Re: [LAU] So what do you think sucks about Linux audio ?

On Sun, February 10, 2013 1:59 am, Dan MacDonald wrote: I agree with most, and they will just take time. (though "sucks" is rather strong in this case) There are some settings that I feel should be user controlled. The SW can't know what use you are going to use it for. There are already some settings that do work with pretty much anything already and they are default. They are a longer latency and just fine for many things. But even one app may use more than one latency setting. For example, when tracking with ardour, I use low latency. Then, when I start mixdown, I relax latency so I have room to add effects, eq etc. Now ardour does allow changing Jack's latency on the fly (guitarix does too) so maybe that is ok. The thing is jack does so many different things, making it plug and play would be easy if we made it dumber... for example take out the firewire interface and the chances of grabbing the right audio device goes way up... especially if the DE sets a correct default device. But then no firewire. The way around this is to make ALSA drivers for FW :) but then no net backend. How about auto select HW monitor based on the card supporting it? No thanks. My card does support it but it also has its own interface for dealing with HW monitoring that includes an (almost) no latency mixer (ice1712). So in my case PnP would be wrong. The reason an analog studio has patch bays is for flexibility. A professional studio is not plug and play. Flexibility is probably one of the number one reasons people like Linux and jack. Some of the things I would like to see in Linux... an ALSA driver for all sound cards. The ability to use any onboard DSP of all sound cards. Pie in the sky stuff right now, but if manufactures used a standard way of a) interfacing with their on board DSP modules (midi/OSC for example) and b) a standard way of loading their proprietary code blocks into their interfaces (MIDI sysexec, whatever OSC uses, one of the old modem file transfer protocols like xmodem that many MBs use for loading firmware) It would actually be easier for the manufacture to maintain support for win/mac let alone linux which would then take care of itself. Yes it would be nicer to know more about the DSP in there so we could do our own code and make even more use of the onboard DSP, but just getting what we paid for without having to give up the flexibility of the linux platform would be nice. Pulse does not allow proper hot swapping of devices, use ALSA_in or zita_a2j to do this. True it is not automated, but then my DE doesn't implement focus follows mind either. UbuntuStudio does not auto open a USB stick just because you plug it in even though xfce has this feature available... because someone (idiot) may have plugged it in during the perfect take. We do try to set things up to work right with the internal audio IF, because that is what most newbies will try first. My experience with this is that as people use all in one apps first (LMMS is popular with windows users) they learn about the flexibility of other linux apps and switch. And if they don't, they are at least enjoying themselves. Bitwig questions would get redirected to the Bitwig site anyway. Things change. So far it has been mostly for the better. -- Len Ovens www.OvenWerks.net