18 replies to this topic

[br85]

Thesis:
boosting bands of digital eq is (fundamentally) a bad idea, especially if you are big on sound quality.

Why?

There are at least two reasons. Discuss.

** I should point out that very expensive dedicated digital eq devices seem to avoid, for the most part, the loss of integrity in a given signal that boosting digital eq causes.

[Edit: I should have said "high quality, well designed digital eq devices, since I'm sure we have all been burned at least once because of deceptive marketing bs somewhere in the audio world, the truth and a lot of audio knowledge is one hell of an asset in the audio world, is it not? I mean who WOULDN'T want superacoustic ultrasonic™ quantumbassmax™ technology in their 2000 dollar amplifier? ]

Edited by br85, 20 February 2008 - 08:56 PM.

[~Spyne~]

by boosting frequencies, you are increases the chances of sending a clipped signal to the amp/speakers.
you should always (if able) decrease frequencies AROUND the point you want to 'boost'

[trism]

exaxctly right..

[br85]

~Spyne~, on Feb 19 2008, 11:59 AM, said:

by boosting frequencies, you are increases the chances of sending a clipped signal to the amp/speakers.
you should always (if able) decrease frequencies AROUND the point you want to 'boost'

BINGO, and there is one other, more complex reason too. Any guesses?

[Iceman_jkh]

phase warping?

[trism]

^^ that was my thinking...

i know that differing xover slopes fcuk with phase, but i wasnt sure on the eq

[20Hurtz]

yeh i thought it was only analouge that effected the phase but i found out it also \ digital as well. Do digital xo cause shift as well?

Edited by 20Hurtz, 19 February 2008 - 03:11 PM.

[~Spyne~]

wouldnt decreasing eq bands also induce phase warping, if boosting does it too?

[Iceman_jkh]

yes. i think so.
And I believe that digital EQ/XO basically simulates analogue XO's.. right down to the phase warping. If you want unwarped phase output, you need to pre-warp the phase in REVERSE prior to running it through the XO. The XO then warps it back (to a phase perfect signal - but with the XO applied).

Also, a single order XO does not phase warp. From then on, the higher the order, the greater the warping (Ive read).

Edited by Iceman_jkh, 19 February 2008 - 03:58 PM.

[Captn_Awesome]

Hrmmm IMO not really a bad idea as long as you adjust your gains to suit. Any form of EQ affects the gain structure.

Whilst I'm the first to admit "cutting" frequencies to improve response is the preffered option frequencies can be boosted (within reason) provided the gain is adjusted to suit

[syd-monster]

i hope this is the 2nd reason you are looking for; trying to give a different perspective on it & something I learnt on my studio days...
When a particular band or set of bands is boosted it cannot take into account the affect it has on speaker responce. Given the idea that any speaker or set of speakers are and inductive (the type of resistance is inductive) item which changes according to frequency.
If were are boosting frequencies we are thus allowing for inductance changes on either side of the boost frequency and on the frequency itself. This then changes the inductance value changing the amount of thd, power etc the amplifier can perform. You end up with a peak on the selected frequency, but also usually a fall on either side. Thus making the responce curve more dipped and peaked than before...
tell me if im wayyyy of the mark.

[~thematt~]

You're waaayyy offf the mark

Two things wrong with EQ. One is the fundamental changes in phase relationships if you arent using a linear phase unit, and the other is the sharp sweet spot that is produced.

Dont forget, EQ is an electronic tool that only operates in a 2D environment. Your taking measurements in a 3D space, and adjusting a 2D device in the hope everything will change. It will, but only favorably at that spot you take the measurements.

With phase, we are NOT sensitive to absolute phase. That is, we CANNOT tell the difference between something like 25deg out of phase, and 180deg out of phase. Sorry folks, but this has been proven time and time again. We are, however, highly sensitive to RELATIVE phase, or the change in phase across the frequency spectrum. The faster the phase changes (like a steeper slope on a Xover), the more noticeable it becomes to us.

Personally, I dont think clipping is a fundamental problem of the EQ. Its a fundamental problem of the person tuning, because they cant level set properly.

Quote

I should point out that very expensive dedicated digital eq devices seem to avoid, for the most part, the loss of integrity in a given signal that boosting digital eq causes

This has nothing to do with the expense, but whether or not the unit is linear phase. If not, you will still notice a difference, no matter how expensive.

Edited by ~thematt~, 19 February 2008 - 06:28 PM.

[syd-monster]

~thematt~, on Feb 19 2008, 07:24 PM, said:

You're waaayyy offf the mark
... and the other is the sharp sweet spot that is produced.

This might have been what I was originally tought, and now understand it in a different way, although maybe not.
+1 on the 3dwave of sound vs 2d. I got that from my studio day when seting up of monitors was critical (ie when 'colouring' of the sound was something to be avoided). Thanks for sharing the info.

[~thematt~]

Could be. From what I garnered from your explanation though, you were referring more to the impedance peaks yeah?

The load you're referring to as well, me thinks is the Reactive load (or the complex component of the resistance), due to the complex back forces produced through inductance of the coil moving through the flux yeah? In which case, I think thats more of a system (amplifier, cable, coil, field flux) problem, then purely an EQ one.

What I meant was, your speaker produces 3D waves, which interact with each other in a complex fashion within the listening space. Peaks and nulls all over the joint, not including doppler distortion for wide-banders. Beaming at higher frequencies, lobing at lower frequencies etc. etc. Where the microphone is positioned (or your ears), and you adjust, you see (or hear) a result. Unfortunately, you may be at a peak point, or a null, so when you move your head (or the microphone), the response is gone.

EQ simply amplifies this problem, and means you are restrained in how far you can move your head before the response changes. This is the #1 reason why EQ is the last resort. Anyone who has ever tuned for dual-seat listening can agree with this one!!

[br85]

You guys all raised some good points, but most of them refer to why eq is a last resort (and not really that great) in general, whereas I was merely talking about digital eq's. Kudos to you guys re. the mentioning of phase issues when using analog eq's. I'm impressed =)

The second thing I was talking about specifically regarding boost with digital eq is the fact that digital eq cannot actually boost at all!. Rather, what it is doing when you use the gain feature is analyzing the behavior of the given frequency band, and synthesizing (sort of) a new waveform based on the selected band's analysis and "gluing" it back onto your signal!!! This might not sound so bad if it's done very, very well, but especially with head deck digital eq's, I'd sooner believe that a snowflake is having a ball of a time in hell =P.