PV string calculation/comparison

My hybrid inverter has two strings of panels. The top string is 7 panels and the bottom is 9. I have been watching the generation of each string and I am a bit confused with the results. My thinking was that if I took the production for string1 and divided it by 7 I'd get a value per panel and then if I took the generation of string2 and divided that by 9 I would get the per panel value of that and the two values would be roughly the same with no shading. What I in fact see is that the values are not the same, the string2 average per panel is always around 90% of the value of string1. Is my thinking wrong or is this pointing to an issue with one of the 9 panels on string2?

If there is a problem with strings2 then it doesn't look like an entire panel is dead as if I divide the string2 generation by 8 it still doesn't match up.

Examples from the GE cloud console today. The roof is not shaded at either of these times:
16:44 (after an hour of solid sun):
String1 - total 1849w - average 264w (7 panels)
String2 - total 2185w - average 243w (based on 9 panels) - 273w (based on 8 panels)

Peak at 15:08:
String1 - total 2227w - average 318w (7 panels)
String2 - total 2629w - average 292w (based on 9 panels) - 328w (based on 8 panels)

The comparison has been about the same since the panels were installed last year and I wanted to get some other opinions before I go back to the installer and ask for some advice.

Are the panels on the same aspect? If not you need take into account time of day.
I have two strings- south and west. Both with 6 identical panels on. The west string never produces the same peak power as the south simply because it doesn’t see the sun properly until later in the day.

The 7-panel string is right above 9-panel string on the same roof so should be getting roughly the same amount of sun.

Zaz ah ok. That doesn’t make a lot of sense. Unless it’s current limiting (which doesn’t make sense either!)

I have a 5kw hybrid inverter and the panels are 375w (6kw in total) so you mentioning limiting is making me wonder if that has something to do with it!? But then I am seeing the same difference from when the whole array is producing less than 1kw up to it most I've ever seen at 5.3kw. I could see limiting as a possibility at peak but not when it is a long way from peak generation!