BrandonZahl.com

Archive for the ‘Speaker Project’ Category

Shop Day

leave a comment

Let’s talk about my first day in the shop. Where the rubber met the road. Where the wood met the saw. It was a big day for me– I had stayed up late the night before, finishing a step by step plan on everything I needed to do. At $100/day for shop time, it would be painful to waste any time. Plus, I was excited to blast through this and get to enjoy the final product. I directed my energy to planning every aspect of the shop day. All the cuts were calculated and written down, each assembly step and test step noted precisely, all the supplies and materials listed. It was going to all go off smoothly, right? I’d be ready to have a speaker listening party the next weekend, I was sure.

So I showed up at the shop with a Home Depot rental truck loaded up with big sheets of wood and other supplies. I found John and got unloaded. I walked him through my plans and drawings. He’d ask a question at some points and then nod at others. Then he just said straight up that he didn’t think it was going to work. The ground fell out beneath me and I couldn’t believe what he said. I started feeling indignant. How did he know? He’s never built something like this before. People had used this same technique before and it worked for them. I’m glad I just shut up and listened though.

John’s counter-proposal was that instead of bending big sheets of wood along the curve, we would cut long, square wooden sticks and line them up vertically along the curve, like a curved fence. I imagined all the gaps in between the wood sticks and cringed at the potential of sound leaking out through them. I felt sure he was missing this key point but he was making other good points about the sheets of wood I was trying to bend. They were just too floppy and even if we clamped it down with straps to bend, there would be places in between the straps where the wood would bow out again. Those guys on the internet seemed to make it work, but the curves they were bending weren’t as tight as mine. Plus maybe they did have imperfections that just didn’t show up in the photos. My plans were starting to unravel.

I still remember exactly where I was in the shop as my head tried to make sense of what to do next. My stomach clenched and my thoughts felt clumsy and detached. It was so hard to abandon the plan that I’d meticulously put together, but it was starting to seem foolish in the light of John’s expertise. He’d spent his whole life woodworking but did he have any expertise with sound? How could I have confidence that these wouldn’t be musical sieves? Was it worth abandoning a plan that people on the net made seem simple and successful?

Finally, I figured out we could fill the gaps between the sticks with Bondo (autobody repair putty) to prevent the sound energy from leaking out. I didn’t know how much it would help. It wouldn’t be as good as solid wood because of the inevitable gaps where the Bondo didn’t reach but would that really matter? Would the tiny gaps filled with air cause some kind of vibration? I wish I understood physics a lot more but I figured this was worth a shot.

So I agreed with his approach and we scrapped my plan. Part of me reeled as those words came out of my mouth but after that moment, it was too late.

Thankfully we didn’t have to abandon the whole plan. The construction of the frame that the sticks would attach to was the same and so I had all those steps lined out. I was glad to start out on something more straightforward and John helped me get familiar with the tools and procedures I’d need to use.

We started out cutting the large pieces necessary for the fronts of the speakers and for the subwoofer (almost forgot about that, right?). He was trying to explain the difference between different types of cuts on the saw and showing me the right techniques but it just wasn’t soaking in. I think a combination of lack of sleep and stress from the plans getting scrapped had rendered me dumb. Thankfully he was patient and made most of the cuts himself. This is all part of the other service he offers– his mentorship.

To riff on that a bit– the hands-on mentorship approach is what makes his shop, IsGoodWoodworks, unique and what kept my project on the rails. From his re-write of my flawed design, to his patient instruction of tool usage and all the little tricks he showed me along the way– his mentorship was invaluable to this project. I had his lifetime of experience at my service and I probably wouldn’t have turned out much without it. Thank you, John!

So I came to appreciate his mentorship pretty quickly that first day. We didn’t end up making it very far, but I did leave feeling like I was in good hands. Which was great because the project had more twists and turns to throw our way. This design hadn’t ever been done before and there were unseen problems lurking down the road. More next time!

Written by admin

June 10th, 2015 at 10:47 pm

Posted in Speaker Project

Forming the Plan

leave a comment

Ok, so in my last post I talked about how hard it was for me to lock in on a speaker that would sound good. Eventually though, I found at least a basic, well-tested framework of what makes speakers sound good. Hooray! Armed with this information, I felt confident that I could get a good value for my money and not fall prey to marketing hype or personal opinions. Plus, I could probably save enough money to buy a couple solid silver USB cables! :)

So my approach relied on getting specific acoustical measurements on each speaker. Turns out that manufacturers don’t actually publish detailed data on this. It’s just too much information for the average buyer to take in and most likely they don’t want to make people second guess their carefully crafted marketing message. Some speaker review sites publish this data but then you’re at the mercy of what speakers they want to review.

Instead of scrounging for data on mass produced speakers though, I ended up turning to a do-it-yourself design. A guy by the name of John Krutke had been testing and designing speakers for decades as a hobbyist. One of the designs he had come up with caught my eye, as it touted to be “a couple of the best performing speakers at any price” and he had the measurement data to back that up.

The little lines are the frequency response at different angles, varying from dead-on to 60 degrees off center. I’ve never seen a plot that has looked this good.

So it’s called the Zaph ZRT and it had glowing reviews from other people on the net. Reading those good comments helped, even with my confidence in the research study I talked about last post. The DIY guys like to get maximum value for their money and so if the ZRT wasn’t any good, they wouldn’t be afraid to critique heavily.

One problem with going this direction is that I’d have to go in blind (or should I say deaf?). I’d be paying for them without hearing them. That’s completely opposite from the way most people recommend buying speakers and it was putting my faith in the research study to the test. Should I throw all that conventional wisdom out the window and bank on what these guys at Harman said? Was I ready to put down all that money, simply based on the measurement data? There wasn’t any turning back once I got the kit and assembled it– no returns, refunds or exchanges. I could try reselling but I’d lose over half the money I put into it, not to mention all the time I’d spent building it. If this didn’t turn out, I would be back at square one and need to take an entirely different approach.

Maybe it was more the engineer in me than the music lover that decided to take risk. If the principles are solid, there’s no reason for an engineer to doubt the results of his project. Somehow those skyscrapers do hold up in real life, like they do on paper. While the laws of acoustic perception are still much fuzzier than laws of physics, I decided the Harman research and measurement data from the ZRT was solid enough to commit to. Late one night, I pulled the trigger and bought the kit. Now it was real and the project had begun.

While I was waiting for the kit to come in, I started planning the woodworking details. See the kit just came with the electrical components necessary to build the speakers– the boxes (cabinets) these parts go into was up to me to procure. The cabinet plans were all drawn up, but I had never done any serious woodworking so I started by looking at pre-made cabinets. I wasn’t really happy with the looks though and started reading stories of how other’s built their ZRT cabinets. Essentially it was just a long box with some holes cut into it, but to me the biggest challenge appeared to be getting the pieces cut exactly the right dimensions. Anything off and you’d get an ugly and perhaps under-performing speaker cabinet.

So I established a test project for myself — building a subwoofer. Subwoofers are the black square boxes that play really low notes. And they’re the lovely things that wake up the neighborhood at 2am when some punk drives by. Building a sub wasn’t nearly the time or resource commitment required to build the speaker cabinets and if it turned out a little wonky, that was fine. It’d be hidden in a corner of the room anyway. If it turned out really bad, then I’d fall back on the pre-made cabinets for the ZRTs.

I also had to figure out where to build these things. I had run a table saw in my apartment to build some stairs but that ended up coating everything in a nice layer of sawdust. I’d need much more than a table saw for this project and so finding a real shop was a must. I Google… err Bing’d (Binged?) for wood shop rentals and found IsGoodWoodworks. They are a co-op woodshop that rents out their space and expertise to the public. It seemed a bit spendy ($100/day) but that was the only option and I figured I could get it done in a day (haha!). Plus they offered mentorship so I’d have somebody to help teach me the basics of woodworking.

So I thought I had it all figured out. I had my plans for the shop day all written down and my Home Depot shopping list ready to go. But as I waited for the kit to come in, I started getting bored. Bored of the boxey cabinet design. I’d be living with these things my whole life and I wanted something that would be cool to look at year after year. The commercial speakers look much better than these DIY designs and that started sinking in. Those beautiful B&Ws from that day back in college had set the bar not only for sound but for looks.

What I really wanted was a curved speaker. Curved along the sides and rear, just like the B&W 803. The desire gripped me and took me off the charted course. I searched the net for DIY curved designs and came up with some decent examples. None of them were like the 803s but I figured I could use the techniques. Essentially their method was to build a skeleton with curved shelves along its length and then bend thin sheets of wood on top of that.

Speaker of my dreams, the B&W 803!

Here’s how I planned on getting that sexy 803 curve!

So I updated my plans for a curved cabinet instead of a box (monkey coffins, as they’re sometimes called). Even though they were more complicated, I figured I could still get the parts cut out in a day and then assemble it at home. I was even more excited now that I knew they would both look and sound awesome.

So now that everything was planned out, it seemed like it was just a matter of getting to the shop and cutting out the pieces. Little did I know that these plans would be scrapped almost immediately when I got to the shop, sending me into the unknown. You’ll find out all about that in the next post!

Written by admin

May 25th, 2015 at 3:52 pm

Posted in Speaker Project

Audio Genesis

leave a comment

Let’s start back at the genesis of this whole project. Like way-back “In the beginning…” style. So yeah: Big Bang, buncha stuff, and then I was in college checking out a place to live. It was a place being rented out by my friend Jeff’s older brother Tim. After the tour was done, he showed us these new fancy British B&W speakers he just bought and I was blown away. I had never heard music sound that good before and didn’t even knew it could sound that way. Something magical had just been revealed and I wanted to spend hours just listening. It didn’t even matter what — it was all so good.

So I wanted those B&Ws pretty badly but didn’t have the space or money, so I just thought about them and listened to things in the $100 and under category. Usually with an itch to to go back to that magical place again.

Fast forward to this year, where I had some money and was ready to board the train to wonderland. I started the search like any other purchase decision– see what’s popular, read reviews on the web, you know the drill. Normally this works great and a landscape of clear options quickly comes into focus. Then it just becomes a matter of whether you really need that extra 8GB of memory or something that was designed in Switzerland by people with expensive glasses.

The problem is that speakers are indeed magical music boxes and there’s no way to describe in an Amazon-style 5 point bulleted list what you’re getting. There are plenty of reviews and opinions but the more you read, the less sure you feel about getting the right thing for you. Somebody says this $1500 dollar speaker sounds as good as this other $5000 speaker and then other people trash it for being junk and say you have to spend at least $3000 to get something that sounds good. What do you do with that? There’s no list of features either– just a bunch of nebulous words like “transparency” and “richness”. Speaker manufacturers know shoppers are looking for something more concrete so they try to help out with a semi-scientific term “Extended Wide Bandwidth soft dome tweeter” but there’s no standard set of terminology and each manufacturer wants you to think they’ve got the most advanced product out there (wouldn’t you like to have have dual 5-1/4″ CMMD high-output woofers for powerful, low-distortion bass???)

So I quickly discovered that the world of audiophiles (the gang of nerds who are looking for the ultimate magic music box) has no consistent framework for determining value in a speaker. Without listening, you just can’t really tell anything about the speaker and once you do listen, its really hard to describe what you hear and even harder to get a consensus on that. Your experience is also influenced by myriad factors all outside the actual speaker itself– things like the room you’re in, the amplifier, the wiring between the amplifier and the speaker, the quality of the source signal. Put this all together and it means that reading somebody’s glowing review of a speaker is even less reliable than reading a Yelp review for that new restaurant you’ve been wanting to check out. And we all know how crappy Yelp food reviews are.

Because there is no consistent framework for what makes speakers sound good and so many factors that affect it, there is a lot of “woo-woo” type stuff. Browsing thru that is both fun and a bit morally offending. Stuff like a “shootout” of $600 USB cables (hello, it’s a DIGITAL signal) and site selling 4″ thick maple blocks hand-planed by Amish craftsmen to ward off vibrations from entering your precious audio equipment (I’m not making this up). Inciting the placebo effect has got to be a multi-million dollar business here!! Even in something as simple as the wire connecting the speaker to the amplifier, you have people who think anything above $20 is overkill and then you have cables with an asking price over $20,000 (AudioQuest Everest)! For a fun read, check out the Amazon reviews for this $6,000 power cable. Anyway, this is the only area I know of where people are willing to spend so much money for things that nobody really can prove makes a noticeable difference or not. It’s madness.

Oh wow, we’re six paragraphs in already. Hopefully you’ve been entertained. I am probably coming off as overly dramatic but this area really is crazy and what I desperately wanted at the time was to cut the crap and get to something real.

Thankfully, I found a nugget of gold amidst the mountains of glittering nonsense. Back in 2004, a researcher at Harman International (parent company of audio brands like Harman-Kardon, JBL, Infinity) did a large double-blind study on hundreds of people where they correlated acoustic measurement data from 70 different speakers with the preference ratings of participants. From that data, they developed (and patented) a model predicts how a trained listener would rate a speaker’s overall sound quality on a scale from 0 to 10. Their model predicted preference with an amazing 86% accuracy and more importantly, they solidly established a set of basic principles that quantify what makes a speaker sound good.

Turns out that the magic isn’t so complicated after all. What they disclosed of their formula (remember the full thing is patented) is that there are three main factors driving listener preference. The first is called frequency response which basically means that if you ask the speaker to play a single, very short note, how loud is that note when it comes out of the speaker. It doesn’t matter really how loud it is by itself, what matters that it the loudness is even from the lowest notes all the way to the highest notes. If it’s not even, then you get things like the high notes sounding too loud and thus a drummer’s cymbals sound harsh or maybe the mid-range notes being too quiet and a singer’s voice not coming out clearly. Makes sense.

The second thing is that this frequency response needs to be even not only for the sound waves coming straight out from the speaker but for the sound coming out at angles. This sound bounces off the walls in your room and gets back to your ears and so if it’s uneven in loudness, you have the same problem as before. And the third thing is how low the bass goes. You want a speaker that can play the canon shots in the 1812 Overture or the epic Skrillex bass drops.

So that made sense to me and was enough for me to lock in on a speaker. Which one I locked in on and why… well that’s for another time. Let me know if you liked this story and I’ll get that yarn a’spinnin.

Written by admin

May 17th, 2015 at 1:16 am

Posted in Speaker Project