FabFilter

I'm sharing this incase it provides helpful insight for others.

I set ffprol2 on the master track following the UA Fairchild 670 plugin, with no other plugins.

I set my track gains (currently vocal and guitar) to achieve some FC670 compression at its default settings.

Between the tracks and FC670 I target -1dbtp with the output gain and 1dbfs input gain and I target -14lufs. If I'm peaking too much but suffer low lufs, I craft track compression just enough to attenuate the peaks, which allows more lufs. And I'm done with the master track.

For low latency I record and perform using the Transparent algorithm with tp limiting 0.1ms look ahead and 4x over sample. When exporting I can try newer algorithms and bump 4x to 8x.

I saved this as default.

For my audio track inputs I added ffprol2 first on each track (in a BlueCat PatchWork rack) to limit output to -7dbtp with input gain at 7dbfs.

I use tp limiting, transparent algorithm, 0.1ms look ahead and a custom lufs target of -21.

Then I set my interface gains to achieve around the lufs target level, which also establishes input peaks around -5dbfs which are caught by the ffprol2 "track" preset.

I let my UA Oxide Tape plugin meter (at default settings) choose the -7 dbtp, as it's strong but doesn't slam the meter. This is confirmed by the UA Century Tube Channel strip inputs (also set at default).

I confirmed the ffprol2 -7 dbtp output levels with a default ffproq3. Then I removed the ffproq3.

Ableton's peak readings match those of the ffprol2 within allowance for the true peak limiting.

This means all of my song projects will share a standardized lufs/tp foundation.

I've found no exceptions. Ffprol2 values match Ableton values as expected, if I inverse match the input and output gains.

Today I'll begin pushing my Pro L2 as my analog track input VU/leveler for my vocal and guitar track.

Last recording went well (But better is always better, right?) My UA Volt476P tends to 'warm up' getting stronger to a point al the way up to -2.5dbfs peaks. My initial -7dbtp could be stronger in my opinion so I plan to try two things at the track chain level... 1. bump up the TP -7dbtp output and 2. bump up the LUFS from -21dbfs to something more. I'm looking to thicken/saturate the mix and push my vocal up into it using my UA Century Tube and Oxide Tape at default levels.

So I will be relying on the Pro L2s at the analog tracks to perform some magic and I'll post how it turns out here. Wish me luck.

Correction, three things ["1. bump up the TP -7dbtp output and 2. bump up the LUFS from -21dbfs to something more"] and push the input gains at track and master levels reasonably above output level settings.

Tonight's findings for singer-songwriters like myself...

I set the track Pro-L2 at Transparent with a 0.2ms look ahead; gain at 8dbfs, output at -6dbfs; turned off TP all together; LUFS target at -14. It doesn't typically get there on vocal or guitar track. To do so would crush any dynamic range.

As before I verified the output with a default Pro Q3 then removed it. At -6dbfs I don't need to worry about true peak clipping.

O/L lights on my UA Century Tube channel strip and UA Galaxy Tape delay blink in loud passages which is acceptable for those plugins.

Sounds good and Track outputs aren't reaching 0dbfs (as set through the chains). Track inputs are liking the 8dbfs gain bump and don't trigger limiting too often.

I'm only using the Pro-Q3 on the guitar track to duck the 250hz high shelf 1db for the vocal track.

The Pro-L2 Master is presently set at a gain of 2; output limit at -1dbtp; TP enabled; Transparent with a 0.1ms look ahead and -14dbfs LUFS target which it's reaching.

Overall I'm getting a strong sound envelope free of transient spikes. If I'm not done I'm close to it.

Also... no oversampling on the tracks and 4x with 8x offline for exporting on the master track.