Levels, LUFS, SoundCloud, Mastering (WTF)

[aivoryuk]

The one thing I will add is that how the end loudness of your master is dependant on how balanced your mix is. If you're in Chris Lord Alge league than your mixes will stand up to a higher loudness level.
If you find that your mixes struggle while achieving a moderate level then either back off the master level or find the element in the mix that is causing your issue.

 

[Cirrus]

Right? It's pretty crazy that all a limiter is doing is putting clicks and pops into our music and he makes a great point about what is this doing to our society if this is in all of our music. Definitely something to ponder, and something that's made me consider how I finalize my own music or any other music I mix.

I feel like this is a perfect example of why null tests aren't as informative a thing as they seem.

Limiters limit peaks and are used as tools to bring up average volume. I know you know this, I'm not trying to patronise you here. They increase the RMS volume relative to peak volume, the subjective, artistic effect is to create a denser sound that's more squashy. They'll reduce the dynamic variation between song sections too.

If you do a null test against the same track without the limiter, volume matched, yes you get clicks and pops. But the limiter isn't adding clicks and pops, is it? It's doing what we know a limiter does - reacting to peaks above a threshold and squashing them down into the body of the sound. There are no clicks and pops until you conduct the null test and create a signal of clicks and pops that doesn't exist in the original.

 

[Will Chen]

@Will Chen I appreciate where you're coming from, however I encourage you to check out these two clips. This first vid is this song when I released it in 2020. LUFS was -9 or -10:



Not bad, but very compressed and the vocal really gets slammed as a result.

Now reference that to the remaster I recently did to shoot a little video for it. Other than the drum samples being improved, the main difference between this and the above track is that my LUFS-I for this was -13.8, so very close to the -14 target.



To my ear there's much more space between the instruments, the guitar has more high end, the vocals sit more comfortably in the mix, the beat strikes harder, and the bass drum / bass guitar have more thump. It's also perceptibly a touch louder than the first track as well when played through Youtube, though to your point Will, the .wav file for the first is perceptibly louder on my PC due to the perceived loudness and limiting.

Just food for thought, and I'm not saying I'm right by any means. 99.999999% of music listened to is streamed, and giving yourself some headroom can make a world of difference.

This is the output stats for the second video for reference:
View attachment 12498

Am I 100% doing it the right way? IDK, but I'm getting better results than slamming it as I did 3 years ago.

I think you should do what you feel is best. This "battle" has been gong on for decades really. To the 2nd video's low end being louder than the first. IMHO that potentially points to balance issues in your mix where the low end is louder than mids/highs. So when compressed/limited, it hits there first (just a guess, not going to run it through a spectral analyzer). There is a ton of R&B/Hip Hop content mastered to -10 to -8 with massive low end.

And I'm not saying I'm absolutely correct either, simply that I feel I shot too low based on being worried about the arbitrary -14 lufs limit. When I was mixing, I used Youtube as a reference and felt I was OK. However, when comparing my Spotify uploads IMHO I made a mistake. C'est la vie.

 

[Will Chen]

I feel like this is a perfect example of why null tests aren't as informative a thing as they seem.

Limiters limit peaks and are used as tools to bring up average volume. I know you know this, I'm not trying to patronise you here. They increase the RMS volume relative to peak volume, the subjective, artistic effect is to create a denser sound that's more squashy. They'll reduce the dynamic variation between song sections too.

If you do a null test against the same track without the limiter, volume matched, yes you get clicks and pops. But the limiter isn't adding clicks and pops, is it? It's doing what we know a limiter does - reacting to peaks above a threshold and squashing them down into the body of the sound. There are no clicks and pops until you conduct the null test and create a signal of clicks and pops that doesn't exist in the original.

Null tests are great and I agree with many (really most) of the concepts presented in that video, but I question the null test analysis he presents. A null test is purely mathematical and presents the differences between 2 files. One might think the result is the impact of the limited track, but that's not totally true. Since the peak info is essentially missing from the limited track, a null test would give you the peak info from the original track along with a pop or click as it's been separated from it's context. Those pops/clicks aren't necessarily caused by the limiter but as a result of the null test. That said, pretty common knowledge limiting will create some distortion as a result, which the null test should also deliver (and given we guitarists love distortion and saturation is big in the mixing/mastering community, not sure some distortion from limiting is having the negative impact on society he alludes to...)

 

[fretworn]

Yeah, that null test result didn’t seem right to me.

I concur

 

[Cirrus]

Yeah, that null test result didn’t seem right to me.

I concur

I mean, my hair does basically look *exactly* like that these days.

 

[marsonic]

Wow, so....there's a lot of weirdness in this thread.

Did anyone ever ask for links to the OP's tracks on different platforms? That would probably be a first-step, honestly.

Anyway....streaming playback levels are not targets and aren't supposed to be targets. Make it sound good and let them do what they're going to do - they can't hurt it just by turning it down. If it sounds better more limited, then it sounds better more limited. The key is to make the decision after you've level matched and to do it based on actually listening, not following blind guidelines.

FWIW, basically no mastering engineers or labels actually follow the guidelines in the rock, pop, or metal worlds. Yes, levels have been coming down a bit...but not that far.

There are a bunch of other subtleties going on that could impact general "sound quality". And peak level does have something to do with it, especially if platforms are doing SRC (some do when they need to, some leave it up to the computer) or data compression (which they all are). The order these things happen in matters.

As far as the limiter deltas sounding like cracks and pops....that generally implies that you're actually just clipping off transients and it sounds like cracks/pops because the amount of time over the threshold isn't long enough for you to really hear a pitch. It's not actually that instructive compared to level-matching before and after the limiter and then doing a blind listening test for which you prefer.