Announcement

Collapse
No announcement yet.

A Year Later, X-Fi Drivers Still Horrific

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by amorpisseur View Post
    I could, but it's expensive and it's not helping the CPU. The Audigy is helping the CPU, does hardware mixing, and has god output quality.
    That is sooooooooooooooooo outdated. Look at any benchmarks comparing the two in recent years and you will see that if there is any increase at all it typically results in a <1% increase in framerates. And dmix handles mixing very nicely nowdays.

    Comment


    • #17
      Originally posted by deanjo View Post
      That is sooooooooooooooooo outdated. Look at any benchmarks comparing the two in recent years and you will see that if there is any increase at all it typically results in a <1% increase in framerates. And dmix handles mixing very nicely nowdays.
      That's indeed so true. Even embedded devices with much less cpu power
      can handle multiple streams resampled and mixed in software.

      Comment


      • #18
        Well, I used to love my GuS (hardware mixing in the 90s) and I'm not paying 150€ for a card who can't even do hardware mixing. Software mixing works, until your workstation is heavy loaded.

        I say no to sound cut !

        I really prefer to give away 30€ on eBay and get an Audigy 2 ZS which will do the job well.

        Comment


        • #19
          Originally posted by amorpisseur View Post
          Well, I used to love my GuS (hardware mixing in the 90s) and I'm not paying 150 for a card who can't even do hardware mixing. Software mixing works, until your workstation is heavy loaded.

          I say no to sound cut !

          I really prefer to give away 30 on eBay and get an Audigy 2 ZS which will do the job well.


          you're free to do what you want of course mine was just a suggestion

          Comment


          • #20
            Originally posted by amorpisseur View Post
            Well, I used to love my GuS (hardware mixing in the 90s) and I'm not paying 150€ for a card who can't even do hardware mixing. Software mixing works, until your workstation is heavy loaded.

            I say no to sound cut !

            I really prefer to give away 30€ on eBay and get an Audigy 2 ZS which will do the job well.
            Heh, my systems run at extremely heavy loads and still no issue with sound and software mixing. (Usually handbrake running in the background pegging out the cores)
            Last edited by deanjo; 10-09-2008, 04:43 PM.

            Comment


            • #21
              Originally posted by deanjo View Post
              Heh, my systems run at extremely heavy loads and still no issue with sound and software mixing. (Usually handbrake running in the background pegging out the cores)
              Software mixing is not very resilient to the huge cocktail of sound servers and API's currently in use. Its all fine and dandy if * on your system uses Pulse or dmix, but there are many programs that use OSS emulation (UT99, many linux games) as well as full duplex direct access over ALSA (Skype). Wheres your wonderful software mixing then huh? Also as you can imagine, near transparent resampling EATS CPU like the dog it is. The X-Fi's hardware mixing SRC would be very nice to have in Linux. The lower latency is extremely needed. No cheap software mixing can do that with minimal resources.
              Last edited by spikestabber; 10-09-2008, 05:50 PM.

              Comment


              • #22
                I got rid of the X-fi card I had by trading it for a SB Live 5.1 card. It was just terrible -- only time any good sound came out of it was when I switched it to the stereo recording mode... Otherwise it only managed to completely fsck up the sound of anything it played... And I'm not counting the weird clicking sounds.

                Comment


                • #23
                  Originally posted by spikestabber View Post
                  Software mixing is not very resilient to the huge cocktail of sound servers and API's currently in use. Its all fine and dandy if * on your system uses Pulse or dmix, but there are many programs that use OSS emulation (UT99, many linux games) as well as full duplex direct access over ALSA (Skype). Wheres your wonderful software mixing then huh?
                  Works fine here. dmix and aoss co-exist quite nicely.

                  Also as you can imagine, near transparent resampling EATS CPU like the dog it is.
                  Not sure what era you living in. Seriously your gripes are long outdated. Sure HW mixing would be nice but truthfully it is not needed anymore with todays hardware. Go ahead and support a company that probably has worse linux support out of mainstream suppliers. Good luck getting that line in going too.

                  Comment


                  • #24
                    Originally posted by deanjo View Post
                    Works fine here. dmix and aoss co-exist quite nicely.



                    Not sure what era you living in. Seriously your gripes are long outdated. Sure HW mixing would be nice but truthfully it is not needed anymore with todays hardware.
                    Todays hardware being a 3.6ghz core 2 quad?

                    Having 1 core at 20% cpu for a top quality software SRC isn't what I call outdated. And thats for a single stream @ 48khz. It just gets much worse when dealing with 96khz and 192khz. The only other option is using a crappy quality SRC, or using plain direct access, but there goes your sound card to anything else that wants to make a sound. This happens in Windows as well, its not just Linux, the Windows Kmixer is junk, which is one of the big reasons behind the X-Fi being superior sounding there.

                    Also what about the people who don't want a crapload of audio API's on their system and a simple setup instead? Hardware mixing works wonders there and the possibilities of annoying bugs that break your sound are much less. I use Ubuntu on my laptop and sound is often a hit or miss there, not so on my desktop, hardware mixing works. The audigy2 SRC isn't very good however, but resampling music to 48000hz with a high quality SRC bypasses any issues.

                    Hardware mixing continues to work, even when <insert next promising Linux audio mixing API here> tries to take over.
                    Last edited by spikestabber; 10-09-2008, 06:10 PM.

                    Comment


                    • #25
                      Originally posted by spikestabber View Post
                      Todays hardware being a 3.6ghz core 2 quad?

                      Having 1 core at 20% cpu for a top quality software SRC isn't what I call outdated. And thats for a single stream @ 48khz. It just gets much worse when dealing with 96khz and 192khz. The only other option is using a crappy quality SRC, or using plain direct access, but there goes your sound card to anything else that wants to make a sound. This happens in Windows as well, its not just Linux, the Windows Kmixer is junk, which is one of the big reasons behind the X-Fi being superior sounding there.
                      Try a 4200+ doing realtime resampling and encoding to AC-3 5.1utilizing jack. Works with games like UT99 as well with the system still being very responsive.

                      Comment


                      • #26
                        Originally posted by deanjo View Post
                        Try a 4200+ doing realtime resampling and encoding to AC-3 5.1utilizing jack. Works with games like UT99 as well with the system still being very responsive.
                        Not all resampling quality is the same. I'm also curious how much latency that adds to your setup. Don't say its zero either as there's always some, especially encoding to AC-3 realtime.

                        Last but not least, if you're happy with your software based soundcard/setup, why continue to comment in this thread? People who are allowed to like hardware mixing for various reasons too. Both sides certainly have their advantages.

                        Comment


                        • #27
                          Originally posted by spikestabber View Post
                          Not all resampling quality is the same. I'm also curious how much latency that adds to your setup. Don't say its zero either as there's always some, especially encoding to AC-3 realtime.

                          Last but not least, if you're happy with your software based soundcard/setup, why continue to comment in this thread? People who are allowed to like hardware mixing for various reasons too. Both sides certainly have their advantages.
                          Yeah, I love it but I didn't say it: I love a clean and minimal setup, and not depending on a bunch of daemons for managing the sound of my workstation is a big plus.

                          Comment


                          • #28
                            Originally posted by spikestabber View Post
                            Not all resampling quality is the same. I'm also curious how much latency that adds to your setup. Don't say its zero either as there's always some, especially encoding to AC-3 realtime.
                            Not enough to be noticeable during use and that's really all that matters.

                            Last but not least, if you're happy with your software based soundcard/setup, why continue to comment in this thread? People who are allowed to like hardware mixing for various reasons too. Both sides certainly have their advantages.

                            Just pointing out that HW mixing is not the "must have" feature that it once was and for the vast majority of people is not needed.

                            Comment


                            • #29
                              Originally posted by bulletxt View Post
                              just a quick note for everyone:

                              at moment, alsa cannot manage which default sound card to use while the kernel is booting. read alsa wiki for more information about it.
                              Alsa can't, but udev can. I have this in a file in /etc/udev/rules.d on Debian:
                              DRIVERS=="HDA Intel", KERNEL=="dsp*", NAME="dsp1"
                              DRIVERS=="HDA Intel", KERNEL=="adsp*", NAME="adsp1"
                              DRIVERS=="HDA Intel", KERNEL=="audio*", NAME="audio1"
                              DRIVERS=="HDA Intel", KERNEL=="mixer*", NAME="mixer1"
                              DRIVERS=="HDA Intel", KERNEL=="pcmC*D0c", NAME="snd/pcmC1D0c"
                              DRIVERS=="HDA Intel", KERNEL=="pcmC*D0p", NAME="snd/pcmC1D0p"
                              DRIVERS=="HDA Intel", KERNEL=="pcmC*D1c", NAME="snd/pcmC1D1c"
                              DRIVERS=="HDA Intel", KERNEL=="pcmC*D1p", NAME="snd/pcmC1D1p"
                              DRIVERS=="HDA Intel", KERNEL=="pcmC*D2c", NAME="snd/pcmC1D2c"
                              DRIVERS=="HDA Intel", KERNEL=="controlC*", NAME="snd/controlC1"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="dsp*", NAME="dsp"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="adsp*", NAME="adsp"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="audio*", NAME="audio"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="dmmidi*", NAME="dmmidi"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="admmidi*", NAME="admmidi"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="amidi*", NAME="amidi"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="midi*", NAME="midi"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="mixer*", NAME="mixer"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="controlC*", NAME="snd/controlC0"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="hwC*D0", NAME="snd/hwC0D0"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="hwC*D2", NAME="snd/hwC0D2"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="midiC*D0", NAME="snd/midiC0D0"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="midiC*D1", NAME="snd/midiC0D1"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="midiC*D2", NAME="snd/midiC0D2"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="pcmC*D0c", NAME="snd/pcmC0D0c"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="pcmC*D0p", NAME="snd/pcmC0D0p"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="pcmC*D1c", NAME="snd/pcmC0D1c"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="pcmC*D2p", NAME="snd/pcmC0D2p"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="pcmC*D2c", NAME="snd/pcmC0D2c"
                              DRIVERS=="EMU10K1_Audigy", KERNEL=="pcmC*D3p", NAME="snd/pcmC0D3p"
                              As you can see, I have an SB Live which works well.

                              Comment


                              • #30
                                when i thought about a well-known hardware company that didn't care about linux - i usually thought about lexmark and "support" for their printers under linux.

                                when somebody had a new lexmark printer and couldn't get it working with cups - nobody could help him.

                                now creative seems to be taking over.
                                Last edited by yoshi314; 10-10-2008, 05:27 AM.

                                Comment

                                Working...
                                X