Le Tue, 26 Jun 2012 08:56:01 -0700 (PDT), Bill Unruh <unruh@xxxxxxxxxxxxxx> a écrit : > On Tue, 26 Jun 2012, Dominique Michel wrote: > > > Le Mon, 25 Jun 2012 14:08:07 -0400, > > Jerry Geis <geisj@xxxxxxxxxxxxxxx> a écrit : > > > >>> The best things to do would be to use something like a dbx, a > >>> feedback suppressor. The problem is than they are expensive pieces > >>> of hardware. The principle is simple: above a given (adjustable) > >>> input level, you get an output level with a gain of 1, and under > >>> this given input level, you get a gain (adjustable)< 1. That way, > >>> the feedback will never begin, so it will never get the > >>> opportunity to become louder and louder. The dbx are mostly used > >>> to suppress feedback on stages. > >>> > >>> It should be very easy to implement such dbx as a software > >>> effect, but I don't know if something like that already exist for > >>> linux. > >>> > >>> Ciao, > >>> Dominique > >> Dominique, > >> > >> I did some quick searching and didnt see anything like that. > >> Is this something you can implement? > > > > Unfortunately not. I am good with hardware, but not a programmer. > > > > If using jack, you can try 2 or 3 effects into the jack-rack. > > > > One is the crossover-compressor. The other is the RMS expander. > > Maybe you can find some setup with them that would do the trick. > > > > Another way to suppress the feedback would be to re-tune the audio. > > Generally, feedbacks are beginning at a given frequency and into a > > very small bandwitch. By changing the pitch just a little bit will > > suppress the feedback. You can do that with the jack-rack if > > rubberband is > > Well no. It it is really feedback, it indicates an insability of the > response and that instability will manifest itself always, triggered > if nothing else by noise. Well, this is not a random instability but a sound loop that will occur only when the gain is high enough and at a given frequency. In reasonable conditions (a guitar amp at full volume is not a reasonable condition but a power operational amplifier in open loop mode... -:) you must first consider the acoustic: - try another mic - avoid direct sound from the speakers to the mic - avoid reflections by damping them if possible Otherwise, in electronics, it is at least 4 ways to suppress the feedback, that is to break the audio resonant loop: 1) with a parametric equalizer that will filter out the resonance frequency of the audio loop 2) like in the old analog dbx (expensive stuffs), by amplify less the audio when the level is lower than a given threshold The noise gate is something similar, but it will completely cancel out the sound when its level is lover than the threshold. (you can maybe find an old one for a few $ in some good guitar shops) 3) with a combination of notch filters (modern dbx, expensive too). This is the same than 1), but with many filters per channels and adaptive algorithms to pilot them. So, this is maybe not so easy to make a software implementation of such a box, 2) would be more easy to implement. 4) by changing the audio pitch. In many cases, this will be sufficient to break the audio loop. This is used in some cheap devices on the market. The result is not as good than with a dbx, but you can also use it as an audio effect. For the voice, the changed pitch should not be a problem. Of course, the efficiency of those 4 solutions will be dependent of the sound level, or of the gain of the audio equipment. It is possible to produce larsen at very low volume if the preamp gain is high enough with a good guitar amp. So the sound level is not all. Of course, this is more fun at high level... but that's another story: at high level, you will be able to control the loop more easily, that is to vary its frequency. That imply that, in our case, if the level is high enough, the feddback will be able to occur at almost any frequency, and the only thing that will work to suppress it is the noise gate (or to switch off the mic when not speaking). > You need to decrease the amplification of > that frequency band. This is most easily done by decreasing the > amplification overall, especially as which frequency feeds back can > depend on exactly where you position your head, or your hands, or the > wall, or whatever. Ie, Unless the physical environment of the speaker > and microphone is exactly fixed, the frequency at which the feedback > occurs will depend on that environment. (I have no idea what your > geomentry is so feedback could even be coming via the computer case, > rather than through the air.) Anyway, decrease the overall amplitude > of either the mic input or the speaker output. > > > > installed. It come with a LADSPA plugin called Rubber Band Pitch > > Shifter in the jack-rack. > > > >> > >> Thanks, > >> > >> Jerry > >> > >> ------------------------------------------------------------------------------ > >> Live Security Virtual Conference > >> Exclusive live event will cover all the ways today's security and > >> threat landscape has changed and how IT managers can respond. > >> Discussions will include endpoint security, mobile security and the > >> latest in malware threats. > >> http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > >> _______________________________________________ Alsa-user mailing > >> list Alsa-user@xxxxxxxxxxxxxxxxxxxxx > >> https://lists.sourceforge.net/lists/listinfo/alsa-user > > > > > > > -- "We have the heroes we deserve." ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Alsa-user mailing list Alsa-user@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/alsa-user
