When do you Low Pass?

For me, coming from a live background and dealing in less than ideal rooms it’s mostly about problem solving and keeping crap out, no real theory just grab it sweep it and try to keep as much air and harmonic content while keeping crap out. Just like lowend content background HF noise can eat up headroom during tracking so you need to be careful, also remember analog and analog like filters are going to create phase issues so you always need to juggle that against the frequency based cut.

For live recordings, done with utility preamps, I like to lowpass at 12k. I once read a forum discussion or article (can't remember) that explained how due to the RIAA curve, most any music released before the digital era would be lowpassed around 12k. I tried it and I liked it!

I know this is not set in stone and I also haven't explained it really well, so if anyone can chime in with more accurate information, I'd like to hear it. The thing is, it worked for my tracks.

I was able to remember and track the info, it's on this forum thread, posts by "Yep":
forum.cockos.com/showthread.php?t=29283&page=13

BTW, I don't know who that guy is, but that whole thread is amazing.

I think you mean LP..

But I agree with all these frequencies for the most part for low passing. 

An MP3 is a downgrade well aligned cassette without the wow and flutter. But remember when there was a short lived time when they were pushing micro- cassette as a hifi format 🤮

Whoa.  Yeah that's filled with some BS..

I could talk about analog filters all day.. It's kinda what I do, so I know a lot about the pros and cons of different types, etc..  And what he says is pseudo-informed but not really useful beyond academic discussion or during design.  He is right that the anti-aliasing filter can hurt your audio if it's not designed well, and that a 44.1K recording *can* sound as good as a 96K recording if the filters are fucked up in the 96k version..  But for most designs that's just not a valid concern as the higher sampling rate will still enjoy higher precision reconstruction inside the filter passband, aka: "details".

And no, an external clock is not necessary.  There is no need for synchronous sampling between tracks nor between converters these days.  Maybe if you're running timecode for large format projects across multiple types of recording equipment it might be useful, but for ITB work, it's just no longer necessary.  Especially if you consider that physical trace delays for distant converter ICs on a PCB could be more than sample differences at higher rates, you'll probably never see true sample accuracy between I/O anyway.  

He is right that any internal clock is better than an external clock though. 

Hmm.. this has me thinking I might try a few of the suggestions Ward mentioned.

Svart, I've heard about internal clocks being better than any external clock, but when I've heard the new Black Lion Audio clock, it gave a wider soundstage that was more 3D than my Apollo 8. Why would that be?

Ok, I'm sure there's still reasons to use external clocks to get around issues like yours if you haven't been able to figure out the underlying cause, but for the most part I was talking about sample alignment.  Most folks talking about using external clocks think they need to have all track sampling perfectly synchronized for some reason, which is just not the case these days.

I use the Abbey Road trick using a HPF at 600Hz and LPF at 10K for my vocal plate or chamber Aux. If you cut everything, there is very little left. So, I tweak for what is pleasing. I use a 30HZ HPF on my GML 8200 on my mix bus. I use filters on other things too, but have no set frequency spots.

Maybe.  

I'm talking about the reasons..  People tend to think that they're going to get better fidelity with external clocking, and that synchronizing all external sources will end up with higher fidelity.  They confuse fixing the pops/clicks with getting a better result..

Now, for the pops and clicks, that's a matter of how devices are set for master/slave and how each device can sync and reclock to each input, etc.  

This is a topic that is HUGELY misunderstood.

I recently got an album to mix, Upon listening, I though it horribly dull sounding. I dropped it through a silver bullet to add a bit of "air" back in. A couple clicks later - nothing. I added more, and then more and then,,,,still nothing.  There was virtually nothing above 14-15k, no matter how aggressive I got up there.

I called the mix engineer - he's 30-ish, and asked him if he had anything on the 2 Bus cause there was no HF on the track. He said "no, I removed all my plugins off the 2 Bus". I asked him if there were any LPF's involved, and he said "OF COURSE! I did like TLA suggested and LPF'd EVERYTHING at 14-15k to make it sound like TAPE. And I tracked it that way so I can't remove it.". I almost gagged. I told him my tape decks - when I had them - went upwards of 40K. He felt embarrassed. I wanted to strangle anyone with the initials TLA. It took heroic efforts to save that CD.

Waaaaay back in the digital console days - I had a visit from the Japanese engineers of the Ramsa digital boards which were - at that time - getting rave reviews for their "analog-like" sound. They were interested in seeing my studio and how I worked, and spent a lot of time taking pictures of the backs of my racks. LOL Anyway, I asked them the secret of the Ramsa digital boards and they told me though their translators - LPFiltering. They had a setting on every channel where they were filtering out HF. It was set, and non-removable. But here is what they told me. and I took it to heart, and it's worked very well for me over the years :

Start with the HPF wide open. Solo the track with good monitors. Slowly bring down the LPF until you hear a bit of HF disappearing. Nudge it back up a tad, and move on to the next source. IT WORKS. Digititus gets beat down pretty quick with this approach. It's not creative, and it doesn't make broad mix changing moves, but it cleans out the un-necessary HF content just like a HPF does, albeit more gently and with a less obvious imprint. THEN, I'll go back and listen to the entire mix in context and make more serious cuts on stuff that can sustain it like Bass, kick, some LF synths, etc..

It works, it's cheap, it's kind of a PITA, but it will take a lot of the digital zing out of your mixes. I have no "SET" freq.'s that I cut at. IMO, that's a rule that doesn't need to be made. Use your ears if you still have HF hearing.

If you're working with hardware and doing this while tracking, I would applaud that.  A couple of good options are the Missing Link and the Chop Shop.  The Chop Shop being my favorite for it's  resonance and slope options, and it's more tweakable - yet still simple like the ML interface.

So, which HPF's and LPF's do you guys prefer?

I just use Logic's track EQ and roll off lows. I haven't tried rolling of highs though, and look forward to trying that to help with my digititus. Occasionally on drums I use a Pullet or API plug.