What if we could generate two datasets with same reads, one with good quality and one with bad quality?

Is your feature request related to a problem? Please describe.
In order to evaluate my error correction tool, I though to build a confusion matrix by filtering one dataset containing bad reads and use as reference one dataset containing the corrected version of the first one reads.

Describe the solution you'd like
I tried to generate two datasets using the same parameters as input (also, the seed in fixed in order to maintain a deterministic generation of sequences) but with quality and glitches values differents. The result is two different collections of reads, coming from different locations of the reference genome.

Describe alternatives you've considered
What I need are two datasets with the same reads, one with perfect quality and one with glitches and errors. In this way, I could compare the perfect dataset with the one corrected by my tool and creating the confusion matrix.

Additional context
I'm honestly not sure if this is already possible, but I have not been able to find the solution to my problem in the documentation or on the internet.

No files, tests, or entry points are identified. Start by reviewing Badread's documentation and existing dataset-generation options, then determine how deterministic generation could preserve identical reads while varying quality and glitches; done means producing paired datasets with matching reads and controlled quality differences.