Page 16 of 24

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sat Oct 28, 2023 00:55 CEST
by maxovrdrive
jw_ wrote:
Fri Oct 27, 2023 13:27 CEST
maxovrdrive wrote:
Fri Oct 27, 2023 11:41 CEST
That intro bit, I agree. I just failed (or forgot) to address the issue as it also bothered me. Unusual though that it's not noticeable while on my DAW (reaper), but I do notice it when I'm reviewing the mix through VLC, foobar2000, or my phone. In any case, if I do proceed to R2, that will be addressed. I will most probably adjust levels instead of touching my 2-bus compressor, as making the attack faster will lessen the low-end "bloom" I have.
If you happen to have an opto-style compressor on the bass, might be worth checking out. Those can be notoriously slow on the first hit and may let through more initially than later on when they are "settled". Even some plugins simulate that behaviour. Had that same issue myself with an la2a and ended up just turning down the bass.
Personally, I think it's a combination of bass, guitars, and possibly even the overheads, that cause the buildup on that first hit. I will definitely review and act accordingly.

Thank you for your comments. Much appreciated.

MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sat Oct 28, 2023 22:00 CEST
by Mister Fox
:arrow_right: The "Overview of Submissions" PDFs have been uploaded.
You can check them through the upper post (post #132).



The Statistic Sheet is used to give an overview of all entries, and whether or not they are within given parameters (loudness, sampling rate, bitrate, proper filename). This is adding to the overall learning process of the "Mix(ing) Challenge". Please take note that creating this sheet is not a fully automated process. I am using Wavelab 10's "Batch Analysis" tool (EBU R-128 specs / equivalent to ITU-R BS.1770-4), but the overall layout and highlighting issues/mishaps, not to mention triple checking files and time stamps, is time consuming. If you do not find yourself on this list, please let me know.

:arrow_right: Please keep in mind, your mix being "tagged disqualified", does not mean "you're out of the game" (exceptions do apply). For more information on the Statistic Sheet and the "Wild Card" Mechanic, please consult the following addendum thread:

Mix Challenge - Addendum: Statistic Sheet and Wild Card Mechanic



:arrow_right: Statistics Addendum:
 ⚠ Moderation Message from Mister Fox  
We have a disqualification rate of 23,75% for October 2023 (80 entries, 19 disqualifications, 1 "Out of Competition" submission not counting).

This is about 1/4th of all participants. Many of these disqualifications are due to carelessness errors: wrong bitrate, loudness specs, not allowed re-uploads, and not following the given file name template as mentioned in the provided "TL;DR Rules.txt" file.

On average (28 months of tracking), we currently have a disqualification rate of 26,00% (compared to last game's 26,08% avg), the rate is slightly declining again. The disqualification rate over the course of the last 12 games (December 2021 to October 2023) is about 25,04% (previously 25,32%), and is also on an slight decline.

I strongly recommend all participants to pay more attention to details in future games.



:arrow_right: A commentary on this month's entries:

I would once again like to extend a warm welcome to all new participants who have found their way into our little community. I don't know how you found our small place, but I do hope you enjoy your stay, had fun, and could learn something from this experience.


Unfortunately, creating the Statistic Sheets took longer than expected yet again. Apologies, I currently have very limited time.

The biggest standout to me this month, is using the correct filename template. It is true that the template changed early into the game (7th October). However, it's not like I've posted multiple times over, that there has been a change. Some of you actually adjusted the filename on your file space. Yet there still were things all over the place.

You will notice me introducing a new color on the Statistic Sheet: in the form of two shades of "purple".

The lighter shade is actually where I was very generous this month. I let both the original and the adjusted filename template count. The entries that used the old template, merely got a warning to please pay more attention to details. Those that got marked with a darker shade actually did not adhere to the template, or there was something else wrong (like: weird spaces, a hybrid of the old/adjusted template, etc). These entries have been "tagged disqualified".

This topic is the most frustrating one to me, as this results in a lot of extra work while creating the Statistic Sheet. Additionally, allowing the filename adjustment has been an exception for both MC093 and MC094. If we are to continue this, then I have to triple check links once each game has ended, and that results in more time used for this investigation. The Mix Challenge isn't the only game where a certain filename template is mandatory. It is provided in multiple forms (yes, including the Mix Pack bundled "TL;DR Rules.txt" file). I've been talking about this since way before October 2022.

Please - pay attention to detail!


The next topic is "Drum Sample Triggering / Replacement".

At least two users disqualified themselves with going that route (@mvd83 and @jinjer93). The Song Provider actually has the assumption that more participants went that route, but omitted it from the documentation. The rules have been clear on that topic - this results in an instant "tagged disqualified". Which is even more baffling, considering that this month offered you access to the most clean drum recordings we've had access to in the over 9,5 years of doing mixing games. We never(!) have multi-tracks where drums have been recorded so bad, that you even need to go that route (an EQ and a compressor still goes a long way!).

Additionally, why did some users see a need to "remove" drum tracks (@Kolvanna), or change the arrangement / feel of the song by adding content that wasn't there before (@jinjer93 -- not only using drum samples, but adding a tambourine?! Did you even read the rule book?). The "Mix(ing) Challenge" is not a "producer competition". Your task is to mix! And please let me not get started on a the handful of entries, that omitted either one of the bass tracks (it was mandatory to use both this month!).

Needless to say... as of this moment, that is not(!) a criteria for being "tagged OUT". But should you be selected for Mix Round 2, you need to correct these adjustments.


And finally... once more the topic of "Song Length":

You will also notice that there are two shades of "yellow" on the Statistic Sheet. And not just a handful this time around like usual. I still understand that some of you "cut as close as possible" without leaving room for follow-up edits (fade in/out). Something I do not recommend (as mentioned in the rule book).

Cutting the song "too short" results in various possible issues. For example: playback devices not responding well to the initial transient (same goes for follow-up editing, by the way). Or in case of this song once more, changing the feel of the song by "reducing" the fade out of the guitars at the end of the track. The original entry was listed as 2:55min. You can technically "cut" the length down to 2:53min. But after that, the changes are too drastic.

While "song length" is not yet a criteria for disqualification (exceptions might apply in the future), this topic is something I need to bring up every month now as well. This is why the runtime of your mix is also listed on the Statistic Sheet. To see if your entry is at the same as the original mix (+-1 second tolerance). Keep this in mind for the future.


Other than that... I unfortunately had to pull out two participants from the game, due to re-uploading one's entry (those being: @mvd83 and @rijavia). These entries are "tagged OUT" and can not advance into the next Mix Round.

And while the "other issues" are the usual-usual "careless mistakes" (sampling rate, bitrate, loudness specs), I did notice that a lot of the entries might have actually been mixed "loud" (probably pre-mastered even), and then pulled down in loudness to adhere to the specs. My assumption comes from plenty of mixes that are around -19,0 LUFS ILk to -16,0 LUFS ILk, yet the transients to do not exceed -6,0 dBTP. This could result in a loss of dynamic and transients - which does give the production the "dynamic feel" to begin with. See this video by Matt Mayfield from 2006. By the way, you can still utilize various summing bus treatment to "glue things together", while not(!) having to "push for loudness" to be impactful (do not trust the "wild west" of Youtube Tutorials that "mixing/mastering to -xyz LUFS is nonsense!", as if anything released between 1970s and early 2000s, with "reasonable loudness" due to medium limitations, never existed!). Just something I've noticed this month.



:arrow_right: To close this out.

Both @BenjiRage and I are happy to see this amount of participation. Although I am not a fan of having to extend the deadline for two games in row. Without the time extension, we would have reached about 51 participants. This isn't bad per se. However, this makes me question if people don't receive enough information on games (newsletter and social media). It also doesn't help, that most of the entries basically arrive during the final 48 hours.

This topic does have an effect on the yearly statistics of course, and is something for me to figure out over the course of the coming months. If there will be future Mix(ing) Challenges, as our Song Pool is empty! And no, this is not a "but you always found a solution, Mister Fox" type of deal. There is legit nothing available for 2024. MC095 in December would be the last game as of this moment.


A huge thank you to everyone that invested time with this month's game. I hope you had fun.

See you in the next one. :headphones:
.




:information_source: I now (re-)open the field for everyone to give each other feedback (highly encouraged). This will be independent to the client feedback however. Though please note - criticism on the rule set and angry posts will be deleted without further notice.


Please watch this spot for the client feedback and Mix Round 2 participant announcement.
This will ideally happen sometime within the next 14 days (starting tomorrow).

We (as in: the Song Provider and I) will keep you updated, and of course send out appropriate reminder newsletters!



EDIT: 29-OCT-2023 13:35 UTC+1/CET - slight text corrections

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sat Oct 28, 2023 23:13 CEST
by JeroenZuiderwijk
Hi Mr Fox, I think there is a mistake in the overview of the loudness specs. I checked and just now double checked and my mix peaks at -1.2dB max! I paid extra attention to it with my mix. Could you please check and comfirm that I am not disqualified?

Thank you, Jeroen Zuiderwijk

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sun Oct 29, 2023 01:25 CEST
by BenjiRage
JeroenZuiderwijk wrote:
Sat Oct 28, 2023 23:13 CEST
Hi Mr Fox, I think there is a mistake in the overview of the loudness specs. I checked and just now double checked and my mix peaks at -1.2dB max! I paid extra attention to it with my mix. Could you please check and comfirm that I am not disqualified?
Hi there Jeroen

I've just given your mix a quick check myself. While I don't use the same software as Mr Fox (I use the TC Electronic LM2N plugin), I'm still getting readings of -1.1 dBFS for the left channel and -0.8 dBFS for the right, which seem to match the statistics sheet.

I wouldn't worry too much though as you're not out of the game totally, it just means you'll join the wildcard club if you get through (I'm already a member :lol:)

Best regards

Ben

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sun Oct 29, 2023 01:29 CEST
by Mister Fox
JeroenZuiderwijk wrote:
Sat Oct 28, 2023 23:13 CEST
Could you please check and comfirm that I am not disqualified?
I just triple checked your entry.

Code: Select all

Wavelab 10.0.x (used for the batch analysis)
Digital Peak L/R:   -1,2 dBFS / -1,2 dBFS
True Peak L/R:      -1,1082 dBTP / -0,8882 dBTP
LUFS:               -16,0 LUFS ILk

NUGEN Audio VisLM (ITU-R BS.1770-4 preset)
True Peak max LR:   -0,9 dBTP
LUFS:               -15,9 LUFS ILk

Youlean Loudness Meter 2 Pro (v2.3.1, ITU-R BS.1770-4 preset)
True Peak max LR:   -0,8 dBTP
LUFS:               -15,9 LUFS ILk

iZotope RX7
Sample Peak Level:  -1,20 dBFS / -1,20 dBFS
True Peak L/R:      -1,10 dBTP / -0,86 dBTP
LUFS:               -16,0 LUFS ILk
Your DAW is therefore showing you dB Full Scale ("Sample Peak"), not dB True Peak.
To be on the safe side, please use a dedicated signal analysis tool.

This list also shows why I opted to introduce a headroom for allowed offsets (see accuracy between metering tools)

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sun Oct 29, 2023 01:33 CEST
by JeroenZuiderwijk
Mister Fox wrote:
Sun Oct 29, 2023 01:29 CEST
JeroenZuiderwijk wrote:
Sat Oct 28, 2023 23:13 CEST
Could you please check and comfirm that I am not disqualified?
I just triple checked your entry.

Code: Select all

Wavelab 10.0.x (used for the batch analysis)
Digital Peak L/R:   -1,2 dBFS / -1,2 dBFS
True Peak L/R:      -1,1082 dBTP / -0,8882 dBTP
LUFS:               -16,0 LUFS ILk

NUGEN Audio VisLM (ITU-R BS.1770-4 preset)
True Peak max LR:   -0,9 dBTP
LUFS:               -15,9 LUFS ILk

Youlean Loudness Meter 2 Pro (v2.3.1, ITU-R BS.1770-4 preset)
True Peak max LR:   -0,8 dBTP
LUFS:               -15,9 LUFS ILk

iZotope RX7
Sample Peak Level:  -1,20 dBFS / -1,20 dBFS
True Peak L/R:      -1,10 dBTP / -0,86 dBTP
LUFS:               -16,0 LUFS ILk
Your DAW is therefore showing you dB Full Scale ("Sample Peak"), not dB True Peak.
To be on the safe side, please use a dedicated signal analysis tool.

This list also shows why I opted to introduce a headroom for allowed offsets (see accuracy between metering tools)
Thank you very much for the explanation and information. It is confusion I must confess, but I'll see what I can do for the future with another analysing tool.....

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sun Oct 29, 2023 01:49 CEST
by jeffssoloband
JeroenZuiderwijk wrote:
Sat Oct 28, 2023 23:13 CEST
I think there is a mistake in the overview of the loudness specs.
Hi Jeroen
Just curious: what software did you use for your analysis?
—Jeff

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sun Oct 29, 2023 02:38 CET
by 1ktone
2 things I missed this month…feel somewhat foolish… :bang:

The filename template includes double underlines and I have now just realised that this is common in the name templates…copy - paste really is the way to go…bummer….and secondly Logic got me with it’s automatic 48k rendering when I had changed that rate earlier in my session to export to 44.1.

At least I know that the LUFS meter in Logic is accurate. I saw 16 when I rendered and even though that’s close to limit I needed to know how accurate the reading is. So thank you for the measurement feedback.

Oh well, Thanks for this month’s fun. Hopefully there will be more in the future… :tu:

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sun Oct 29, 2023 07:06 CET
by kombainera
I'm really sorry that there are no songs in the song pool, but there's nothing I can do about it. I want to ask the experienced ones how do you actually approach when you start mixing gain stage wise. My way is to normalize all the files in the project to -18 true peak db, in this way I am sure that I give a suitable level first for all plugins then for summing busses and mix bus. It can be a little quiet in general but it's not a problem at least for me because I can amplify the signal in many places so that I can hear it at suitable levels. If it is mixed through a mastering chain its natural got loader when a compressor is added and saturation. If i need to send something for mastering i think its better to have a lot more headroom then to send a mix at -3. Is this the best way or are there better practices? In 64 bit float point summing gain stage is more or less not needed at all since u can adjust at the mix buss at any time but its not the same with the plugins it matters how hard i hit them at least for me.

Re: MIX CHALLENGE - MC094 October 2023 - Mix Round 1 in evaluation

Posted: Sun Oct 29, 2023 07:47 CET
by JeroenZuiderwijk
jeffssoloband wrote:
Sun Oct 29, 2023 01:49 CEST
JeroenZuiderwijk wrote:
Sat Oct 28, 2023 23:13 CEST
I think there is a mistake in the overview of the loudness specs.
Hi Jeroen
Just curious: what software did you use for your analysis?
—Jeff
I check the 'render stats' in Reaper when I render the mix. The readout says: peak -1.2dB. That is why I am confused. I had no idea the readout was not accurate. My mixes where always within the loudness rules, so never had this problem before. So I downloaded Voxengo SPAN this morning.....and indeed, there is another readout. So that is what I am going to use from now on.