Mix Challenge

Franz wrote: ↑

Mon Nov 25, 2019 15:15 CET

Hi !

I come back to the very interesting subject of some effects of microphones and especially that of proximity.
Indeed the choice of a microphone is crucial because it must be appropriate to the source that you want to record.
I don't expand the subject, there are very good advice on the net.

It is true that quite often at the time of mixing that one realizes for example that the voice is very present
and very close compared to other instruments whereas when one listens to the voice alone there is no problem.
Some software was mentioned (which I didn't know) and I would like to talk to you about the software "DOUBLE MS SCHOEPS"
from the manufacturer of excellent microphones SCHOEPS (besides the software is free).
Its use is relatively complex but I find the results quite incredible: changing for example the directivity of
the microphone (cardioid, omni, 8, etc ...) the effect of proximity changes. Since we are in the competition MC060
with the song sung by Eveline, try this software on the track of her voice and you will understand.

Visit the SCHOEPS.DE website

Cheers.

Franz

Hello Franz, I know SCHOEPS plugin I often used it in post-production and surround, but never try in any different situation! I'll give a try on vocal for sure. thanks for sharing.

Hi,

I had forgotten another software for sound placement: "anaglyph" concocted by researchers (a little crazy)
who have not quite mastered the smooth operation of the software. I managed to use it briefly on a voice track
and the result is quite convaicant (based on 3D audio).
If you're brave, try this software at this address: www.anaglyph.dalembert.upmc.fr. (is free)
Without warranty.
Cheers
Franz

Mistake on my part, the internet address is:

analglyph.dalembert.upmc.fr

Cheers

Franz

Henlo.
So some life stuff happened and I'm quite late. However, I absolutely fell in love with the song and decided to finish it. And as I can't post it anywhere else except here without offending owners rights - here we go.


MP3
WAV

The most challenging part this time around was making all elements work together with the vocals, as it was definitely a centerpiece of the production. General edit was cutting around 300-500hz mark with just plain and regular stock EQ, as it seems to have a lot of buildup in this area on all instruments (incredebly boxy acoustic guitars, a bit in the overheads and snare). After that - a little bit of 8k on almost everything, quite a lot of delay and reverb on lead vox from CLA Vocals, CLA-3A on a snare drum, and cutting about 3-6 DB on orchestral instruments around 800-1400 zone, as they were battling the vocals. Also bumped down the super lows on orch. percussion to let bass and kick breathe, and after that a whole lot of panning. To separate two acoustic guitars a decided to put midi guitar closer to the middle with vocals and piano, and live performed all the way to the sides, while filling in the gaps with orchestral arrangement and drums.

To finish the whole mix off I tried to sum it through a little analog mixing board (Soundcraft EPM6) to give it a little bit extra punch and saturation, as well as using Logix's Urei on the masterbus as well as Pultec replica EQ for some air and low-end.


Sorry for being so late,
Elijah.

Elijah,
my 2 cents - it's worthy, the "analog" approach is performing very well, the song sounds lovely.
however, I think there is still some lack of digital swiss-knife tech to get rid of the vocal "bed-room" sound (which was defeated quite successfuly in other examples), and there is a foggy cloud of reverbs that is basically very nice, but perhaps a bit much.
Perhaps there is a tricky trade-off between some natural vintage characteristics of the recording, and the analog mixing paradigm in a digital environment. Somewhere it may overlap and create a bump, somewhere it may cancel out. Just keep on. Good path.

This is my first time participating here and I turned in a really fun and dynamic mix that wasn't squashed and loud. Usually people pick the louder option when given a choice and I will surely lose this contest if it's on (average) loudness!

When the mixes are being judged what is the general approach to level matching mixes for comparison? What are the average and peak loudness targets used during the evaluations? My apologies if this is a secret or has been covered before, I'm just curious!

Thanks,
Mat

there is rich explanation to the rules, besides the rules, and I think it has been covered. to me, als occasional participant with some technical understanding, it seems sufficient.
mixing is a complex task, and a responsible profession, so I think it is required that one is able to study a lot of "papers" also.
the client / song provider of the episode will use his or her own ways of comparing results, but hopefully within the set of rules and guidelines.
we are dependent on the people who gracefully contribute songs and mixtracks. as in real life, the client is king.

Ok, I thought I read that there was some level matching of the wav files for the judging, but it sounds like you're saying they just give them all a listen as posted. I appreciate it!

that's what you read. I expressed my hope that rules are respected and they say one should level-match.

they ought to do their level-matching while comparing mixes, as it is in their own interest and delivers important insight, but styles and ways of working can be very different, and the song providers can't be policed into doing this exactly in a particular way.

letsmix wrote: ↑

Tue Nov 26, 2019 19:32 CET

When the mixes are being judged what is the general approach to level matching mixes for comparison? What are the average and peak loudness targets used during the evaluations? My apologies if this is a secret or has been covered before, I'm just curious!

There is no secret behind this. In fact, I've written an article about this (which can be updated at this point, since there are new tools). You can find that one here:
viewtopic.php?f=13&t=12


How the client (our song provider) does the A/B comparison, is completely up to himself/herself. I can only give recommendations how this process goes. If one goes for loudness normalization, I usually recommend to "pull down" the loudness rather than "pull up", since this involves peak limiting (as in: destructive editing). In case of MC058 where I had to step in, I used Wavelab and pulled the tracks down to -23LUFS ILk before listening to all of the mixes. Then it's only up to your personal taste rather than being fooled by loudness jumps.

This is also the reason why the rules recommend to not exceed certain work/reference levels. This way your content stays dynamic even with slight summing processing (tape machines, "record enable" compressors doing a 2dB GR, etc).


Of course we can not force anybody to do this extra step. But at these days, it's super simple to pull off.