1.  Xpure Vs Namaster

Using 3 mask apodization (via namaster mask apodization routine) : {$ \theta_* = 5, 10, 30 ^\circ $}, we compre the results between Namaster (100000 simulations) and Xpure (~ 300 simulations) :

mean + std
combined std

• Observation :
  • Std with xpure suing same Namaster masks produces the same biased spectra variance, especially the variance 'explosion' at low ell.
  • Same apodized mask produce different stds. Why ? ==>

Spectra using 1 cmb map :

apo=5°	apo=10°	apo=30°

• observation :
  • spectra difference seems to decrease as apodization length increases

1.1  Full

pcl
cl

1.2  gal band (fsky = 80 %)

	apo=5.0	apo=10.0	apo=30.0
cl
pcl

1.3  pol cap (fsky = 80 %)

	apo=5.0	apo=10.0	apo=30.0
cl
pcl

1.4  xpol on cap :

	apo=5.0	apo=10.0	apo=30.0
cl
pcl

1.5  Boost nside for cap mask :

• going from nside= 32 -> nside=128
• lmax goes from 63 -> 255

	apo=5.0	apo=10.0	apo=30.0
cl
pcl

• observation
  • NaMaster gives similar results
  • diff namaster/xpure less important.
  • does xpure suffers from pixelization effect ?
  • BB bias at low ells is stil present for low apodization lenght (known issue since first use of namaster -> bad mask apodization + derivative)