I use the new version of TT DR Meter (which is now managed by MAAT; DRMeter Mk II) along with Ian Shepherd's Dynameter and Voxengo SPAN (for frequency analysis obviously but provides some other useful onscreen info e.g. Phase correlation); the 3 must have's for me. Also the Default Wavelab meters. And occasionally NuGen MasterCheck (really useful for monitoring different codecs and True Peaks via these codecs). I also actually like having a small Klanghelm VUMT sitting in the corner too - just feels... comfortably familiar and dirt cheap!

 

Izotope tonal balance control (not a controller),
In combination with tdr limiter history graph.

 

- TT Meter plugin
- Voxengo SPAN (speccially the K-meter, I use K14 for my electronic music).
- Izotope Ozone (it has its own Loudness meters in LUFS)

I also have some albums that I take as Loudness/Dynamics reference, even if the reference is of a different genre from what I am producing. One album is for example "Protection" by Massive Attack.

 

Is there a connection between LUFS and "DR"? No! For the simple reason that "DR" is not a part of LUFS measuring! And this is nothing but the truth (afaik).
Can you create a connection between these? Yes, you can. Under the prerequisite of a given track (or mainout in the mixer) and the aim to affect the LUFS value with a dynamic module like a compressor or a limiter, you can create this connection.
But I never made this prerequisite. I was talking about the theory, the basic technology and, as you will see, the independency of these two in the state of the mix.
I made these two measurings within the same project, just by changing some parts (not exchanging).

Spoiler: LUFS and "DR"

https://drive.google.com/open?id=1S-B9v0s33sNMv28K4w16-YctHr9TxPbv

I accidentally even reduced the "DR" while lowering the RMS and thus accordingly the LUFS for ~ 10dB. If there would be a direct connection, this would not be possible.

 

I don't measure my DR. However, I measure my LRA. If my LRA is too large I will tame it with a compressor (usually slow attack and moderate to slow release times.) Once my LRA is at a point I am happy with, I will begin working towards gaining loudness (LUFS.) I use clipping and limiting, wideband or multiband.

Sometimes I'm comfortable with an LRA of 2.5 - 3 LU's. This is the absolute extreme and dependant on material.

Since doing my little experiments mentioned here in this Post, I bring my Maximum Momentary Loudness value up to -8 dBFS. I find that the amount of clipping and/or limiting involved to reach that level is totally fine when using quality clippers/limiter(s.)

As a general LRA, I aim for about 6 LU's... Sometimes as high as 8 LU's, and as low as 2.5 - 3 LU's. DR, although I do not monitor it, I believe can be "perceivably" maintained by using slow attack times on compressors and not exceeding -8 dBFS when clipping/limiting. Whatever the final LUFS Integrated measurement is dependent on the material.

For example, while clipping/limiting at -8 dBFS (M), you could end up at -10 LUFS (I). Or you could end up at -14 LUFS (I). Depends on the material.

 

Last edited by a moderator: Jun 18, 2018

Far out! It's a real challenge, putting knowledge into words. I have not done this much. I need to get better at it.

 

Would you say that Momentary Maximum and Short Term values are kind-of-like the new RMS values? Not literally, of course. Or, perhaps more like VU values?

What gets me is the integration times...

300 ms for RMS and VU.
400 ms for Momentary.
2000 ms for RMS.
3000 ms for Short-Term.

I kinda view Loudness Measurements as the newer up-to-date method of measuring audio. As if it is the all-in-one, go to, way of measuring audio. True Peak is obviously included but is also the least important value (generally speaking, in music,) when it comes to loudness.

 

@LeviD Since you can measure RMS in real time and offline, I would say real time RMS is more close to momentary and offline more to integrated.
But where do you got this 2000ms for RMS from?

I agree, LUFS is the more up to date version of loudness measuring, but please, have some indulgence with the old guys.

 

Haha

I actually pulled 2000ms from the K-Metering inside of FabFilter Pro L-1&2. It uses a window of 2000ms. It really is null and void though when it comes to Integration values. As is short-term in loudness metering, because the integrated value ends up being the same whether using smaller windows (1 second, 2 seconds, etc) or larger windows (10 seconds.)

 

Last edited by a moderator: Jun 19, 2018

From my experience, both -8 LUFS and -14 LUFS have their own purposes, and generally which level I will master a track too depends entirely on the genre of the song.

It really comes down to the loudness war (a debate that will never end).

For a while I thought quietness had won the war when Spotify decided that ALL of their songs will be adjusted to -14 LUFS as part of the upload process. While I agree that this was a smart decision on their part, I also agree with Soundclouds choice of not touching the loudness at all.

When something has a certain loudness, that loudness is actually somewhat arbitrary, because it's not hard to just turn up the volume yourself if you deem it not loud enough (regardless of what the track was mixed too)...

So whether something is loud or quite, it doesn't actually matter... in terms of perceived loudness...

So why is it important then?

It is important, because the level that you set your song at will determine your dynamics, which is easily summed up as the distance between your loudest and quietest sound. When you watch a movie, the levels that they master at are somewhere below -20 LUFS. The reason for this is because they want a "whisper" to sound like a whisper, and an "explosion" to sound like an explosion. -20 LUFS is an AVERAGE loudness, meaning that -20 LUFS is more like a max and min of 0dBs and -40dBs

If a song has an average LUFS of -8 dB, then that means that they are somewhat limited to an average minimum of -16 dB as their quietest sound.

When Spotify takes every song and re-adjusts them, what they are doing is simply creating a uniform dynamics for artists to work with, and the loudness that they've chosen (-14) is a loudness that leaves a lot of headroom to play with in terms of quiet sounds and loud sounds.

Songs that will sound MUCH better on Spotify are songs that benefit greatly from having a high dynamic range, these songs include all of your alternative, rock, reggae, pop, jazz, bluegrass, country, etc... Basically any band that is dealing with live instruments will benefit more from -14 LUFS, because they can create a song that will sound more like a live performance, as opposed to being only limited to being loud...

-14 dB is a good level to master too especially if you intend to put it on spotify, however, if you are making EDM style music then I highly recommend that you mix to -8 LUFS. The reason for this is something a LOT of studios overlook.

It actually has nothing to do with dynamics. At this level, your dynamics are very limited, but the biggest (and probably only) benefit to -8 LUFS is that it creates a standard loudness for kicks, snares, and sub-sonic frequencies which is huge...

When you have a song with a massive dynamic range, the biggest problems that you will usually run into is setting the levels for your subwoofers, because if you don't have them set right, and you accidentally set the levels to a "quiet" scene, then an explosion of bass could possibly blow your sub, or at least scare the hell out of you when it clips.

When different songs are being played back to back that are HEAVY in sub frequencies, you can't possibly adjust levels for each song. You can turn down your bass and let the louder songs be loud and the quieter songs be quiet, but for an EDM show, you want the bass from every song to shake your body (without blowing your sub, or having to deal with that aweful clipping sound.) So if all kicks, snares, and basses have a small dynamic window to fluctuate in between, the higher the level that you can set your subwoofer at and not worry about turning it down due to distortion.

This issue is irrelevant to live performing bands, because you can't really "distort" or "clip" the levels of a drum set being played live. -8 LUFS is really only relevant to electronic styles of music, or music that can't really be performed live (except for vocals). When I DJ, I take all of my tracks, turn them down a tad to the same level (usually -3 dB), and then I play them through Ableton Live 10, but I put an Ozone 8 Maximizer on the master, and set the "true peak" limiting on, so that I can turn my subwoofer all the way up, and not worry about it. If a song was "quieter" than the rest, then I'd turn it up past -3 dB, and mash it against the ceiling until the bass levels where good.

I only started doing this after I kept running into occassional songs that had weird dynamic ranges that caused the bass levels to change part way through the songs, causing subwoofer issues...

anyways, this is way too long, but for whoever takes the time to read it, I hope it helps you, and good luck

 

I was reading this post out of interest. Can't help but shamelessly plug ours, we released the Bute Loudness Analyser a few months ago : www.signumaudio.com/bute-loudness-analyser

 

I'm fine with shameless proselytizing as so long as it's a relevant product, and I hadn't heard of Bute before. I grabbed the trial version and it is a nice plugin, but, honestly, the price is pretty steep, especially considering I still have companies/individuals who I have on my Pay-Them-When-I-Can list. Thanks for the heads-up though.

 

You can use/do whatever you want to your "music" so that it sounds good to your ear... but ultimately, if you want to broadcast your audio, it will need to conform to broadcasting standards. You WILL NOT have a choice... that is, other than just walk away from the opportunity.

It is done for a simple reason... to maintain control and balance and ensure that everything is equal to one another when played through small tv speakers or even radios. Imagine how uneven tv would sound without that standard of measure... especially from amateurs that THINK they know what they're doing...

 

Sorry for the super late reply, let's say already Kraftwerk's Computer Welt and Electric Cafe albums. Original masters btw.
I spent a little time studying Computer Welt with a few basics analysers, and noticed a few interesting values, but i'm not sure nor about my Flac version, nor about the visualizers, but esp. nor about my knowledges. I've always somewhat dreamed someone who would point me all the values to watch and match.

 

What do you mean?

 

I'm not sure if the Flac is the original master or some remastered version. I'm not sure about visualizers, as they tend to give different values.

 

Yeah, I understood that, I meant in detail, examples.

 

Last edited: Jul 29, 2018