-
Mastering process question
Geez, it’s been quiet around here lately. I’ll throw up a question that plagued me last night. I’m mastering a song for my son’s band and ran across an issue to figure out.
We have a 48K, 24 bit, wav file at -8LUFS with a -1dB True Peak value. When I opened up the mastered file in RX11 I could see a bunch of energy between10 and 20 Hz that I couldn’t hear in headphones. That’s wasted energy even if you’ve got subs. I figured if I attenuated that energy I might could squeeze some more loudness level. The intended result of applying multiple passes of fading out below 20Hz was successful in lowering the energy in the basement. It looked better.
The unintended result was surprising, they usually are. Like attempting to force lower oil prices by invading and nationalizing an oil rich country’s resources but causing a global realignment of power, backward thinking leaders to chaos, and quickly World War 3.
Fortunately, the unintended consequence I experienced was less fatal. At some point in the multiple passes of EQ ing out that low, low end, my LUFS level changed. How can it measure louder if I’m consciously taking away some of the level? The Google machine’s response was that my low end EQ created a phase response change leading to an asymmetric waveform that reads higher.
If I were to refine my question down to is core, it becomes this… Would using a flat-phase brick-wall EQ below 20 Hz allow me to accomplish my goal of minimizing that unusable energy but NOT induce a phase change nor an asymmetric waveform and therefore avoid being penalized by the LUFS police?
New year, new questions.
-PT
-
4 Replies
-
-
Late-night tutorial-ramble response incoming!
#PT4Prez
💜
-
Dana,
Thank you for the late night ramble. Keen observations as we have come to expect here.
I had in fact, thrown a 48dB/oct high pass filter across my L/R buss as the first process in my mastering chain. My ears said 51 Hz was the spot to park the filter and do no damage. I trust my calibrated headphones. I then took that mastered file into RX 11 to have a look if any of its processes would help. A little de-clicking cleaned up some remnants of what I believe to be the artifacts of limiting. Visually, there was certainly some energy below 20Hz but audible? No.
Nonetheless I persisted and ran multiple passes of the 20Hz and below fade out process to effectively lower the amplitude which caused the LUFS level to change for the worse and sparking the whole conversation.
Part of my brain says to try doubling up on the steep high-pass filter but set the second one down at 20 or 30 Hz to see if that also causes a slight LUFS level change. Another part of my brain then said “Fuck the meter, trust your ears!” My remaining brain cells tells me that maybe we should tamp down on this #PT4prez thing. I have some unresolved legal issues in Arizona, there’s no sense in drawing attention, ya know? Like Obi-Wan said, “I’d like to avoid any Imperial entanglements.”
Certainly appreciate the effort and insights. I’ll get back to you after further experimentation.
-
Hahaha – Obi-Wan quote for the win. We defs don’t want to lose you to Imperial entaglements!
Huh, that is bizarre you’re seeing energy below your cutoff, post-processing. Is there anything else in your signal path between the 48dB/Oct hi-pass and the limiter? Just curious if there’s some kind of ‘analog modeling’ tape or console plugin that could be re-introducing some “authentic analog artifacts” to the frequency spectrum, post-EQ, pre-limiter.
If all else fails, yeah just use those sweet ears of yours, follow your gut, and don’t stress the meters!
-
-
Log in to reply.
