Thursday, September 10, 2020

Templates, or How I Learned to Stop Worrying and Love the Workflow.

 


I've always wanted to be a bit better than I actually am. That being said, it's not necessarily a bad thing, as it has constantly forced me to learn new things and advance in my craft.


What, however is rather bad, is that I've never accustomed to do things like the others. If I see everyone and their friends using method X or sample library Y for fast and efficient results, you can bet your ass off I will turn away, forcing myself to find altenative methods instead. Because surely, if they all do this because it's easy and fast, everyone can do it- and I should be better than that everyone, whoever that is. 

 

And more than once, it has turned out there is a good fucking reason why some methods are more popular than others and I've been mostly just wasting my time on trying to get the same results, only slower and sometimes even worse.

 

 

Templates are a good example of this. For those who don't know what it means, those are DAW (e.g. Cubase) projects which more or less consist of VST instrument tracks which are chosen, balanced and mixed completely in advance. Which means that you're just basically performing into a readily-mixed and instantly deliverable project. Which in turn for me, the eternal adversary and emperor of cynicism, means that "I'd always sound the same and would always just rely on the same sample libraries".  Then again, I'm pretty sure if you didn't know what that is, you may find the rest of this article rather boring. (And for the rest of you, if you're only interested what's my final take on the new workflow instead of my journey figuring that out, there's a new, more technical post coming on the subject.)

So, where was I again? Yes. I've avoided templates my whole career for these incredibly both smart and credible reasons. Instead, I've done it the hard way, as usual. Creating a new project every time from zero, mixing in another and finally mastering in yet another (all being created from scratch every time), has taught me a lot in the process in this neverending quest for getting better, but it's also slow as fuck. And if there is a request for a change after the master has been delivered, at worst I'd have to go back to the original arrangement and start from there again. Not to mention that as we have drifted towards adaptive music in our games during the last years, bouncing separate stems is rather cumbersome if not even impossible to an extent using this method. "Slow, deep and hard" sounds like a great name for a metal album, but not for daily work routines. 

A month ago at work I got a task to do some placeholder music, so I decided to just have some fun and chose a couple of stupid and overclichéd instruments to start with. Add some deliberately unimaginitive C-Am-Dm-G pisstake of a song with ukuleles and whistling and you have that stuff one hears in toilet paper or holiday commercials. However, after a couple of hours I noticed myself being embarrassingly delighted of that abomination of music, but not because of it's tonal values. During the process, I had absent-mindedly added some corrective EQ:ing here and there, added some reverbs and a couple of more tracks of noodling- and realized suddendly that this whole mix sounds very...ready! With a childish excitement I wanted to test something right away, so I made another song in ten minutes using the same tracks and a huge grin spreaded to my already wide face. I could re-use this whole thing in the future by just deleting the old midi data and playing new stuff to replace it every time when a palette like this is needed! And without having ever to do it from a scratch again! Sounds a lot like...yep, I think I had just invented the concept of a "template". Le sigh.

Having now accidentally made my first template ever (retrospectively thinking, finding solutions on my own terms seems to be the keyword in all this) I realized the most important thing I had blatantly misunderstood with the concept: you're not supposed to build a MOTHER OF ALL TEMPLATES but several ones instead. So, I could have a "casual" template, a "light orchestral" template, a "heavy RPG" templ...wait, what? *uses a potion of charisma, orcish pendant of luck +3 and walks to bossman*

 

By the blessing of (Satan) my boss, I started to work with the RPG template. I spent two days on balancing different instruments together I knew I would use, routing different mics into different mixer groups and trying generally to get myself as much flexibility as I could with the close and hall mics of various instruments. Learning from the previous task, I decided that I will not do any more processing that is strictly necessary, but spent a considerable amount of time going through every single new track with Fabfilter's ProQ-3 in order to even out the worst resonances, getting rid of unneeded frequency information and most importantly, placing everything nicely into the same stereo field and space in general utilizing M/S processing. (Pro tip: most of the "roomier" samples you encounter will sit with the rest nicely when you carve enough sides off from the lower mids.) For me, getting 90% there with the sound using only this one plugin made me realize how much sample mixing differs from mixing real instruments.


 

I decided to implement more ensemble patches to my workflow instead of using single sections which I normally utilize when writing orchestral music. The reason for this wasn't only because it would speed up the process, but also because it makes everything sound a bit less superimposed and much more glued together section-wise. And as my aim was more closer to production music anyway instead of a real recording in mind, I didn't see any reason why I shouldn't take advantage of the samples.
It didn't take long, however, to realize that I have absolutely no fucking clue how I should be using them because I had become so accustomed to working with single sections. So I did what any reasonable composer would do and cried for help at the great VI-Control forum, which has become my number one source for anything composing-related during the years, filled with incredibly talented and helpful people. Getting valuable tips right off the bat helped to me to realize that I'm just once again overthinking everything into eternity. 

After getting a grip on how to use all those patches, I marked down all the chords of the piece I am working with in order to remember what I'm actually supposed to play and blasted away. Reinforcing and adding lines with single instruments, adding and EQ'ing more tracks if needed and generally expanding the palette, this whole "doing first, thinking later" kept me having more fun than I've had for a long time. Another lesson learned- no matter how badly you want to develop your skills, don't lose the fun while doing it.

I ended up fiddling with different mic setups still for a couple of days, added linked sidechain bus compression for all exportable stems, rerouted half of the instruments, did some minor mixing for some groups and finally declared victory for the Orcish Horde. I deliberately left all advanced mixing, automation and other tweaking off as I'd still want to have some room left for tailoring the sound suitable for each project separately and also because in mixing, there's are no "one size fits all"-solutions.


My awesome RPG template was now officially finished to the point I could easily use it as the base of any upcoming material, but I still wasn't satisfied. Because the HEROIC MODE was still awaiting for me! Having been first starting with a smaller and simpler template, expanding it into a more complex one, I figured out that it was finally time to do something I could use as the base for all new ones. I had learned some valuable things on what could be done, should be done and what definitely shouldn't be done, and was screaming to taking advantage of all that. And I already had a bunch of ideas what I still wanted to achieve. But let's first see what I've learned thus far....

DO'S

- Many templates! Handle it! Project-specific templates are fresh, fun and useful. Keep on doing more while expanding your repertoire.
- Use those ensemble patches, one-trick-ponies, chord patches and symphobias. Nobody cares. In fact, fucking embrace them.
- In order to fit different sample libraries into the same space and define their individual depths, rely on mic positions and (M/S) EQ first.
- When balancing things together, the best way is to compose a short piece/ snippets where you deliberately use as much stuff as you can.
- Keep all the faders and panners at unity. All balancing should be done either in the VST instrument or using the channel input gain first.
- Plan your routing beforehand. Messing up the balance by changing groups on the fly is NOT fun.
- Hard to decide which timpani miking you should use? Load another to the same Kontakt using a different midi channel.
- Your favourite libraries don't always give you what you want. Don't be afraid to use something else if needed.
- You can stack articulations if it sounds better and gives you what you need. Nobody gives a shit.
- Be organized. Use folders, color coding and comprehensive names. The bigger your template grows, you'll quickly understand why.
- Disable and purge tracks mercilessly. You're not going to need that brushed flautando recorded on a bottle at the edge of silence anyway.

DON'TS

- Don't process too much on the go. You can always do that in the end when everything else is done. If you really have to.
- Speaking of processing, leave all compressors out until mixing. That old setting doesn't work with anything new without tweaking.
- Don't fucking pan that 1st violin room mic from the mixer or you end up twisting the whole room. (unless you're Grateful Dead.)
- No, you can't rely on close mics alone no matter how experienced you are in mixing. (that one took me two day's work to realize)
- Unless you're doing pure classical music, don't fall into the trap of "realism at all costs". Your reference film score isn't realistic either.
- There is a difference between "wrong amount of reverb" and "a wrong type of reverb". Learn to differentiate those two.
- You don't need a stereo widener in 90% of the cases. In fact, you need more narrowing. Woodwinds, I'm looking at you.
- Don't overcomplicate things if it's not needed. If it sounds good, it is good.
- Stop having too many options. Pick something, move on and change if it doesn't work. Don't load up different 8 solo violas "in case you need them". You don't, unless you do. Before that, don't bother.

To be honest, I'm not sure why I even wrote this all down. Maybe because I was just so excited about suddenly finding a completely new way to work? Or to remind myself later why I love my job? (hint: it has a lot to do with excitement and fun!) Or just to use it as a too long prequel to talk about my new workflow later? In fact, I even removed a lot of useless babbling in order to keep it even somehow tolerable and mildly interesting to read at times, yet I expect someone really to care that "HURRAY I MADE AN UKULELE VAMP WHICH MADE ME REALIZE HOW TO EQ TRUMPETS A WEEK LATER". But hey, sometimes that fucking stupid ukulele vamp straight outta toilet paper commercial can lead you to greater things which you weren't even thinking at first.

And the purpose of all this is basically just about that; that I feel I've accidentally found something which I thought could be useful to someone else too, so they wouldn't necessarily do the same stupid stubborn mistakes I have done. Learning is fun, but bullheaded fixation on things makes it quickly less fun and more desperate at worst. And man, am I good on the latter!

In the next part, I will write more about how am I utilizing these ideas when building a new template, articulation-mappings, workflow, exporting things for Wwise and whatnot. Unless I get stuck on building that new gargantuan space opera template with a tarantinoan touch in an imaginary Blizzard universe, in which case it might take a while. (that, or completing the latest Tomb Raider with the missus, whichever happens first. Death to Kukulkan!)

7 comments:

  1. Näitä lisää! Itse studio- ja tuotantohommien kanssa aloittelevana näistä teksteistä saa kyllä paljon inspistä ja ideoita. Piti ruveta selaamaan kaikki aiemmatkin blogitekstit läpi.

    ReplyDelete
    Replies
    1. Kiitos, pitääpä tehdä lisää selkeästi! Seuraava teksti tulossa viimeistään ensi viikon alussa, jossa saa kyllä sopivasti päätä pyörälle aiheesta. "The Art of War, pt II" on myös semmonen missä sivutaan metallituotantoa demovaiheessa mikä kannattaa myös tsekata jos et ole ehtinyt jo!

      H

      Delete
    2. Art of War-tekstit tuli lukaistua niiden ollessa uunituoreet vielä. Luin vielä uusiksi molemmat ajatuksella. Sen pohjalta innostuneena ilmoitin oman bändin tyypeille, että nyt me aletaan jatkossa myös tekemään kunnon demot ennen varsinaisia studio-rupeamia. Saa kyllä sitten varmasti paremman kuvan mitä ollaan oikein soittamassa ja tekemässä.

      Tähän asti ollaan Guitar Prolla tehty biisit valmiiksi, reenattu X määrä (treeniksellä missä ei oikein ota mistään selvää - parhautta) ja sitten nauhoitettaessa ihmetelty että tältäkö tämä oikeasti kuulostaa. Koita siinä sitten miettiä mitä sen tuotannon kanssa haluaisi tehdä. Sitten kun levy on jo about masteroitavana, alkaa tulla mieleen kaikkia juttuja mitä ois voinut koittaa tai jopa jättää tekemättä.

      Delete
    3. Kannattaa tosiaan ehdottomasti tosiaan pyrkiä tekemään demot jo niin loppuun asti mietityksi kuin pystyy! Yleensä sitten äänitysvaiheessa ne tuntuvat vielä paranevan kun voidaan miettiä uusia ideoita siihen päälle tai vaihtoehtoisesti huomata että joku asia ei toimikaan niin hyvin bändin kanssa. Yksi tapahan on tietenkin myös äänittää niitä treenejä ja sitten kuunnella niitä jälkikäteen. Ja haukkua rumpali! Aina!

      Delete
    4. Meillä menee kyllä niin päin, että rumpali (eli allekirjoittanut) haukkuu kaikki muut :D

      Ei varsinaisesti liity juuri tähän postaukseen, mutta onko suositella kokemuksesta jotain hyvää EQ-softaa? Tällä hetkellä käytössä vaan Reaperin oma ja se on jotenkin... kömpelö? Ja saa kaiken kuulostamaan huonolta. Siis siinä mielessä, että ihan kuin Reaperin EQ laittaisi kaiken päälle jonkun "bändin eka demo"-filtterin. FabFilterin EQ tässäkin postauksessa vilahtaa, mutta onko joitain muita ehdottoman tsekkaamisen arvoisia? Vai kannattaisiko suoraan hypätä FabFilterin tuotteisiin?

      Delete
    5. Fabfilterit on kyllä ihan tajuttoman kovia, etenkin se EQ. Suosittelen lämpimästi tsekkaamaan esim. Black Fridayn aikaan jos olisi jotain alennuksia tjsp! Mulla on muutamia vähän ns. characterimpia EQ:ita kanssa mutta tuo on kyllä sellainen aution saaren vehje että ihan äkkiä ei muita tarvitse. Ja jos edelleen kuulostaa sen jälkeen paskalta niin sitten vika on enää lähtösoundissa tai käyttäjässä. :D

      Delete
  2. This comment has been removed by the author.

    ReplyDelete