Első kör

if not on 18/6 cycle, adjust proportionally... 12h is 2/3rds of 18h, so you need 150% (or 3/2nds) power to provide equivalent DLI. 16 to 12 would be 133% more power. easy peasy math.

after 9 weeks time to believe it was a mis-packaged photoperiod or the breeder cut a corner with the autoflower process and got a ratio of photoperiods etc... i.e. human error on the breeder's end. i'm not sure if the mutlip step process to get autos is 100% or if it's some greater than 99% probability. Maybe it's unlucky, but more likely human error. if your light is powerful enough, bump up power 150%, same distance and reduce light from 18 to 12 hours/day. This will provide the same DLI to your autoflowers and yield will not suffer one bit. Hopefully you got some headroom on light power. Or, if you have a second tent and light, put it on 12/12 and remove this plant from the 18/6 auto tent.

each seed is different and you could have gotten a random one that wasnt all the way stabilized and it has the fast flowering trait rather than full autoflower trait. tbh it looks like it may be about to start flowering based on the picture of the top main of the plant in the first picture. Give it a week or you could reduce the light schedule just to induce flowering. it may even react without needing to go full 12/12. ive ran autos with 12/12 for flowering cycle and it didnt affect yield in my experience. ill check out the full diary and send you a message if i see anything else that stands out to me.

Yeah going from MH to HPS for flower is perfect. Plants life cycle wants white light seedlings then blue then red in flower...

2000k might be too "warm". Other wavelengths are still important and not just the red ranges. The other thing to consider-- plants adapt to the light the receive. So the pigmentation and such adapts to CCT and intensity. If you drasticalyl change the light provided, that potentially has some drawbacks if you do it midstream... probably small effects that are diffiuclt to see with eyes and not tracking large samples while freezing other variables and having a control group for camparison to make a rational assessment on such a thing. Don't over think it. 6000k and 2000k are probably both too extreme regardless of whether you use them in vege or flower and the effect of it is virtually impossible to recgonize with your eyes. no matter what, some reconfiguration over time will occur in the leaves and that has a cost of 'something'. wehther it is outweighed by some other benefits is not something any human can resolve by simply doing it with a few plants in their home and wildly variable environments etc. these people are just lying to themselves and reinforcing existing long-held beliefs regardless of shoddiness of evidence used. like mouse said, focus more on providing proper levels of DLI and then play around with CCT options after that need is met.

Yes, if you're planning to switch to 12/12 for flowering then it's a good time to switch your bulb to the red end of the spectrum.