“Of the challenges that GNU/Linux users face when choosing hardware components for any system is the sound card compatibility. ALSA, or officially known as Advanced Linux Sound Architecture, provides much of the audio and MIDI functionality to Linux users and is largely replacing OSS. Today we are examining Linux audio performance in the gaming environment with a slew of various sound cards by examining their effect on frame-rate performance. The contenders are Creative Labs Sound Blaster Audigy 2 Z3, Creative Labs Sound Blaster Audigy 2, Chaintech AV-710, Aureal Vortex (AU8820), and AC’97 integrated audio.”
Linux Audio Comparison
About The Author
Follow me on Twitter @thomholwerda
2006-03-10 1:14 amAdamW
He compared the performance of various games with the different soundcards in question. He specifically stated the review was about performance, not audio quality. The tests showed that even though Linux doesn’t really have much in the way of support for hardware audio acceleration, some of the cards still performed measurably better than others. Therefore it’s a useful round up.
They obviously don’t know about the AV710’s high-quality stereo output mode, though. I sent a comment about that.
2006-03-10 1:28 amhalfmanhalfamazing
———-However, today we had simply performed a small number of benchmarks with five of the hundreds of audio devices supported by Linux. With the lack of Creative X-Fi support under Linux at this time, the Sound Blaster Audigy 2 ZS is certainly one of the better sound cards available with Advanced Linux Sound Architecture support. This article was meant to simply deliver a small insight into Linux audio performance and some of the popular competitors.—————–
And look at the graphs. The ZS2 definately leaves more CPU cycles for other tasks besides sound.
A review site that is seemingly dedicated to linux usage, I frequent the website constantly, very well written.
This article is no different. Bravo.
We’ve definately got a long way to go.
> And look at the graphs. The ZS2 definately leaves more CPU cycles for other tasks besides sound.
How can it?. There is NOTHING these games do for acceleration in Linux. These games simply do mmap and basically it’s the same on all “properly” programmed drivers. 1 stream of Stereo 16bit 48Khz audio should take the same amount of CPU and processing time on any soundcard. If it takes less time, it means the driver is actually running fast.
I use the 4Front OSS drivers as I have an Audiophile 2496 and it simply works better and has more features than ALSA. I just love the fidelity enhance that 4Front drivers have for Mp3s.
But I guess anything GPL beats anything proprietary in the Linux world
2006-03-10 10:07 ammatthew_i
But I guess anything GPL beats anything proprietary in the Linux world
Almost, but not exactly. A more accurate statement would be opens source stuff beats anythin proprietary in the free software world. Although some of the zelots tend to think like that, those of us in tune with the real world have observed that the open source development model generally produces a higher quality end product than the closed source model which is why some prefer it.
Also fighting bugs in proprietary software can be depressing if the developers are unresponsive. The same thing goes for open source software, except sometimes you can fix the bugs your self. I found a bug with my d-link router and firefox (and konqueror), and of course no one in d-link support (seems like no one at d-link) could help or even say “Sure we will pass this on.” It was very frustrating. If it had been an open source project things probably would have been much different. (No, I don’t want to hack any Linksys or Netgear routers, but thanks for asking)
2006-03-10 12:00 pmanda_skoa
There is NOTHING these games do for acceleration in Linux
I am not into audio programming, but isn’t that the job of either the low level API like ALSA or the high level sound API the game is using?
2006-03-10 12:11 pmdarkbrain
You are right! At least someone have studied
Here is my feedback for phoronix guys:
It’s very good to point an interest to audio device for linux, actually situation it’s terrible!
Any way your articole isn’t good
1) You don’t say witch backend is used. Openal can manage OSS,ALSA,SDL and others, the default is OSS so i suspect that you have used that. But 4Front, OSS native or OSS “emulated” by ALSA?
Try to put this lines in your .openalrc (in your home)
(define devices ‘(alsa))
(define alsa-out-device “hw:0,0”)
(define speaker-num 2)
And you’ll get a boost-up!
2) you write: In addition to its needing CPU resources, which resulted in marginally more frames per second in our tests, the Creative Sound Blaster lineup is unique with ALSA and Linux for its hardware mixing abilities as well as its other benefits for audiophiles.
I’m sorry but it’s wrong, it’s true that alsa driver for SB can mix in hw (up to 21 voices) but openal do all in SW no HW acceleration at all. If you want a true Mix in hw go to:
I suggest you to publish some configurations files.
Future CPUs will be so fast, that you will not need sound card acceleration. Probably x-fi from creative is last sound chip for games. Future is CELL CPU and multiple core CPUs.
Does anyone else think this review is worthless when comparing only 800×600 as the resolution? I haven’t ran a game in that resolution in years.
2006-03-10 3:11 pml3v1
You have too choose a resolution which leaves enough headroom to be able to notice and measure the differences caused by the different soundcards.
2006-03-10 5:53 pmrockwell
//Does anyone else think this review is worthless when comparing only 800×600 as the resolution? I haven’t ran a game in that resolution in years.//
Remember, he’s using Linux … 800 x 600 is awesome on that platform. The rest of us Windows idiots have to deal with crappy rezos like 1400 x 1050, while playing lousy-ass games like Battlefield 2.
“Remember, he’s using Linux … 800 x 600 is awesome on that platform. The rest of us Windows idiots have to deal with crappy rezos like 1400 x 1050, while playing lousy-ass games like Battlefield 2.”
I expect they used that resolution so the speed of the graphics rendering was irrelevent. It’s commonly done to simplify the load when testing other components such as CPU or subsystem performance.
Not all Windows users are idiots, despite your impressive example.
2006-03-11 3:48 pmrockwell
//Not all Windows users are idiots, despite your impressive example//
And, apparently, not all Windows users can detect sarcasm, even when applied thickly.
This is evident in *your* impressive example.
This guy is something of a hero since he is almost single-handedly testing out Linux and hardware for speed, gaming, home multimedia, etc., in the way 1001 Winsites do. As others have suggested, I just wish he’d break up his writing into more paragraphs.
The summary seems to be “When it came to the Linux audio performance, the Creative Labs Audigy 2 line-up with both the vanilla and ZS parts faired the best overall. In addition to its needing CPU resources, which resulted in marginally more frames per second in our tests, the Creative Sound Blaster lineup is unique with ALSA and Linux for its hardware mixing abilities as well as its other benefits for audiophiles.” So I guess if you want fast fragging then a Creative Audigy 2 or ZS is the way to go.
If you want comparisions of soundcards, I’m sorry, you’ll have to do better than saying if a soundcard played MP3s or even configured properly. Windows guys atleast do a Audio Rightmark test, give you dirt about recording, dvd playback, gaming and other features of the soundcard. This is just amateurish – like everything else in Linux.
Secondly gaming on Linux doesn’t use any EAX or multi voice features which means you can do fast fragging on a cheapie 7 buck card. Infact none of the features that soundcard vendors talk about in Windows apply to Linux.