2020-MAY-09 Info: Please help spread the word about the Songwriting Competition and help it reach 15 participants per month on average

Mix Challenge - General Gossip Thread

Ask us a question, give feedback, make suggesions
User avatar
Mister Fox
Site Admin
Posts: 1649
Joined: Fri Mar 31, 2017 16:15 CEST
Location: Berlin, Germany

Re: Mix Challenge - General Gossip Thread

#171

Post by Mister Fox »

I've moved the last post from the MC063 thread. Let me please address some points.

bluesation wrote:
Mon Mar 30, 2020 12:34 CEST
CeZar wrote:
Fri Mar 27, 2020 21:02 CET


I'm sorry man, how exactly do you expect me to give 100 feedbacks? How willing will you be to listen to a song you've written 100 times from start to end?
Yes, this we expect from a song provider, feedback. We spent several hours with your song, helping you to get a better sound it has. Former song providers had no problem with that.
The song provider is only bound to give feedback to the top 10 entries. This has always been the case, but has been slightly adjusted in the rule set for MC064 and forward.

That you can get feedback here on the Mix Challenge has always been a special case compared to other communities. That there was feedback for all entries previously was solely due to how the song provider worked and handled a specific situation. And while I do encourage the conversation, "clients" are not obligated to give feedback other than Mix Round 2.


bluesation wrote:
Mon Mar 30, 2020 12:34 CEST
CeZar wrote:
Fri Mar 27, 2020 21:02 CET
A: Correct the problem;
The rules say, that there is no correction (changing the source material) allowed, especially quantize and pitch correction.
Not wrong, but also not that simple either. Even before the update of the Rules and Guidelines, it was made clear that you have to go by your gut-instinct, and that you have to weight in if a "timing edit" and "drastic pitch correction" is really necessary. You should always keep the artistic vision in mind.

Check in again with the rules for the participants and the FAQ. For the latter, most notably FAQ Q/A #10 and #11.
Rules for participants wrote:
Fri Mar 31, 2017 20:41 CEST
Rule set for mix participants
  • ...
  • it is allowed to edit and to a certain extend even mangle the source material to improve the production (e.g. pitch correction within given boundaries, timing/phase correction within given boundaries, re-amping of bass/guitars if DI track were present, vocal comping, noise gating/strip silence, etc), as long as this does not result in drastic sound design and a complete new sound (e.g. morphing one sound into another one - example: flute into trumpet) - unless otherwise stated
  • ...

tumewor
Posts: 6
Joined: Wed Mar 11, 2020 10:46 CET

Re: Mix Challenge - General Gossip Thread

#172

Post by tumewor »

Hello Mister Fox, i hope you're well, healthy and rockin.

with the slight changes you made, its still confusing.
in one point you want us to keep the original sample rate and bitrate

"export in the sampling rate and bitrate the material was provided in"

while in other point you want us to change the original sample rate and bitrate ?

"If a multitrack package was provided in 44/16 - mix and export in 44/24"

and then

"if a multitrack package was provided in 44/24 and the end result shall be delivered 48/24 - mix and export in 48/24."

the "shall be delivered" , what is that suppose to mean ?

its still as abstract as the last rules before an update.
But anyway, those abstractness will surely serve it purpose just like any other established ground rules somewhere else, there's nothing different here.

Peace Out,

User avatar
Mister Fox
Site Admin
Posts: 1649
Joined: Fri Mar 31, 2017 16:15 CEST
Location: Berlin, Germany

Re: Mix Challenge - General Gossip Thread

#173

Post by Mister Fox »

Seriously? It is still not straightforward?

The sentence right before that paragraph clearly states:
unless otherwise stated, export in the sampling rate and bitrate the material was provided in. Bar minimum in 24bit.

I really fail to understand where this is... quite honestly... hard to understand?
At this point, we're at "dissecting every word of a sentence" rather than the whole paragraph. I honestly do not have the patience anymore to regurgitate this over and over again.

But if this will put you at ease, I will refine it even further and clearly state "must be", as in "you have to - else you're out!" - if this is the only way to get the message across.


EDIT (17:40):
updated the rule set once more - I hope it is finally unmistakable now.

User avatar
bluesation
Posts: 32
Joined: Thu Jul 12, 2018 09:38 CEST

Re: Mix Challenge - General Gossip Thread

#174

Post by bluesation »

Notification when starting Round 2

I suggest sending a notification email to all senders after completing the first round. In the last challenge, some missed the countdown. It also happened to me that I almost missed round 2. Sometimes you have a lot to do and don't think about it and some song providers are fast and others are slow with their decision. That forces you to check again and again whether something has changed.
Thank you

User avatar
Mister Fox
Site Admin
Posts: 1649
Joined: Fri Mar 31, 2017 16:15 CEST
Location: Berlin, Germany

Re: Mix Challenge - General Gossip Thread

#175

Post by Mister Fox »

In fact, three notifications have been sent out
  1. through the new forum "mention" feature (see this news entry). I'll use your name @bluesation as example (you should have now been pinged via the forum notification system)
  2. a newsletter has been issued to everyone on 28th March, listing the current running games
  3. announcements of Round 2 on social media (Twitter and Facebook) - as I always do every time Round 2 for the MC, and the voting period for SWC is starting off

As you can see, I already do everything I can. I can't do much more, neither can I be blamed for people not checking in.

Chriswilson83
Posts: 50
Joined: Sat Feb 09, 2019 20:52 CET

Re: Mix Challenge - General Gossip Thread

#176

Post by Chriswilson83 »

I was unaware the sampling rate rule was recently changed, so my bad.

Must say though for me it means I can't work on another 44.1k project here again. My Allen and Heath QU24 does not support it (don't believe me go read the specs and manual). I'd have to upscale it all to 48, then sample back down to 44.1 after, something I wouldn't be happy doing in a real life scenario.

User avatar
Mister Fox
Site Admin
Posts: 1649
Joined: Fri Mar 31, 2017 16:15 CEST
Location: Berlin, Germany

Re: Mix Challenge - General Gossip Thread

#177

Post by Mister Fox »

Chriswilson83 wrote:
Sun Apr 05, 2020 00:24 CEST
I was unaware the sampling rate rule was recently changed, so my bad.
The sampling rate "rule" actually exists for quite a while now. Even before MC056, where we made clear for the first time "due to technical reasons, please only submit in..."


Chriswilson83 wrote:
Sun Apr 05, 2020 00:24 CEST
Must say though for me it means I can't work on another 44.1k project here again. My Allen and Heath QU24 does not support it (don't believe me go read the specs and manual). I'd have to upscale it all to 48, then sample back down to 44.1 after, something I wouldn't be happy doing in a real life scenario.
I've read up on the manual, and some forum posts. So I perfectly understand your concerns (and frustration for that matter).

I'll respond to your PM and try to find a solution.

User avatar
Mister Fox
Site Admin
Posts: 1649
Joined: Fri Mar 31, 2017 16:15 CEST
Location: Berlin, Germany

Re: Mix Challenge - General Gossip Thread

#178

Post by Mister Fox »

I did some more digging on this topic, and in fact a deep dive on a lot of "budget" Digital Mixing Consoles released in the last 15-20 years.

From all the companies I've checked (which were: AVID, Presonus, Roland even down to the old Roland VS2400, Behringer from the old digital Behringer DDX3216 to the recent "Wing", Midas, Yamaha, SSL, etc), Allen & Heath is the one company that cut corners and really doesn't like you using 44,1kHz, or in some cases even 48kHz.

Their QU-series is locked to 48kHz, and their SQ-series is actually locked to 96kHz even unless you use their DANTE add-on card (which offers you 48kHz and 96kHz) or just the "internal drive" system. It's just how they're built, and there is zero chance to evade that. So I completely understand the frustration from participants if you use this (or similar locked) digital mixing console not only as main DAC, but also as mix environment (we're long past plain gate/compression/EQ at this point).



:arrow: What is the solution?

I honestly don't know. I did not expect that "locked" sampling rate in hardware is still a thing - or rather, "a thing" again. And I really do not want to lock out users from participating.

There is of course the option to still mix in 48kHz or even 96kHz - because of your hardware limitation (actual real hardware limitation, not a setup error!) - and render out in 44kHz. But the quality of the sampling rate conversion depends on how great your host handles it, or the offline tool you use (like Izotope RX, SoX - Sound eXchange, AudioMove, r8brain - no tool performs at equal quality). Then there is also the thing with "can I check the export with my equipment that everything is alright?", or can you still use your computer audio output (PC Audio mainly, might not be easy on a Mac).

Then there is also the fact, that if I allow to export in a lower sampling rate to match the source material, people might use this as open door for excuses again. Even though this topic only affects a rare amount of participants.



I started to read up on SoX - Sound eXchange again (I have SoX built into Wavelab and Izotope RX also adapted this code IIRC), since this is cross-platform. AudioMove has been cross-platform as well (apparently, I never knew that). I hoped to find a suitable solution to those that are(!) hardware locked, to provide the best quality after conversion. However, AudioMove is not as "neutral" in terms of sample rate conversion compared to SoX, yet it comes with an UI. SoX on the other hand is more neutral, but it's command-line only and I have zero experience with macOS and Linux command lines.

I'm still scratching my head over how I should handle this in the Rules and Guidelines. :thinking:




:?: Could those being affected by this, please chime in with their feedback?

I'd like to gather some more thoughts and opinions over the weekend, so that I can add possible adjustments to the Rule Book by the next newsletter around the 15th.

Thank you.

White Punk OD
Posts: 148
Joined: Tue Aug 14, 2018 23:58 CEST

Re: Mix Challenge - General Gossip Thread

#179

Post by White Punk OD »

Last year I co-engineered an album. Some songs were recorded years earlier with an outdated harddisk recorder at 48k. All other songs were produced at the newer project studio at 44.1k.
We had to switch the master clock and the focusrite devices to the correct SR a couple of times. Naturally, we rendered at corresponding recording SR. Also, I mixed one song at 48k in my notebook.

The point of this case study is the following:
A) We did all kinds of test conversions, brickwalled or not, for the artist to listen, e.g. on a great car stereo. We pointed out, that our equipment was not invested for mastering, and the quality would be limited, also it happened that throwing the files into the Windows CD burner did not sound great, and it was clearly better to prepare at least with r8brain. But we could decide together, how we wanted the mixes, and the artist took part with meticulous details, and everything worked, though a bit slow.
B) We sent the mixes to an established mastering engineer, songs having different sample rates. He gave very competent and helpful hints and we corrected some mixes. SR was never a topic to talk about. We just said we had this kind of recording history.

And ONLY that way we got versions in EVERY format needed for distribution, from CD to Spotify etc., that had the FULL quality, and we were very happy with every song.
The mastering engineer had the perfect tools and corrections to cross-convert whatever was required.


Workflows in music business can be different. Some artists want to master their own songs, though it is not recommended, but it may be also a matter of budget.
But when the product goes to a mastering engineer, and mixes are done externally somewhere else, then the final quality is a matter between the technical studios involved, and they may do anything that works.
The artist has normally two options: waiver and sign, or demand improvement. In our case he was present anyway because he is the studio owner and had called us to work for him. He will receive standard formats that are usual, including mp3 for the bedroom tablet, and a copy of the mastered CD before he commits to the mass printing. We all agreed on this, and we kept a couple of details in limbo, to wait out what the mastering would do to it, and then corrected a dB or so and took one more roundtrip with that song.

In short, I believe procedures should follow the path of the best possible quality, corresponding to what is available, from recording format, to rendering, to mastering. Intermediate formats will be just that. This must be good for musical judgments, balances, effects.
There should be no loophole or compromise in what is sent to the mastering.
Final judgment, also about fullness of spectrum, is only possible with the final product.

This is one way to get to satisfying results, and I had joined this path.


--
If it is mixing for the sake of mixing, then some incompatibility among participants about SR cannot probably be solved to satisfaction, unless there is a full HQ SRC somewhere in the process, run by someone who is trained with this tool.
From what I have observed, r8brain is very ok when 44.1 > 48, but for downsampling, depending on the material, there might be a little loss, and it is not trivial to make up for this.

User avatar
Mister Fox
Site Admin
Posts: 1649
Joined: Fri Mar 31, 2017 16:15 CEST
Location: Berlin, Germany

Re: Mix Challenge - General Gossip Thread

#180

Post by Mister Fox »

Honestly..., your posts really star to confuse the hell out of me. And I'm sure I'm not the only one.


The question I've asked was for those people with equipment "locked" to a certain sampling rate compared to the provided source material. This is still nothing about mastering at all. Just mixing and then exporting to offer the edit in the same sampling rate and bitrate, as the source material was provided in. How you got there, is temporarily not important. I want people to stay in the game, despite their limitations - and to maybe learn something in the process to evade shortcomings in the wild (as in: outside of the mixing challenges).
  • Upsampling is never an issue, as you can fill the remaining frequencies with things like delay, reverb, saturation (through various forms of distortion), etc.
  • Downsampling is indeed the issue with sample rate conversion tools working as different as how they can get.

Pretty much all affordable or even free/open source sample rate conversion tools are discontinued.
  • r8brain is Windows only
  • AudioMove hasn't been updated since 2009
  • SoX hasn't been updated since 2015, but got implemented into Steinberg Wavelab and Izotope RX
  • Hosts and VST/AU/AAX plugins are existing a-plenty, but there is no guarantee of them being offering "clean results"


I am not looking for an all-around solution, but a solution for a special case. Neither do we need to constantly bring up the "mixing / mastering" topic-blend that has been going on in this thread for quite a while now.

:arrow: 98% of ADC/DAC (recording modules, digital mix consoles, etc) can be pulled into the appropriate sample rate needed to edit and export the material - so for those users, that is an absolute non-issue other than a slight workflow adjustment. There are still the 2% remaining where this is not possible (see Allen & Heath Digital Mix Consoles), and this is my main focus of my last post.

If I had the knowledge as to how command line tools work on macOS, I'd provide a detailed post and corresponding conversion code for SoX (as this is the most neutral among the bunch). As of this moment, I can only do that for Windows (as I am a Windows user).

But that still leaves the possible problem for those users having that particular hardware-side samplerate locked device, that they are not able to cross-check other sampling rate material after the edits. Unless they have a second DAC (may it be a second recording/playback device, or a PC Audio Chipset on the mainboard).



So again - this topic mainly addresses those people that are truly affected by the limitations of their equipment.
What can I do, as host of this place, to keep these participants engaged rather than sending the impression "you're locked out, sorry"?


Let us try to keep it simple for once. Please!

Post Reply