1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Equalization: Further exploration - Minimal phase vs. Linear phase

Discussion in 'Headphones (full-size)' started by wind016, Sep 15, 2011.
  1. wind016
    I'm not an audio engineer so I hope someone with the knowledge could help explain things to me in laymen's terms or correct me. That would be really helpful. I'm glad that equalization has started to become more popular recently and that people are starting to understand that equalization CAN sound good, but I wanted a thread that would further the discussion away from simple frequency response adjustments.
    The benefits of equalization as a quick fix to neutrality is already common knowledge, but I'm not sure the deficiencies are quite as known. Explanations tended to be a bit technical for me so I only have a general idea of what's going on.
    Minimal phase equalizers
    People know that the most common equalizers, minimal phase equalizers, cause a lot of phase shifting and the steeper the slope of the filter used, the greater the phase shifting which would cause unwanted interference in the sound waves. The effect of minimal phase equalization when cutting or adding to frequencies would also change the soundstage according to the frequencies adjusted. Vocals or instruments in those frequencies would either be pulled in or out of the soundstage resulting in a smear of the soundstage and the other frequencies. A small cut or gain on an minimal phase equalizer can have a global effect on the sound; soundstage change and interference.
    Linear phase equalizers
    From what I understand, linear phase equalization while increasing phase shifting also delays the frequencies keeping the wave signal transparent and without interference. Frequency balance can be changed without a change in soundstage or transients. However, linear phase equalization causes a slight dispersion of the signal through time. Ie: Adding gain on the bass may cause the bass to hang in there a bit too long which would result in that "whooshing" feeling due to sudden loudness of the bass. However, using shelving techniques and keeping the frequencies down seem to make the delay in signal non-existent. It seems that using bell-curves, especially sharp ones, tends to increase further time dispersions, however, I have been fine reducing frequencies rather than increasing.
    How I EQ on my Stax:
    The fewer the bell curves and adjustments, the more transparent it sounds.
    Lowering the preamp should not resolve the problems above, so it should be best to reduce frequencies instead rather than resorting to lowering the preamp. Adding gain on the pre-amp, AFAIK, causes a lot of signal lagging issues in linear phase equalizers. I'm not certain of the effects of changing the pre-amp on minimal phase equalizers but I try to keep the pre-amp untouched.
    Personally, I find linear phase equalizers to be superior. More expensive equalizers generally have the side effects reduced, but linear phase advantages outweigh minimal phases'. Supposedly, Algorithmix has reduced the time-shifting to inaudible levels, but I haven't heard it yet.
  2. wind016
    No one cares about the deficiencies of equalization? We can try to avoid audible problems in EQ while also working to improve the sound.
  3. paconavarro
    I do like EQ, but I dont have a descent EQ at the moment... :frowning2:
    I dont EQ my DT990s, but I do EQ the ATH-ANC23 and HD515
  4. wind016

    There are some good free VST linear phase equalizers that you can find. The better ones start costing more than headphones, such as the $1500 Algorithmix Red, but a lot of good ones can be really cheap. The PSP Neon HR 2 had a deal for $99 from $250 awhile ago, but the deal ended after about a month. I check out Gearslutz mastering forums for comparisons between equalizers quite often.
  5. SLaRe
    The reason of why this really interesting thread is almost desert and lacking attention is just because someone who buy supposedly "high-end" equipment doesn´t care about room correction, EQ and all that stuff, because he believes that if a headphone has a $1000 price tag will sound without any flaws straight out of the box without any EQ, so I assume that you are not going to receive a lot of replies.
    I´ve seen very expensive sets of equipment in rooms with reverbarations, resonances and that sort of thing without being treated with acoustic stuff, but the owners sweared that the equipments sound really good because were very expensive (and were really really expensive).
    It´s a pity, because with a bit of EQ you can get great results, both in headphones and speakers.
  6. MaciekN
     Ringing is casued by filtering signal content above 22.5kHz (in a 44.1kHz standard) and transients, the steeper the filter (the faster it cuts any content above filtering frequency) the more artifacts it produces. Linear phase filters, while being phase coherent, distribute this ringing both before the signal that caused it and after, in an equal amount. To do so they introduce some delay in which the signal is processed (that's how ringing can happen before the signal that caused it). This is unnatural, as every real sound we hear has only post-ringing (or echo). That's why minimum phase filters were developed, they shift the phase of higher frequencies but maintain all the artifacts after the main signal, which should make it sound more natural.
    Both types of filters have some advantages over the other, it's a matter of what's accpetable for you and what is not (as with ringing distribution vs phase shifts).
     I think it's well explained here: http://www.gearslutz.com/board/mastering-forum/569408-linear-phase-versus-minimum-phase-eqs.html
     Also, here are some pictures and theory: http://www.google.pl/url?sa=t&source=web&cd=1&sqi=2&ved=0CBwQFjAA&url=http%3A%2F%2Fwww.ayre.com%2Fpdf%2FAyre_MP_White_Paper.pdf&rct=j&q=%22ayre%22%20phase&ei=wE5zTuC7Go-hOrKblaEM&usg=AFQjCNF9frL5zkZ3mJkyCpb5smk2ZM5D6g&cad=rja
    EDIT: This is not only a problem of EQing, if someone resamples 44.1kHz redbook to 96kHz/192kHz to reduce the impact of filtering then there are both minimum and linear phase resamplers :wink:
  7. wind016


    I'll test it some more and do some more reading of comparisons this weekend. I'm not really getting much ringing has long as I shelve and reduce frequencies. BTW, PSP Neon HR 2 has oversampling and it does sound better with it activated.
  8. wind016
    I read through some of the Gearslutz thread. While minimal phase is probably better for mixing, I don't think it is for music listening. It's an informative thread but I am much less likely to hear a time shift with my techniques than even a small phase shift.
  9. paconavarro
  10. MaciekN


    According to what you've wrote in the first post I thought it would be other way around, linear for mixing (setting soundstage etc) and minimum for listening (natural distribution of ringing in playback), although, some poeple at gearslutz prefer minimum phase for both.
  11. wind016

    No no. Linear phase has ringing, though I haven't found it necessarily to be audible when you just reduce frequencies. Minimum phase changes soundstage and affects all frequences even with the slightest changes.
  12. MaciekN
    Minimum phase has ringing aswell :wink: Just less audible (theoretically). Don't you think then that mastering with one filter and then listening with the other ruins much of the work if localisation and frequency change?
  13. wind016
    I'm not sure what you mean because I use equalizers to compensate for the FR issues in headphones, which is the ultimate goal in using the EQ or just to listen the way I want while trying to avoid as many issues involved with EQ as much as possible. I find a good linear phase is detailed enough that using it isn't that big of an issue and usually more enjoyable to get the exact sound I want.

Share This Page