RMAA Testing
RMAA Testing
We are using RMAA 3D to show the CPU usage over buffers of the tested cards. I am testing the Prelude 7.1 against an Auzentech X-Meridian 7.1 to illustrate the boost in gaming the Prelude can produce. I am also using an Audigy 2 Hardware card for contrast. The X-Meridian 7.1 is a software based effects card and while it can perform EAX 1 and 2 in games it is done through software and is more taxing on the System. The Prelude contains a 20KX DSP processor and all gaming effects are off-loaded to the DSP chips on the card and the Prelude is much less taxing on the system. Your measurements may vary as system load is also a factor. The Audigy 2 and Prelude data was very close in my measurements except the Audigy 2 only has 62 buffers in Direct sound 3D+EAX. This test doesn’t take into account higher levels of EAX. These measurements were done under WindowsXP.
I am a little surprised with the overall results of the measurements and only put so much stock in the results. To illustrate my point you can clearly see how the hardware acceleration helps. The Preludes results are very low showing it is very CPU light when gaming. The X-Meridian results were extremely high, I believe this has something to do directly with the high buffer measurements as with this card they are all done through the systems CPU and RAM. I was also surprised with the results presented by the Audigy 2 card, showing slightly lower CPU usage then the X-fi Prelude. I think this is the result of the measuring methods and possibly driver maturity. The tests are to be used as guidelines and can vary from system to system and are not meant as absolutes.
RMAA Measurement tests
I am using this application to measure various parameters of the soundcards being tested for the review. This application and the data it produces are a bit of a controversy because there are factors that can throw off the measurement and accuracy. I tested in both internal and external loop back modes. The internal loop back mode shows the quality of the internal routings as the signal never goes to the output stage. The external loop back mode is test that is solely based on the playback and recording capabilities of the cards in question. A card may have excellent playback and alright recording but it will show as decrease in playback quality. These numbers are submitted as guidelines only and could possibly vary from system to system. There are a lot of unseen and unavoidable variables.
Shown graphs are for 16/44.1, 16/48, 24/96 and 24/192. Remember testing varies based on the cards input also, so all tests should be used as only a guideline. A device can measure great and sound bad and measure bad and sound great. Audio is really subjective. These tests are really just examples. These tests while not able to show us how good this card sounds, they can illustrate to some degree how good the cards hardware actually is.
Frequency Response
The straighter the better, it shows the card outputs have even signal level for each frequency.
This test was performed to test the output characteristics of the card. Some soundcards outputs can vary greatly which colors the sounds quite a bit. The linear nature of the results is very good. When there is variance there is color added to the sound. The brickwall limit in the 24/192 sampling is most likely the upper limits of the recording input.
Noise level
The lower the better. This graph shows the card has a very low noise floor. Usually noise is expressed in a ratio of sound to noise or SNR.
Dynamic Range
This graph show the card has a very wide dynamic range.
The dynamic range is the difference between the loudest and quietest signal the card is capable of producing the larger the range the better.
THD+N
This is the total harmonic distortion plus noise. The lower the measurement the better
This shows the amount of harmonic distortion present in the signal. Harmonics present themselves as overtones in the original signal. These are good results.
Stereo Crosstalk
This is the bleed through between channels, the lower the better.
This test shows bleed through between the left and right channels. These are very good results. There is very little channel interaction.