It's 2019 and I'm back to scare you about the LFE bug again, and SACD rips in particular.
We're moving beyond physical media. Ponder the case of digital files with LFE, in 2019. They might be
lossy DTS, AC3...
lossless MLP, DTS-MA HD, DolbyTrueHD... and DSD (and its relative DSF)...
any of which might have been converted to WAV (PCM), and thus maybe FLAC too...
All (but one*) of these are supposed to be recorded such that the LFE channel is down 10dB in level, and they expect the AVR to give their LFE a +10dB boost, after decoding, but before any bass management.
But 5.1 and this whole LFE -10dB offset stuff began with Dolby (and then DTS) and for awhilein back in the Oughts, and referring here to digital transmission to the AVR (i.e., over HDMI), only Dolby (and DTS) data got proper LFE boost, and only if they were decoded to PCM in the AVR. AVRs were too dumb to realize that incoming PCM LFE (from player-decoded DD, DTS, and from DVDA) also needed the boost. Some of them also seem to have been doing the boost *after* bass management, a terrible idea.
Stipulate that by now , AVRs are smart enough to do this right. AVRs now (and for quite some time) just bass boost all LFE (hopefully BEFORE bass management). So you can stream a raw DTS file as .dts, or stream a decoded DTS file as .wav, and the AVR will do the right thing: boost the LFE by +10dB. Ditto all the lossy and lossless flavors of DD and DTS, Bluray and DVDA too. (NB: The 'right thing' assumes those files were recorded with the proper -10dB LFE offset in the first place...**)
So what about SACD? Why does anyone like this stupid format? Here's another reason to hate it. It was the only one* whose official spec said NOT to apply the -10dB offset to its LFE, leave it full-range. That always messed things up....I'm told SACD players eventually got around it by lowering their *PCM* LFE outputs by -10dB (AFAIK all players had to option of converting DSD to PCM before HDMI output..some cheaper ones *only ever* did that). And maybe (I don't know if this is true, maybe) some AVRs got smart and when they saw raw DSD incoming, they 'knew' not to boost its LFE? But ha ha too bad , because it seems some SACDs do NOT adhere to Scarlet Book spec, and *did* lower their LFEs. That would mess up raw DSD LFE output from such 'DSD smart' AVRs. They'd work perfectly on 'DSD dumb' AVRs.
And now that we can rip SACDs and stream them as data files to an AVR, do you see the issue in 2019?
- If the SACD followed Scarlet Book spec, you either need an AVR that is smart about DSD data, or you need to convert the rip to PCM and lower that LFE by 10dB. (The recent versions of foo_input_sacd, btw, have an option to lower LFE by 10, leave it 'as is', or raise it by 10...that also applies to creating a PCM file from the DSD.)
- If the SACD doesn't follow SB spec, the LFE is already lowered correctly (I hope), so you need an AVR that *isn't* DSD-smart, or, just convert the DSD file to PCM (that does the trick whether your AVR is DSD-smart or DSD-dumb).
But...how can you know which spec was followed? You'd have to compare the LFE of the DSD version to the same mastering in DD, DTS, or DVDA. Fun! And expensive.***
I say, better not to have an LFE at all on music discs, really, and especially not on SACDs!
And of course, problem solved in 4.0.
*two really, if you count first generation DTS music discs, which had full-level LFE, but even they corrected it , I've read
**I suspect one or two DVDAs I've heard fail here
*** on another thread I'll show a recent comparison I did of a Genesis DTS 96/24 DVD , to its SACD counterpart. Spoiler: the DSD (converted to PCM) LFE 'as is', was the same level as the decoded DTS LFE. Think of the implications.