## xine-devel

 [xine-devel] Re: [xine-user] Re: libac3 trouble From: Guenter Bartsch - 2001-05-12 15:04:35 ```Hi Bill, On Thu, 10 May 2001, Bill Fink wrote: > > BTW: do you have any ideas about proper ways to do downmixing? simply > > adding the channels and dividing isn't always the best solution as it > > often results in a very low-volume center playback > > (left_channel = l_surr + left + 0.5 center / 2.5 or something like that) - > > I guess we'll need some logarithmic scaling here or something of that kind > > ... I hope this won't harm if we do downmixing first and then do the > > imdct... > > I've been wondering about this. Is it actually correct to treat all 5 > inputs equally, which is what: > > left_channel = (l_surr + left + center/2) / 2.5 > > effectively does. Or would it be more correct to treat total left, > center, and total right equally for 3 basic inputs, with total left > and total right each having 2 subcomponents. This would make the > equation something like: > > left_channel = ((l_surr + left)/2 + center/2) / 1.5 > > which simplifies to: > > left_channel = (l_surr + left + center) / 3 As a stated in my earlier posting, I really think this is a question of "taste", not so much of "do the right thing". So, in fact I think we should experiment with different dowmixing functions here and perhaps offer two or three different downmixing methods for the user to choose from, perhaps by using a command-line option and/or a config file entry. Cheers, Guenter -- time is a funny concept ```
 Re: [xine-devel] Re: [xine-user] Re: libac3 trouble From: Michel LESPINASSE - 2001-05-16 20:37:22 ```On Sat, May 12, 2001 at 05:04:22PM +0200, Guenter Bartsch wrote: > As a stated in my earlier posting, I really think this is a question of > "taste", not so much of "do the right thing". So, in fact I think we > should experiment with different dowmixing functions here and perhaps > offer two or three different downmixing methods for the user to choose > from, perhaps by using a command-line option and/or a config file entry. The AC-3 standard actually specifies how the downmix should be done, so that music studios know how their stuff will sound whatever the number of speakers you have. So I think by default we should just follow the standard. Being able to override this is a possibility, but for 99% of the time we want to stay with the standardized default. (You dont see remixing controls on your high-end standalone ac3 decoder, and if you really wanted to do remixing you'd probably want to start from the multitrack source material not the 5.1 downmixed version) -- Michel "Walken" LESPINASSE Of course I think I'm right. If I thought I was wrong, I'd change my mind. ```
 Re: [xine-devel] Re: [xine-user] Re: libac3 trouble From: Guenter Bartsch - 2001-05-15 10:49:32 ```Hi Michel, On Sat, 12 May 2001, Michel LESPINASSE wrote: > The AC-3 standard actually specifies how the downmix should be done, > so that music studios know how their stuff will sound whatever the > number of speakers you have. So I think by default we should just > follow the standard. yep, of course - but I still think we need some options for people with different tastes or setups. > (You dont see remixing > controls on your high-end standalone ac3 decoder true, but such high-end equipment can safely assume that it's connected to equivalent speakers and set up in a room where people want to enjoy the movie as it was meant to be (don't get me wrong here, this is exactly how I like it, too :-). But - we're doing software decoding and that means we have to care about low-end too, people with just 2 speakers on their desk who want to have a quick glimpse at the DVD they want to watch entirely in their home-cinema in the evening. You don't want full dynamic sourround sound in the office, you need flat tv-like sound there. BTW: other software decoders do have downmixing options - I know for cinemaster in particular which has options like "dialog normailzation" or "enable LFE decoding". Cheers, Guenter -- time is a funny concept ```