FYI
- Posted by admin on August 9th, 2009 filed in Sleep Recordings
- 2 Comments »
Many apologies for the stream of posts suggesting recording/posts were imminent. This is not true. I’m experiencing equipment difficulties that require cash to be thrown at the solution. Basically, due to bedroom / computer separation I need to have a (phantom powered) condenser connectected to a ~20m lead to the computer. Unfortunately, either the ebay lead I have is low quality (likely) or phantom power over that distance is not a good idea. Either way, I’m looking for either a (non powered) mic that can do the job (the pickup of most non powered mics would make them useless) or I’ll have to attempt another lead. If you have any suggestions, please post in the comments below.
October 31st, 2009 at 1:25 am
Greetings fellow slashdotter, I’m an audio producer so maybe these tips will help improve your recordings.
The lead is the problem – you should not be having a problem with phantom powered mic at 20m. Stick with condensers – much more appropriate for what you are doing.
Make sure desk and computer are plugged into the same wall socket to avoid ground loops that will introduce noise.
If you have hard surfaces in your recording room consider hanging a wall rug or putting sound absorbing tiles in the corners of the room to reduce reverberations that make it harder to identify speech. Natural reverb will introduce your speech several ms after it’s produced – making a crappier waveform.
Cardiod mics are better but only if you can predict where the mouth will be – consider a phantom powered omni-directional if you can only use one mic.
Better though – have you considered two mics recording into a single mono channel. Looking at your set-up you may find you are actually recording a reverb of the walls rather than a straight path – also try this set-up with the mics at right angles to each other. Inverted waveforms will cancel each other out and reduce the volume.
Do not attempt to reason with sound recording – it is often counter intuitive.
Recording a noise channel *outside* of your recording room gives you scope to filter out noise you are not interested in. Invert the noise channel, apply it to your source channel – hey presto silence – except for your speech. Obviously use a mic with similar characteristics – preferably from the same batch – and balance the processing or recording gain.
After that you may want to consider checking out sonic visualiser as it will help you to filter out noise/silence and concentrate on waveforms that are speech.
Final thing you may want to consider using a software noise gate set to the frequency range of your voice to filter on your voice to reduce the amount of dead space you have to listen to and then process that into a high bitrate ogg/mp3.
At that point you may have a good enough recording to put the sleep talking through voice recognition.
Hope that helps, I’m actually doing something similar but just because I wanted to know how loudly I snore. I’m fortunate because the studio gives me access to a fairly tricky array of mics.
Somnography never really occurred to me – but you have inspired me to try it.
Best of luck – I’ll track your progress with interest.
November 10th, 2009 at 9:52 pm
Thanks for your comment.
Regarding the long lead. There were very few (cheap) sources for such a long XLR. At the stage of your commenting, I had already caved and successfully put a behringer mic preamp (bringing it back to 6.3mm) in the signal chain. Much easier for me to control gain and I can be guaranteed of decent quality signal recording. Somewhat overkill, perhaps, but now I don’t have to run my audio machine for recording aswell as my server, I can simply use the phantom power on the behringer preamp and record straight to mp3 on my fca202 on the server.
The mic is a studio projects B1. Cheap, but definitely adequate for the job. Has taken plenty of hits during sleep and still soldiers on.
As for acoustics, unfortunately, my biggest problem is that I live in Australia and I not a native, I am hot all the time – In summer I often need two fans on me just to be able to sleep comfortably. That much moving air over the mic (and the basic static like noise of the fans) is my only issue. Big one, though. I did try to webcam my sleeping for one night, but having a light on was hugely disruptive. Some things I have to live with, I guess.
As for processing, I have that down fairly well, I do use a noise gate, and therefore have a very easy time going over each recording. Really, most of the time is taken doing the noise gate over the entire signal after the fact and whatever compression is required. I wish I could post the real dynamics of my talking / shouting, unfortunately that isn’t going to make me many friends. I must admit, my lack of posting at the moment is due solely to the amount of time it takes to sanitise these files, go through them and then post the entry. A lot of work when you leave it till the point you’ve got, say, 30 recordings to go through, with maybe each noise gate taking 2-5 minutes.
Interestingly, I also felt as if I wasn’t talking as much lately, due to a fair bit less turmoil in my life, which could very well diminish the crazy. We shall see if that is actually true.