| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
#include "third_party/blink/renderer/platform/audio/biquad.h"The new tests in this file don't seem to use the BiquadFilterHandler at all. Would it make sense to move them to a new platform/audio/biquad_test.cc file instead?
// (n-2)*log(r) + log(|c1*(n+1)*r^2+c2|) - log(eps)Let's update this comment
```suggestion
// (n-2)*log(|r|) + log(|c1*(n+1)*r^2+c2|) - log(eps)
```
fdlibm::log(fabs(c1 * (n + 1) * r * r + c2)) - log_eps;Perhaps this should be `std::abs` too, since we're working with `double`s?
return 0;Let's add a comment here saying something like we don't want to keep calculating tail time for filters with non-finite coefficients.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
japhet@ PTAL at t/b/r/p/BUILD.gn
Thanks in advance!
#include "third_party/blink/renderer/platform/audio/biquad.h"The new tests in this file don't seem to use the BiquadFilterHandler at all. Would it make sense to move them to a new platform/audio/biquad_test.cc file instead?
Done
Let's update this comment
```suggestion
// (n-2)*log(|r|) + log(|c1*(n+1)*r^2+c2|) - log(eps)
```
Done
fdlibm::log(fabs(c1 * (n + 1) * r * r + c2)) - log_eps;Perhaps this should be `std::abs` too, since we're working with `double`s?
Done
Let's add a comment here saying something like we don't want to keep calculating tail time for filters with non-finite coefficients.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
Build file change LGTM - didn't review the logic change
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
| Commit-Queue | +2 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |