Help generating an event rate per allocation period plot

I am running the following code to plot event rates as a function of allocation time. Unfortunately, python crashes without any error when I call histo for the first time. Looking at it as far as I can, it seems the error is occurring on the C side. I guess it's a memory issue since this doesn't occur when using fewer bins. So I have a few questions.

1. Is this the right way to produce this information?
2. If it is, is there a preprocessing step I could/should perform to reduce the amount of data required? I considered the ROIFilterAlgorithm, but I assume I would need to normalize the x and y values to avoid allocating the full sensor width and height.

from metavision_ml.preprocessing import histo
import numpy as np
from metavision_core.event_io import RawReader
import matplotlib.pyplot as plt
from tqdm import tqdm

file = input('Enter the path to the input file: ').strip()
rates = []
dt = int(1e6) # 1s
record = RawReader(file)
# record.seek_time(1e6) # skip first second if needed
events = record.load_delta_t(dt)
events['t'] -= events[0]['t'] # important! almost all preprocessing use relative time!
height, width = record.get_size()

tbins = np.arange(100, 1000, 100) # Time bins from 100us to 1ms

for tb in tqdm(tbins):
    volume = np.zeros((dt // tb, 2, height, width), dtype=np.float32)
    histo(events, volume, tb)
    # combine the two polarities
    im = volume.sum(axis=1) # sum the two polarities
    # average the time bins
    im = im.mean(axis=0) # average the time bins
    # flatten the image
    im = im.flatten() # 1D array
    # Calculate the event rate
    rates.append(np.count_nonzero(im) / im.size)
    del volume # free memory