31 band eq and polarity questions

  • Thread starter Thread starter polarity
  • Start date Start date
P

polarity

New member
I have been working with a couple friends lately that have been recording for years. They have a good sound and are always looking to improve. He recently purchased a 31 band EQ and we started messing with inverting the polarity (I think that's the correct term) to basically make all the tracks perfectly clean. Seems like it can weed out any part of the sound you're not happy with.

He was working on it this morning and figured out he could eq in only the frequencies he likes best for each instrument and then mix together like normal. My question, is there a reason NOT to do it this way. Will we work on this for hours and then figure out that the mix comes out thin or has other problems? Or is this how it's done by those who know best?

Thanks
-Will
 
Are you talking about a 31 band graphic EQ? There's not much use for those in recording. Also, I don't think you're using the word "polarity" correctly, I might be wrong. Reversing polarity on an instrument doesn't really accomplish anything unless it's out of phase with another track. Simply reversing polarity doesn't really accomplish anything, and ceratinly has nothing to do with making anything "perfectly clean", but like I said, I might not be understanding what you're trying to say by "reversing polarity".

On top of all that, reversing polarity and a 31 band EQ pretty much have absolutely nothing to do with each other.
 
Yeah I really don't know the proper terms. Basically taking a recording, I thought the term was reverse polarity but if you were to do that and play the original track the sounds should cancel themselves out? Really hard to explain when it's all just started and we don't know the terms. Basically it makes you hear nothing. Then you can use an EQ to boost a frequency, now that frequency no longer matches the original so you hear it again. So if you have any sounds in your recording (a hum or something) you basically try to avoid whatever frequency it is at on your EQ (leave it in perfect alignment with the original track) so you no longer hear it.

The part with the 31 band is just for more places to change frequency instead of like the 7band I have in my DAW. My thought is if I do this with say an electic guitar.. Then only change frequencies that I REALLY like on the guitar (make them the only audible frequencies) then do that with every other piece would the mix be lacking in the end?

Sorry for bad terms, I probably should have done googling first
 
You won't get anything useable doing EQ that way.
It looks like you can....just pick a band and boost/cut until all the freqs cancel, and then just put back the ones you want...but it doesn't work out well trying to do that.
There is overlap in all of those freqs, even if it's a 31 band EQ...so when you move one, there are side freqs being affected too...and RAMI is right, 31-band graphic EQs are not the tool of choice in recording use. Not saying you can't put one to work, but usualy only as a "got-nothing-better" option.

The more you mess with the EQ, the more issues you create...you want to EQ as little as needed when mixing.
Not saying that dramatic boosts/cuts are never done for effect...but if you *need* that much EQ to fix something...it's probably been messed up at the tracking stage.

During mixing....you want to use EQ more in cut than boost mode, since cuts don't hurt the audio as much as boosting can...and you want to be just making small adjustments overall, but again, sometimes a track/sound just needs more cut or even boost, so it's not unusual to do it, it's just that if you find yourself needing to do that across most of your tracks, most of the time...then look upstream for the reason.
 
That's what I was looking for Miro, thank you. Did I have the correct terms and process or was I way off? Either way it's something I won't keep trying, just wanted to make sure I get my terms straight if it was wrong.
 
You were using the term 'reversing polarity' correctly, you were just doing something out of the ordinary, so we weren't sure that's what you meant.

Besides your method not really working like you think it will, there is another reason why you wouldn't want to do this. Most instruments use a large part of the frequency spectrum, if you take most of it out, you will lose the sound of the instrument. For example, the main thing that makes it so you can tell a flute from a violin playing the same note is the harmonics the instruments give off. If you were to cancel out the right batch of frequencies for each of them, they would sound the same, but they wouldn't sound like either a flute or a violin. It's the natural balance of harmonics that an instrument gives off that make the instrument sound like it does. The last thing you want to do is mess that up.
 
Very true, the guy I was working with just started trying it this morning and was happy with the sound he was getting but I figured by time he did this to every track and tried to mix it all back together it was going to be missing something.

I've seen someone talking about doing something similar to this to cancel out room noise or something, but I don't know where it was or the details.
 
polarity said:
I've seen someone talking about doing something similar to this to cancel out room noise or something, but I don't know where it was or the details.
Click to expand...

There's a technique where you record somebody on live monitors, then record just the monitor mix bleeding into the mic without moving it. Mix the monitor bleed track at the same level with the polarity inverted and the bleed is gone.
 
As for 31 band eq, handy for live sound but I can't remember the last time I used one for recording.

Polarity on its own doesn't really change the sound much. It might be noticeable on signals with asymmetrical waveforms. But when there are multiple mics picking up the same thing it can change the sound quite a bit. It won't necessarily make it technically correct but one polarity setting may sound better than the other.
 
If you've got room noise in the recording, get rid of it before hitting the 'record' button!
 
You also have to realize that eq's work on the principal that you are attempting to use. when you cut something, the eq is using phase shift to cause cancellation at the frequency it is set for. When you boost, it is using phase shift to add to the frequency that it is set for.

All you are doing is making a really phasey eq out if an eq and a bunch of messing around. You can seriously do the same thing with a parametric eq with at least four bands.

There is a reason why no one knew what you were talking about. It's kind of like trying ti drive from Iowa to Nebraska and going through Texas on the way.
 
Makes sense. He was getting a great result on the guitar track he was working on, but like I said I figured the mix would be missing something when he was finished.
 
This is a simple matter of overthinking/overusing the equipment at hand.
 
Thanks guys. I figured if I could get a good answer about it anywhere it would be here.

Can I ask though, why is 31 band bad for recording? I would think anything that gives you "tighter" control would be better. IE instead of being able to move 4 or 7 things in the EQ and the rest being changed because of how it has to flow to the next one.

I'm not doubting that you're right, please don't misunderstand. I was just hoping someone could give me the why? Someone did say earlier if you are having to EQ out that much that you need it there was a problem with the recording and you should do it over anyway. If that is why I completely understand.

And thank you all for the responses, I swear I'm not intentionally dense.
 
Probably the best reason not to use a 31 band EQ is that it gives you too much temptation to tweak everything. But we usually don't use graphics much in general because they are extremely limited - all of their parameters (center frequencies, bandwidth...) are fixed. A parametric EQ is so much more useful. If you only really need to mess with one specific frequency, it makes more sense to just turn the knob to that frequency than split the whole thing into 31 bands and pick the closest one.

I want to disagree with the thing about EQs using phase shift to do what they do. It's my understanding that phase shift is a side effect of the way the EQ works, but not exactly why they work. In the analog world a frequency range which is being attenuated can almost be thought of as being "shorted out" at some point in the circuit. If you add gain to what's left after that, you get boost.
 
31 bands are kind of clunky. The frequencies that they move are the same 3 notes in 10 octaves. For example, the one that I have starts at 20hz, 25hz, 31.5. That is an E, G and B (not exactly, but it helps illustrate my point if you just go along with me). The next three are 40hz, 50hz, and 63hz...E,G,B an octave up. It just goes all the way up the spectrum that way, the same three notes in every octave. So it really isn't as versatile as you would think.

Also, if you can't get something to sound good with a 7 band parametric, you need to rerecord it. Nothing should need that much EQ.

Ashcat: I suppose I could be wrong, but that has been my understanding.
 
Yeah, but the 31-band graphics look cooler. You can do all kinds of designs and smiley faces. :D
 
You must log in or register to reply here.
Back
Top