/largest-contentful-paint/#sec-largest-contentful-paint-interface, the values are not touched in Browsertime/sitespeed.io, they are picked up directly from the browser.
Your question
FCP is greater than LCP, that doesnt make sense