do not go beyond a 6db cut or boost with software eq's?

[jugalo180]

i read in a few threads that it is reccomended that you not cut or boos past 6db with software eq's due to thier converters. if this is true then is it also not reccomended to do roll off's or band passes with software eq's? yes, or no, please explaine.

thanks

 

[James Argo]

...Gee, very good question!!! Make sense. I dunno, anyone ?

 

[ambi]

I've heard this many times. It's a good idea. If you have to then go for it. I usually cut like -18 sometimes in the sub 30hz area if it's WAY too boomy.
But i try to never boost more than 3, maybe 5 max.

 

[littledog]

i read in a few threads that it is reccomended that you not cut or boos past 6db with software eq's due to thier converters. if this is true then is it also not reccomended to do roll off's or band passes with software eq's? yes, or no, please explaine.

No. Assuming for the sake of argument that it's a valid rule (and in the end I prefer to just do what I need to do to get the sound I want regardless of the rules) it was not intended to include the applications of hi pass and low pass filtering, notch filtering, etc.

Plus, every particular brand of software EQ will have it's own particular strengths and weaknesses.

 

[Toki987]

Ok, so where the issue with the converters, and what would the effects of making -6 db cuts at say 4 different frequencies below 100hz? what would that do to the converters?

 

[jugalo180]

cool

thanks everyone for the interest, and input in this thread.

 

[MONTE]

Piss on rules...does it sound OK? Trust your ears. I'm not saying the rule is untrue, but rules are broken all the time in the recording industry. If it gets you where you need to go, do it!

 

[pipelineaudio]

like monte said, and I dont see how converters have anything to do with it.

 

[volltreffer]

I read in an article on recording/mixing (dunno by whom it was) that the guy recommended not to use any cut/boost with more than 6db (except low pass and high pass filters) for the first 50 songs that you mix. He assumed that the recording thing went wrong if you need to...

aXel

 

[Toki987]

Thats interesting. With you saying his statement being based on an assumption, then perhaps he assumes that if you have to do a lot of eq cutting, the initial setup for recording the track needs is wrong if it requires that much cut or boost. Like maybe mic type or placement, onboard eq on an instrument, or something else not setup properly initially for the recording. That sounds more like a training issue, or a lack of confidence in software eq engines. I have always heard that 3db is the smallest readily distinguished change in volume that most on average can detect, my assumption would be that it should apply as well to change in amplitude of frequencies that compose a sound as well. Of, course a trained ear would more likely detect smaller increments. I could be wrong, but then again I believe we are dealing with assumptions here, and I`m just thinking out loud casue its supposed to be good exercise.

 

[Middleman]

From what I have studied on this, all based on various internet experts. And I am open to correction in the pursuit of accuracy...

The convertors ability to respond is only as good as the quality of the convertors so their results are kinda fixed. The software's ability to cut and add data (1s and 0s) is the critical factor.

Cutting should not be a problem with digital as you are just throwing out 1s and 0s. If there are averaging functions in the code of the plug in then this will come into play. If there are critical tolerance levels, in the code of the plug in, they may be exceeded at -6db, (The convertors may dictate the tolerance levels and may be taken into consideration by the way. Probably not however by your averge plug in coder)

Boosting is a problem because now you are asking for the software to make up digital 1s and 0s plus there may be averaging functions that effect the final request that is sent to the AD convertors. This is purely best algorithm wins in creating digital data.

That is why it is better to cut than to boost in the digital realm. It is also why some plug ins suck and some are stellar.

 

[jugalo180]

that's what i'm talking about

ask and ye shall recieve.

 

[Toki987]

now that makes a lot of sense. Subtraction from known values should be done with no problem from the computational ability of the processor. However, to create unknown values based on known values, it would take a software algorythm. SOme may no be written as well as others and the results may vary accordingly. That sounds like a good logical perspective Middleman. I buy it.

 

[bleyrad]

i would think that recommendation would have been made with phase problems in mind, not because of this digital "making up numbers" stuff - which, by the way, i have my doubts about the validity of.


6db is maybe a rough guide, it's not like 6.1 is suddenly so much worse. and play around. there's no set way. it's ridiculous for someone to say "only do this your first 50 mixes"... you don't learn anything that way.

the same problems occour in analog EQ's too.

waves has a linear-phase EQ you can use to avoid this, but it introduces massive latency, so it's not too useful on individual tracks. i always use it for post-mix EQ though.

 

[crosstudio]

I heard that rule about not cutting more than 6db several years ago, when i was using like cakewalk 8 or something.

back then all of us DAW users were recording in 44.1khz/16bit and the effects were 16bit as well.

now, I record in 48khz/24bit and the effects (I use Waves) and Mastering software (SEKD Samplitude Master) algorithms utilize 32bit floating point for dynamics processing.

so I no longer worry about that 6db rule.

besides have you ever actually cracked open your "analog" gear? most of that *&#$%^ is just a circuit board and a wall wart.

 

[jugalo180]

good

man i was feeling closed in for a minute there. thanks everyone. hey cross i have waves also. i read that it was regarded as like the best software for mastering. you mentioned that you are using SEKD Samplitude Master also. what is offered in there that waves can't handle? what do you know about the psp mastering software?

 

[crosstudio]

In addition to doing CD burns and doing a good job with the sample rate and bit depth conversion, SEKD Samplitude Master has some nice filters and a spec analyzer.

Since SampMaster (ie.. RedRoaster) uses 'objects', you can put a different set of directx components on each object.

Now of course, you could do all of that in Sonar (or whatever) and then take the converted wave file that contains all of the tracks on your CD and put it into a CD program to add the track indexes, but I like being able to do all of my mastering in one program.

before I got Waves, I use to use the multiband compressor that is built into SampMaster.

 

[jugalo180]

cool

hey cross, i'm using waves lab analyzer right now, and it's helped me out a lot in the two weeks i have used it. is sample master more accurate? user friendly? you mentioned that sample master has a nice bit depth converter, is it better than waves L1? i need some of your advice ob1. i use waves mastering to bring my mixes out, for example the multiband compressor which is really impressive. i have no wish to do my own mastering, i just want to get my mix as close to what i want it to sound like, and the mastering tools are the only thing (i) have found to work for me. is this a frowned upon practice? i also record and edit in acid pro, which i find a lot easier than acid video, i have a lot more ease. i read that acid has it's own particular sound which wasn't regarded as bad, but it wasn't regarded as good either. would i get a better quality sound from cakewalk, cubase, or the sample master you use? if yes, is because of the way the software converts it's information? i know i'm asking a lot of questions, and if only one get's answered that would be more than enough.

thanks

physical therapy

 

[Boray]

i would think that recommendation would have been made with phase problems in mind, not because of this digital "making up numbers" stuff - which, by the way, i have my doubts about the validity of.


6db is maybe a rough guide, it's not like 6.1 is suddenly so much worse. and play around. there's no set way. it's ridiculous for someone to say "only do this your first 50 mixes"... you don't learn anything that way.

the same problems occour in analog EQ's too.

I agree with that, bleyrad!

/Anders

 

[Middleman]

Yeah, Iv'e done some more research on this and the 6db thing came about from coments made by one of the engineers in the book Mixing Engineer's Handbook. I can't remember who it was just now but was reading the book last night at 3am (couldn't sleep).

bleyrad , you are right on, it comes from the phasing effect from some analog gear from years past. The problems of digital make up is also a real issue however, not necessarily at 6db but I have read some stuff over at musicplayer.com on the Massenburg's forum where they go into this in detail if you are curious.