Thank you for the feedback. Looking into baffle diffraction today. Read a good
paper by Jeff Bagby on diffraction. Then ran a diffraction simulation in Vituixcad and looking at the results to see what is correlating and what isn't. I've loaded it into ARTA as a target curve just to visualize.
View attachment 1420716
Red curve is the simulated diffraction, and the maroon is my original enclosure, and light blue is the new enclosure with all the work we've applied. The diffraction chart is modeling the original enclosure. (See chart below for the diffraction curve simulated for the new enclosure.)
The main observation is that we have a bump at 600hz and a dip at 1200hz, and one final bump at 2khz. I'm not sure if the simulation is actually accurate because according to Jeff it should have more and more random bumps and dips as you go up in frequency due to harmonics, which I don't see in this simulation. But, taking it for what it is, it seems that there is a good amount of correlation 600hz-1800hz, so that likely explains the dip at 1khz as you say, good intuition there.
But when I model the diffraction of the test enclosure I built. I get a simulation that predicts a boost at 1000hz not a dip. This diffraction curve should correlate with the blue FR curve.
View attachment 1420720
So I'm not really sure what to think at this point. One would think if it models accurately for one it should for the other as well but I don't observe that. I doesn't seem correct to accept results which agree with my intuition and then reject the very same results when they disagree with my intuition.
Since I'm working within the confines of an existing 3-way tower I don't have any control over its baffle, until I build a new speaker, but I'd rather keep these and just have it work with updated drivers.
The irony here is that my original enclosure which I was complaining about is still flatter than the test enclosure which I built with everyone's feedback.
Should I start over and make another/different test enclosure? How would you change it?
