Mix Challenge - Gossip and Discussion

How to do this, that and the other. Share, learn, teach. How did X do that? How can I sound like Y?
Post Reply New Topic
RELATED
PRODUCTS

Post

Compyfox wrote: @No_Use:
Any recommendation in terms of providing a bundle to download for host A/B?
You mean in terms of where to host it ?

Private webspace would be best probably, but I don't have any.
Dropbox can lock downloads if certain amount of traffic is exceeded.

Hm...dunno...Google drive maybe ? (if noone can offer private webspace)

Post

Hosting is a non issue, I was more like thinking "what format", No_Use.
But I think WAV and/or FLAC is the way to go.


We're still on internal debates, but I think we will start something new with MC02 - which will cost us initiators more work, but will pay off on the long run: Loudness Normalization prior to the voting process.

I just tested this in Wavelab and it works fairly fine. So we will not only listen to the tracks the next days and might comment on certain rule violations (learning curve!!!) in the process, but we (or better said I) won't disqualify anyone this time around. Unless we found out that you resubmitted a track - which won't happen since I snagged all files within the first 2-3 hours of their release. So if you say "wait, but that's not my track", we will know... ¬_¬ *cough*


But to get back on topic:
The main advantage of Loudness Normalization is that we (the listeners) won't be swayed (influenced) by the good old "loud = better" phychoacoustic trick. Those tracks that were pre-mastered (not part of the rules btw!), will of course sound less dynamic due to the missing transients. But... it will be easier to A/B the tracks this time around. We (the listeners/bystanders/participants) can actually focus on the sound of the production and therefore judge it more objectively. It will also be easier to spot things that you (again, the listeners/bystanders/participants) might pick up as inspiration for your own future creations.

Due to this, we (unfortunately) have to make it a mandatory rule from now on to also provide a WAV or FLAC compressed WAV file, since lossless files are simpler to process without sound degradation. I hope this is reasonable.


The typical technical bit of information (and before anyone asks):
This process is pretty much non-destructive. The tracks won't be adjusted "upwards" to a more higher loudness, but down to a more reasonable level. So no compression or limiting involved. Just transparent gain adjustment. Probably down to -16LUFS (SLk, EBU R-128 specs), which is somewhat around -18dB RMS.

H-man wrote:Yep, that's where I'm coming from :wink: I've just checked my Wav against some of the others and I'm way too loud so apologies for that ....won't happen next time.
This is actually the spirit of the KVR Mix Challenge:
You stay in practice, you can learn new things, and you can improve on your current skills.

The rules might look strict, but they actually contribute to the learning factor. One of the main reason we started this in the first place, and why some of us on KVR are dedicated to teach new things. Or at least try offer a different viewpoint.

MC01 for example gave me ideas I've never though of pulling off in a production - personally I went more risky with MC02, but I have to wait until the voting is over if it payed off. On the other hand I tried to share how to properly set up signals and mix at a reasonable signal strength since day one. So far it seems to pay off for some users. :tu:
[ Mix Challenge ] | [ Studio Page / Twitter ] | [ KVRmarks (see: metering tools) ]

Post

Now that I've listened to all the mixes, I have to find that some mixes must be disqualified because they have not complied the official rules. This concerns the following mixes:
Headphone: Total alienated by dramatic sound design. It is a completely different song.
3ee: In the intro, the kick drum is missing. I do not know if it is allowed to mute sections? I could not read anything in the rules.
wesleyt: His mix is ​​not named in this format "Name of the Artist - Name of the song - KVR MC <nr> - contest participant".
Some participants have not complied the loudness specification. Why not?
All in one, very good mixes here. But even bad mixes with poor mono compatibility. Had they read the insights of the other participants, they would have to be noticed.

Post

@mwaudioprod:
Very good ideas ... you should make a list of what the mixers are allowed to do and what not ... that would enhance the creativity of the mixers ... so we can standardize the mixes ...

Post

Headphone wrote:@mwaudioprod:
Very good ideas ... you should make a list of what the mixers are allowed to do and what not ... that would enhance the creativity of the mixers ... so we can standardize the mixes ...
I do not need to make that, there are already official rules:
http://www.kvraudio.com/forum/viewtopic ... 3#p5798473
I have not made them. ;)

Post

mwaudioprod wrote:Some participants have not complied the loudness specification. Why not?
All in one, very good mixes here. But even bad mixes with poor mono compatibility. Had they read the insights of the other participants, they would have to be noticed.
I'm hoping that the normalization method Compyfox is implementing will take care of perceived volume differences.

Post

Compyfox wrote:Hosting is a non issue, I was more like thinking "what format", No_Use.
But I think WAV and/or FLAC is the way to go.
Ah ok, sorry for misunderstanding.
I agree of course, wav or flac is best.

Post

Headphone wrote:Very good ideas ... you should make a list of what the mixers are allowed to do and what not ... that would enhance the creativity of the mixers ... so we can standardize the mixes ...
The goal is to emulate what would happen in a real session. If you make the song unrecognizable or completely leave out instruments, it's not too likely that the client will accept the mix.

Post

Regarding the rules, the thing I don't understand is why aren't we allowed to replace a once submitted mix with a new one (until deadline) ?

The problem I see a bit with this rule is that late submitters could have a potential advantage as they might draw some good info from the other already submitted mixes and their descriptions, while the early submitters can't draw any advantage from this info (they have to mix 'more blindly' so to say).

Any opinions on this ?

Post

No_Use wrote:Regarding the rules, the thing I don't understand is why aren't we allowed to replace a once submitted mix with a new one (until deadline) ?

The problem I see a bit with this rule is that late submitters could have a potential advantage as they might draw some good info from the other already submitted mixes and their descriptions, while the early submitters can't draw any advantage from this info (they have to mix 'more blindly' so to say).

Any opinions on this ?
I see it the same way. That's not fair. We should change the rules.

Post

No_Use wrote:Regarding the rules, the thing I don't understand is why aren't we allowed to replace a once submitted mix with a new one (until deadline) ?

The problem I see a bit with this rule is that late submitters could have a potential advantage as they might draw some good info from the other already submitted mixes and their descriptions, while the early submitters can't draw any advantage from this info (they have to mix 'more blindly' so to say).

Any opinions on this ?
It's not the real world. When a client gives you a job, you can't just submit revisions ad infinitum. There are times when it can happen (once, maybe) but you really want the one you turn in to be as close to finished as possible.

If you don't like this rule, the solution is to simply not turn your mix in until mixes are due, that way you can do as many revisions as you like in the meantime. That goes against the spirit of making this feel real world (not to mention giving you an unfair advantage over those of us who don't have hours and hours of free time to spend on our mixes :P) but at least it's not explicitly breaking the rules.

Post

I kind of have to step in there.. because the rules on that behalf are more than fair.

It is a two edged blade if we overhaul that rule.
Early entrants could indefinitely change the track, while late entrants are basically stuck. On the other hand, the early entrants are somewhat ticked off by those sitting on the fence. There is no ultimate solution.


In this particular case, we simulate an actual competition (like a remix competition, where you can't send in more than one entry either), not a real live scenario. You have one entry, that entry is final. You have 15 days for the mix, he rule book even RECOMMENDS to take a day off and listen again.

In a C2B (Client to Business) relationship, communication is key. So one or several revisions are indeed possible, however not indefinitely. But I think the rules are pretty clear in terms of the simulated scenario: you have a virtual client, he/she expects a great mix from you. You have one chance.

I understand the negative vibes towards the late entrants. But actually... this is another thing that counts to the learning curve - ignoring the other users. Satya, Eric and I unfortunately can't afford ignoring the tracks, since we're basically the initiators of this challenge. So we changed the rules a bit for MC02 to analyze the tracks after the challenge, and then decide what to do rather than doing it at the beginning.

On that behalf, I set up a strict rule for me this time around to actually not listen(!) to any track, and work on the content if i have time. Since I'm basically doing "paper work" and giving "general suggestions", and I'm stuffed with off-KVR work as well... that happens to be at the very last minute.


So all I can do is to back up Eric here with saying:
Hold the track back as long as you can, edit it to your needs - feel inspired by early submissions. But ask yourself in the process - do you want to find your own style, or just copy others?
Last edited by Compyfox on Sun Jul 20, 2014 2:32 pm, edited 1 time in total.
[ Mix Challenge ] | [ Studio Page / Twitter ] | [ KVRmarks (see: metering tools) ]

Post

Thanks for giving your thoughts on this.

Uncle E, Compyfox I see your points and I can go along with them, no problem.

Though this isn't actually always the case:
In this particular case, we simulate an actual competition (like a remix competition, where you can't send in more than one entry either), not a real live scenario. You have one entry, that entry is final.
I've joined a few competitions here http://www.recording.de and there you ARE allowed to update your (re-)mix until deadline.
You send the link of the new version to the moderator and he updates it in the first post of the thread where all mixes are collected (at least that how it worked in the past when I joined competitions there).

Just pointing out, as said, no need to argue about this rule from my side anymore, as I can see your points / reasoning for it.

Post

I am aware of other competitions where you can resubmit. But most of the time that isn't the case.



We're still in the middle of discussing some things internally, but a couple of corner stones for the current MC and future MC's:
  • as of this moment (and for the foreseeable future) we're still not disqualifying anyone(!), we however suggest that you take the outcome as learning curve
  • the next voting process will be the first blind voting (like the OSC in the instrument section), we're currently in the middle of overhauling the voting rules and let you know in the next voting thread how this is working from now on.
  • the voting thread will get a link to a global SoundCloud folder, where all tracks are uploaded for everyone to listen to (since DropBox can be troublesome) - we have two tracks that were submitted in 32bit float, since we don't know if SoundCloud can handle that, these tracks will be Dithered down to 24bit (also for a possible FLAC distribution)
  • the voting thread will also have a DL pack containing all the current entries for you people to A/B within your host (as suggested by No_Use)
  • the tracks will be loudness normalized - but that doesn't mean that this is a loophole for future contest entrants to work outside recommended level (yes, I am aware of higher reference levels than -18dBFS!). This is still a mix challenge, not a mastering challenge
  • due to the new introduced Loudness Normalization process, we have to make it mandatory to provide a WAV or FLAC. In the future, I might even consider HD-AAC Lossless or even Wav Pack as suitable lossless submission (but I have to dive a bit deeper into these formats first)
  • there is planned to be a PDF with statistics of the current entries. Like: submitted track's average signal strength, max average signal strength, max digital signal strength, comments on rule violations. A PDF is more streamlined than writing that down in the thread

With that said... we hope to start the voting process by Wednesday latest. Taking note of the new challenge schedule, It will last for 7 days again.
[ Mix Challenge ] | [ Studio Page / Twitter ] | [ KVRmarks (see: metering tools) ]

Post

Compyfox wrote: [*]the voting thread will also have a DL pack containing all the current entries for you people to A/B within your host (as suggested by No_Use)
Thanks for taking my suggestion.

Soundcloud does conversion to 128kb/s mp3 for streaming (probably well known anyway) so this package is nice (and I'd recommend) for easily listening the tracks in original quality. :tu:

http://help.soundcloud.com/customer/por ... -distorted

Post Reply

Return to “Production Techniques”