What do id do with the rest of the eq bands after i dial in the suggested ones

  • Thread starter Thread starter Chris Jahn
  • Start date Start date
SonicAlbert said:
Is it just me, or does it seem like Chessrock is getting more confrontational lately?

Regarding the original topic, use a parametric eq instead of a graphic for what you are doing. Use only the number of bands you need in order to create the sound you are looking for. In other words, start with one band and then add them as necessary.

I'm just curious if anyone knows the answer to this: In a graphic eq plugin, are the bands that are set to zero disabled or still active? In other words are they bypassed at zero, with no processing occurring, or are they still active and processing the audio, but at no change?

I believe when a plugin loads, it is going to use it's full cpu cycles, regardless if you use one or all bands. It HAS to be this way for the simple fact that having to engage more cpu cycles in real time if you decide to try out a band would probably wreak living hell in a DAW!!!
 
Ford Van said:
I believe when a plugin loads, it is going to use it's full cpu cycles, regardless if you use one or all bands. It HAS to be this way for the simple fact that having to engage more cpu cycles in real time if you decide to try out a band would probably wreak living hell in a DAW!!!

I think this is true for memory allocation, but shouldn't be true for CPU, which should be dynamic. I know for shure that my CPU load changes as I engage or disengage EQ bands in the Samplitude mixer. For plugins, it may vary from one plug to another, and the quality of the code writing.
 
Robert D said:
I think this is true for memory allocation, but shouldn't be true for CPU, which should be dynamic. I know for shure that my CPU load changes as I engage or disengage EQ bands in the Samplitude mixer. For plugins, it may vary from one plug to another, and the quality of the code writing.

Yeah, after thinking it through a bit more after posting, I came to that same conclusion. My bad. :)
 
Robert D said:
I think this is true for memory allocation, but shouldn't be true for CPU, which should be dynamic. I know for shure that my CPU load changes as I engage or disengage EQ bands in the Samplitude mixer. For plugins, it may vary from one plug to another, and the quality of the code writing.

In a parametric you can kick the bands in and out, which changes the CPU load as you've mentioned. But in a graphic eq aren't all the bands active all the time, i.e. the CPU load does not change as you bring bands up or down from the zero position.

I can see your point Southside. Any band that is zero will have the bits go in and come out the same. However, while this may seem academic, those bands will still have gone through the eq algorithm, correct? In other words, they've been processed, and in the process the plugin calculated that no processing was needed, or processed the audio with zero changes.

That's a bit different than having no processing, i.e. a pure bypass.

The thing is, I can notice some subtle difference in tone sometimes when a plugin is engaged, even if all the settings are at zero, or whatever the "blank" audio pass-through state would be.

So a graphic eq plugin with 30 bands of eq always on sort of scares me.
 
SonicAlbert said:
Any band that is zero will have the bits go in and come out the same. However, while this may seem academic, those bands will still have gone through the eq algorithm, correct? In other words, they've been processed, and in the process the plugin calculated that no processing was needed, or processed the audio with zero changes
That's a bit different than having no processing, i.e. a pure bypass.
Weeeeellllllll...processed digital data with zero changes is exactly identical to unprocessed digital data. In such a case the fact that the data was "run through" a process is irrelevant. If I wrote a program that takes a number and adds zero to it, and you ask me to give you the number, it's completly irrelevant as to whether I give you the pre-processed number or the processed result.

What I'm wondering is if you're dealing with 24-bit or 32-bit data, and you're only writing the plug with 24-bit floating point registers somewhere along the way, rounding errors can occur that would be significant enough to be audible. But even then, if the input value is truely zero - meaning that it's also zero to any signifigant number of decimal places - there should still be no errors (unless 32-bit is being truncated to 24-bit, but you'd know about that.)

For this reason, I'd suspect (but don't promise ;) ) that any coloration caused by the introduction of a plug into the digital chain is not necessarily caused by a lack of band "bypass", but rather by the overall architecture of the plug itself or perhaps by the extra processing contraints put upon the CPU/FPU itself by the introduction fo the plug. But the slider position remains irrelevant in that regard. It doesn't cost any more or less processing power to "process" at 0 gain than it does to process at +18dB. It's still the same instruction set with the same size of data, the only thing that's different are the values being plugged into the equation, and they in and of themselves don't have an effect on processing speed (000000000000000000000000 costs just as much in CPU cycles to churn through the circuitry as 111111111111111111111111 does.)

G.
 
SouthSIDE Glen said:
As hard as this may seem for you to grasp ... nobody gives a rat's ass what an anonymous sales copy boy has to say about anything. Don't you have a Snickers commercial to pull?

:D What does that even mean? ? :confused:

I do like Snickers, though, I must admit. Mmm.

Going back and reading your answer to this EQ band question ... I have to admit, I'm rather impressed.

I think people who criticize you for your lengthiness don't give you enough credit for the amount of words you can pack in to an answer.
 
chessrock said:
Going back and reading your answer to this EQ band question ... I have to admit, I'm rather impressed.

I think people who criticize you for your lengthiness don't give you enough credit for the amount of words you can pack in to an answer.

You SHOULD be impressed in that at least he HAS some answers to give people!

Your brevity surely does not equal quality!!! ;)
 
chessrock said:
I think people who criticize you for your lengthiness don't give you enough credit for the amount of words you can pack in to an answer.
I can think of a two word answer for you...

:D

G.
 
SouthSIDE Glen said:
I can think of a two word answer for you...

:D

G.

Kiss me? :p





























....................(sidestepping this mess)
 
Ford Van said:
You SHOULD be impressed in that at least he HAS some answers to give people!

Your brevity surely does not equal quality!!! ;)


I dunno. Maybe my eyes are deceiving me, but I could have sworn the guy got his answer in the first freakin' response. Some newb asks what to do with his leftover bands, and JohnnyC replied: "You don't have to do anything with them. If your ears are telling you it sounds good only moving two or three bands then leave it at that."

Why the hell does anything else need to be said about the subject? What on God's green earth could conceivably be added to or expounded on after that answer? The only thing I can think of is that either no on bothered to notice that his question was already answered after the first reply ... or people just like to read their own posts and marvel at how much more dense they can be with their verbage.
 
chessrock said:
Why the hell does anything else need to be said about the subject?
Yet you found the need to step in with something totally off-topic, absoutely unproductive, and making a redundant point already made by RD. Physician, heal thyself.

G.
 
Last edited:
Can You Feel The Love Tonight......


I'll point out that some plugs, particularly ones designed to emulate famous hardware, will intentionally change the sound going through them even with a flat setting, just as their HW counterparts do.
I'm guessing (an educated guess) that there's no more reason that inactive bands in a parametric EQ can be switched out than in a graphic EQ.
A graphic EQ is a a set of fixed center freq, fixed Q filters, arranged so that their half power points line up to leave no holes in the spectrum. In a hardware 31 band graphic, there's 31 circuits that are identical except for differing component values to determine the operating freq of that filter. Software usually just emulates hardware, and the code for a 31 band graphic EQ probably has 31+ modules, each describing the action of each of those filters. It should be that if there's no cut or boost in a band called for from the gui, there's no call to the module. The main piece of code would execute a series of IF THEN ELSE statements. IF a cut or boost in band x has been called for by the gui, THEN call module X, ELSE check next band.
 
chessrock said:
Why the hell does anything else need to be said about the subject? What on God's green earth could conceivably be added to or expounded on after that answer? The only thing I can think of is that either no on bothered to notice that his question was already answered after the first reply ... or people just like to read their own posts and marvel at how much more dense they can be with their verbage.
So does that mean you think my suggestion to try a parametric instead was unnecessary?
 
MadAudio said:
So does that mean you think my suggestion to try a parametric instead was unnecessary?
Obviously, MA, you only posted that because you love the sound of your own voice, just as I only posted the ear training w/EQ technique because I have nothing better to do that to listen to myself give the same advice that I've given in a dozen threads before. :rolleyes:

First chess complained that I made too many posts, until everyone discovered by looking at the server stats that not only is his posts-per-day average is practically the same as mine, but also that he's been averaging that number of posts for a LOT longer than I have. And the majority of those posts are just shooting spitballs at others.

Now he tried complaining that my posts were too long, until it was pointed out to him that they were at least relevant and useful, unlike his constant irrelevant and unhelpful whining and sniping.

So then he tried laying the "redundant" thing on me, even though he's consistantly guilty of the same thing - even within this thread.

Give it up, chess. If you don't like it, you're welcome to leave anytime you like.

G.
 
SouthSIDE Glen said:
I only posted the ear training w/EQ technique because I have nothing better to do that to listen to myself give the same advice that I've given in a dozen threads before.

I must have missed that one. Will look it up.
 
Back
Top