26-Aug-08: Successes and failures VII

Shellshocked is probably the best way to describe my state of mind this morning. This was the weekend of the August Bank Holiday and, traditionally, it's time for relaxing, time for musing on the passing of summer, and a chance to contemplate the slow, steady progression towards Christmas and a new year. That's the theory, anyway.

3 pm, Friday afternoon: I shut down the computers, switch off the soldering iron and prepare to take it easy. It's been a tough couple of weeks, no doubt about it, and I need a break.

Then the phone rings. "Hello" says the voice at the end of the line. "Have you seen my e-mail? My customer has a problem with his sequencer. Can you sort him out?".

Oh $%^&*$%^&*%^&*!!!!!!.

This is a new customer and a new dealer so it wasn't a problem that I could safely ignore until next week. The error code was a mystery. I've not seen one of these before and I had to go look it up. F3 = Bad sequence record. Huh? Weird.

I quickly sort out a recovery kit - some instructions, some software and some diagnostics that will at least tell me more about the problem, which I mail off to the customer. It's either a flat battery, duff hardware or something we haven't seen before. F3 means that something, somewhere, destroyed a sequence record. Usually, it's a flat battery but this is a new machine with a new battery so that's not a likely senario.

I spend a nervous hour or two waiting for a reply but 5 o'clock comes and goes without a response and I need to make dinner.

After dinner, I get back to work. I decide that I need to get another couple of CPU boards sorted, in case I need to ship a replacement out to the customer very quickly. I pick a board out of the store, test it, download the factory presets and load it up with the latest software build. Then it's into the studio for a quick workout. Nothing unusual in any of this.

Problem. The board generates a constant series of MIDI Data Errors in the receiving synths. I plug the board output into a MIDI data analyser and... it's junk. A long, steady stream of corrupt MIDI messages.

I check the same software build in a known good system. Perfect data, so it's the new board.

I pick another board out of the store and go through the set-up routine, and then plug it into a test synth. More of the same. MIDI Data Error.

I try a third. It's identical. Rude words issue forth. Lots of them.

I start to break the problem down into chucks, looking for the most likely cause, write some diagnostics and manage to eliminate the more obvious problems.

The clock rolls around to 8 pm and I'm exhausted. Then I google up an ancient post on an obscure German web site and spot a symptom that looks familiar. Babelfish does a reasonable job of translating the page, enough so that I can see where this is going. An idea forms.

Up in the workroom, I start hacking software and, 20 minutes later, I have a workable utility that will tell me if the problem is what I think it is.

At first, the fix doesn't work but I persist. I'm sure that this is what's going on. I turn and turn the encoder wheel, which is changing a calibration value on the chip itself until, suddenly, the test synth bursts into life. I plug the output of the test board into the data analyser and discover that I can switch the data errors on and off by turning a knob. I've found the problem. It's a factory calibration issue with the microprocessor itself.

The fixed up code runs for the rest of the night though I wait before adding the fix to the full release. I need time to calm down.

An e-mail arrives from the customer the following morning - the recovery kit did the trick and I discover what caused the bad sequence record - he accidentally sent ZEIT a very large MIDI file, which the sequencer should handle but, in this case, it didn't. Weird. I can only imagine that some part of the MIDI record looked like a ZEIT file, which caused ZEIT to go into capture mode. I've checked the code but I can't see what might cause this. But at least he's up and working again. That's good.

I spend Sunday morning adding the fix to the main code. It too works and the useless lump of metal on the floor of the studio is suddenly producing some rather sweet music again.

I'm relieved but this has been a real wake-up call. The problems with the CPU board have been known about for around 18 months and we've been working on a replacement board for around a year now but these solutions have been little more than temporary fixes. The real solution needs to be a bigger board with more power and more potential.

This episode is the last straw. Time to jump ship and start afresh.

I already had a couple of suitable candidates in mind but, as of this morning, I have a purchase order for two new boards. They're both OEM parts, which means we can buy them off the shelf and, if they don't work, we can just send them back and get a replacement. Both boards are affordable and the development tools are available for OSX. And they're public domain too. No more developing on a rusty old PC. No more expensive development kit. No more attempting to solder chips that don't fit their PCB footprints. Bliss. They run Linux too, straight out of the box.

We've christened the project ZEIT II. It will be an evolution of the existing machine rather than a revolution and we won't support the old CPU board after the end of the year. We'll close this chapter of the company's history and open a new one.

I can't wait.

22-Aug-08: Successes and failures VI

I'm writing this entry a double time because I'm under a couple of stupid deadlines, thanks to the long weekend here in the UK, so it won't be the usual polished tosh that regular readers have come to expect.

Yesterday, I talked about the track Logoscape, which had just been licensed for use as a backing track in a pod cast to promote Bones Burnt Black, a Sci-Fi novel by Stephen Euin Cobb. Intrigued, I did a little bit of ego-googling and quickly arrived at Mr. Cobb's web page.

A few minutes later, I began listening to "The Future And You", a pod cast hosted by Stephen, a science programme covering the kind of material that you just don't hear on radio or TV any more, not since the days of the BBC's much lamented Tomorrow's World or The Burke Special and harks back to a time when Horizon was a programme about science and not just an excercise in dumbing down. (Tell you what. Here's a guarantee. I'll pay £100 to a charity of your choice when Paris Hilton does the first voice-over on Horizon).

Anyway, I'm listening to the pod cast and well... Good Lord Above!, here's Logoscape and the Bones Burnt Black promo. My lil' tune! I'm thrilled. Wow!

The rest of the pod cast is essential listening. Check it out for yourself but be warned - there's a high geek factor here. But then this website is a High Geek Zone. :)

P.S. Here's the promo for Bones Burnt Black. Hope it doesn't infringe somebody's copyright!

Speaking of the esteemed Miss Hilton, I was trawling through my Junk Mail folder and a piece of spam caught my eye. I snapped it because it made me laugh out loud. Enjoy.

Frankly, I doubt Miss Hilton could even spell Large Hadron Collider. She probably thinks that Hadron is something she get to play with in those iffy videos that are all over the net.

21-Aug-08: Successes and failures V

On Tuesday, I received a message from those nice people at Magnatune - another Ion track, Logoscape, had been used in a licensing deal, this time as part of a podcast to promote Bones Burnt Black, a Sci-Fi title by Stephen Euin Cobb.

Logoscape was a throw-away track that was, quite literally, thrown into the album at the last minute, just to push the running time over the magic 50 minute mark. It's really just a bunch of atmospheric drones and a minimal lead line that, in the album, serves as a bridge between the title track, Future Forever, and the highly melodic Minerva. Its main job is to change the energy from quiet intensity to soothing mellow and I never really heard it as a scary piece. However, in the context of this podcast produced by Singularity Audio, it works very well indeed, enhancing the sense of menace and something-not-quite-right-here mood of the novel.

I sent a short e-mail to Shaun Farrell, the sound guy at Singularity Audio, thanking him for choosing Logoscape. This is the reply I received by return:

Thank you, David! I love the music, and I love magnatune. To have such great music available to license for little projects like this with reasonable prices is a wonderful thing. I liked so many tracks on your album that it was difficult to choose, but I think Logoscape is definitely the right choice.

Keep up your great work.

Shaun

So, a little bit closer to the goal. One step nearer to getting there.

If you want to hear the podcast then it's over at the Singularity Audio website under the Services tab.

A couple of weeks ago, a local school decided that it no longer had any use for an aging and apparently non-functional Perseus telescope. They wanted to know if our local astronomical society, Newcastle Astronomical Society had any use for the instrument.

I picked the instrument up from the school and it quickly became apparent that the instrument had suffered a lot over the years. Various attempts had been made to repair the instrument, though these had not been at all successful. In fact, these attempts at repair amounted to little more than well-intentioned vandalism.

Anyway, I've put right what I can though the whole stepper-motor gearbox assembly looks shot through. Originally, the scope wouldn't drive at all. This was because the gearbox had been re-assembled with one of the gears sitting off-centre and, whilst it didn't take long to spot the problem, it kinda made you wonder about the technical competence of the individualwho took it upon themselves to fix this instrument.

Last weekend, I stripped the whole mount so that the drive was easier to get at. It was clear last night so, as an experiment, I took the optics outside and pointed the beast at the Moon. After much pissing about and general confusion, I centred the Moon in the main field of view and.... Pow!!!! Amazing. Clarity was a step above the 5" Maksutov I usually use for the Moon. Contrast was good, though the field of view was painfully bright at times. I need a lunar filter for sustained viewing.

In short, the instrument clearly has a lot of potential and is worthy of spending just a bit more time and effort on restoring it to a useful condition. It is, however, a bloody heavy instrument. I need to find a way of moving it from A to B without risk either to the instrument or to my vertebrae.

With Jules now managing my day-to-day workload, I've had more time to work on sequencers as well as related hardware projects that have been on the back burner for too long and one of these projects is ZEIT II. There are other projects, which I'll discuss later, but ZEIT II has my complete focus at the moment. The whole machine has been given a make-over recently and now sports a far simpler go-faster design though the electronics is much the same. Recently, our main focus has been finding a successor for the old Mega128 CPU board that has served us well for many years but is now looking a little long in the tooth.

I've found a strong candidate and, if I can get it to work then ZEIT II will look like a very impressive machine indeed - more memory, an order of magnitude more processing power and a set of interfaces that are enough to make ya drool. It satisfies one of my main goals with the whole ZEIT project - that we would never make the machine obsolete. Rather than just announce a new model every couple of years, we decided that we would develop and support the machine for as long as possible, adding new software and optional hardware modules to build a complete sequencing system. I'm confident that we can still do that. The new CPU is also an OEM component, meaning that we can buy the fully-tested part off the shelf and, if it doesn't work, back it goes and the sellers send us a replacement. It's one less issue to worry about. My job is easier and the end product is much more reliable so everybody wins.

Furthermore, this new CPU has enough processing ooommmmpph! to run audio applications and that's the next goal. There's a lot of competition in the hardware sequencer market at the moment and my gut feeling is that we should expand our activities into noise-making equipment, as was the original goal of Infection Music. I'm currently drooling over the possibilities of building a 24-bit, virtual analogue synthesiser with a number of novel and unique sound creation techniques. FM fused with Wavetable synthesis, multi-mode filters, novel effects routing, that sort of thing.

So far, progress has been excellent. I've been able to install the development tools to the iMac and cross-compile some noddy applications for the target board. I've also been able to find suitable slave boards that I can develop from the iMac too, thus escaping the tyranny of the PC Development system. The slave boards will be used to interface the fast CPU to slower devices like the VFD and external hardware.

But that's for another day. The day-to-day business rolls ever onwards. We shipped the first retro rack sequencer last week. Another machine will go tomorrow and another will ship early next week. Suddenly, the enormous backlog of machines now looks manageable again.

19-Aug-08: Successes and failures IV

There's one frightening meme that we all have in common. It's universal, fixed, immutable and it leaves me completely mystified (and scared) whenever this phenomena manifests itself.

We all think we're the best Goddam driver on the planet.

I first read about this belief system in a snippet by the American Humourist, Dave Barry. Every guy I've ever known has been the same. Every driver I've ever met says exactly the same thing. They're the best driver there is and everyone else is a moron.

And I'm just as guilty. Only last Saturday, I bitched endlessly at the old fart who carved me up on the motorway whilst, at the same time, I executed a 3 lane jump that would make Evil Knievel scream in terror. Okay, maybe not. That was a bit of an exaggeration but you get the point. I'm as big a moron as the next moron. But at least I'm willing to fess up to this character defect.

Will you?

But the driving delusion is just one of many other delusions that I cling to and that hold my delicate psyche together. I honestly believe that I am halfway decent looking despite the fact that I don't shave often enough, I don't take care of my skin and I still sport the same haircut that was popular in 1979 when Punk was on the wane. I honestly believe that personality matters more than looks and that women will be wowed by my witty conversation and jovial banter. Hence, in my mind at least, I can safely ignore my ever expanding waistline because that's just so surface. I am, truthfully, becoming Homer Simpson.

I also believe, or at least I did, that I am, in some way, a competant Manager.

Yeah, that's right. A Manager.

I'm sure there are lots of good managers out there. I'm sure there are lots of people out there who would make excellent managers, if only they were given the right opportunity. Thing is, I'm not one of them.

Like the driving delusion listed above, this realisation took a long time to take hold. As a driver, I am, at best, crap. As a manager, I suck.

I don't have an organised mind. I don't have the discipline. It don't have the attention to detail or the ability to connect multiple different strands together, to see a clear path through a complex and difficult set of problems, and to emerge from the other side with a product or milestone that isn't weeks, months or even years behind schedule.

Like I said, it's taken me a while to realise this simple truth. Call it ego. Call it a whatever you want but just don't call me a manager. Please.

Last week was really, really difficult. Difficult customers, difficult or impossible technical problems, failing equipment and suppliers who don't know the meaning of the word urgent, all collided into a heady melange of ugliness that made me wonder what the fuck I am doing here when I could be earning big bucks as a programmer for some faceless corporate giant.

I became convinced that we can do better. I am still convinced that we can do better. The problem is/was that I'm just not good at deadlines and priorities. I procrastinate. I waste time. For want of a better phrase, I prefer to bugger about with minor tasks, housework and blogs rather than settle down to the task in hand. And I've known this for a long time. Many, many years in fact, but I laboured under the delusion that this was just what managers do and that this is how managers spend their time. It isn't. Not by a mile.

Hence, I decided to hand over the Project Manager's title at Infection Music to Jules. She's a natural manager. Disciplined, focussed, precise, logicial and with an uncanny ability to multi-task. In other words, all of the things I'm not.

The transition will be difficult (basically because I'm still learning to delegate) but I have to accept that managing deadlines and priorities is not my best quality. I'm a technical guy, a geek, a nerd. That's my strength. On the positive side, I am now free to concentrate what I do best, which is design stuff, build stuff, dream stuff. So long as I hit my goals then I'm on safe ground.

It's just a shame that it took me six years to realise this.

09-Aug-08: Successes and failures III

My cousin Ted got in touch the other day with a web page problem. He wanted to know if there was a way he could embed a single menu file in multiple web pages, so that he could change the menu file and all of the others web pages would pick up this change at the same time. "Sure", I said "Server Side Includes will get you sorted."

Good advice you might think. But here's a confession. This is one piece of web technology that I've never, ever been able to get working on our One and One Server.

Adding SSI's is, theoretically anyway, very easy. Firstly, you're supposed to modify your .htaccess file to tell the server to parse your web page sources for server side directives. No problem there. I use the same file to control access to these pages and specifically to exclude certain domains that are the source of frequent attacks on my contact forms.

The next step was to write an include file. Again, no biggie. You just cut and paste the sections you want to replicate from your source.

Finally, you add the SSI include directive to your source, which in my case is nothing more than the line include virtual "menu.html". Eagle-eyed viewers will note that I have not reproduced the exact text because that would simply include the menu file all over again.

Simple, right? Sure.

Could I get it to work? Could I bollocks.

I tried everything. I flushed caches, tried different browsers and different commands. I went through tutorial after tutorial, hunting for some elusive nugget that would block the code but, strangely, nothing worked.

Finally, I called the technical support staff at One and One. After a lengthy wait, a voice answered the phone, I explained my problem and he went away for a few minutes to consult with his colleagues. After a few minutes he returned. "Sorry, Mr. Hughes. That's not the kind of thing we support but... I've changed the line in your file and you'll find that it works now.

And, bugger me, it does.

I thanked the support guy for his help and put the phone down. Then I compared my file with his modified version and was astonished to find that............ they're identical!

I'm forced to conclude that some level of server-level hand-waving and a small amount of smoke and mirrors was involved here. Is it possible that they had to push a button at their end before it would work properly? I wonder... :)

Never mind. It works. As proof of the pudding, these blog files now use SSI's and you didn't even notice, did you? :)

06-Aug-08: Successes and failures II

There have been many successes in the studio in recent weeks. I'm sitting here this morning feeling pretty good in that I've added a couple more pieces to the huge melting pot of ideas that will, ultimately, form the next couple of Ion discs. There was some discussion last week of a brief appearance on Radio Newcastle to discuss the whole Generations project but that has, for the time being, failed to materialise. We'll see. These issues always take time to move from the initial brief into a fully formed idea so we'll just sit tight and see where it goes. Then again, this is Radio Newcastle. Expect the unexpected.

As I said below, I've avoided using my precious ZEIT sequencer for the past couple of weeks. It was becoming just too easy to lose oneself in one sequence after another and, as I must stress, we're not a sequencer band. Instead, I've concentrated on getting to grips with a couple of other sequencers. I gave up on the sequencer inside the Yamaha EX5. That was too much like torture. I know Cubase like the back of my hand but, every now and again, it throws out something useful which I hadn't noticed before. The same is true of my SCI Pro One except that, last night, it delivered up one of the most astonishing sequences I've ever heard. Alas, the memory on that machine is volatile and sequences disappear as soon as the power goes off, and so I frantically recorded several lengthy passages of just this sequence into the VAIO in the hope that I could capture some of this magic.

Another instrument that I have tended to ignore has been my Yamaha RM1x. I'm not fond of this instrument because it breaks the cardinal rule of every instrument. "If you need to use the manual then the instrument is a wash."

The problem with my RM1x is that it too suffers from a problem with its memory. It's fully battery-backed but it seems to forget or randomly update its sound set from one day to the next. Say, for example, you come up with a killer bass patch on Tuesday. Come Wednesday, that killer bass patch will be subtly different. Come Thursday and it will be different again. On Friday, it might be something else entirely or back to your original patch. So it tends not to be used for anything except guide rhythms when you're putting a track together and little else.

Last night, I started to dick around with some of the front panel knobs and discovered new ways to add articulation and variation to bass and percussion parts which, previously, I'd ignored. A real surprise and a very welcome addition to the sound palette.

Over in Cubase, I was dicking around with some choir sounds, layering up one voice after another to create a huge wall of choral excellence, when I came across a very simple, very basic chord progression that worked just perfectly with the whole Ion sound. I then found the most delicious melody I've ever heard to work along side it. Instant bliss.

We'll see how it develops later today. Occasionally, stuff that sounds great one night will sound like kaka-doo-doo the next day so I'm anxious not to be too enthusiastic for this piece of noodling in case it really is crap.

04-Aug-08: Successes and failures

Jules has a little postcard hanging on the wall of the garage. It reads "Into every life, a little rain must fall" and there's a cute little picture of a hippy chick waving an umbrella around in the midst of a shower. It's cute and funny and makes me smile though usually this is because it's right next to my own contribution to the wall, a cartoon I snipped from a gent's magazine more years ago than I can remember. It depicts a gentlemen of advanced years wearing a 'virile Viking strap-on phallus' with the caption "Brace yourself, Wench". Okay, maybe it's not that funny but the world needs more knob jokes, imho. If you can't laugh at a good knob joke then you're in pretty poor shape, frankly.

But I digress. As usual.

"Into every life a little rain must fall". We run three businesses here. IML is the sequencer/instrument side of the operation. Infection Media does all of the web pages. Thinking Metal looks after all of the music. We keep them separate but, inevitable, one feeds the other in a symbiotic/parasitic manner and this means that the stronger businesses can support the weaker businesses when times are hard.

And, trust me, times are hard. No doubt about it. This mini-recession looks like becoming a full-blown tsunami and I fear that there are going to be a lot of casualties. I just hope that one of them isn't going to be us.

With this in mind, we've done what we can to diversify, added extra companies and sought out additional business to make sure that this little adventure continues unhindered.

However, some companies don't make this easier for us. Some companies take six months to complete an order and then send us the bill carefully and thoughtfully labelled "30 days overdue" before the goods have actually been delivered. Some companies ask you to fall over backwards to help them out of a tricky tax situation and then sit on your bill for six weeks whilst they enjoy an extended 3 week vacation in some sunny part of the world whilst you're stuck here in the pouring rain wondering if you can afford groceries this weekend.

This is what kills most UK businesses. It isn't greedy, corporate fat cats passing the buck from one desk to another whilst attempting to avoid the blame for the latest screw-up. It isn't down to government policy though they don't help matters one iota. It isn't down to customers voting with their feet. It's down to lazy, idle accountants who prefer to wait until a company is either going under or has actually gone under so that they can avoid a liability. That's not the way to do business. It's wrong and someone needs to start a record somewhere of all of those companies who can't or won't own up to their responsibilities, both to their customers and their bosses.

02-Aug-08: Yet again... it's the return of the blog... (Yawn!) :)

Last month, I decided to reduce the number of updates to the blog, mainly because the number of people reading the blog didn't justify the effort. Or so I thought.

However, over the last month, I've watched the number of visitors to the site fall, consistently, so that as of this morning, we're down to about 10% of June's head count. Then again, June was a crazy month, with two number ones over at Magnatune and a lot of airplay around the world so the number of people coming to the site should have been higher than average. That's to be expected. I wasn't expecting the visitor numbers for July to fall away so quickly and, worse still, those people who did visit the site didn't stay very long. So, I think that it's safe to conclude that the blog was one of the reasons why visitors were coming to the site and sticking around for longer. With that in mind, I've decided that, once again, the blog will become a routine fixture, maybe not updated every day but certainly once or twice a week.

July was just as crazy - two weeks of frantic, back-breaking work followed by 2 weeks of holiday, spent largely driving around the north east of England or in the garden. I did spend a fair bit of time working on another instrument design and got as far as writing some test programs so that the thing makes a noise but it's a long way from a finished product. That was very satisfying.

Today, I'm back at work full time and it's a strain because there is so much to do - deadlines I missed, machine's not tested, web sites not finished. However, I needed the break. I really did. Too much work dulls the tools and we all need to set some time aside to sharpen the saw.

I have a job log, which is an essential tool for keeping track of what needs to be done. Each job has simple description, a priority and the amount of time I think it will take, which I usually under-estimate. I can then quickly sort these jobs in Excel, firstly in order of priority and then in order of duration. This gives priority to quick wins and, yes, there's a danger that tasks which are lower in priority and take longer to complete will never be serviced. Hence, I also schedule tasks for specific times during the working week, so that they will always get serviced. This isn't rocket science. The method is based on techniques outlined in "The Seven Habits of Highly Effective People" by Stephen Covey, which is a very important reference work in the business world. If you haven't read it, do so. It will save you a lot of time and effort.

Everything goes in the "Job Schedule", from "Build Sequence X1 for Customer XXX" to "Buy milk", from "Order Replacement Power Supplies" to "Post Gas Bill". That way, nothing gets missed off. Posting the gas bill might not be a business activity but if the gas bill wasn't posted then the red letters would start appearing and this really would have a major impact on the smooth operation of the business. More than anything, it would mean that we could fire up the kettle and I couldn't face the working day without a hot cup of something or other.

Speaking of which, I decided to stop drinking coffee last month as part of a diet. I wondered if I was one of those people whose weight is affected by caffeine and so I replaced my usual 2 cups of Uncle Joe per day with tea. A couple of friends suggested that I might experience some withdrawal symptoms but, mercifully, there were none. After a month of tea, I didn't feel or look any slimmer and so I went back to coffee, only to discover this morning that my weight has actually dropped by 5 pounds. That's not bad going so maybe there is something in this caffeine theory.

But I digress. Again.

I've been working on the Awakenings set ready for the September concert. It's been going reasonably well with new ideas appearing every night, some of them useful, some of them not immediately relevant but it's still early days. I've tried to avoid using certain tools, so that they don't become a crutch for a lack of creativity. One such tool is my ZEIT sequencer, because it's just too easy to switch it on and bliss out on one fantastic pattern after another. Ion is not a sequencer band. We use sequencers but we prefer not to let them dominate every single track. They're not our raison d'etre.

In another musical side project, I've set aside some time to see if I can make my beloved iMac into something more useful. It has Garage Band and Reaper but I want to see if I can record a full album using nothing more than free VST's and a plug in keyboard. It's a challenge and one that is fairly exciting but not high priority. The idea came about when I saw a couple of laptop jockeys at a gig recently and I wondered if they were producing crap music because they were using laptops or if they were just crap musicians. I suspect the latter, frankly.

That's all for now.

 

Copyright 2008 David J. Hughes and Thinking Metal Music