- Written by Robert Kozel
- Published on 26 December 2013
The Marantz AV8801 SSP On The Bench
My standard benchmark tests were done using Direct Mode so that all digital signal processing was off. On tests using the XLR input I measured the XLR preamp output of the AV8801. On tests using an RCA input I measured the RCA preamp output of the AV8801.
At 1 kHz into the XLR input, THD+N was 0.002852%. We see a number of even and odd harmonics as well which continue out into the spectrum, which average about 95 dB below 0 dBFS. As expected, the distortion numbers are low since we are measuring a preamp output. It is interesting to note that the third harmonic dominates the second harmonic on this test.
At 1 kHz into the RCA input, THD+N was 0.009480%. The major difference between this and the previous graph is related to the voltage difference. For a power amplifier with the typical gain of 20 (26db), a 2 VRMS RCA input is required to produce 200 Watts out into 8 ohms. With 2 VRMS balanced as shown in the previous graph, you only have 1 VRMS single ended which is what runs internally inside the AV8801. In this test, we have 2 VRMS single ended which accounts for the higher distortion and higher distortion spurs.
At 10 kHz into the XLR input, THD+N was 0.006671%. We see more distortion spurs at the higher frequencies with the largest peak being the second harmonic at 20 kHz which is almost 80 dB below 0 dBFS. The THD+N value is slightly higher but overall the spectrum is well behaved.
At 10 kHz into the RCA input, THD+N was 0.017012%. We see more distortion spurs at the higher frequencies with the largest peak being the second harmonic at 20 kHz which is 70 dB below 0 dBFS.
The IMD measurement using the XLR input was 0.001833%.
The IMD measurement using the RCA input was 0.001475%.
Here are the results for 19 kHz, 20 kHz combined test frequencies using the XLR input. There is a small visible B-A peak at 1 kHz about 87 dB below 0 dBFS. We see distortion spurs throughout the spectrum with the largest peak being at 18 kHz and around 89 dB below 0 dBFS.
Here are the results for 19 kHz, 20 kHz combined test frequencies using the RCA input. There is a large visible B-A peak at 1 kHz about 68 dB below 0 dBFS. We see significant distortion spurs throughout the spectrum with the largest peak being at 18 KHz and around 44 dB below 0 dBFS.
The frequency response of the AV8801 was ruler flat past 20 kHz and starts to roll-off at 44 kHz.
Here's the test that was the most surprising of all. When Audyssey MultEQ XT32 is enabled, the AV8801 downsamples the input signal to 24 bit, 48 kHz which results in all information beyond that frequency being discarded. This behavior is not specific to the AV8801, but to all current Audyssey receivers and processors. The issue is not related to the Audyssey algorithms themselves, but rather a limitation of the DSP chipsets used in current products like the AV8801. So does this really matter? It really depends on how much high resolution content you listen to and whether the benefits of room correction outweigh any loss in resolution. The good thing about the AV8801 is that you can turn off Audyssey and perform an A-B comparison for yourself and decide your preference. I would like to give Secrets Dr. David Rich kudos for discovering this limitation by researching the AV8801 service manual. I tested this for him when he was publishing his work on Anthem's Room Correction (ARC) which does perform room correction out to 96 kHz.
On the video side of things, while the AV8801 passed all of our standard deinterlacing tests, there were some problems with loss of fine detail with the 4:2:2 color space and HDMI. The component video processing side of the AV8801 had some other issues, including the inability to pass whiter-than-white. There was also some loss of fine detail in the burst patterns with component video. If you want the best video results with the AV8801, I would recommend disabling the video scaling.