How to build a DIY dub reggae inspired sound system

Start at the start, Dave. Why did you do this thing?

Bass. Sub bass specifically. This is the story of how building a simple DIY sub blew out of all proportion. Because more bass is better bass, and the best bass is more.

B.A.S.S.

It also serves as an [im]practical guide how to hack together a fully working dub reggae inspired sound system in your attic. With the mere application of a few years, many hundreds of hours effort, and some £s.

To recap; The journey started here. I had almost no tools, knowledge or sense. But I had a shed, an attic and an internet. Which is all you need to build a sound system, it turns out. That, and a lifelong love of dub reggae and bass music. And whisky. Lotta whisky.

So, I had a sub as a starting point. Now I needed a vision. And that vision was B.A.S.S. …with the rest of a sound system attached to it. With scant regard to sanity, wisdom or personal freedom, I started reading everything I could about sound systems and speaker building. And made plans. Many, many plans. Like this:

Interestingly (to me) the system ended up looking and performing pretty much exactly as I planned it. Who knew. WAIT! mind blown… B.A.S.S = Build. A. Sound. System.

I digress…

I needed wood (fnurk). Turns out brand name DIY stores only stock really poor quality plywood. As I found out when building my first MVP sub. Why plywood Dave? Good question. Because MDF dust is carcinogenic. And plywood is prettier. But, as with just about everything in this project, it turned out to be much harder than you might first think (fnurk), to get good wood (gnurk). But I did. And it was glorious (gnargle).

One of the things about being an idiot, is that you learn how to cut unfathomably straight lines with fundamentally inappropriate tools. I built this entire thing using a cheapy jigsaw with some old blades that didn’t fit properly, with no jigs or guides – all freehand. I like to think of it as quintessential British pluck. Or idiocy – your shout.

B.A.S.S. – Yes, bass. I had a sub. It worked. It proved a principle, and it was ace. I knew how to build more subs. Problem was – drivers. I needed drivers. Drivers made from pure unobtainium. I mention this now, as I’ll come back to this story later. Spoiler alert – B.A.S.S.

A dub reggae sound system consists of subs (sub bass), bass (kick bins), mids (vocals) and tops (tweeters). All covering off specific parts of the frequency range.

I figured I’d work from the bottom up (Gnurgk). I wanted to design and build everything from scratch. Because being anything other than a complete masochist would be a cop out. I already had a sub, so I started with bass bins (kick bins). That’s when I realised I had bit off more than I could chew. So I figured I’d add some constraints.

My MVP sub used a little 6.5″ driver that made a BIG sound because of the clever chappies that designed the enclosure. Sealed boxes were simpler to build and more fault tolerant in their design. Perfect for numpties with ideas above their station, just like me. So I searched around for 6.5″drivers that might fit the bill. Using a couple of android apps that matched drivers to box volumes, I found Visaton ws-17e drivers. They were (crucially) cheap as chips, and even more crucially, available from a stockist less than a mile down the road from where I live… sweet serendipity.

Wait – android apps, box volume calculators… the actual f*ck?! What I hadn’t realised (before it was too late) was that to build speakers, you need a basic understanding of carpentry, joinery, maths, physics, sound engineering, acoustics, electronics, amplification and Thiele Small (TS) parameters (don’t @me, even now I couldn’t explain). The beauty of the internet is that you don’t need to know any of this stuff. Just find the right calculator[s] online, mash in the numbers with your pudgy meat sack fingers et voila – instant genius.

I’m not going to dive in to the technical details of the size and shape of these things. Keep it light and breezy, that’s the way (Leave a comment if you want to explore the nerd life). Suffice to say I built 4 kick bins – 2 bigger boxes and 2 smaller boxes, using the same drivers – to see what the differences were like. It was all a massive experiment.

Trivia – speakers are generally stuffed with wadding to stop unwanted frequencies reverberating around the inside of the enclosure and escaping through the driver – muddying the sound. The boxes you see above are all stuffed with polyester wadding. These are very simple, reinforced sealed boxes with a driver at one end, wired to a terminal cup on the other end (where you stick in your speaker wire from an amp).

Issue now was, how the hell do I funnel the correct frequency ranges to these things? I had to start buying hardware. And learn how to use it. And make somewhere to put it. So I stopped building speakers for a bit, and built an amp rack. To put my new amps and crossovers in.

Trivia – A passive crossover is a simple network of capacitors and coils, that allows certain frequencies to pass through a circuit, but not others. An active crossover is a piece of hardware (or software) that allows you to dynamically choose what frequencies are assigned to individual outputs. A passive crossover requires diligent planning, design and a strong knowledge of electronics. An active crossover costs £55 from Amazon.

At this point I had acquired three amps – one for sub, one for bass and one for everything else. I also had two active crossovers, because one went bang and I got a replacement. (you get what you pay for…). Proper sound systems have separate amps for each part of the stack, so I was keeping it real. At least that’s what I kept telling my wallet.

Then I started thinking about mids and tops. The tweeters turned out to be mercifully simple. You don’t even need to put them in a box as the dome is sealed. But I built some funky boxes anyway. Then threw them away and left them open. I used Monacor DT-300s with WG300 wave guides.

Because I had active crossovers and multiple amps, everything was essentially modular, so I could grow and test the system organically. For instance, instead of having the tweeters separate and everything else (vocals and bass) coming out of the kick bins (and sounding really muddy), I could use the smaller two kick bins as dedicated mids instead. With the vocals and mid-bass separated out, it sounded loads better. Separating out the sound so each different section does one thing, and one thing well was a win. Those particular Visaton drivers had a particularly wide, usable frequency response which allowed me loads of options.

Then it happened. Two Exodus Anarchy 6.5″ woofers came in to my possession. The drivers the original MVP sub design used, and pure unobtainium in the UK. Their story on its own is mental. Involving airline pilots, more air miles than Tyler Durden, layovers in Utah, San Francisco, New York, London, Kent and finally Bristol. 9 months I think it took in total to get them after I cheekily asked an American friend of my wife to buy some for me.

By that point I had run out of the juices (hrurk) required to build those bonkers tapped horn sub boxes (gneek). So I figured I’d use the last of my good wood (grurk) to build two more Mid boxes for vocals, and slap the two anarchy woofers in a plank. But hooked up to the amps so they could be broken in a bit. Which I did.

Everything was all getting a bit techno by this point and I was questioning what the f*ck I was doing with my life.

To the point where I damn near sold the lot and jacked it all in. Spoiler alert: I didn’t.

But I did have a long break from building stuff and did something more interesting instead. Though I can’t remember what.

Then it was time. I had come full circle. I had put together a fully functional (albeit rudimentary) sound system, but I was missing the final piece of the puzzle.

B.A.S.S

So I built two more subs. This was one of the most challenging things I’ve ever done. It took months. Blood, sweat, tears, the lot. What a journey. But boy was it worth it. I learnt so much. From building your own tooling and guides, through to sanding and finishing with amazing beeswax. All done with zero screws, all glued – for proper geek points. They are beautiful things, and provided I’m not a dick and blow the drivers, will last a life time.

So that was it, mission accomplished.

Or so I thought.

Turns out that with so many different elements, tuning it is a nightmare. Also, as soon as anything goes wrong, tracking down the issue is bloody hard. And don’t talk to me about wires. Never mention the wires. So. Many. Wires.

And then there’s my attic. Which is acoustically, a twat.

But I did it.

I. bloody. well. did. it.

There’s a couple of minor issues. One – I can’t actually listen to it. I have neighbours, and small children. It’s fecking loud. And one of the things about bass, is that it travels. Like REALLY travels. You know whales… gentle undersea leviathans? They talk to each other over hundreds of miles using sub-sonic low frequencies. Two – I can’t get it out of my attic. It’s too heavy, and too complicated to move. So I’m never going to really hear it at its best. Did I mention the acoustics in my attic are wack?

It IS very pretty to look at though.

And the kids dig an attic rave.

Additional 15th Jan 2020:

A DIY sound system is never ‘finished’. I ended up building a 4th tapped horn sub for a nice round number and putting a bunch of extra effort in to adding extra driver covers and making it properly nice to look at.

Which is nice. Because I rarely get to turn it on. 

There’s always the thought “Maybe I should re-build those bass cabs to improve the [excuse to tinker more]…” 

Update: Not sure if I mentioned this but I stripped and rebuilt the first MVP sub and built another sub (both with the 8 Ohm Tang Band drivers) so I have 4 x ‘Insubnia’ subs in this rig spread around the attic. It makes sick wubs, bro.

 

Building Jill the Peg. A Step by Step Walkthrough

I enjoy solving difficult problems, especially in animation, so attempting to bring a sketchbook drawing to life of a lady with 3 legs certainly fitted the bill. I wanted to keep the vibe of the original image as much as possible, so tried hard not to do any extra drawing to make life easier for myself, such as adding an extra buttock for the 3rd leg. I’ve put together the following video and supporting notes to show the process that I have developed over the years for every animation that I produce. It took 2 full day to make this cycle and I’m really pleased with the outcome. A nice way to ease myself back in to the world of animation and scratch an itch with my Life Drawing I’ve been meaning to scratch for years.

Pre-production

The source image was created by tracing the original black pencil crayon drawing from my sketchbook with Dip Pen and Indian Ink on Watercolour paper, scanning and processing in Photoshop to give a clean high contrast line, importing to Illustrator and vectorising the image before finally pulling into Flash and stripping out everything apart from the black line ready to be split apart, made into symbols and animated.

Step 1

The first thing to do is to put together 4 rough keys for the cycle – two passing and two contact points and work out the timings. I picked an arbitrary 40 frame cycle to start with (10 frames per key) and put together the 2 little tests on the bottom left. Perspective is hard in cycles so the ground lines you see reflect the viewpoint of the image and is reflected in the roughs (n.b. Blue for left, Red for Right). I’ve always kept it simple with cycles (see ‘standard’), but after reading the Animators Survival Kit by Richard Williams  I wanted to see whether his approach would be suitable for this walk – he keys in a slightly different way, but is too cartoony for what I’m trying to achieve, so I stuck with my simple route. I then apply the rough keys to each of Jill’s legs, do a quick spine and head on a 1-2 up/down loop with the shoulders rolling as they would on a 2 legged walk. The last stage is to work out how to stagger the legs correctly. Rolf Harris’s sketch Jake the Peg had him walking around using the middle leg forward followed by the outside pair simoultaneously (see video below) – I didn’t want to directly copy this as I was after something a little more subtle, so the outside pair are offset a little here. Trial and error to the rescue until it felt right.

Step 2

The first thing I noticed is that the cycle seemed a little long – she’s walking too slow. I removed 8 frames to test, so we’re down to 8 frames per key for a 32 frame cycle. This instantly felt like more the right sort of speed so I moved straight on to rough in-betweens for the left leg. Once that was done, I copied, pasted and re-coloured the same animation for the other legs so I wasn’t redrawing that same thing over and over again. Once I was happy this was working I in-betweened the spine and head on simple up and down cycles to match.

Step 3

So, that’s the hard bit done, its like putting together a blue-print for the rest of the process. Now its clean-up, rigging and filling in the gaps. Timing is always the most important thing to nail at the start. Making sure you get the foundations right at the roughs stage is the only way to do it. You are a lot less emotionally attached to rough scribbles, so hacking them around and changing the timings / motion is a lot easier and less demoralizing than if you just spent ages making clean art work, only then to realise its not quite working properly. The first thing to do here is split the legs in to symbols for thighs, calves, ankles and toes, then match the symbols to the roughs and motion tween. At this point you have to trust your roughs. You can see where this approach breaks down if you look at the knee joints – you can see them pull apart on the ‘up-stride’, but we’ll fix that later. Last stage is to put in the entire body and head on a simple up/down cycle so you can see the whole thing without blue pencil roughs.

Step 4

The inefficiencies of this approach are made apparent by the ankles on the middle and right legs here. As soon as you make the legs opaque, so you can get the layer ordering correct and place it on a coloured BG you can see the problems. I’ve already started to test fixing the left leg – you can just about see that I’m stretching the calf symbol on the up stride to keep the gaps between knee and ankles more even. The tricky bit at this stage is how to split apart and rig the body so it moves nice and naturally. The knack is to use as few symbols as possible to give nice, subtle motion to the entire body so you don’t drive yourself mad with a symbol for each line. At this point you need to start using ease-ins and ease-outs on all the body symbols to smooth it all out and its very easy to get over complicated. I did play with animating her boob jiggling slightly, but decided not to use it in the end as I didn’t want it to be too distracting. Its a weird enough thing as it is and very easy to try too hard. Looking at it again now I kinda like it, but you can spend forever tweaking things and that way madness lies…

Step 5

We’re just about there. Its now about refining the body symbol animation and getting the layer orders correct before filling in the gaps on the legs and tying up the loose ends. Here’s a close up of the hips as they are in the final spot:

Step 6

This is a perfect example that you are almost finished and should stop fiddling. I’ve tried to finesse the spine cycle to fit with the hips a little better, but ended up killing the feeling of the up and down motion. If you look at the base of the spine it looks almost static. If I was to animate this across the screen, your eye would read it as floating. Not cool. So back to the original version in Step 5 we go.

Step 7

This is the finished cycle. You can tweak things until the cows come home, but the real skill lies in knowing when to stop and move on, as evidenced in Step 6.

Step 8

The last stage is to test that the cycle works as it would be seen in production. Its all well and good working as a static test, but the real proof of the pudding lies in whether it works moving along a path. Its more difficult to get walking across the screen right than you’d think. Perspective makes it even harder. Getting the feet absolutely planted is the goal and it just about works here. Again I could have played with it for ages to get it perfect as she imperceptibly speeds up towards the end and the foot very slightly slides at the start. But that’ll do for me :D

I hope that’s a little window into my workflow – If you have any questions or comments please fire away!