Page 1 of 2

Q&A - Mix Challenge - rule set and reference/target levels (split from MC35)

Posted: Sat Aug 19, 2017 00:50 CEST
by Tbase2000
 ⚠ Warning Message from Mister Fox  
This is a split topic from MC35. Let us please continue the discussion about "reference and target levels" in here. It can be more detailed this way.

Additional to this thread:
I will fine tune the "target level" topic with a rule book maintenance update prior to MC36 (e.g. simplify)


I had a little trouble understanding the loudness specifications.
I've done some reading and think I have the basics down. I just would like to be clear.
Here are two notable quotes from the reading:
1) "The mix challenge shall not exceed an average signal strength of:
-15dBFS (RMS meter) or +3 VU (VU meter with a reference level of -18dBFS)
The maximum peak can automatically be -3dBFS if we consider a dynamic range for transients of up to 12dB.
I do not see where this is confusing in any way."

2) "Probably down to -16LUFS (SLk, EBU R-128 specs), which is somewhat around -18dB RMS."

If your DAW allows you to change your master bus levels from 0 to -18dbFS reference then you make the
change and mix as usual trying not to run in the red I assume. I have no idea how you would
get a reading of -18db RMS if your max level is -18dbFS. Only your loudest signals would
ever reach -18db without peaking. Where would you even get an RMS reading from?

In my case I don't have the option to make the mixer conversion so I used MLoudnessAnalyser
to meter my mix. I mixed as usual then turned down the master with a gain plug-in to fit
the rules and adjust the MLoudnessAnalyser results since they are pre-fader.
There is nothing on that meter that says anything about -18dbFS. You can
set a target LUFS reading and it will show you your peak levels, as well as your average levels
but they don't say RMS. I found the second quote after I did my mix which said to
target -16LUFS. I set my target to -18LUFS because I didn't know what I was doing. Screen capture in my folder link.

I read through some of the online stuff you have but honestly it's like reading a calculus book. Most
of the online reading talks in LUFS as a target and not dbFS.
If your DAW doesn't have the capablility to speak in -18dbFS reference levels...how would you
use something like MLoudnessAnalyzer that only speaks in LUFS? What is your target LUFS? I'm
assuming -16LUFS from the second quote but I have 5 other mixers who have no idea how to get
the correct level for the contest.

Re: MIX CHALLENGE - MC35 August2017 - Submissions until 21-08-2017 11:59pm GMT+1/CEST

Posted: Sat Aug 19, 2017 01:36 CEST
by Mister Fox
I try to summarize and simplify. :educate:

We talk maximum signal strength (in dBFS or dBTP) and average signal strength (in dBFS RMS realtime/Dorrough specs or LUFS/ITU-R BS.1770-x specs). This is important so separate. And I tried to do that in the recommendations. RMS, VU and LUFS meters measure the average signal strength; PPM, digital and True Peak meters measure the maximum signal strengths. All within their limits / range of accuracy of course.


You should ideally shoot for (especially on the master bus):

Average Signal Strength:
Either -18dBFS RMS to -15dBFS RMS, or 0VU to +3VU

Maximum Signal Strength:
At these avg values it's even hard to reach -3dBFS max, but if you have such strong signal peaks, try not to exceed -1dBFS max


You usually mix with a reference level already (a lot of analog modeling plugins are using -18dBFS/0VU as reference). That means that your maximum signal strength can exceed up to -6dBFS (if not more), while your VU (especially lowend heavy snares) barely hovers around 0VU avg.

Clearer now?

Re: MIX CHALLENGE - MC35 August 2017 - Submissions until 21-08-2017 11:59pm GMT+1/CEST

Posted: Sat Aug 19, 2017 01:51 CEST
by Tbase2000
Not really. My DAW doesn't speak in -18dbFS/0VU or have RMS, LUFS average values. MLoudnessAnalyzer and others only talk in LUFS. Target Spotify, target YouTube, target Mix-Challenge. What's the Mix-Challenge target in LUFS?

Cubase 9 LE (Mini) I would not recommend it. 16 tracks max and no -18dbFS or LUFS option for monitoring. Sigh.

Re: MIX CHALLENGE - MC35 August 2017 - Submissions until 21-08-2017 11:59pm GMT+1/CEST

Posted: Sat Aug 19, 2017 02:53 CEST
by Beeble
This is very confusing stuff!
Loudness and signal strenght is like temperatur and distance. It can not be convertedto each other.
dBfs and dBu have nothing in common. The digital peak scale is not equivalent to the analog RMS scale.
You can never match dBFS and dBu.

I for myself understand maybe 50% of that stuff ^^

Q&A - Mix Challenge rule set and metering/reference levels

Posted: Sat Aug 19, 2017 04:41 CEST
by Mister Fox
This is a split topic from MC35.

Let us please continue the discussion in here.

Re: Q&A - Mix Challenge - rule set and metering/reference/target levels

Posted: Sat Aug 19, 2017 04:45 CEST
by Mister Fox
 ! Bonus Info
The following post might get a little bit technical. However it discusses "metering tools" and part of the Mix Challenge rule set. Apologies in advance


Tbase2000 wrote:
Sat Aug 19, 2017 01:51 CEST
Not really. My DAW doesn't speak in -18dbFS/0VU or have RMS, LUFS average values. MLoudnessAnalyzer and others only talk in LUFS.
Okay , first you need to get out of your head that "LUFS" is the ultimate answer to everything. LUFS is for mastering and broadcasting (stream distribution). It is a less error-ridden loudness (or average signal strength) measurement concept, no doubt. But if you're already struggling between Digital Max Peak (maximum signal strength) and RMS or VU (average signal strength), then LUFS meters will only confuse you more.

Sadly, MLoudnessAnalyzer doesn't add to understanding purposes either.


Tbase2000 wrote:
Sat Aug 19, 2017 01:51 CEST
Cubase 9 LE (Mini) I would not recommend it. 16 tracks max and no -18dbFS or LUFS option for monitoring. Sigh.
Personally I'm using Cubase Pro, so I'm unsure what LE/Elements offers in terms of metering tools. Though a quick check revealed, that the Mix Console lacks the "master meter" section from the so called Control Room. So I understand what your problem is. But I can offer a (hopefully) working solution (aside from recommending a host, that offers more than 16 channels!).


Solution 1 - Visual Aid:
Would be to setup some color codes for your host. You can do this via this entry in the Cubase 9 LE manual. Setup "Channel Meter" to have: green from -inf to -18dB, yellow from -18dB to -6dB, and red from -6dB to +inf. For the "Master Meter", set up the same but change the red-zone to -3dB to +inf


Solution 2 - Proper metering tools:
Ones I can recommend and are affordable:

For VU's: SleepyTimeDSP - StereoChannel (discontinued, find it on Bedroom Producers Blog, Free), Klanghelm VUMT2 (standard version is enough, costs 14EUR, offers VU, PPM, RMS), alternatively also PSP Audio TripleMeter and zplane's excellent PPMulator. I do NOT recommend TBProAudio's meter, as it's single UI mode is unusable for VU usages with stereo channels especially (does some weird "summing" of the signal). And while Hornet Plugins VUMeter is an excellent concept, the measurement time frame is a bit wonky and often reads lower vwlus than e.g. VUMT or PPMulator would show.

For RMS meters: an alternative to the (IMO overpriced) Waves Dorrough is actually Martin Zuther's K-Meter in Normal Scale/RMS mode. Or alternatively in K-14/RMS mode (same behavior, shifted scale). It's free, it's a bit picky with certain hosts. But for this price, it can't be beat

For LUFS meters: here is a list over on KVR Audio



:arrow: But back to your host for a moment...

The meters your host is using as default, are digital peak meters. So... maximum signal strength. If you only have access to this type of meter, try not to exceed -1dBFS as absolute maximum on the sum. To get a better overview of what your signal is actually at (perceived loudness), I do recommend to use a metering tool like a VU, a RMS meter or (if you really want/need to) an LUFS meter (I recommend the SLk ballistics) to check for your average signal strength.

A combination of both a digital peak meter, and a VU will get you really, really far!


Tbase2000 wrote:
Sat Aug 19, 2017 01:51 CEST
Target Spotify, target YouTube, target Mix-Challenge. What's the Mix-Challenge target in LUFS?
Spotify currently uses -14LUFS, Youtube anything between -13LUFS and -11LUFS. If you want to use an ITU-R meter (EBU R-128 meters are just "presets" of those), try to shoot for -16LUFS avg (+-3LUFS, SLk) for the Mix Challenge.

Remember - the focus is on mixing!


Beeble wrote:
Sat Aug 19, 2017 02:53 CEST
This is very confusing stuff!
It wouldn't be, if users would explcitly declare what they mean by dBFS vs dBFS for example.


Beeble wrote:
Sat Aug 19, 2017 02:53 CEST
Loudness and signal strenght is like temperatur and distance. It can not be convertedto each other.
dBfs and dBu have nothing in common. The digital peak scale is not equivalent to the analog RMS scale.
You can never match dBFS and dBu.
Technically absolutely correct (see Sengpiel Audio), practically the lines have blurred for years at this point.


Loudness "inside" the PC these days means "average signal strength" or "perceived loudness". Actual "Loudness" (as in, environment noise, or a "sound pressure level") is measured differently, and it's either measured in Sone or dB SPL (dB a-weighted or c-weighted).

LUFS, RMS, VU (inside the box)... are all based around dBFS (sample accurate). The "type" of meter is however declared by the ballistics (inertia/rise&fall), the reference point (e.g. -18dB) and the weighting filter (VU's and RMS realtime meters are z-weighted/unweighted, LUFS meters are k-weighted). To distinguish the meters, you then use the appropriate abbreviation (PPM, VU, RMS, LUFS/LKFS, etc).

You can create a "somewhat relation", but this would go super technical, and you need proper test tools to set that up. So let's not go there.


Beeble wrote:
Sat Aug 19, 2017 02:53 CEST
I for myself understand maybe 50% of that stuff ^^
It's a science in itself, no doubt. I've been doing this for years at this point (see my various posts over on KVR Audio), so it looks easy for me. I never got around creating a proper tutorial series on that subject though. Doesn't mean that it's a forgotten topic. And I still try to simplify things for everyone.

Re: MIX CHALLENGE - MC35 August2017 - Submissions until 21-08-2017 11:59pm GMT+1/CEST

Posted: Sat Aug 19, 2017 05:09 CEST
by Tbase2000
Thank you very much for that explanation. Cheers!

Re: MIX CHALLENGE - MC35 August2017 - Submissions until 21-08-2017 11:59pm GMT+1/CEST

Posted: Sat Aug 19, 2017 10:38 CEST
by Mister Fox
Tbase2000 wrote:
Sat Aug 19, 2017 05:09 CEST
Thank you very much for that explanation. Cheers!
If there a re further questions, just ask and I try to get those sorted out for you.

Other than that... I do recommend to at least upgrade one Cubase version, as this would enable you to use more than 16 Audio Channels (I just saw that LE got drastically limited on that behalf, though AI with 32 tracks isn't any better - unless you can still get AI/LE5, which is drastically outdated!), or either get Cockos Reaper, Presonus StudioOne Artist (as this can use third party VSTs) or similar budget priced hosts. Especially if you want to enjoy more of upcoming challenges.

Re: Q&A - Mix Challenge - rule set and reference/target levels (split from MC35)

Posted: Sat Nov 06, 2021 18:04 CET
by AN/DI P&R
Hey there. Currently working on MC082, this'll be my first mix challenge on this site.

Quick question about 'undershooting -24 lufs integrated'

The start of my mix sits under this for a few bars, there's only a choir and a flute during this, will this be ok or will that disqualify my entry? The rest of the mix ends up at -18 lufs integrated fyi.

Thanks in advance!

Re: Q&A - Mix Challenge - rule set and reference/target levels (split from MC35)

Posted: Sun Nov 07, 2021 05:01 CET
by Mister Fox
Welcome to the forum!

AN/DI P&R wrote:
Sat Nov 06, 2021 18:04 CET
Quick question about 'undershooting -24 lufs integrated'
"LUFS Integrated" means, that this is a value for "the whole program stream". Or in our case (mixing challenge), for the mix from start to finish.

You don't analyze individual sections of the song like with old RMS avg or the K-System (v1), but the whole song. And since the other values (SLk / Short Term Loudness and MLk / Momentary Loudness, both k-weighted) can vary depending on the density of the mix, the "Integrated" value for the whole mix provides a more "neutral" and definite readout of the loudness / program stream.


So if you run your finished edit through either available loudness analysis tool like Youlean Loudness Meter 2, NUGEN Audio MasterCheck or VisLM, Izotope Insight, zplane Fennek, Melda Production MLoudness, Tokyo Dawn Labs Limiter 6's metering section, etc, then the Integrated Loudness value (ILk) for the whole mix should be between -24,0 LUFS (absolute lowest) and -16,0 LUFS ILk (absolute max).

I am using Wavelab 10 at the end of each game to check all entries. It's analysis is using the EBU R128 specs. So to get similar readouts, I highly recommend using a EBU R128 or ITU-R BS.1770-4 preset in your go-to analysis tool.