It's difficult to say what to suggest. You have a dispersion problem from what I can discern from the information you present. Poor dispersion could originate in the dry blending stage of your process. Or it could be that you are not processing the dry blend well, getting enough shear/mixing at the twin to distribute the TiO2 evenly and then enough smear/transport to break down the TiO2 agglomerates and disperse the pigment.
None of these are uncommon w/ TiO2. It's a difficult pigment to disperse and distribute throughout RPVC.
Thing to look at:
Dry blend sequence – When do you add the pigment? How long after adding the TiO2 do you run the mixer to reach drop temp? I strongly urge you to cool your dry blend to about 105-110 F before trying to process it, but that not going to effect mixer efficiency. It may help your extrusion process.
Extruder – Observe the compound at the vent. During good processing, this material should not be powder, nor should it be melted completely. It should look like hot oatmeal, slightly melted, but lumpy and porous in texture. If it’s melted, your extrusion condition is to extreme and hot. Back off on the heat, oil, or shear. If the vent is powdery, you need more work in the material. Add heat to the zones before the vent, up the oil temp, or increase the shear via screw speed and feeding.
Finally, if your problem can’t be solved at the extruder, then you will need to address it at the mixer. TiO2 dispersion and distribution is problenatic, but can be helped by; adding the pigment early in the mix cycle, changing to a surface-treated version of pigment which is easier to disperse, raising the drop temperature to prolong the mix time after pigment addition, pre-disperse the TiO2 in PVC powder 50/50, and use it at 2X the required level, like a master batch.
Detail on your dry blend formulation (full phr levels on all ingredients) and current mixing cycle details would be helpful in troubleshooting the problem.