Re: MIX CHALLENGE - MC103 April 2025 - Mix Round 1 in evaluation
Posted: Sun Apr 27, 2025 16:46 CEST
I'm glad you found it helpful, I hope it will make a difference. This is still thing that have to be used with caution and referenced agains other monitoring situations.juhu wrote: ↑Fri Apr 25, 2025 23:52 CESTSo to complete my investigation regarding crossfeed, I've found this Youtube clip that explains what CanOpener is really doing behind the scenes and even shows a plot of the mid-side EQ curve involved:juhu wrote: ↑Fri Apr 25, 2025 18:54 CESTReading your comment made me curious about Harman curves, so I tried to educate myself by googling a bit and then tried matching my MDR-7506 phones to a few variations of the Harman curve using https://autoeq.app/ - and I'm quite impressed about the results I got listening to the reference snippets they have provided there! I let it fit to a parametric EQ with 1x low shelf + 4x peaking bands + 1x high shelf and set a few constraints on a few peaking bands manually in order to get the best match. Now I need to also educate myself on the crossfeed / Can-Opener technique. I'll definitely try this combo approach on my next mix once I figure out the crossfeed as well, so thanks for the tip!Pentatonicdave wrote: ↑Fri Apr 25, 2025 12:20 CESTI use mid side eq based on can opener curve + custonm eq curve based on harman curve. it works wonders. in addition I use realphones.
https://www.youtube.com/watch?v=ktKD8SYGUiI&t=283s
Naturally, I then also replicated this CanOpener mid-side curve on top of the Harman-matching curve for my MDR-7506 (described in my previous post quoted above) and now mixes on phones (with matching curves enabled) do sound astonishingly close to how they sound on my HS80M's (with matching curves disabled)! I wasn't expecting to ever get this close tbh! Nice.
One additional note regarding the Harman compensation curve fitted using https://autoeq.app/ though: the Q-values provided by the curve fitter do not directly match to Logic's Channel EQ's understanding of Q-values (regardless the state of the Q-Couple toggle button) so I had to manually adjust the Q's in Logic visually to get the plotted curve as close as I could.