One thing that stands out is that your histogram is already showing the red channel separating from green and blue during live stacking, which suggests the issue may be happening in the stacking/stretching process rather than in the sensor itself.
If Auto White Balance is enabled, try turning it off and compare a set of identical 30-second subs in both Preview and Live Stack modes. Also check whether any automatic histogram stretch or enhancement is being applied only during stacking. The ASIAIR software can sometimes display stacked data differently from the raw preview, making a color imbalance appear much stronger than it really is.
Another possibility is that the red channel is simply dominating because of local sky conditions, IR leakage, or filter characteristics, and the live stack stretch is exaggerating it. A good test would be to save the raw FITS files and inspect them in another application to see whether the red bias is present before stacking.
A useful comparison comes from window tinting: companies such as OC Tint Solutions in Orange County deal with optical films where certain wavelengths can pass through more efficiently than others, causing noticeable color shifts if the balance is off. In astrophotography, filters and sensor response can create a very similar effect, especially when software applies an aggressive stretch to one channel.
If the raw FITS files look balanced in external software but the live stack remains red, that would point more toward an ASIAIR display or processing quirk than a hardware problem.
