> From: "Maluvia" <terakuma@xxxxxxxxxx> > Subject: Re: Companies Refusing to Release/Permit > Linux Drivers > To: linux-audio-user@xxxxxxxxxxxxxxxxxx > Message-ID: <200602260828200450.009F9B4F@xxxxxxxxxxxxxxx> > Content-Type: text/plain; charset="us-ascii" > > >We don't need to worry about things like this. There is a Free > >Engineering movement forming and getting stronger. I believe > >closed-source and propriatery stuff gets sterile and dies after a while > >simply because it's not fun to work at, so this should be a > >self-regulating problem. > >Carlo > > I am aware of "Open Hardware' and have been keeping an eye on it. > There is nothing we would like better than to build our own audio interface and processor, but I have not yet seen anything in the way of sound cards that approaches the level of a Hammerfal DSP - yet. I've seen some diy converters with spdif outputs that look ok. I guess the big deal is making a multichannel computer interface to plug them into. > > What I *have* been wondering for a long time is why the AD/DA technology is stuck at 24-bit? > I've already heard all the brush-off arguments: nobody needs anything higher, you can't hear the difference, it takes too much disk, space, blah-blah-blah - which is a lot of bunk. There are not any true 24bit a/d converters yet. There are not even any true 22 bit a/d converters, perhaps in some labs. I think the best converters in the world manage about 20bit. (120db dynamic range.) If a converter had a 24bit dynamic range (144db) and full scale was +7db, then it would have to be able to resolve differences of 10 nano-volts (10 one billionths of a volt). That's perhaps possible with cryogenics. Remember each extra bit *doubles* the dynamic range! Real 24 bit recording should resolve from below the brownian noise floor of air molecules hitting your ear drums to beyond the threshold of pain. That's why we are stuck at 24bit. > Disk space is cheap, I *can* hear the difference, and bit-depth is far more critical to audio fidelity than sample-rate. > I want to build a digital recording system that has the same fidelity as a 24-track reel-to-reel, and I believe this can in fact be achieved at high enough resolutions. Best noise floor you will get out of tape is about -70db. That's why we had all those dolby-A boxes. You can do that with a Soundblaster live. Tape also has enourmous amounts of jitter and distortion. What makes tape sound nice is not the fidelity! You will get useful information up to around 30Khz with a good tape machine though. Recording is about creating illusions, not fidelity. If you record an acoustic guitar in a totally dead room with the flattest most accurate mic and pre, in to best a/ds in the world, it sounds... ok. Put some reverb and top end on it, a little compression, perhaps add a little distortion with an aural exiter, or recording to tape, and people will say 'wow, what an amazing fidelity guitar recording!' :) > > I assume the blockage here is related to patent issues w/re to Cirrus logic's Crystal Semiconductor Corporation which owns the '483, '841, and '899 patents and has been agressively pursuing and winning infringement cases w/re to this technology. > http://www.ll.georgetown.edu/federal/judicial/fed/opinions/99opinions/99-1558.html > http://www.cirrus.com/en/press/releases/P36.html > > There is also the smell of the RIAA and Hollywood here > http://www.uspto.gov/web/offices/dcom/ olia/teachcomments/motionpiccomments.pdf > - as ever, psychotically paranoid about piracy (and, I have always believed, concerned about competition from the independent sector.) > I don't understand all the technology, or the legalities, but methinks something's rotten in Denmark. > Why have CPU speeds and RAM and HDD speeds and capacities leapt ahead at such an incredible pace, while we are still stuck with 32-bit PCI buses and 24-bit converters? > > I do hope Free Engineering can change all this - but this patent stuff is intimidating - they seem to enjoy going around smashing fruit-flies with sledge-hammers. > This is patent abuse - wielding patent law not merely to protect legitimate rights to income from an invention, but to quash any and all possibility of competition in the marketplace. > > Obviously, a totally new technology - not based on Crystal's - will be required to get out from under this cloud of restriction. > Surely we can come up with something even better. > To me the whole weakness and vulnerability of Open Software and Hardware arises from simply trying to RE technology then adapt it, rather than designing something completely new, then using Open Licensing schemes to keep the bullies from appropriating it for anticompetitive purposes and restricting consumer access to useful technologies. > > - Maluvia
