Thoughts/questions about Fabfilter Pro L2

Marcel

I am new to fabfilter Pro L and L2, I bought L2 a few days ago.

I have two issues with it:

Short version:
1. What is the point in using L2 when you master a track that has a high audio volume level to start with?
2. Any thoughts about the -14 LUFS level standard? To me it feels too low.

Same questions, longer version:
1. If the volume level of the track you are mastering is at or higher than the target level in L2, is it wise to lower the input level? For instance by lowering the volume slider of the track in Auria Pro. If you don’t lower the input level there seems to be no point in using L2, makes me wonder: why use L2 when audio input level is at or above target level, what benefit does L2 have? Is there some compression quality in it that raises the quality of the music? Does it smooth out unwanted sounds even when gain is set to 0 (while audio level is at target level)?
2. In the online documentation I think there is a general consensus about a target level of -14 LUFS for streaming, for instance SoundCLoud. To me, when I listen to that level, it sounds too low. Any thoughts about it? When I put something on SoundCloud the volume level of it is way higher than – 14 LUFS, and to me it feels a bit weird to conform to the -14 LUFS norm. Thoughts?

jwmmakerofmusic

1. Yep, I usually do this via an EQ like Pro-Q, even if I don’t actually use the EQ. I just lower the output so as to have better control of the signal. See the answer to the second question for answers about “compression quality” and the like.

2. -14LuFS is the ultimate solution to the loudness wars. Instead of trying to squash everything, a loudness standard is introduced instead. Every track on Spotify has been set to play back at this level, no matter how much a track was squashed in the original mastering stage. Wouldn’t it sound more pleasing to the ear to simply add some multiband compression to glue a mix together, EQ it to balance it out, and bring it to a pleasing -14LuFS without having to squash the fuck out of it with a maximiser? Oh, sometimes the occasional peak will still try to pop through above unity gain, but that’s where L2’s limiting function should come in - that is, rarely.

Now, to answer some questions from the first question - if you really wanted to squash your track to, say, meet the insane -3dB RMS levels that Spinnin regularly imposes on its releases, L2 can do that transparently to a fault. However, despite little to no distortion, you’re still sacrificing how dynamic your track will end up sounding. Unfortunately, I sometimes have to use two limiters on the master track - one set to -14LuFS, and one set to -3RMS. I always render my -14LuFS versions, but if I DO have to begrudgingly send a “loud master”, I simply switch one limiter off and the other on.

Not trying to be “Sanctimonious Sam”, but I know what aesthetic is pleasing to my ears versus what the industry still at war with loudness demands.

richardyot

Just in case you haven't seen my video on Pro-L2:

To answer your questions:

If you're aiming for -14LUFS then you are likely to need to lower the input levels, not on the Master volume though (because that is post-insert) but on the channels, groups or busses. Much more efficient to use the groups as per my video.

On most tracks even with the relatively quiet target of -14LUFS there will still be some transients that go above unity, and the job of the limiter is to stop that from happening, so Pro-L2 will be doing its job by catching those transients and bringing them down so they don't clip.

Not all streaming platforms normalise tracks to -14LUFS, SoundCloud being the case in point. On SC a track mastered to -14LUFS will sound quieter than other tracks. The real benefit only happens on platforms that do target -14LUFS (YouTube, Spotify, iTunes): on those platforms tracks that are mastered to be "loud" will be played back quieter so that they conform to the standard. The reason those streaming platforms do this is to avoid sudden jumps in volume from one track to another when people are shuffling playlists. So if you are just mastering for SoundCloud it's OK to master somewhat louder, but on on all the mainstream platforms it's a losing strategy because it will make your tracks sound both quiet and un-dynamic. Hope that makes sense.

Tarekith

What they said

Marcel

@jwmmakerofmusic and @richardyot thanks a lot for your answers! Have read them fast, but need more time to understand them fully, will do that later. Thanks!

Marcel

Hi @richardyot, Richard, Thanks for you answer again, it makes sense, and thanks for your specific comment about Soundcloud (and other platform) levels.
I did watch your vid indeed, but that was before I bought L2, so it didn’t make much sense at the time, but now that I’ve worked with it a little it is a helpful vid!
A question arises in my head after watching your vid: what if you have a track with obvious soft parts and loud parts....the average won’t be a very good compass then to work with, cause the loud parts probably are then subject to too much limiting. Answering the question myself: perhaps the best thing to do then is to take the average of a loud part.......good idea? I guess using the average of a whole track as a starting point for setting the gain right only works when there is not too much variation of soft/loud parts within a track.

richardyot

@Marcel said:
Hi @richardyot, Richard, Thanks for you answer again, it makes sense, and thanks for your specific comment about Soundcloud (and other platform) levels.
I did watch your vid indeed, but that was before I bought L2, so it didn’t make much sense at the time, but now that I’ve worked with it a little it is a helpful vid!
A question arises in my head after watching your vid: what if you have a track with obvious soft parts and loud parts....the average won’t be a very good compass then to work with, cause the loud parts probably are then subject to too much limiting. Answering the question myself: perhaps the best thing to do then is to take the average of a loud part.......good idea? I guess using the average of a whole track as a starting point for setting the gain right only works when there is not too much variation of soft/loud parts within a track.

I'm pretty sure the streaming platforms calculate the average for the whole track, so that's the best way to approach it. Besides if you have dynamics (quiet and loud parts of the song) surely you want to preserve them? Otherwise deal with them at the mixing stage.

Tarekith

LUFS does indeed calcuate and compensate for loud and quiet parts in it's calculations, including how long or short they are. That's what the "integrated" means when people say LUFS is an integrated measurement. That's also one reason you need to play the entire track start to finish if you want an accurate LUFS reading of it. Some desktop tools like Myriad and Triumph will calculate this for you offline to speed things up.

richardyot

@Tarekith said:
LUFS does indeed calcuate and compensate for loud and quiet parts in it's calculations, including how long or short they are. That's what the "integrated" means when people say LUFS is an integrated measurement. That's also one reason you need to play the entire track start to finish if you want an accurate LUFS reading of it. Some desktop tools like Myriad and Triumph will calculate this for you offline to speed things up.

Thanks for this additional clarification - super useful.

ImortalTrashcan

@richardyot said:
Just in case you haven't seen my video on Pro-L2:

Very informative video, thanks!

[Deleted User]

Thanks all for the info! Video and extra info is very good.