Raw

I've not done any more recording recently, but I have been listening through the last batch of semi-final mixes on an mp3 player with unexceptional earphones. One track in particular sounded well out of kilter, far too much high end on the percussion. But played back in my little home studio, it was actually the upper bass which dominated. Yesterday I finally got around to doing a few measurements. First off I tried using this Ubuntu laptop for signal generation and analysis but had no joy getting input from the cheapo USB ADC/DAC that came unexpectedly bundled with my mixer. Once more it pained me but I had to revert to the WinXP desktop I use for recording (when I got the soundcard a couple of years ago, I had no joy getting the drivers to work on Linux).

So I wound up using the excellent open source, multi-platform Audacity to generate a handful of test tones: pink and white noise, swept ('chirp') sine waves, all recorded close to full-scale and saved as mono .wavs. These I dragged into the DAW software I use (EnergyXT - low cost, strongly recommended) and set up to play through channels 1 and 2, with the input from channels 3 and 4 coming from a cheapish measuring mic I bought ages ago. To do the analysis I used Blue Cat's FreqAnalyst, a free spectrum analyzer VST plug-in .dll. This gives you a pretty real-time view, with persistence of the peak values if you want (which I did).

I quickly confirmed that I had blown one of my tweeters, and what's more the other had a buzz around 220Hz. Yuck, the A below middle C is the last kind of place you want distortion. The monitor speakers are Tannoy Reveal 6s, which otherwise sound good to me - to give them the benefit of the doubt I'll blame the problems on my overloading of them, although I'd have expected them to have handled the output of the amp I use. But as it turned out, both the sweep and pink noise suggested the response was actually suprisingly flat (well, call it about +/- 10dB assuming the mic response was flat) apart from gentle error curves around the low bass and mid/upper treble. The latter could explained by the bust tweeter, the former I was able to cleanly compensate for by tweaking the level/crossover freq of my subbass monitor. I should have done that ages ago! (for future ref I got the flattest response with subbass at it's lowest crossover freq and just under max volume). To tweak the bass I made another sweep tone, 1-200Hz, which did give quite a few peaks/troughs which from hearing the room's response were due to the parallel walls (there's no trapping and virtually no damping in there, just a bit of dispersion from the furniture). I was impressed by the frequency range, there was quite smooth visible response below the lowest and way above the highest I could hear. Alas I didn't appear to have anything on hand to take a screenshot, I'm a bit lost on MS software these days...

I did recently invest in a pretty decent set of headphones (which sound wonderful) so they should cover the hi-fi angle of mixdown, with the freshly adjusted monitors to check the bass end. I still haven't got a pair of genuine iPod earphones, but have got some reasonable ones which should give me an idea of response on that kind of player. I do want to do a pile of heavy-duty dance stuff before long, but apart from sketches I think I'll leave that until I can more accurately simulate PAs (with a smallish PA, which can double up as my sound system in the new house).

Oh yeah, one other thing this setup revealed quite incidentally, with the mic on but no signal through the monitors, was how incredibly high the noise floor is in there. I have occasionally noticed fairly wideband noise on recordings made in the room, but I'd kind of assumed it was from circuits not environment. For voice and amp'd guitar I've been using cardiod mics and relatively high levels so it must have been largely masked. With the omni measurement mic though, the spectrum display was dancing away merrily without any intentional source. Recording and playing it back, the noise source was obviously the fans. Guess I need to build some kind of soundproof box for the PC (something else I should've done ages ago), must read up on approaches that still allow some airflow.

changing tinymce settings to try and stop it mangling URLs

this should be http://dannyayers.com

this should be http://dannyayers.com/

this should be http://dannyayers.com/css/

this should be http://dannyayers.com/post.html

this should be http://www.energybulletin.net

this should be http://www.energybulletin.net/

this should be http://www.energybulletin.net/stories/2010-12-15/hand-powered-drilling-tools-and-machines

I've added some fresh tag-handling code to this blog, and in the process broken its rendering of tags, I'm finding generating XML in Scala a bit confusing. This will hopefully locate the problem.