2020-JUL-01 Info: Check out the poll results regarding a dedicated "Rule Book" sub-forum, and read up on what will happen on the forum the coming days

Mix Challenge - General Gossip Thread

Ask us a question, give feedback, join surveys, make suggestions
Mork
Backer
Posts: 103
Joined: Fri Apr 14, 2017 21:55 CEST
Location: Hamburg

Re: Mix Challenge - General Gossip Thread

#191

Post by Mork »

Ok, I also got disqualified, but because of too many LUFSs.

I will obey next time, but I have to say that I don’t think this is helping anyone. What does that have to do with anything other than housekeeping? Why -16? At least make it -14 which is the most relevant level nowadays (Spotify). YouTube, Amazon, iTunes are even -13. Also keep in mind that every meter gives you different results. Pro Tools told me I was at -15 (which I thought was target level), Jorge told me I was at -14, Pro L2 tells me something different as well as RME Digicheck. You can hear how „exact“ this works on Spotify, where most songs are hovering around -14 but almost never hit it (according to my meters). But then again, who cares? It‘s close enough. My point being, what difference does it make for a mix? Absolutely none! So why this obsession with levels, this is not TV Mix Challenge? The mixes still have to be volume matched for comparison, no matter what. A too hot, undefined, squeezed mix will stand no chance against a properly dynamic, punchy, open mix.

My suggestion is to just stick to a level recommendation (-14db LUFS at -1dbTP, being a „real world“ level). That would also decrease work load for the judges, so they could focus on the important stuff (the actual mixes) and not some unnecessary analyzing work. Sure, it’s not that hard to achieve, I will turn down my mix 1db more in the future, but why all that unnecessary fuzz about it?

doobop
Posts: 5
Joined: Tue Apr 14, 2020 19:33 CEST

Re: Mix Challenge - General Gossip Thread

#192

Post by doobop »

Thank you all for your answers but I'm not sure how to get into incognito mode.
Could anyone try to download and tell me if I done it correctly please??
I don't want to be part of the next phase of the competition, I just want you to check so I won't be disqualified again for the same reason..

https://www.dropbox.com/s/ogdo1bmnldwqp ... p.wav?dl=0

Regards

User avatar
Jorgeelalto
Posts: 89
Joined: Sat Apr 01, 2017 00:41 CEST
Location: Madrid, Spain
Contact:

Re: Mix Challenge - General Gossip Thread

#193

Post by Jorgeelalto »

doobop wrote:
Wed May 06, 2020 13:32 CEST
Thank you all for your answers but I'm not sure how to get into incognito mode.
Could anyone try to download and tell me if I done it correctly please??
I don't want to be part of the next phase of the competition, I just want you to check so I won't be disqualified again for the same reason..

https://www.dropbox.com/s/ogdo1bmnldwqp ... p.wav?dl=0

Regards
I can download that file now :smile:

To get in incognito mode, if you are using Google Chrome, you can go to the menu on the top right of the screen (the three vertical dots, next to your user image) and click New Incognito Window, or press Ctrl+Caps+N. In Firefox, it's New Private Window in the same top right menu (this time, three horizontal lines one in top of another), or Ctrl+Caps+P.

User avatar
Jorgeelalto
Posts: 89
Joined: Sat Apr 01, 2017 00:41 CEST
Location: Madrid, Spain
Contact:

Re: Mix Challenge - General Gossip Thread

#194

Post by Jorgeelalto »

I'd like to talk about documentation, and the documentation requirement in the MC.

I feel like the current obligations around documentation when submitting a mix are insufficient and, although ideally they should make sense, in practice they don't. I came back as a client this past month and having to follow these rules was too subjective and I think it benefits nobody.

This is the first thing you see when reading the rules and guidelines as a participant: "Furthermore, the concept is to post documentation of what you did with the song. Documentation can be as simple as a screenshot, or an in-depth listing of plugins/modules and settings being used." I found several users did one thing or the other, but both alone are meaningless. A few screenshots of the session without any explanation on what's happening is not useful - and then, explaining a whole mixing session is probably too much, and not needed. On the other hand, posting a list of plugins you used (even if they are sorted by track) without talking about the settings used on each instance is, again, useless for the client (as they don't know what's happening) and, therefore, time wasted for the participant. The same "screenshot documenting style" is again mentioned further down ("You can also provide screenshots, but they should show of your mix console and your used insert.").

What I agree with about documentation in the R&G are these lines:

- "It is mandatory to document the mix/production. At bar minimum, focus on one particular "sound" you have enjoyed creating, and how you went there." This here should be the focus. There are a lot of participations, and what would be cool is for each submission to be accompanied by a few lines, written by the submitter, about that one element they are proud of, and how they achieved it. Or those drums that were all over the place, how did they controlled them.

- "This is adding to the learning factor of understand possible mistakes and/or for others to benefit from your findings, ideas, not to mention gather some new inspiration from your type of mixing. Nobody can improve if you can't learn anything." This is the spirit of the MC!

What I want to say, maybe the R&G should be more clear about documentation. For example, a guide for a bare minimum documentation, from which lazier users like me could create step by step their post to follow the rules, and from which clients could differentiate between submitters that post documentation and submitters that don't, more clearly. Ideally, this minimum doc should be meaningful and tell something specific about your mixing process and your submission. Not all the mix, not what plugins you used, not a screenshot of the mixer.

What this new bare minimum documentation should be? I'm not so sure. I would aim to something the likes of:

1. Choose a main element (drums, bass, vocals, keys, guitar, synths).
2. Why was it difficult / what was it interesting?
3. How did you solve it? What tool did you use?

And these three steps, at least for one element. I think this would make the documentation requirement much more interesting and easier to implement for participants, and easier to validate for clients.


I want to know your opinions as participants, as clients and as organizers (Fox) :grin:

User avatar
Mister Fox
Site Admin
Posts: 1693
Joined: Fri Mar 31, 2017 16:15 CEST
Location: Berlin, Germany

Re: Mix Challenge - General Gossip Thread

#195

Post by Mister Fox »

While I appreciate the feedback, Jorge, I can certainly invest way more time into better documentation. Ultimately my problem is that nobody seems to be reading the Rules and Guidelines for each corresponding game, even stating that "I didn't know of them". Or the other likely scenario, that people interpret it like they want.

So in order to make clear that this community is serious on this topic, we have to step up and clean up the field of rule breakers. Ideally, this is the job by the Song Provider, with the hosts (the staff of the community) as backup. This is why I am way more active on that behalf now - even if that is extra work for me (but looks like I should have done this sooner).

I am currently considering contacting longtime community members to help me out on that behalf.





:arrow: On the topic of "do not exceed -16LUFS ILk"...
Mork wrote:
Tue May 05, 2020 13:51 CEST
I will obey next time, but I have to say that I don’t think this is helping anyone. What does that have to do with anything other than housekeeping? Why -16? At least make it -14 which is the most relevant level nowadays (Spotify). YouTube, Amazon, iTunes are even -13. Also keep in mind that every meter gives you different results. Pro Tools told me I was at -15 (which I thought was target level), Jorge told me I was at -14, Pro L2 tells me something different as well as RME Digicheck.
Actually, the readouts were as following:

Code: Select all

Wavelab 10 Batch Analysis	=	-14,7 LUFS ILk (EBU R-128 specs)
Orban Loudness Meter		=	-14,4 LUFS ILk (ITU-R BS.1770-1)
Youlean Loudness Meter 2 Pro	=	-14,7 LKFS ILk (ITU-R BS.1770-4)
Nugen Audio VisLM2		=	-14,6 LUFS ILk (ITU-R BS.1770-4 specs)
Nugen Audio VisLM2		=	-14,6 LUFS ILk (EBU R-128 specs)
Klanghelm VUMT2			=	-13dB RMS avg / -8,3dB RMS max (+3dB AES-17, z-weighted)

Maximum signal strength		=	-2,7dBTP
Looking at the RMS values alone, you were pushing it a bit. This even goes against the reference level recommendations. And I quote:
Rules and Guidelines for Mix Participants wrote:It is recommended to work with a certain reference level. We recommend to work at -18dBFS = 0VU (EBU R68 alignment level) - please consult the Wikipedia entry on Peak Programme Meters, and/or the forum entry on "Gain Staging with a VU Meter". Working and exporting within the range of -21dBFS (RMS avg or -3VU) for mezzo-forte to -15dBFS (RMS avg or +3VU) for forte-fortissimo passages, while not exceeding -2dBFS maximum signal strength (digital peak), is considered "good practice"


Which brings me to the the following:
Mork wrote:
Tue May 05, 2020 13:51 CEST
It‘s close enough. My point being, what difference does it make for a mix? Absolutely none! So why this obsession with levels, this is not TV Mix Challenge?
No, the Mix(ing) Challenge is not out to create "TV mixes" (which are -23LUFS or even -24LUFS, depending on the country. Netflix even going down to -27LUFS Dialog btw!). Neither are we talking "streaming release" at -14LUFS (btw: Tital and iTunes are -16LUFS, Youtube is currently shooting for -14LUFS, Spotify still can't decide on a suitable value like everyone else).

The Mix Challenge is first and foremost just that:
Rules and Guidelines for Mix Participants wrote:This is a mix challenge, NOT a mixing and (pre)mastering challenge. The main focus of the game is is to create a balanced mix, which can later be used for a possible mastering process
And if we purely focus on mixing, incl. people utilizing reference levels, we (ideally) never reach -15LUFS ILk as absolute maximum. In fact, most mixes I've done in the last 5 years never really exceeded -16LUFS. If I overshot that value, I just reach for a gain plugin and pull the signal down. I've never adjusted "up" unless I've mixed and (pre)mastered / prepared for a distribution for a specific target medium (like Social Media .. which is a completely different scope compared to the Mix Challenge).

These values don't come out of thin air - they result from experience with various types of meters and sheer endless content that was ran them (personal experience, and the experience by those that created these specs in the first place). If you have a track that runs at 0VU +-4 (VU reference being -18dBFS, 300ms rise/fall) or at -18dBFS RMS avg +-4 (Dorrough specs, 600ms, which are equal to the K-System v1 specs), you quite literally never exceed -16LUFS ILk if we talk music. Unless you've robbed the production of any dynamic (loudness range).



What I try to teach since 2014, is to learn how to rediscover proper gain staging and suitable project setup setup again. Easy integration of hard and software (with more and more tools these days having reference levels for saturation), mixing in such a way that you never have to adjust the output signal strength, or try to prevent possible signal overs with a limiter (unless you went crazy with transient designers).

I other words: I try to offer you different perspectives on "how to mix" - and only that.

Creative treatment of the summing bus (like gluing the signal together with a summing compressor, or a tape machine, or a specific console summing bus, etc), is one thing. But this is still not pre-mastering or "setting the stage for broadcast". If anything, it's setting the stage for the next person in line. May it be the mix engineer at the video editing studio that integrates your prepared soundtrack into a scene. Or the mastering engineer who is taking care of the final touches (fairy dust, maybe signal compacting, checking the stereo field, pushing the production to the right loudness target).

Am I really the only person that doesn't find this so difficult to understand?



Mork wrote:
Tue May 05, 2020 13:51 CEST
My suggestion is to just stick to a level recommendation (-14db LUFS at -1dbTP, being a „real world“ level). That would also decrease work load for the judges, so they could focus on the important stuff (the actual mixes) and not some unnecessary analyzing work. Sure, it’s not that hard to achieve, I will turn down my mix 1db more in the future, but why all that unnecessary fuzz about it?
There was never any real fuzz about this. This only started earlier this year after I have drawn a clear line in the sand.

:!: Prior to a couple of games ago, this was but a mere "recommendation" to (ideally) not exceed -15LUFS.

It turned out to be a general mess with releases being all over the place. People thinking "I have to do mixing and mastering". After feedback from various Song Providers, and some general frustration of me constantly having to step in, I've decided to be more strict and locked the absolute maximum allowed loudness to -16LUFS (which is more than achievable - see above). Once more - this is not a random number and/or decision - there is both a certain concept and a lot of experience behind this.

This maximum target level is to (once more) push the focus on to mixing only.

I am still looking for suitable affordable/free tools that can be used for accurate analysis after the mixdown. With keeping a certain allowed tolerance in mind (which will be less than the spec known/allowed +-1LU). Ideally tools that also offer drag and drop of WAV files, that are updated on a regular basis and/or provide presets that are at the pulse of time. I will get to that in the near future.



The Songwriting Competition on the other hand, merely has the recommendation to not exceed -14LUFS (music doesn't have to be squashed to be impactful). However, that game has a completely different focus. Please do not mix up the rule set.

Mork
Backer
Posts: 103
Joined: Fri Apr 14, 2017 21:55 CEST
Location: Hamburg

Re: Mix Challenge - General Gossip Thread

#196

Post by Mork »

First and foremost it is not my intention to discredit any of the educational background. I think we both are doing this long enough to know how important proper gain staging is. -18 dbFS as 0 VU is a good recommendation being the agreed upon standard here in Europe (I think you mentioned that). However, not every system is aligned the same and not everyone works the same. My converter puts out +4dbu at -16dbFS so that's my 0 VU. Then there is a different school of thought that adjusts clip gain or plugin outputs to have the channel faders around 0 for best resolution while automating, with only the main elements hitting 0 VU (TLA does that for example). Who's right, who's wrong? Do whatever floats your boat. The outcome is what counts, imo. But I agree, you should at least know where your 0 VU is and why.
Mister Fox wrote:
Mon May 11, 2020 07:24 CEST
Actually, the readouts were as following:

Code: Select all

Wavelab 10 Batch Analysis	=	-14,7 LUFS ILk (EBU R-128 specs)
Orban Loudness Meter		=	-14,4 LUFS ILk (ITU-R BS.1770-1)
Youlean Loudness Meter 2 Pro	=	-14,7 LKFS ILk (ITU-R BS.1770-4)
Nugen Audio VisLM2		=	-14,6 LUFS ILk (ITU-R BS.1770-4 specs)
Nugen Audio VisLM2		=	-14,6 LUFS ILk (EBU R-128 specs)
Klanghelm VUMT2			=	-13dB RMS avg / -8,3dB RMS max (+3dB AES-17, z-weighted)

Maximum signal strength		=	-2,7dBTP
Looking at the RMS values alone, you were pushing it a bit. This even goes against the reference level recommendations.
Yup, as mentioned I aimed for -15(ish). And yup, it's pushed, like I mentioned in the description. The level of push has been painstakingly tweaked to give me the desired sound I was looking for. Take away that sustained 808 and it's a whole different story RMS-wise. This is far from being crushed.
Neither are we talking "streaming release" at -14LUFS
Spotify is using replay gain "+3", which should come close to -14 LUFS in most cases (as they say) and you're right about Youtube (which is great) and probably the others.

Isn't this supposed to mimic a real life mixing scenario? And real life is that probably 99% of the (real life) mixes done by the participants here will end up on some kind of streaming platform. And real life is that not every project will have the budget for mastering. If you are aiming for -16 now your mixes will translate well on Tidal and iTunes, but will be 2db quiter than the rest on Spotify, which will result in a very unpleasent suprise for most clients. If your mix is at -14 it will be in line with rest on Spotify and Youtube and simply be brought down 2db on Tidal. -14 at -2dbTP gives you 12db dynamic range to work with, which is plenty for popular music. For classical or more "serious" music there should be a "custom" recommendation like -16 or -20.

The TV reference was a joke because they are so very strict with their levels (which makes absolutly sense in that case).

In the end it doesn't really matter at what level I upload a mix here, I just have been frustrated to get disqualified for being at a level which is far from unreasonable. I liked the song a lot and put lots of effort into it. Just my 2 cents

Cheers
Mork

Post Reply