How do Digital EQ's differ?
-
- KVRian
- Topic Starter
- 649 posts since 30 Oct, 2004 from Brighton, UK
Since we're talking about great EQ's, have to mention Tritone's Hydratone, which is just awesome. Check out the dmeos at www.tritonedigital.com (mac only for now)
-
Lawnmower Of The Damned Lawnmower Of The Damned https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=29783
- KVRian
- 850 posts since 16 Jun, 2004
A big part of the sound difference comes from the way that the EQs handle phasing. So far as I am aware, all EQs, whether they are hardware or software operate on the principal of phase cancelation.
In your EQ, let's say you make a 6 db cut at 600 hz. What the EQ is actually doing, is making a copy of the sound running into it, shifting the phase of the selected frequency, and combining the modified copy with the original.
Different EQs handle the phase shifting differently. Different curves, different methods of shifting, etc. And that, more or less, is why EQs all sound different.
In your EQ, let's say you make a 6 db cut at 600 hz. What the EQ is actually doing, is making a copy of the sound running into it, shifting the phase of the selected frequency, and combining the modified copy with the original.
Different EQs handle the phase shifting differently. Different curves, different methods of shifting, etc. And that, more or less, is why EQs all sound different.
Excuse all the blood.
-
- KVRAF
- 11378 posts since 3 Feb, 2003 from Finland, Espoo
I've got about 5 hours of experience with the Weiss EQ-1 MK2 (with all the options added, including internal double precision upgrade, aka 96khz internal mode). It's fantasticly transparent. It simply boosts/cuts whatever you throw at it without adding any smearing/phasing around the 'edges'. I preferred it in the non-linearphase mode, aka, minimal-phase or whatever that mode was called.citizenchunk wrote:
ps - i've heard nothing but good things about the EQ-1, and its LP/DYN versions. i can't wait to get my hands on one. i'm already blown away by the DS-1. were you totally blown away by the EQ-1?
Today I played with the Algorthmix Orange and Red Linearphase EQ's and frankly, I do not really understand how some people compare them to the weiss. IMHO the weiss is far better at being 'transparent' which is what the algorthmix EQs try too. However, these directX plugins are some of the smoothest plugin EQs that I've heard (but nothing incredible, I'm sure Aleksey Could code a monster linear-phase eq if given enough CPU cycles).
I'll have to test them a bit more but at 1200 euros per plugin they are kind of out of my league.
The new Smooth mode in Voxengo HarmoniEQ makes it the smoothest non-convolution VST plugin eq that I've heard to date.
On wednesday my Sintefex FX8000 6ch version is here so I'll give you guys some juicy pictures/previews of what the box can do.. you guys jealous yet? Well, if not, then let me add that I got it at an amazing price too.. jealous yet?? huh?? HUH?????!!!??
- bManic
-
- KVRAF
- 11378 posts since 3 Feb, 2003 from Finland, Espoo
Yes, they will.fractalism wrote:btw, will the free plugs be availbale for PC?Ocean Zen wrote:TriTone Digital's Eqs have just blown me away after trying the demos.
They're the fattest smothest sounding EQ I've heard yet
Check out the demo for Hydratone & ValveTone at www.tritonedigital.com
-
- KVRer
- 13 posts since 22 Mar, 2004
the main variables are:
Filter type (IIR, FIR, etc) and, hence, phase behavior.
Filter Q and Pole/zero locations and behavior (constant verse dynamic)
There is a lot of trade offs in DSP so you have to combine and adapt to meet your goals. Given that one filter has numerous possible configurations, EQ are certainly not all the same (nor should we want them to be!).
The Oxford EQ does some different things such as move the poles of the filters based on the frequency. This has to do with basic filter design, some filters designs do better at different frequencies (particularly at the high and low end ranges) so changing the poles can improve the sound greatly. They also measure the Q at a different spot than many do (at the half db-gain points). Type 3/4 Oxfords have a Q that changes based on gain amount. In those cases, the Q is wider at lower gain structures but more narrow at higher gain structures. The last two components surrounding the Q are the easy part, it is the choosing of zeros that is the art.
Funny enough you can get more correct EQ in the digital realm. A sophisticated Butterworth is maximally flat with no resonance peaks, however, it is those peaks that impart much of the color of analog EQs/filters. Bessel filters make good EQ filters as they are smooth offering gentle roll-off. Correct and flat is not what many want. They want character. Whether said character appeals to the user is another thing but by no means restricted to the digital space. Most EQs use Bessel and Elliptic designs. Oxford uses a Lattice filter which offers Butterworth, Bessel and Elliptic filters.
Some suggest that on higher frequencies (say above 12k) adding small amounts of even harmonics distortion (polynomial waveshaping) provides a better sense of "air" than basic band boost. I've not played with that method, so i can't say what the effect would be. Another topic for another day.
The point being there is many to approach filter (and hence EQ) design. Not all equal nor easy to replicate.
Filter type (IIR, FIR, etc) and, hence, phase behavior.
Filter Q and Pole/zero locations and behavior (constant verse dynamic)
There is a lot of trade offs in DSP so you have to combine and adapt to meet your goals. Given that one filter has numerous possible configurations, EQ are certainly not all the same (nor should we want them to be!).
The Oxford EQ does some different things such as move the poles of the filters based on the frequency. This has to do with basic filter design, some filters designs do better at different frequencies (particularly at the high and low end ranges) so changing the poles can improve the sound greatly. They also measure the Q at a different spot than many do (at the half db-gain points). Type 3/4 Oxfords have a Q that changes based on gain amount. In those cases, the Q is wider at lower gain structures but more narrow at higher gain structures. The last two components surrounding the Q are the easy part, it is the choosing of zeros that is the art.
Funny enough you can get more correct EQ in the digital realm. A sophisticated Butterworth is maximally flat with no resonance peaks, however, it is those peaks that impart much of the color of analog EQs/filters. Bessel filters make good EQ filters as they are smooth offering gentle roll-off. Correct and flat is not what many want. They want character. Whether said character appeals to the user is another thing but by no means restricted to the digital space. Most EQs use Bessel and Elliptic designs. Oxford uses a Lattice filter which offers Butterworth, Bessel and Elliptic filters.
Some suggest that on higher frequencies (say above 12k) adding small amounts of even harmonics distortion (polynomial waveshaping) provides a better sense of "air" than basic band boost. I've not played with that method, so i can't say what the effect would be. Another topic for another day.
The point being there is many to approach filter (and hence EQ) design. Not all equal nor easy to replicate.
-
- KVRAF
- 2460 posts since 3 Oct, 2002 from SF CA USA NA Earth
Any filter, EQ included, has some sort of effect on the phase of the signals going through it, as well as the frequency spectrum. Philosophical differences about how to deal with phase relationships are about the only major differences you'll encounter among FIR EQs (and most EQs with more than, say, 5 bands, are FIR).
Those phase-relationship design decisions are what lead to:
Those phase-relationship design decisions are what lead to:
If I use certain EQ's to boost a kick drum somewhere around 55-80Hz, I'll get a muffled and gutless sound which just goes *mmmpphhhhh* and that's it. If, on the other hand, I choose some other EQ's for the job, I'll get a real kicker that kicks the ribs through the lungs.
Don't do it my way.
-
- KVRist
- 415 posts since 8 Dec, 2003 from NYC (born & raised in Philly)
goddamn! _bp, you just said a mouthful. where did you get all this inside information on these company's filter designs?
== chunk
== chunk