Any benefit to separating sounds?

[Randy Robinson]

Hi everyone,

I have searched for this and came up with nothing, so I am asking the experts.

I am wondering if there would be any benefit from separating out certain sounds from an audio track?

To clarify more, would it sound better to have ONE driver play multiple sounds at once ? or would it sound better to have one driver play the vocals and one driver play the guitar and one driver to play the drums and so on.... I am not sure if this would have any adverse effects on the sound ? I am not an audio engineer or do I know much about the technical side of audio, so I thought someone could chime in.

Thanks

 

[j_garcia]

Difficult to say since there are benefits to different designs. One driver is a good choice because you don't need an x-over inline. One driver can't do everything perfectly though, so to me, perhaps a 2-way is ideal. Breaking it down further lets you extend various ranges to drivers best suited to them, but then the x-over can start to become complex and driver interactions start to occur. So to separate those even further into specific sounds would seem to be adding much more complexity for potentially little or no benefit. It may work theoretically, but would the benefits be worth the complexity?

 

[Randy Robinson]

I know this is sort of hard for everyone else to understand because I am the one who is thinking all of this.
One factor I am thinking of is this : Intermodulation Distortion
Intermodulation (IM) is the interaction of two or more frequencies. This interaction creates new frequencies that are the sum and difference of the reference tones, e.g., 100 &1000 Hz, may produce sidebands of 1100 and 900. Intermodulation adds non related, highly dissonant frequencies that are far more objectionable than harmonic distortion, and audible at levels lower than any other nonlinear distortion. Music is made up of many simultaneous frequencies making IM distortion a major concern. It is consistently agreed that IM is more detrimental than other forms of nonlinear distortion.

I know that this can be slightly avoided by a 3-way loudspeaker design, but then you have crossovers to consider and those provide a whole different level of complexity.

Side note... do they even really make 5-way loudspeakers anymore ?

 

[lovinthehd]

How do you imagine you would split out particular sounds/instruments from the recording? Without having the multi-track original and a mixing board, don't think you'll get too far with this.

 

[Randy Robinson]

That was my idea, when mixing custom tracks on my computer. I can keep them separated out.

 

[Randy Robinson]

I just image it would sound better if you had 2 drivers on each side one playing the vocals and lets just say the other driver playing all the instruments, to keep it simple... I would think the vocals would play clearer that way vs one driver having to play the vocals and the instruments all together... but I do not know what that would do to the sound. Would it make it sound to weird ? I am sure it would all depend on placement, but in this case the drivers would be placed close to each other in the SAME loudspeaker cabinet.

 

[lovinthehd]

I just image it would sound better if you had 2 drivers on each side one playing the vocals and lets just say the other driver playing all the instruments, to keep it simple... I would think the vocals would play clearer that way vs one driver having to play the vocals and the instruments all together... but I do not know what that would do to the sound. Would it make it sound to weird ? I am sure it would all depend on placement, but in this case the drivers would be placed close to each other in the SAME loudspeaker cabinet.

Not a speaker designer but sounds like it would make designing such a speaker a royal pain. How many drivers would it need? Why mix instruments together?

 

[Randy Robinson]

well without getting into the design aspect of it.... I am just looking at getting some feedback from audio people on how that might sound. It is sort of like making a speaker with lets say a 10-way crossover, but without any crossovers, so you could get ultra hi fidelity and have uber control over how the music sounds. Just image how a LIVE band is set up, like that but obviously more compact and refined.

 

[shadyJ]

If you want to reduce intermodulation distortion, the thing to do is narrow the frequency band that the driver is tasked with playing, not simply reduce the amount of content that is sent to it, like what you are suggesting with a speaker for every instrument. But of course, that brings in all sorts of other problems.

I think your idea does show a slight misunderstanding of driver operation. As long as it is operating in a linear manner, you don't have to worry about IMD or other types of distortion. Some drivers can exhibit outstanding linearity. They can reproduce all kinds of complex material without adding anything significant. Furthermore, the audibility of IMD is not well understood, but, generally speaking for loudspeakers, if you have low THD, you will have low IMD. Just keep the drivers operating within their comfort zone. When you push them hard, they will run into all kinds of distortion, so if you want an absence of distortion, just use speakers that keep their drivers well within xmax.

 

[lovinthehd]

You'd need crossovers, active crossovers would be easier in such a system than passive I'd think but could get expensive with electronics...

 

[Randy Robinson]

the eq for each track would be set at the software level. before it gets to the speaker

 

[j_garcia]

There's a lot that goes into the speaker design to get it to sound good. The source material does not determine how the speaker sounds. If you want to pursue this, I'd start with an understanding of how speakers are designed.

 

[Bucknekked]

What we need here is a @TLS Guy response on speaker and crossover design. Since the OP mentioned a single speaker handling each instrument or voice (one voice, one speaker) I would expect a picture of that speaker designed back before I was born that does it all but looks like a used hotplate that TLS Guy has. In fact, I think he actually has one of those speakers.

I am not worthy to post a reply as I barely understand the benefit as compared to the pain-in-the-ass-factor of putting the solution together. I am always interested in a good discussion however.

 

[Randy Robinson]

I will describe in more detail later...

 

[Randy Robinson]

just imagine you have a live band setup but at home. lets just say for argument sake that each speaker has 7 drivers. maybe 2 tweeter 4 full range and one for low end. my custom audio track will have 7 channels, from my computer the signal will go out to a two 7 channel receivers and then the speakers will hook up to that. So all the sound adjusting is done on the software level the same as it would be if you where to mix all those tracks own to a 2 channel recording. I figure there would be some advantage from keeping the tracks separate and not mixing them. Better fidelity, better sound stage, less distortion.... I set up a test with only 2 speakers on each side and I did send the vocals to a separate speaker and everything else to another speaker, so I only had 2 speakers on each side, but to me it sounded better and the vocals where more clear.

 

[William Lemmerhirt]

I think another issue would be asking for example, one driver to cover only one instrument. Each instrument usually covers a broader range of sound than one driver can cover. A drum kit for example can range from the high 30's all the way up to the limits of some tweeters. Check out this chart. You'll see that multiple drivers even for a single instrument is just going to be a requirement. Also, I think there would be time alignment issues associated with all the drivers involved. Maybe with the right software that could be corrected. Fwiw, I may just not understand where you're going, so keep that in mind. Also, speaker/driver tech is a very mature market and you may have a solution for a non existent problem. A Rebel without a cause perhaps.

 

[William Lemmerhirt]

http://www.independentrecording.net/irn/resources/freqchart/main_display.htm
Here's a link to the interactive chart.

 

[Randy Robinson]

some interesting feedback... I am now looking up
Fundamental Frequency and Harmonics as well as some other info.
Like you said William, maybe I am just a rebel without a cause..

 

[Randy Robinson]

So, here is reference to one main point my setup would theoretically eliminate.

For the full article. http://www.homestudiocorner.com/dealing-with-low-mids-and-a-muddy-mix/

For a short reference from that article.

......So what does Kent’s quote have to do with us? Well, when mixing a song, we’re combining a bunch of tracks. If the fundamental frequency for each of these tracks is somewhere between 250 Hz and 500 Hz, then we’re obviously going to have a huge build-up in that range.

This, ladies and gentlemen, is where the mud comes from. You know what I’m talking about. Each track you recorded sounds amazing when you solo it, but then when you bounce a mix of everything, it’s all muddy and cluttered, and it gives you a headache. (I’ve done mixes in the past that literally give me an instant headache…sigh.)

Dealing with the “mud range”
So how do we take this information and put it into practice? Here’s my advice: next time you’re mixing, and things are sounding muddy, don’t immediately reach for the high frequency EQ knob and turn it up. Instead, spend some times cutting out some of that low-mids.

It’s been said many times, but I’ll reiterate it here. EQ is meant to be acorrective tool more than it’s meant to be an effect. Removing problem frequencies (doing an EQ cut) is almost always more effective than boosting other frequencies.

The easiest way to do this is to take one of your EQ bands in the low-mid area and do a big boost. Next, sweep the frequency up and down until you find where the “mud” is coming from. (I have a sneaky suspicion it’ll be between 250 and 500 Hz.) Once you find the frequency, turn the gain down until you’ve done about a 3 dB cut and take a listen.

Oftentimes a 3 dB cut is all it takes to clear things up. Use more or less as you see fit, and hopefully there will be much less mud in your future.

 

[j_garcia]

OK, but as it stands right now, it will be difficult and costly to implement the way you are describing. If you eliminate it at the mixing stage, why would we need to do anything to the speaker?