I was thinking about the 2 styles of mixing to the Lfe channel. Especially with all the errant masters with full range audio in the Lfe track sneaking out recently.
There are two styles of working with an Lfe channel in a mix.
Treating it simply as a discrete channel like any other audio channel vs. using it as an extension to add to the mix in the mains channels.
Anyone remember the original surround panner in Protools HD?
There was a dedicated Lfe send. The intended workflow was to add (ie duplicate) bass having sources to the Lfe for a bass "extension". Said source would still be directed to the mains channels per the panner. The Lfe send would also send it to the Lfe channel. That would duplicate the bass content going to the mains channels and also send that source to the Lfe per the send level.
Discrete mixing is more straightforward. No kludgey duplicating things. You route your mix elements to the channel/channels you please including the Lfe. You might only mix some bass element to the Lfe channel.
There's no frequency band restriction "built in". Again, you do what you please while being aware that the format states that the Lfe will not have any frequency content above 120Hz.
Obviously the send/extension method ends up sending full range audio to the Lfe channel when you do that! In this scenario it's just assumed that any playback system will restrict the Lfe to bass only. The subwoofer speaker itself can only reproduce so high and there is likely a crossover built into it for protection anyway. This is pretty fast and loose!
So... what happens on playback with different systems? Does this alter a mix from system to system?
Short answer is actually: Most of the time "no".
When mixing using the extension technique, your monitor system sub channel is going to be frequency restricted the same as anyone else's. At least grossly. You're duplicating sub bass that is also in some of the mains channels every time but you're still dialing in the mix to taste and you only get the sub bass reproduced by the Lfe channel speaker.
If the listener has the same setup with a 1:1 speaker array the mix is heard as intended.
If the listener has a small top speaker managed system redirecting the mains bass content to the Lfe, the Lfe still restricts the audio from the Lfe channel. The bass from the mains channels is added in and the mix is still heard as intended.
If the listener has a big mains, no sub system redirecting the Lfe channel content to the front L/R speaker... Oh oh! Unless the speaker management is low passing that Lfe channel content first, now we have a giant mix alteration if that Lfe channel has full range sound!
The discrete method has no worries. All cases of different speaker management end up with the same mix balance as intended.
At a glance, the extension method seems fast and loose and sloppy as illustrated. If the system has a 1:1 speaker array though, no one from mix engineer to end listener is the wiser. It simply works. Most subwoofers are going to be low passed with a crossover around 120Hz anyway. Yes, there could be alteration if you had a sub that put out something above 120Hz! But most cases still translate correctly. There's nothing inherently wrong with duplicating bass program in multiple speakers either. And again, it translates.
As evidenced by the older surround panners with the Lfe send (and you can still find panner plugins that have this feature today), this was intended to be a mix workflow style thinking of the sub as an extension. The discrete style is obviously fully intentional and bulletproof. It's just another channel. I'm mixing a source element to it and doing any low pass with intention myself.
Note, you could still be fast and loose and let your subwoofer crossover restrict the channel and send full range audio there! But you probably aren't if you are dialing up discrete things with intention.
I'm not in any way going to suggest the 'extension' method is wrong! It appears to be the first workflow style and the panner plugins with the Lfe send prove it. But I will suggest that this can really skew a mix if you have a speaker managed system and depending on the speaker management routing! If this was caught in mastering though, it would be a simple fix (low pass the Lfe channel at 120Hz) and now it translates to any system again.
I think our boy Steve Wilson is mixing on Protools and then no one caught the full range Lfe channel in mastering on a few releases!
If you are speaker managing the Lfe channel into your front L/R on a 'big mains' system, you might want to add a low pass eq at 120Hz on the Lfe to your speaker management. That would be insurance that any fast and loose 5.1 master with an errant full range Lfe track removed the unintended higher range audio there that was not heard in the studio and not intended to be there.
Have fun with that!