Testing heteroscedasticity statistically

Hi everyone,

I conducted a Breush-Pagantest and it is unsignificant, so there is no heteroscedasticity. My residual plot however looks pretty heteroscedastistic to me. So my question is, if the statistical test says there is no problem, can I trust this? Or should I do a White's test as well? If the two tests are not significant, can I reliably interpret my regression output?