MIX CHALLENGE - MC098 June 2024 - Mix Round 1 in evaluation

[Mister Fox]

Please keep it civil, everyone.

I'll also quickly step in and clear some questions.

I do not know how in the world I got a bit rate of 15 bits, so I should definitely review my mix bus, but I was aiming for 16 as it was the format provided, ...

A "dropped bitrate" (e.g. 23bit from 24bit, or 15bit from 16bit) usually hints at the available dynamic range not being fully used. For example: if your track is in a reasonable loudness (like -18dB RMS avg, or in case of the Mix Challenge -16,0 LUFS ILk max), then your maximum signal strength (your transients, which basically reflect the "punch" of a production) is allowed to peak up to -1dBTP. Although a more realistic value is usually -3,0 dBTP, unless you have some really strong / rogue transients in your mix. If the maximum signal strength falls below a certain threshold, analysis tools then show a "drop" in bitrate. This isn't necessarily a bad thing - just additional information.

If it is the other way around, e.g. showing 25bit, even though file is in 24bit -- then your signal strength is way too high (clipping).

Your submitted entry is indeed in 16bit, not 24bit.

...I am new to the Mix challenge and only now discovered that the "rules" files attached to the download is specific to the song rather than just a txt copy of the rules on the forum.

Fine-Print: The bullet points written in the TL;DR Rules.txt are indeed the same as the TL;DR Rules on the forum.

However, the "Important Stats to check" section is adjusted for every game - listing the required specs, maximum allowed loudness, and the filename template. This TXT was finally added in 2023 as an absolute baseline for every participant - for the simple reason that not everyone reads what is written in the forum, which led to endless, repetitive, tedious debates that I no longer have the patience for.

However it makes no sense to convert a 16bit sample to 24Bit in post production phase and there is furthermore no need for 24 bit in this case.

It doesn't matter if it makes sense to you or not. The rule book clearly states in the Rules for participants of the Mix Challenge, sub-section "Upload and Submission Guidelines", bullet point 13 (3rd from bottom of this section)...

Unless otherwise stated, export in the sampling rate and bitrate the material was provided in. However, bar minimum in 24bit. Examples:

• If a multitrack package was provided in 44/16 - export in 44/24
• If a multitrack package was provided in 44/24 - export in 44/24
• if a multitrack package was provided in 44/24 and the end result needs to be delivered in 48/24 - export in 48/24

I also made that once more clear in post #001, the reminder in post #003 and the aforementioned Mix Pack bundled "TL;DR Rules.txt".

Only having access to 16bit recordings this month, was due a hardware limitation of the equipment the song provides were using. We might run into this topic again in a future game. These days, however, follow-up processing (pre-mastering / finalizing for distribution) is done in either 24bit integer or 32bit float, and colleagues in this area will always ask for 24bit bar minimum.

Not seeing the error on my file name. Any suggestions? No spaces even though there are spaces in the name?

You sadly didn't adhere to the provided filename template.

Code: Select all

Filename Template: MC098__Arnwyn__Robot_Baby__ForumUsername.wav
Entry Filename:    MC098_Arnwyn_Robot_Baby_quaint twang.wav

Technically, you didn't even use an underscore between "quaint" and "twang", which can result in access errors depending on the server the file is posted. I posted about the reason for this enforcement multiple times on the forum, most notably in this news post from October 2022.



Both of the just commented on entries are not "tagged OUT", merely "tagged disqualified". If the Song Provider offers you a spot for Mix Round 2, you can still advance with the help of the "Wild Card Mechanic", and fix your small mistakes in the process. Please consider this a learning experience.


This should cover the most common questions.
Thank you for reading

[zed999]

Anyone care to comment on dealing with the stick hit?
On many, perhaps the majority of the uploads I've listened to, it sounds very, very separate not just to the kit but to the entire track, like someone is in the room with me tapping on something.

[sthauge]

the audio rendering engine in a DAW has to do gazillions of multiplications, and therefor the highest possible internal math resolution should always be used, and from there you can dither down to 24bit, and the quality will be better than at 16.
engineering means to know what an engineer knows. Let us all strive for good and rich knowledge. I seek to know more myself as well.

topic to move:
should we deactivate hyperthreading on windows DAW systems? does it need a change in BIOS settings?

Here's is an explanation I did in an other forum about Hyper-threading(HT):

"If you have a Intel CPU, one thing you can do to reduce XRUNs is to turn off CPU hyper-threading. HT is Ok if using eg. MS Word or browsing internet, but have no place in audio processing. Audio is number crunching. I have a Intel i7 CPU. It has 8 threads, but two and two threads share one arithmetic unit, this means that before or later you will get constraints on the arithmetic unit and BOOM, you'll get XRUNS. My experience in a sessions with high DSP and XRUNs, I got 20% less DSP consumption and no XRUNs if turning HT off."

A metaphor could be: you have a road with two parallel lanes(two threads). It's traffic in both lanes with cars(audio that needs processing) going at the same speed and they can't brake. Then the two lanes suddenly become just one lane(shared arithmetic unit). Since the cars in both lanes are moving at the same speed, sooner or later they will crash(XRUNs) where there's only one lane. The more cars, the more crashes(XRUNs).

In Linux I can use a system command to turn it off, but turning it off in BIOS might also be an option.

Steinar

[Davias]

My apologies for the sparse documentation, I posted the song in a rush believing I was past the deadline.

A bit more details about my mix :

Tracks :
- Gain Staging
- Gating to clean certain stems
- SSL Console saturation (Satson)
- EQ to remove unwanted frequencies (Kirchoff)
- Compression/Saturation/Stereo Imaging when needed
- a bit more SSL flavor with BX Console SSL 4K G
- Limiting/Clipping if needed (LoudMax/GClip)

Busses in general :
- SSL Saturation with Satson Buss

Drum Buss :
- Clipping
- Vari-Mu Compression (Pulsar Mu)
- Tape Saturation (ReelBuss V3)
- Limiter

Bass :
- Usual cleaning then
- Multiband Compression
- Bass Boost
- API style Compression (Lindell SBC)
- Top end saturation

Space :
- Valhalla Vintage Verb on vocals and some instruments
- Valhalla Room for drums and some other instruments

Mix Bus :
- SSL Bus style Compression (SSL Native Bus Compressor)
- Pulltech EQ (Warmy EP1A)
- Tape Machine (TRacks)


About the Stick, since someone asked how to gel it with the rest, I hope I did that well, I tried to use a transient shaper to give it more tail, sent to the room reverb to give it a space, and then hope that the vari-mu compression glue it well with the rest.

It was a pleasant track to mix, with most of the material clean, the bass was the worst offender with heavy resonances

[zed999]

About the Stick, since someone asked how to gel it with the rest, I hope I did that well, I tried to use a transient shaper to give it more tail, sent to the room reverb to give it a space, and then hope that the vari-mu compression glue it well with the rest.

Thanks for taking the time to explain your approach.
I am interested in how others tackled it and what they thought about it.

I've just listened to your mix and it sounds as I described earlier to me - very loud compared to the rest of the kit and overall mix but you clearly liked that and there is wrong or right. I think this was the most difficult thing to judge level in this mix. Usually we might struggle to get an instrument to cut through the mix and this was exactly the opposite - quite difficult to prevent it sitting on top of the mix and sounding like a click track.

I should not ask without saying how I tackled this - I only used eq, high passed the sub clutter, low passed at 6KHz/12dB/octave, 6dB bell cut at 3.5K, 6dB bell boost at 200Hz and a little more reverb than the rest of the kit. After that I found myself adjusting the level down until it was almost lost. I found listening at quite low level but from my kitchen really helped with this because to my ears it overwhelmed until it almost disappeared so I settled for that in the end. Who knows?

I make no claims to success and of course, there is no correct, it's taste.

[Strange]

Anyone care to comment on dealing with the stick hit?

This was my first challenge:
First I lowered every syncopated stick about -6dB gain to enhance and vary the groove.
The balance between the hihat and the stick is also important for the groove.
Then I used the same Neve-Channel Plugin for all drum tracks, but I treated them differently.
All drums are grouped and got the same soundcity-drumroom.
I hope it worked...

[BenjiRage]

Anyone care to comment on dealing with the stick hit?
On many, perhaps the majority of the uploads I've listened to, it sounds very, very separate not just to the kit but to the entire track, like someone is in the room with me tapping on something.

Hey Zed

To be honest, I can't say I had any issue with how it sounded. From a drummer's perspective, it's a pretty standard cross stick! It did have more low frequency than it needed, so a bass rolloff was the first thing. Little bit of compression and a boost of the highs, followed by some reverb with a shimmery tail in addition to the standard drum room verb and it I found it slotted into place.

A consistent, quality reverb with modulation helps to get it sitting right and not sticking out as much. This goes for all drum elements, you want them to sound cohesive like they've been performed together in the same room - I always use Valhalla Room on my drums for this task, it's golden.

In your mix though I think the stick sounds fine, nothing really out of place about it. One thing that does seem out of place though is the right crash, it's two bars behind where it should be!

Best regards,

Benji

[zed999]

In your mix though I think the stick sounds fine, nothing really out of place about it. One thing that does seem out of place though is the right crash, it's two bars behind where it should be!

Best regards,

Benji

Oh heck, my deaf right ear missed that, I have no idea how it happened, I'll have to investigate. and
Hopefully the song provider likes my mix enough to give me a chance to correct this clumsy error. Thanks for pointing it out.

[elements]

Anyone care to comment on dealing with the stick hit?
On many, perhaps the majority of the uploads I've listened to, it sounds very, very separate not just to the kit but to the entire track, like someone is in the room with me tapping on something.

In the end I pulled out mostly around 500 - 1K, boosted about 1.5db at 4.6K with a dynamic eq and put it through a blend of the logic distortion plugin. I find that the distortion plugin helps to round things out a little while giving it a little presence. Actually one of my favourite things to add on snares as well.
I wasn’t trying to emulate a kit as such….

[zed999]

Anyone care to comment on dealing with the stick hit?
On many, perhaps the majority of the uploads I've listened to, it sounds very, very separate not just to the kit but to the entire track, like someone is in the room with me tapping on something.

Hey Zed

To be honest, I can't say I had any issue with how it sounded. From a drummer's perspective, it's a pretty standard cross stick! It did have more low frequency than it needed, so a bass rolloff was the first thing. Little bit of compression and a boost of the highs, followed by some reverb with a shimmery tail in addition to the standard drum room verb and it I found it slotted into place.

A consistent, quality reverb with modulation helps to get it sitting right and not sticking out as much. This goes for all drum elements, you want them to sound cohesive like they've been performed together in the same room - I always use Valhalla Room on my drums for this task, it's golden.

In your mix though I think the stick sounds fine, nothing really out of place about it. One thing that does seem out of place though is the right crash, it's two bars behind where it should be!

Best regards,

Benji

Just had another listen to your mix now I've got over my cymbal blunder. Like many musicians I've dabbled with drums when I got the chance and my son was a keen drummer at one point with a kit set up in our basement so I like to think I have a handle on this. What I hear in your mix is that you're one of relatively few you got that stick hit in proportion, it's not on top of everything, you got the level "right" in my humble opinion.

Thanks to everyone else who replied on this aspect, I see many different approaches to eq/comp etc, all valid of course.

I think the level was the most important aspect.

Side topic would be judging transient levels on drums perhaps. Very difficult in headphones and very difficult with loud playback. For me the best check is to turn the whole mix down until it's almost silent. As you approach silence what do you hear? Another is to listen from a distance, for me that's my kitchen - it's a similar tactic - if the snare (very often) or in this case stick hit dominates, it's too loud. For my sensibilities the stick hit can't be louder than that snappy snare hit, it doesn't make sense. I worked back from the toms, levelled the snare then the stick. I do like drums to sound like a kit.