VOGONS


First post, by keropi

User metadata
Rank l33t++
Rank
l33t++

Hiya!
inspired from the discussion here (Audigy vs. Onboard Realtek HD Audio -- which is better?) I decided to run a series of tests with the PCI soundcards that I have available in order to determine the impact they have in cpu usage...

The system used was my recently acquired p4 system. For the tests I underclocked the cpu from 2.4ghz to 1.2ghz (100mhz fsb instead of the normal 200). The rest of the system is based on a Gigabyte GA-8IK1100 mobo , an AGP GeForce4200ti with 64MB vram , 512MB DDR1 RAM running in dual-channel mode and the OS is WinME with the unofficial SP2 and DirectX 8.2.
(3DMark01 score went from 11.150 to 7.070 points after underclocking)

And now the sound contestants! 😁

a. Turtle Beach Santa Cruz TB400 (crystal CS4630-CM based)
b. SoundBlaster Live! 5.1 SB0060
c. the motherboard's onboard AC97 solution based on the ALC658 chipset
d. Creative's SB PCI ES1371 (rebadged Ensonic AudioPCI)
e. WaveForce 192-XG (yamaha's own implementation of the YMF724E-V chipset)
f. Terratec Integra, an OEM card based on the ForteMedia FM801-AU chipset

16k6m1x.jpg

22au06.jpg

... there is an obvious lack of cards here, most notably the Vortex based ones but unfortunately these are the only PCI cards that I kept (since I never really cared about PCI sound for retro-usage except from live! cards)... 😒
Always wanted a vortex2 but never really found one at a nice price... I had a couple of ESS Solo-1 cards too but I donated them to friends for amiga use...

Anyway, the drivers used are the latest WDM ones for each card and if available EAX/DirectSound3D settings were enabled.

And now I present for your viewing pleasure the benchmark results! 😊

4hf4h2.jpg

(and a direct link in case the forum crops the image: http://i41.tinypic.com/4hf4h2.jpg )

It is apparent that the onboard ALC658 chipset is the overall most cpu hungry solution.
The SB Live! is best in 2D sounds and scores very good in 3D ones.
I also suspect that chipsets scoring better in DS3D than live just produce an inferior 3D quality sound (ie not much processing done).

Either way, it was fun doing the benchmark, I wish I had more cards 😁 😁 😁

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 2 of 13, by Malik

User metadata
Rank l33t
Rank
l33t

Yes! That is quite explosive! Though I can't digest the so many numbers, they must mean something! 😁

And, uh... if not asking too much, is it possible to highlight the best performer and worst performer? (either at the result column or at the name column?)

And I also wonder how the Audigy2 and X-Fi will perform in this setting.

Anyway, great work there, keropi! I'm sure people can refer to this chart for own research and application in practice 😉

5476332566_7480a12517_t.jpgSB Dos Drivers

Reply 3 of 13, by keropi

User metadata
Rank l33t++
Rank
l33t++

thanks guys!
Sadly I no longer have a pci x-fi or an audigy card... For the x-fi I doubt I could test it in winME due to driver lack.
The worst/best is already colored Malik, in each line there are red numbers (worst) , black numbers (mid-range) and blue numbers (best) , so in each test you can easily see the best and worst performer 😀

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 5 of 13, by swaaye

User metadata
Rank l33t++
Rank
l33t++

Did you guys know an AWE can supposedly accelerate a few Directsound static buffers? It can store samples in its RAM. But this is practically useless AFAIK.
msdn.microsoft.com/en-us/library/window ... s.85).aspx

AWB99 is interesting but I'm not sure if it's entirely accurate. Take those ES1371 results above. That chip can only process two streams at once in hardware. It also only has the most basic effects.

Last edited by swaaye on 2012-02-07, 19:34. Edited 2 times in total.

Reply 7 of 13, by keropi

User metadata
Rank l33t++
Rank
l33t++
leileilol wrote:

Try MDK2 benchmarks with the sound enabled, I dare you 😜

I shall! does the retail MDK2 disc has said benchmark? Also what is your opinion leileilol, do I underclock again to 1.2ghz for MDK2?
I am curious about the outcome of this, audiobench does not necessarily proves anything, real gaming benches are best 😊

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 8 of 13, by leileilol

User metadata
Rank l33t++
Rank
l33t++

All versions of MDK2 have the same benchmark, even the demo! Though the latest patch adds a option to disable sound for benchmark in addition, but what I mean by MDK2 benchmark is PCI vs onboard sound

apsosig.png
long live PCem

Reply 10 of 13, by keropi

User metadata
Rank l33t++
Rank
l33t++

I did some tests today with MDK2 and the system running @1.2ghz.

Except from AC97's worst sound quality there is no performance impact in the game whether one uses the onboard AC97 sound or the Live! or even no sound at all.
That's right, no difference at all.
I used 1280x1024/32bit color, no hardware T&L and Vsync OFF from ForceWare drivers.
I get 95-102fps on the falling stage in the start and ~100-112fps in the 1st level room while moving around.

IMHO mdk2 just maxes out on this system. I don't have a p3 system running with onboard audio, I always pick mobos without audio...

Anyone remembers another game that features a performance display or most preferably a benchmark option? (Like Devil May Cry 4 has, that would be ideal)

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 12 of 13, by keropi

User metadata
Rank l33t++
Rank
l33t++

swaaye I am looking on Serious Sam atm, can't find the benchmark option?
also downloading 3dmark03 to check it too since it could be better in providing a report

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website