[Dean Olin]

I think I read somewhere on this site that a new version would be coming out in late 2021? I’m wondering if the grading of the quizzes will be updated at all? It seems like the current grading just looks at how many parameters are correct, and then derives a score from that. The problem with this is that quite often, my sound is very similar to the hidden patch, but I have arrived at it in a different way, so I get a low score. I think that using some sort of audio analysis using a waveform similarity measure such as cross correlation or some type of audio fingerprinting might work better, since it would just be comparing the actual sound produced as opposed to the parameters used, and, after all, it’s only the sound that we’re after. Is anything like this in the works?

[Joe Hanley]

Hi Dean

2.0 will be out in the summer. We’re adding improvements to the scoring process that will help make scoring more consistent, especially when there are only subtle differences. We’re also adding a light red coloring to indicate when a control is technically wrong, but the sonic difference is so subtle it doesn’t realistically matter.