It appears this has never gotten its own thread here, but it should. It's something that's come up for discussion a lot recently on AVSforum regarding surround audio for high-definition video (HD and Blu-ray). But long before that it had an impact on DVD-A and (possibly) SACD surround mixes. It's the likely cause of the sporadic complains of 'low bass' even from properly calibrated systems.
The problem is that some AVR/player combos do not automatically adjust the level of the LFE channel correctly when it's passed as PCM. In most cases the LFE needs a boost of +10 dB at output -- this is as per Dolby Digital spec for LFE signals, which DTS also tends to follow. This boost also has to happen BEFORE any bass managment takes place, to keep the bass in balance. In effect, this means that when you audition the same 5.1 mix in Dolby Digital and DVD-A formats, the DD may have more bass..and the DD levels are the correct ones.
Virtually all gear performs the LFE boost correctly for bitstream lossy sources -- Dolby Digital/DTS 5.1 passed digitally. Passage as PCM happens in a number of situations, though. When passed as analog, bitstream must be converted to PCM in the player first. DVD-A is always passed as PCM , whether analog or digital. SACD* can be passed digitally via HDMI 1.1 after conversion to PCM. HD and Blu-ray audio is currently always passed as PCM. So the LFE bug potentially affects a variety of sources.
How to compensate? Boosting the subwoofer channel in the AVR would seem one option, but it only works if it is done BEFORE bass management. That's rarely the case for digital sub channel boost. Some AVRs do offer a user an option to boost the analog subwoofer input, to compensate for it when using analog connections -- this catches the LFE before BM. But very few AVRs allow bass management or DSP at all when the analog multichannel panel is used. A few DVD players (like the Oppo970HD) allow an LFE channel boost by the user, which affects both analog and HDMI output; this makes BM & DSP possible downstream for the HDMI connection. I've tried this and it seems to work well for DVD-A (though it suggests the Oppo is already lowering the levels of ALL channels by default, to allow enough digital headroom for the user to apply as much as 10dB of boost to any channel)
Ideally the correct levelling would happen in the AVR, before BM, for all sources passed digitally -- this means being 'smart' about sources that DON'T need the boost (SACD might be one; some old DTS music-only discs are another) , or at least, offering the user the option to turn the boost off (so there could be one preset for SACD/old DTS, and another for all the other formats).
The issue is outlined in detail in this AVSforum thread (first post), along with a growing list of which gear does what:
http://www.avsforum.com/avs-vb/showthread.php?t=748147
Some brands seem already do LFE right (e.g., Denon). Some do it right but disable further DSP. Some mfrs (liek Pioneer) are now offering firmware updates to their AVRs to correct the LFE PCM bug.
*(I'm uncertain whether SACD LFE 'expects' a 10dB LFE boost as a matter of spec, or optionally at the discretion of the mixer, or ever. Some posts by Bob Katz on prosoundweb indicate it doesn't, but he wasn't certain either.)
The problem is that some AVR/player combos do not automatically adjust the level of the LFE channel correctly when it's passed as PCM. In most cases the LFE needs a boost of +10 dB at output -- this is as per Dolby Digital spec for LFE signals, which DTS also tends to follow. This boost also has to happen BEFORE any bass managment takes place, to keep the bass in balance. In effect, this means that when you audition the same 5.1 mix in Dolby Digital and DVD-A formats, the DD may have more bass..and the DD levels are the correct ones.
Virtually all gear performs the LFE boost correctly for bitstream lossy sources -- Dolby Digital/DTS 5.1 passed digitally. Passage as PCM happens in a number of situations, though. When passed as analog, bitstream must be converted to PCM in the player first. DVD-A is always passed as PCM , whether analog or digital. SACD* can be passed digitally via HDMI 1.1 after conversion to PCM. HD and Blu-ray audio is currently always passed as PCM. So the LFE bug potentially affects a variety of sources.
How to compensate? Boosting the subwoofer channel in the AVR would seem one option, but it only works if it is done BEFORE bass management. That's rarely the case for digital sub channel boost. Some AVRs do offer a user an option to boost the analog subwoofer input, to compensate for it when using analog connections -- this catches the LFE before BM. But very few AVRs allow bass management or DSP at all when the analog multichannel panel is used. A few DVD players (like the Oppo970HD) allow an LFE channel boost by the user, which affects both analog and HDMI output; this makes BM & DSP possible downstream for the HDMI connection. I've tried this and it seems to work well for DVD-A (though it suggests the Oppo is already lowering the levels of ALL channels by default, to allow enough digital headroom for the user to apply as much as 10dB of boost to any channel)
Ideally the correct levelling would happen in the AVR, before BM, for all sources passed digitally -- this means being 'smart' about sources that DON'T need the boost (SACD might be one; some old DTS music-only discs are another) , or at least, offering the user the option to turn the boost off (so there could be one preset for SACD/old DTS, and another for all the other formats).
The issue is outlined in detail in this AVSforum thread (first post), along with a growing list of which gear does what:
http://www.avsforum.com/avs-vb/showthread.php?t=748147
Some brands seem already do LFE right (e.g., Denon). Some do it right but disable further DSP. Some mfrs (liek Pioneer) are now offering firmware updates to their AVRs to correct the LFE PCM bug.
*(I'm uncertain whether SACD LFE 'expects' a 10dB LFE boost as a matter of spec, or optionally at the discretion of the mixer, or ever. Some posts by Bob Katz on prosoundweb indicate it doesn't, but he wasn't certain either.)
Last edited:
