by Alec Muffett
Has anyone reading this used Bias Inc’s Soundsoap ?
I have a lot of automatic-gain-control-related hiss on the video from my digital camera; I’d like to clean it up a little, and Soundsoap seems a good candidate. I am not looking for hi-fi quality, but want to remove the artefacting and annoyance from these digital equivalents of 8mm cine.
Ideally I would llike to try it out; the sample noise-removal on [www.bias-inc.com] is rather impressive for a cheap’n’cheerful consumer solution…
I have used SoundSoap a lot for cleaning up recordings made in a large auditorium. As they were made with the internal mic on a PowerBook they were full of hard drive noise and, on very hot days in the summer, fan noise.
SoundSoap does a very good job if the noise is mild. It requires little to no knowledge of audio technology. Other apps I have used were much more complicated.
The stronger the noise gets (i.e. fan noise) the more artifacts will be in the cleaned audio. Artifacts normally sound like the burbling of a very bad mp3. It really can get unbearable if the material used had already been encoded in a lossy fashion.
Obliged for that, Jo, thanks.
The audio off my digicamera is (of course) compressed into AVI, and what you say does make me wonder – it is loud hiss when the background is silence, and falls away dramatically when there is actual signal to record. This makes me suspect an automatic gain fault in the camera, and may make matters “interesting” to correct.
In that case the audio is most likely not “AVI”-encoded, which is merely the container format, but compressed with ADPCM. This is a rather primitive, low-quality codec that compressed about 1:4. Post-processing in this case is rather sub-optimal. You may want to try using a so called “compressor”-filter.
It is normally used to reduce the contrast in loudness in a piece of audio. I have used it on the recordings mentioned above to get the more quiet parts of the recording to be understandable. These were the students asking questions. The professor being closer to the mic was louder. Using the “compressor” the difference was reduce by amplifying the students’ voices automatically to about the same level as the professors voice.
Some of these “compressor”-filters can work the other way, too. Thus reducing the quiet (in your case noisy) parts to silence. There is some primitive noise filters that work that way. These are then called “noise-gates”.
I just checked it, using QuickTime’s movie properties editor to unpick the AVI; apparently the IXUS’s audio is 11024Hz mono 8-bit unsigned int without compression (!) – so I would think from that there *is* hope for SoundSoap?
I’ve been fortunate enough to locate a colleague in the USA who has got a copy, so we’ll give that a whirl and check back in a few days. 😎
Leave a Reply