Mix Challenge

Al B. wrote: ↑

Sat Sep 22, 2018 23:46 CEST

I like the R1 mix of Marc Clement. His mix sounds very clear, loud and in the face. He tamed the voice very well. I would be very interested to know exactly how he made his vocal sound.

Thanks Ali.B,

I had a look today and here is the vocal chain

VOCAL CHAIN

Dbx 160 75% wet about 10db max reduction

SSl G.... 4db at 2.37k, -1.4db at 450hz. tickle with comp
Slate SSl, I prefert the high end...1 db high shelf at 1.9k
Slate fg401 75% wet -3 db reduction max
de esser

Vocal Stereo Group

Eventide H910 9% wet
Abby RD chamber 12% wet
Soundtoys Microshift 10 % ish
Melda dynamic eq for the saturation

Sent to rear bus

Hope that helps

Marc Clement

Hello Mister Fox,

Would it be possible to have a link that includes all the songs of the challenge at the same time? with all normalized volumes would be the best to compare?

Kind regards,
Patrick

vintage wrote: ↑

Thu Sep 27, 2018 10:25 CEST

Hello Mister Fox,

Would it be possible to have a link that includes all the songs of the challenge at the same time? with all normalized volumes would be the best to compare?

Kind regards,
Patrick

I would second that. I downloaded the mixes I liked initially and then normalised them. This is quite revealing I must say. I suppose the judges get the normalised files, right?

My best,
FP

They will of course judge normalised songs, a general download link of the challenge would be practical, we can of course normalize everything to listen to it ourselves...

Hi,

regarding the normalization, I would like to ask, what kind of normalization you are talking about? Just simple peak normalization? Wouldn't make much sense, IMHO. So, what kind of normalization is applied? Which software, processor, plugin? Give 5 engineers a song to apply loudness normalization to, and you end up with 5 results sounding differently. Having a link to all mixes normalized the same way the mixes are provided to the artists, would be very cool. An explanation on how to apply the same normalization ourselves would be perfect. Thank you. A very interesting topic, by the way. Please don't get me wrong on this. I've listened to almost all mixes myself and there are mixes, which are nearly mastered already. That's why this is a really interesting/challenging topic. Doing just peak normalization, the mix with the most squashed dynamics would always appear to sound "the loudest" and as such would be judged to be the "best" sounding one whereas the mix with the most dynamics would always sound "the quietest" and would not be able to compete, if the judges do not adjust that volume potentiometer themselves. I am not complaining about anything here. How would you compare all those mixes without being aware of how to use that volume potentiometer of your hardware? What options do we have, if we really want to provide 60+ different mixes having an equally perceived mastered loudness without the judges having to adjust the volume themselves for every single mix?

This is the average level measured in LUFS; ideally, control should be provided in EBU R128 (-23LUFS). This can be done by many software programs such as Wavelab or any other small dedicated program. The idea is to compare all mixes at an average level +/- identical. Indeed most of the mixes proposed in this challenge are not good for mastering, the dynamic (DR) is too weak, often too compressed.

The ideal for me to deliver a mix in 24 bits / EBU R128 -23LUFS (medium level) -6dbTP (peak) DR14 (dynamic) to give way to the mastering engineer.

To know the current level of mixing, the best plugin for me is: Maat DRMeter mkII, there are many other very good ...


Small correction on my post I meant -6TP and no -06TP big difference ... I'm sorry

Beautiful day

For approximate normalization (which would be more than sufficient), the listener can just use any DAW — add all of the competing tracks and use automatic target-based gain-set plugins to even out the average levels.

E.g., €5 Hornet VU and LU meters would do the job.

vintage wrote: ↑

Thu Sep 27, 2018 10:25 CEST

Would it be possible to have a link that includes all the songs of the challenge at the same time? with all normalized volumes would be the best to compare?

To answer these questions real quick:

Collecting all mixes and dropping them in one post is a whole lot of work, and actually not possible as the forum has a link/quote limitation. Keep in mind, currently one person handles everything(!) resolving around the Mix Challenge. it is (currently) easier to do for the Songwriting Competition, but it's still manual work.


Regarding Loudness Normalization - people can do that themselves.

I wrote a simple tutorial as to "how" in the production techniques sub-forum:
How to: Loudness Normalization (Manually and Automated)

vintage wrote: ↑

Fri Sep 28, 2018 13:59 CEST

The ideal for me to deliver a mix in 24 bits / EBU R128 -23LUFS (medium level) -06TP (peak) DR14 (dynamic) to give way to the mastering engineer.

A bit Off-Topic, but since it got brought up:
The values are not correct, or rather highly dependent on the final medium you want to distribute. If it's "Broadcast", then EBU R-128 (-23LUFS ILk/-1dBTP max, although France requests -24LUFS ILk/-2dBTP max) might be the route to go. If it's modern day streaming, then it's -16LUFS to -14LUFS (at -1dBTP max).

The "Dynamic Range" (Maat DR-Meter mk II) is mostly irrelevant in this case, and a fairly outdated way of measuring program material. It's is also highly dependent on program material "density" (if you don't have strong peaking transients, then your DR value will be low! I've also written plenty about this on KVR Audio).


The focus of the Mix Challenge is to just mix, not go into fine details with ITU-R BS.1770-x specs. Mix with a combination of a Digital Meter and VU Meter / RMS (realtime) Meter, work with a reference level of the known -18dBFS (EBU convention R68) or -20dBFS (SMPTE RP 0155). Your mixes will naturally be within -21dB RMS to -15dB RMS (or -18dB RMS avg) with a healthy signal peak that can(!) reach up to -1dBFS max.

Mastering is a different topic/game.

Thanks for these clarifications, indeed the standard Broacast EBU R-128 is equal to -1dbTP max (and not -06dbTP as I said) if I have the chance to continue ... I will submit a file format ITU-R BS 1770 thanks for your reply and your precious time

Another post of interest - which explains "Gain Staging" and the main difference between Mixing and Mastering

viewtopic.php?p=1355#p1355