An instructional video on minimum vs. linear phase

Author
bitflipper
01100010 01101001 01110100 01100110 01101100 01101
  • Total Posts : 26036
  • Joined: 2006/09/17 11:23:23
  • Location: Everett, WA USA
  • Status: offline
2013/04/19 14:36:29 (permalink)

An instructional video on minimum vs. linear phase

Dan Worrall's done it again with yet another great video, this time clearly illustrating the concepts and applications for minimum- versus linear-phase filters. Check it out: EQ: Linear Phase vs. Minimum Phase. Although it's a promo for FabFilter Pro-Q, the information is generic enough to be applicable to any equalizer that's capable of both modes of operation.


All else is in doubt, so this is the truth I cling to. 

My Stuff
#1

9 Replies Related Threads

    batsbrew
    Max Output Level: 0 dBFS
    • Total Posts : 10037
    • Joined: 2007/06/07 16:02:32
    • Location: SL,UT
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/19 15:17:45 (permalink)
    excellent.

    Bats Brew music Streaming
    Bats Brew albums:
    "Trouble"
    "Stay"
    "The Time is Magic"
    --
    Sonar 6 PE>Bandlab Cakewalk>Studio One 3.5>RME BFP>i7-7700 3.6GHz>MSI B250M>G.Skill Ripjaws 4 series 16GB>Samsung 960 EVO m.2ssd>W 10 Pro
     
    #2
    batsbrew
    Max Output Level: 0 dBFS
    • Total Posts : 10037
    • Joined: 2007/06/07 16:02:32
    • Location: SL,UT
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/19 15:21:29 (permalink)
    you know, i use the sonitus eq in individual tracks on almost eveything...
    but when i want specific sound and surgical effect, i use Waves Linear Phase EQ.....
    which i really like...

    i imagine a lot of folks that do funky things like copy tracks and slide them, get all kinds of funky artifacts that they do no understand.....

    that's why whenever i feel a need to use the polarity switch on...... anything, i always feel...........dirty.
    LOL

    Bats Brew music Streaming
    Bats Brew albums:
    "Trouble"
    "Stay"
    "The Time is Magic"
    --
    Sonar 6 PE>Bandlab Cakewalk>Studio One 3.5>RME BFP>i7-7700 3.6GHz>MSI B250M>G.Skill Ripjaws 4 series 16GB>Samsung 960 EVO m.2ssd>W 10 Pro
     
    #3
    bitflipper
    01100010 01101001 01110100 01100110 01101100 01101
    • Total Posts : 26036
    • Joined: 2006/09/17 11:23:23
    • Location: Everett, WA USA
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/19 18:32:40 (permalink)
    i imagine a lot of folks that do funky things like copy tracks and slide them, get all kinds of funky artifacts that they do no understand.....

    Yes, they do. And we all do parallel compression, parallel distortion, parallel whatever. Automatic PDC has made it easy for us to split and recombine tracks, rarely considering the possible ramifications. For example, a plugin that misreports its own latency (a common problem) can cause bizarre issues when used in a parallel signal path, and when that happens you'll have a devil of a time diagnosing the problem.


    This got me thinking about the common scenario of a tempo-synced delay. The delays are fed back and recombined with the original signal, usually after passing through a LPF. I wonder now if this isn't why I've always preferred non-synced and gated delays. Or why simulated tape delays and modulated delays are so popular.



    All else is in doubt, so this is the truth I cling to. 

    My Stuff
    #4
    The Maillard Reaction
    Max Output Level: 0 dBFS
    • Total Posts : 31918
    • Joined: 2004/07/09 20:02:20
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/20 07:47:02 (permalink)


    The video suggests that linear phase pre ringing occurs because of look ahead and plug in delay compensation.

    That explanation left me with a "wait a minute" moment.

    Is that really why we get pre ringing? It seems like if that was the only reason that it would be predictable and that a methodology for negating the pre ringing could be applied.

    Is it correct to think that pre ringing only occurs with dsp or does it also occur in analog circuits that are *almost* linear such as the Pultec design?

    Is it really caused by look ahead processes?


    best regards,
    mike


    #5
    drewfx1
    Max Output Level: -9.5 dBFS
    • Total Posts : 6585
    • Joined: 2008/08/04 16:19:11
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/20 13:14:51 (permalink)
    mike_mccue


    The video suggests that linear phase pre ringing occurs because of look ahead and plug in delay compensation.

    That explanation left me with a "wait a minute" moment.

    Is that really why we get pre ringing? It seems like if that was the only reason that it would be predictable and that a methodology for negating the pre ringing could be applied.

    Is it correct to think that pre ringing only occurs with dsp or does it also occur in analog circuits that are *almost* linear such as the Pultec design?

    Is it really caused by look ahead processes?


    best regards,
    mike
    The non-technical reason for why you get pre-ringing is because that's the way a linear phase filter works.
    A linear phase filter is a symmetrical FIR filter with identical ringing on both sides of a signal. The delay compensation just shifts the whole thing over in time to get you back where your unfiltered signal started. 


    If you try and get rid of the pre-ringing, you will end up with a very different filter. A quick example of what happens to a filter if I get rid of the pre-ringing:






     In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
    #6
    The Maillard Reaction
    Max Output Level: 0 dBFS
    • Total Posts : 31918
    • Joined: 2004/07/09 20:02:20
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/21 15:15:06 (permalink)

    Thanks.

    Would a analog FIR filter just have the symmetrical ringing and a delay?

    best regards,
    mike



    #7
    drewfx1
    Max Output Level: -9.5 dBFS
    • Total Posts : 6585
    • Joined: 2008/08/04 16:19:11
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/21 15:44:55 (permalink)
    "Analog FIR" is a bit of a misnomer here, as analog FIR's are impractical for audio applications.

    But a filter is a filter and the math/theory works the exact same way in analog or digital. 

    The advantage of digital here is the limitations come down to the math/theory itself and how much processing power and precision you want to throw at it. But analog also has physics to worry about.

     In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
    #8
    The Maillard Reaction
    Max Output Level: 0 dBFS
    • Total Posts : 31918
    • Joined: 2004/07/09 20:02:20
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/21 16:09:10 (permalink)


    Thanks.

    I guess I was guessing that in analog the symmetrical ring would occur but that there is no super easy way to compensate for the delay the way it is in a buffered digital process.
    I guess a delay that was sort of a pain to deal with getting synced back up to every thing else would be impractical.

    best regards,
    mike


    #9
    Jonbouy
    Max Output Level: 0 dBFS
    • Total Posts : 22562
    • Joined: 2008/04/14 13:47:39
    • Location: England's Sunshine South Coast
    • Status: offline
    Re:An instructional video on minimum vs. linear phase 2013/04/26 14:53:12 (permalink)
    bitflipper
    Yes, they do. And we all do parallel compression, parallel distortion, parallel whatever. Automatic PDC has made it easy for us to split and recombine tracks, rarely considering the possible ramifications. For example, a plugin that misreports its own latency (a common problem) can cause bizarre issues when used in a parallel signal path, and when that happens you'll have a devil of a time diagnosing the problem.


    Unless when you are using a parallel signal path/sends you are aware of possible downsides of using Automatic Processing Delay Compensation beforehand and cater for them accordingly of course.
     
    I find an ADC disable switch along with a sample accurate line delay a great comfort and help.
     
    Here's a good one at the right price.
     
    http://www.voxengo.com/product/sounddelay/
     
    and an even more comprehensive audio alignment suite again at the right price.
      
    http://varietyofsound.wordpress.com/2011/05/02/prefix-final-teaser-and-release-info/
     
    post edited by Jonbouy - 2013/04/26 14:55:46

    "We can't do anything to change the world until capitalism crumbles.
    In the meantime we should all go shopping to console ourselves" - Banksy
    #10
    Jump to:
    © 2024 APG vNext Commercial Version 5.1