Volume levels for the GTG
With a digital signal you can't go higher than 0 dBFS. If you look at most movies using a spectrum analyzer, the peaks are usually -20 dB to -10 dB. Music can be mixed right up to almost 0 dBFS. Archaea's Deadmau5 clip had peaks that were 8 dB higher than his Shambala clip. Since the songs varied so much, replay gain
was used for the playlist to make them the same relative volume. The person whose playlist was playing could then adjust the volume as he desired.
The receiver used (Pioneer VSX-53) can actually go to +12 on the master volume without any calibration. The 0 is just an imaginary number until calibration is done. I wasn't sure of the gain structure of the receiver so I put it at 0 instead of +12. However, +12 was probably a full un-attenuated signal. The reason for having 0 at -12 is to give users the ability to adjust their speaker levels without thinking they are losing dynamic range through the receiver's usually lower quality digital attenuation. The software used, JRiver, does volume attentuation at 64 bits and is basically lossless. With Carp's receiver, I think I could have gone to +17
which means it was still remembering his calibration and hadn't been reset.
With the receiver at 0 and JRiver at -17 dB, there was 34 dB of gain available when I set JRiver to 0 for its reference level. It was set to 0 using Omnimic sweeps and making the overall average frequency response at about 83 dB. I'm not sure of the Omnimic , but most calibration sweeps are at -20 dB. This would put a 0 dBFS signal at around 103 dB with 34 dB of gain still available. The receiver's amp wouldn't have handled these volume levels without clipping so by leaving the receiver at 0, I was capping the volume with the first three speakers at 127 dB if
the signal had peaks at 0 dBFS. If there were peaks this high, it still would have clipped the amp. Since an amp produces voltage instead of watts and the speakers' have varying impedences, the amp could have clipped with one speaker and not the other even though the speakers were still rated at the same efficiency. I prefer to use high power amps and judicious use of the volume control. The amps for my speakers all output about 56 volts which is 392 watts at 8 ohms.
The DR250 could have played up to about 135 dB in this system, but there wasn't enough power available. Bill Fitzmaurice has a maximum voltage setting recommended for all his speakers and you shouldn't exceed that voltage or you will cause damage. I'm not sure what the voltage is for the DR250's but I figured we didn't have enough available on the Pioneer receiver anyway. However, the receiver could have clipped at the higher volume levels.
Since Carp has such high efficiency speakers, his receiver lets him adjust the subwoofer level much higher. With lower efficiency speakers you don't have as much headroom. If, for example, you had a 87 dB speaker connected to Carp's receiver then his master volume at 0 is actually at -7 dB. He would then only be able to turn up the overall volume by 7 dB. What happens if he wants to run his subwoofers hot? If the receiver is smart enough (like it should be), it increases the sub channel by 7 dB and then attenuates the other speakers by 3 dB. Let's say before he listened at -3 dB and now he listens at 0 dB and thinks he is listening at a louder level. Since the other speakers had to have their volume pushed down, 0 dB is now exactly the same for the speakers as -3 dB was before and he is actually listening to his sub 7 dB hotter than he was before. I think this happens a lot and people adjust their subwoofer gain but don't have the digital headroom and it ends up pushing everything else down. If they don't recalibrate they are no longer at "reference" with the volume at 0.
With 97 dB speakers and 100 watts, the maximum volume the amp could produce per speaker would be around 116 dB (126 dB for the DR250). With the listening position at 15 ft, it would have been only 102 dB. With all the speakers and subs playing this would have been louder, though.Edited by desertdome - 8/6/12 at 2:23pm