How it came about

Back in the early 90′s I was happily making music with an Atari ST, Roland MT32, Casio CZ-101 and a Roland TR-707. It was quite a powerful system for the time given it’s relatively low cost. As I have always done, I had converted a spare bedroom into a makeshift home studio to be able to setup my ever expanding collection of musical equipment. I was writing some of my own original music, was co-creating original music as part of a synth-pop duo (the duo was called “Gotcha” - with the very talented Steve Austin) and was also creating some covers. One of the covers I created to an almost obsessive degree was “Enjoy the Silence” by Depeche Mode.

Roland MT32

Their album, Violator, had been released a year or two before-hand and was in heavy rotation amongst my playlist, which in those days meant I had made a mixtape on cassette… yes, that was when mixtapes were recorded on actual tape! Much to my girlfriend of the time’s annoyance (sorry Rachel!) I would spend hours locked away in the spare bedroom busily recording music using Steinberg Pro24 running on my Atari ST. I had recorded quite a few different covers, and the ability to record MIDI performances and edit them to build up multi-part arrangements was the perfect solution to my lesser playing skills at the time.

Roland TR707

Placing it in the wild

Violator - Depeche Mode

Not too long after this, I started using (BBS) bulletin boards to get in touch with like-minded people from just about everywhere. This lead fairly quickly to finally getting onto the Internet. This was still the mid 90′s, so the Inernet was a very different place back then. In fact there were still WAIS and Gopher servers hosting documents as the WWW was still relatively new. Eventually, many music and MIDI related sites began to pop up, including several MIDI song collection websites. Amingst them was one of the best, Neil Burton’s “MIDI File Central”. So, in 1996, I dug out my old Atari disks and go the MIDI file transferred onto the PC and uploaded my version of “Enjoy the Silence” for inclusion in the site’s files. It got 4 stars… woohoo! Having put it on the site, it meant it was now available to anyone to download, listen to and use. Sadly, the site no longer exists, but you can view parts of it in the Internet Archive (click here to see the cached copy of the Depeche Mode section of the site). Since that time the file has travelled far and wide around the internet, popping up on hundreds (if not more) of sites from general MIDI collections to Depeche Mode fan sites.

Self portrait of me in my studio

I’m following the suggestion to make 1st Aug “How I got started in ColdFusion” day (How I got started in ColdFusion Day).

To set some context, I was employed at The Daily Express (I was there for 14 years) and moved into some web based applications as we were looking at replacing our in-house Unisys A-Series mainframe based advert booking system with a bought-in system. As part of the bidding process we created a proof-of-concept enhanced in-house system. Part of this job involved creating an intranet based set of data retrieval pages that replaced some of the out-going mainframe functionality. With all the data held in a Unisys DMSII database on the mainframe, we had to look at accessing the database from within the mainframe system. To this end, one of our very capable contractors set-up an ALGOL based HTTPD service (from UC Berkeley I think) into which was plugged some COBOL based FCGI-Bin modules to handle fulfillment of data requests. The query forms were created as static pages on a SPARC based Sun Netra web-server which passed the form into the ALGOL web-server which routed the request to my FCGI routines which queried the database and built-up HTML table based responses making extensive use of the “STRING WITH POINTER” command in COBOL-74. It was quite an education but gave me a real insight into the internal plumbing of  a web server. I also lead to me later producing a web-based internal phone book search system using nothing more than a preformatted text file and a DOS “FINDSTR” command as a CGI-BIN routine.

In the end, the company bought in a client-server system based on Sun servers and Windows desktops. The web based query pages I had created had proved useful, so we set about recreating the facility against the new system. I was asked what middleware I would recommend and I plumped for Allaire ColdFusion. Unfortunately, they were not as safe a bet as Sybase were and so it was decided to go with Sybase PowerDynamo instead. I duly created the desired system using the tools I was given.

At a later time I eventually persuaded the management to buy in a copy of ColdFusion and so I finally got my hands on Allaire ColdFusion 4.5. Instantly development went much more quickly as the syntax was rather more mature and much more feature-rich than PowerDynamo.

I have used every version of ColdFusion since v4.5 and, like many dedicated CFML coders, have been part of the ColdFusion beta programs for the last few releases or so. I have been running my own software company since 2001 and have several web applications that I now sell into the Newspaper and Publishing industry… all primarily based upon ColdFusion.

However, no matter what I did, I could not get rid of the broken image problem on all of the networked blogs, only the main “parent” blog was working correctly. In the end, I managed to track the issue down to the PHP script that acts as a proxy for all media file folders within the blogs.dir directory structure for each distinct network blog (wordpress/wp-includes/ms-files.php). I discovered that when serving an image via the proxy script, the content size was 2 bytes longer than when serving the image via it’s direct URL. This led me to the solution. Something in the process within ms-files.php is writing 2 bytes of data to the output stream before loading and streaming the target media file, so by clearing the buffer before streaming the content fixes the problem.

Here is the change that needs to be made, just add lines 82 and 83 to flush the buffer before streaming the target content:

74
75
76
77
78
79
80
81
82
83
84
85
86

if ( ( $client_last_modified && $client_etag )
    ? ( ( $client_modified_timestamp >= $modified_timestamp) && ( $client_etag == $etag ) )
    : ( ( $client_modified_timestamp >= $modified_timestamp) || ( $client_etag == $etag ) )
    ) {
    status_header( 304 );
    exit;
}
 
ob_clean();
flush();
 
// If we made it this far, just serve the file
readfile( $file );

The general idea is to give myself an outlet for all the rants and ravings that have hitherto been solely delivered to my long-suffering cat, Bowie. I think publishing these ramblings might prove to be a bit more cathartic and, just possibly, a little bit entertaining to anyone unfortunate to stumble their way onto this site  between the pub and the kebab shop. It will also please my cat no end! It will no doubt cover my rantings about current events, but will also include musings and ramblings about my job, my passions and my persuits, so expect the odd snippet of music and the occasional photo.

I’ve published a few older articles and posts I’ve made to other forums and these will appear as though they were published here on the same day they were published elsewhere… just to keep things consistent chronologically.

I installed a heatsink on the regulator in the volume control board, that issue is now solved. I picked up a new voltage regulator board for the PSU from the kit supplier Conrad. Installed that and power is now working correctly.

I also picked up a programmable remote control on eBay for the price of a couple of pints to give me a smaller and rather more ergonomic remote control than the one that came with the Chinese kit. That is now programmed and working well.

Finally, I decided that rather than just splitting the output of the volume board to the output select board, the sub-woofer and cue-mix feeds, I would use a pair of “buffer” preamps to feed the sub and cue-mix outputs. I sourced a couple of stereo preamp kits from Maplin, modified them to allow use of the OMEG potentiometers for level control (as opposed to the kit’s supplied PCB mounted mini-trimmer pots) and have now installed these too. It should mean that the main output volume will not be affected by whatever volume I set for the sub and cue-mix feeds.

Here’s a few photos of the finished beast.

Next was a test fit of all the components to ensure the drilling was accurate or if any adjustments need to be made. The volume knob is a temporary one until the big aluminium knob arrives from Hong Kong! After a full day, most of the interconnect wiring is done.

Yesterday, I completed the wiring and fired it up for the first test. It didn’t go well… or at least I’m fairly sure that’s what the burning smell and smoke indicated!!

After some inspection, I’m fairly sure it’s just a rubbish regulator board. I struggled to find adequate documentation for the board and so I’ve ordered a different one which should be here in a couple of days.

In the meantime, I hooked up my bench PSU and tested the rest. Turns out I had the LEDs of the output channels in the wrong order, so that when I selected channel 1, channel 4 lit up… a little swapping of lines in the DIL header plug sorted that one. Further testing was good, the IR remote control correctly turns the volume knob and selects input channels as expected, though I do seem to have a “sticky” input channel 4 switch/LED.

A nice pair of OMEG dual gang pots arrived this morning which will be volume controls for the feed to a sub-woofer (for “pleasure listening” as opposed to mixing) and the cue feed to the headphone amps.

One addition I’ll need to make is to put a heatsink on the voltage regulator onboard the attenuator board. It wouldn’t normally need one, but for the fact I’m running a 12V supply within the unit and hadn’t realised the voltage reg is an LM7805, meaning it’s reducing a 12V supply to 5V… hence the amount of heat it’s generating. Again, the almost non-existent documentation with the Chinese kit didn’t include any such caution.

It will have 4 selectable inputs, where the selected one can be routed to 4 selectable outputs. In the middle of this routing is a volume control. I’ve opted for a stereo balanced motorised ALPS attenuator with remote control. I’m modifying some Chinese kits I bought from eBay and intend to change the original design to allow the use of illuminated push buttons for channel selection as opposed to the default rotary switch design.

The only parts I’m currently left to source are a nice 2U shallow depth chassis (the 1U I originally bought won’t be tall enough!) a big knob (ohhhh Matron!) and a suitable toroidal 240/12-0-12 transformer for the PSU.

The project consists of 4 PCBs:

• 2 selector boards employing relays to switch the audio path, one is for input sources, the other for output destinations.
• Voltage regulator board, set for 12v – 0 -12v. This will be fed by the above mentioned transformer.
• IR remote control attenuator board which has the ALPS motorised fader mounted on board.

I’ve taken this opportunity to make use of my new camera (Canon EOS 550D) and have shot HD footage of me assembling the first PCB (the input selector board) demonstrating how I prepare components, mount them on the board, solder them and finally trim the leads. The video is below followed by a selection of photos of the 4 finished boards. I’ll post more as I move through the assembly process.

Just as a little note, I had to colour-correct the video and I’m not quite happy with the resulting colour cast. I did have the perfect temperature lighting for this (see the photos), but it was partly supplied by mains LED lighting and it was strobing when being filmed! I had to use a halogen bulb instead but the colour temperature was way too low giving a heavy yellow colour cast.

and now the photos:

Having spent sometime researching what MIDI controllers, kits and pre-built modules are available, I finally settled on a neat little board from a US company called Hale Micro Systems. I bought one of their UMC32 modules. Essentially it’s a converter board that takes 32 digital or analog inputs and converts them to MIDI messages which are transmitted over a USB connection using your OSs built-in class-compliant drivers.

To cover the majority of EQ plugins, but primarily the UAD Cambridge (and due to the fact that it emulates the Sony Oxford EQ, it also applies to the Sonnox R3 EQ plugin which I also use), my design would need controls for 5 parametric bands with Freq, Gain, Q pots and an Enable switch plus a Freq pot and Enable switch for both High cut and Low cut filters. In terms of aesthetics, I also wanted to echo the colour-coding of the controls which meant endless Googling of potentiometer knobs on an endless stream of component manufacturer websites.

Having found the pots, switches and knobs I was going to use, I was then able to design the layout and therefore choose a suitably sized case for the project. The front panel legend was made up in CorelDraw. I printed a copy on normal paper and taped it to the top of thew case and used it as a drilling guide. Once the drilling was completed, the legend was printed onto water-slide decal paper, and then after soaking in water was transferred onto the prepared case. After drying overnight it was then coated with several layers of clear matt acrylic lacquer.

The next job was to wire up the pots and switches. All were connected to a common ground and +5v supply from the UMC32, then the middle wiper of each pot and opposite terminal of each switch was wired in turn to 24 of the UMC32′s inputs.The UMC32 itself was then stuck to the inside of the case simply by using double sided sticky foam pads so that the Mini-USB socket protruded into a pre-drilled hole allowing the unit to be plugged into a standard USB port on my PC. Using the UMC32′s config application, each channel was configured as a digital switch or analog input and assigned to specific MIDI continuous controller numbers.

The more or less finished control surface

I have had this one planned for a while… I was just waiting for something suitable to turn up on eBay and last week I managed to snag a used EPOS keyboard for £5… woohoo! It’s a USB keyboard and is the TM_KMR128A made by Tipro (product info). All the keys are programmable and each can be programmed to issue a single character or a string of keystrokes.

To the right is a picture of the more-or-less finshed article. The keypresses have been programmed to map to the major shortcut-keys used by Cubase. I searched high and low for suitable icons to use under the transparent keycaps, but I couldn;t turn up anything useful, so I went ahead and created my own set using Microsoft Visio.

The main point of the contoller is to present dedicated keys for the kinds of commands you useall the time day-in day-out within Cubase and so, once you learn the positions of those keys, between that and the big icons, it should hopewfully speed up your workflow. I have noticed an improvement in my own worflow, though do find I still turn to my Frontier Tranzport for all transport control funstions, especially since it can easily be carried around the studio.

Here’s a couple more pics.