My AVR-3806 Audyssey experience (measurements inside)
I acquired a Denon AVR-3806 receiver for demo purposes in order to test out the Audyssey Room EQ functionality this past weekend.
My room, although reasonably well treated with bass traps and first-reflection panels, suffers from some very strong modal peaks at 30Hz, 45Hz, and 90Hz. Parametric equalizers (good quality ones) are expensive, and usually have only balanced connections which introduce thier own problems when trying to connect to a consumer (unbalanced) system. Also, they can (and do) introduce phase errors in the system. (The Audyssey system which uses FIR filters can in theory mitigate this)
So, the idea of upgrading my receiver and getting 5-channels of time-aligned EQ to boot seemed like an idea worth pursuing.
My main speakers are PSB stratus Golds (which can play down to around 26-28Hz by themselves). I also have a Velodyne HGS-12 subwoofer. I tried both running the main speakers as "Large" with no subwoofer, as well as trying to cross the subwoofer over for the lowest regions according to where the Audyssey decided they should be crossed over.
I am attaching my before and after graphs. Some general comments:
- I tried measuring with 1 mic position, 2 mic positions, and 6 mic positions, both with sub and without. Generally speaking, the measured response afterwards was remarkably similar. So I have only included one "after" graph and this is fairly representative of all the results.
- When I added the subwoofer into the mix, the system consistently chose to make the main speakers "Large" anyways. While this may be considered a valid mode to some, its not IMO the typical way a subwoofer would be integrated into the system. If you turn the main speakers to "Small", however, you have basically overridden the Audyssey EQ
- As is clearly shown from the "after" graph, the Audyssey EQ for some reason wants to take *A Lot* of the bass energy out of the system. I saw the same approximate frequency response for every different time I ran the algorithm, and it was also similar between the "Audyssey" and "Flat" settings ("Flat" was a little better). No doubt this is what nearly everyone is experiencing when they say that the Audyssey system sucks the life out of the bass.
- Not only that, but I dont see any significant improvement in the "smoothness" of the bass. Its still just as peaky before and after.
- There is a "Manual EQ" setting in the receiver. It offers an option to "Copy the Base Curve" from the flat setting. I had hoped this would allow you to take the "Flat" Audyssey curve and hand tweak it. However, this mode seems to disable the Audyssey EQ entirely. If I copied the "Flat" curve and made no further changes, the resulting frequency response looked nothing like the response I got when the Audyssey EQ was engaged.
- The system seemed to do a pretty good job from about 100Hz upwards. My room is well treated so the "After" response is actually a little bit worse, but its still +/-3dB - very respectable.
Overall, the system clearly did not seem to offer any improvements for my scenario. For a completely untreated room it might offer more improvement, I cant say. A year or so ago, I experimented with the PC-based software called "DRC" which does a similar thing. My feeling then was my feeling now - I'm not sure these "Auto" systems can be trusted to calculate a true inverse of the room properly, and if theres no means to manually adjust it the tool just is not that valuable.
Comments are welcome. I'd be curious if anyone else has measured the before and after, and what kind of results they acheived.
Thanks,
Andy K.
I acquired a Denon AVR-3806 receiver for demo purposes in order to test out the Audyssey Room EQ functionality this past weekend.
My room, although reasonably well treated with bass traps and first-reflection panels, suffers from some very strong modal peaks at 30Hz, 45Hz, and 90Hz. Parametric equalizers (good quality ones) are expensive, and usually have only balanced connections which introduce thier own problems when trying to connect to a consumer (unbalanced) system. Also, they can (and do) introduce phase errors in the system. (The Audyssey system which uses FIR filters can in theory mitigate this)
So, the idea of upgrading my receiver and getting 5-channels of time-aligned EQ to boot seemed like an idea worth pursuing.
My main speakers are PSB stratus Golds (which can play down to around 26-28Hz by themselves). I also have a Velodyne HGS-12 subwoofer. I tried both running the main speakers as "Large" with no subwoofer, as well as trying to cross the subwoofer over for the lowest regions according to where the Audyssey decided they should be crossed over.
I am attaching my before and after graphs. Some general comments:
- I tried measuring with 1 mic position, 2 mic positions, and 6 mic positions, both with sub and without. Generally speaking, the measured response afterwards was remarkably similar. So I have only included one "after" graph and this is fairly representative of all the results.
- When I added the subwoofer into the mix, the system consistently chose to make the main speakers "Large" anyways. While this may be considered a valid mode to some, its not IMO the typical way a subwoofer would be integrated into the system. If you turn the main speakers to "Small", however, you have basically overridden the Audyssey EQ
- As is clearly shown from the "after" graph, the Audyssey EQ for some reason wants to take *A Lot* of the bass energy out of the system. I saw the same approximate frequency response for every different time I ran the algorithm, and it was also similar between the "Audyssey" and "Flat" settings ("Flat" was a little better). No doubt this is what nearly everyone is experiencing when they say that the Audyssey system sucks the life out of the bass.
- Not only that, but I dont see any significant improvement in the "smoothness" of the bass. Its still just as peaky before and after.
- There is a "Manual EQ" setting in the receiver. It offers an option to "Copy the Base Curve" from the flat setting. I had hoped this would allow you to take the "Flat" Audyssey curve and hand tweak it. However, this mode seems to disable the Audyssey EQ entirely. If I copied the "Flat" curve and made no further changes, the resulting frequency response looked nothing like the response I got when the Audyssey EQ was engaged.
- The system seemed to do a pretty good job from about 100Hz upwards. My room is well treated so the "After" response is actually a little bit worse, but its still +/-3dB - very respectable.
Overall, the system clearly did not seem to offer any improvements for my scenario. For a completely untreated room it might offer more improvement, I cant say. A year or so ago, I experimented with the PC-based software called "DRC" which does a similar thing. My feeling then was my feeling now - I'm not sure these "Auto" systems can be trusted to calculate a true inverse of the room properly, and if theres no means to manually adjust it the tool just is not that valuable.
Comments are welcome. I'd be curious if anyone else has measured the before and after, and what kind of results they acheived.
Thanks,
Andy K.