Digital Bass Management and Clipping Distortion

Discussion in 'Audio Hardware' started by multiformous, Jul 16, 2017.

Thread Status:
Not open for further replies.
  1. back2vinyl

    back2vinyl Forum Resident

    Location:
    London, UK
    I just did a quick Google search and it appears that this is a specific, and well-documented, problem with Oppo players. As I said before, if this was a problem with the whole concept of bass management, it seems extremely improbable that it would not have been recognised until now.

    See this thread on Oppo players

    I can easily understand that this is a potential problem with bass management but clipping protection is widely used in digital audio and I would guess that it is built into any decent A/V receiver, to be deployed whenever bass management or other processing is turned on.

    I would also assume that there is a convention among disc manufacturers that they have to allow sufficient headroom for bass management when manufacturing 5.1 discs. (This would not apply to CDs because they do not normally undergo processing.) Evidently this was not the case with the test disc used by the OP.

    So my guess, and it is only a guess, is that the OP has stumbled upon this problem only because he is using an aberrant disc in an aberrant player. My guess is that this is simply not a problem for the vast majority of people using bass management in multichannel audio.

    An interesting topic though.
     
    pdxway likes this.
  2. multiformous

    multiformous Active Member Thread Starter

    Location:
    Newfoundland
    This is not a subwoofer clipping issue. This is not an Oppo bass management issue (yes, I saw that thread when I tried researching this issue -- the clipping discussions always centered around the sub output, not the distortion in the main channels which is at issue here). I recorded from the analog output my Oppo for convenience. I could have also used the pre-outs on my receiver to show the difference between bass managed and full range modes. The problem occurs when I play music from any source that is bass managed by either the player or the receiver (I've tried multiple receivers of different brands, but the Oppo is the only player I have that does bass management for analog outputs). I got Chris Kyriakakis of Audyssey to confirm it some time ago. I have spent plenty of time testing and investigating different settings, equipment, and signals. It's not a problem with this particular album, though I will say it's "easier" to hear with this piano music mastered to near odB. Which is why I used it as an example. The distortion may be masked by other types of music with a lot more going on. But I've heard it elsewhere:

    Here's another example you can test.

    Why use bass management? Because with my subwoofer I can get a much flatter response in my room down to 20Hz. With just the speakers, response begins to dip below 50Hz which may or may not be an issue depending on the program material. But I like that deep bass and I want to hear all the frequencies! Without the subwoofer, there's just something missing from a track like "House of Cards" by Radiohead.

    One should also not assume that multichannel audio is necessarily mastered with any more headroom than contemporary stereo recordings. Take a look at "Lucy In The Sky With Diamonds" from the recent Sgt. Pepper's remix: waveform / amplitude statistics). Not a lot of headroom in those main L/R channels. And this is the 96/24 presentation from the Blu-ray.

    32-bit floating point processing is all well and good until those signals exceeding 0dB need to be converted to analog. My Oppo UDP-203 "features a premium 32-bit DAC from AKM, the AK4458VN, which is an 8-channel DAC with support for formats such as 192 kHz / 32-bit PCM and multi-channel DSD."

    The only solution I've found to this problem (aside from disengaging bass management) is to create a little headroom myself by turning the volume down when playing music from my computer connected to my receiver via HDMI or optical cable (say, in foobar). A few dB is enough; I go with -6dB in foobar to play it safe.

    Unfortunately, a player like the Oppo doesn't have an option to turn down the digital output volume. It does however let you adjust the output level of the analog outs. So in theory, I could use the Oppo in full range mode and output that analog signal at a low enough volume so that my receiver could then do an ADC and bass manage that signal with headroom. But that extra conversion doesn't interest me.
     
    Last edited: Jul 19, 2017
  3. multiformous

    multiformous Active Member Thread Starter

    Location:
    Newfoundland
    An electronics engineer designing a bass management system might say to a mastering engineer "leave us some headroom so our processors have space to work with." To which a mastering engineer might reply "why are you processing my master at all?" Or perhaps the more likely retort is "we can't do headroom because everything has to be as loud as everything else." "Not my problem mastering engineers don't leave consumers with digital headroom for processing at home..." "Not my problem that consumers at home want to process my final mastered audio with whatever fancy doohickeys and whiz-bang 'features' hardware manufactures are using to move more product..." Not saying these conversations happen(ed), but you get the idea.

    And yes, I'm well aware of the (almost total?) lack of information regarding this issue elsewhere on the nets. When I first heard this happen on my system, after an initial freak-out-worry phase I went looking and didn't find anything on the subject. The references were always to subwoofer signal clipping due to how certain processors sum the channels and boost LFE and all that. This is not what's going on here: The clipping is audible in the main channels and is due to filter effects causing signal amplitudes to exceed the digital maximum. I thought it was odd that nothing relevant was coming up in the Google machine, and briefly thought myself crazy or incompetent. But I kept checking and testing and tinkering, and I was already sure of my findings before asking this guy on Twitter about it. And since there is such a dearth of discussion surrounding this issue, I decided to create this thread so that others might one day see in their Google search results.
     
    Last edited: Jul 19, 2017
  4. acdc7369

    acdc7369 Forum Resident

    Location:
    United States
    The only way a high pass filter could cause clipping, even if the digital source is a 0 dBFS square wave, is if it has a gain of more than 1 at any point along the frequency spectrum. So obviously those digital bass management filters are applying gain when they filter out the low end.

    There is no excuse for that. It's just bad software implementation. Have you contacted Oppo support about this? They might be able to fix it in a future firmware release.
     
  5. pdxway

    pdxway Forum Resident

    Location:
    Oregon, USA
    I do agree that bass management does the thing you claim. With pure direct stereo using my Pioneer, the piano untouched clip is very clean even at very loud volume. Once I engaged bass management, no matter I engaged 32bit or not, I can hear slight distortion. But, only very slight, much better than your Bass Manage Clip. I can live with that in my system to gain bass down to 20 Hz.
     
    multiformous likes this.
  6. back2vinyl

    back2vinyl Forum Resident

    Location:
    London, UK
    Does that Sgt. Pepper track produce clipping distortion on your system? If not, then the next step would be to ask why not.
     
  7. back2vinyl

    back2vinyl Forum Resident

    Location:
    London, UK
    What you say is intuitively true but in reality, the OP is correct - applying a high pass filter to almost any audio file does in fact create new, higher, peaks. Here's the explanation:

    Why a high pass filter creates clipping

    You can try it yourself in Adobe Audition or any other DAW. As the OP demonstrated earlier in this thread, applying a generic high-pass filter to any audio file creates new, higher peaks, so if there is not enough headroom to accommodate them, you get clipping.

    I accept this part of the OP's argument. What I'm still struggling to accept is that the OP has stumbled upon a problem that no one in the audio industry or the audio community has ever identified. So at this point I'm still assuming that bass management systems in good A/V receivers incorporate clipping protection to deal with this obvious problem. I'm sending an email to Anthem to see what they think.
     
    multiformous likes this.
Thread Status:
Not open for further replies.

Share This Page

molar-endocrine