Crossover Slopes

[Phoenixcolt]

What do the folks who use deck crossovers typically set your deck slopes to?

I have an Alpine that allows me to go -6, -12, -18, or -24 db/octave. I am mostly concerned with the front and sub crossovers. The rear gains are turned way down through the deck crossover since I barely use them.

Right now my front is set to 80Hz hp, 18 db/oct, and -2 db. My sub is set to 80Hz lp, 6db/oct, and 0 db. Do you usually just adjust the slope to what sounds best to you or is their a method to setting the appropriate slope?

I am wondering if I should turn the front crossover gain back to 0, and just adjust the slope downward until I get the desired removal of door vibrations , and then just leave the sub slope at 0, or if there is a better method. I know I will only be able to get so much out of stock speakers, but for a little longer, that is all I have to work with.

[Francious70]

Just go with what sounds best to you. That's really all that matters with audio, what YOU like. There is a scientific way to adjust these things, but if you like it the way you have it then who cares.

The different slops have different phase changes such as 90* for -6dB and 180* for -12dB and 270* for -18dB and 360* for -24dB.

Check out http://www.bcae1.com

[lashlee]

To add what Francious noted, that you generally want to cross smaller drivers like tweeters with a steeper slope to help protect the driver. Too shallow on a larger driver may allow some "bloating" or bleed over and cause some muddiness. Depending on the driver, install and desired effect I would start somewhere around a 12 db/oct and play with it and see what sounds best.

[smgreen20]

Xover points and slopes are ALL driver and placement dependent. Dynaudio uses a 6 dB/Oct slope for their tweeters, so a stepper slope for a smaller driver is irrelevant. Things to look at are the drivers freq response and at what rate they drop off. You shouldn't go outside of those parameters. Start at the lowest point and slope that's safely within the drivers freq response and adjust until it sounds good to YOU. Do that for all drivers you can.

[stipud]

Nah, what sounds good to you isn't always what sounds right. Screw what your ears think until you learn what "sounds good" means first.

For example 6dB/oct is probably way too shallow of a slope for your subwoofer. It will be playing gobs of midbass with that setting, which makes it sound boomy and easy to localize. Rule of thumb is that the human ear can only localize sounds above 60Hz or so, so in general the lower the crossover frequency, the better (for imaging). Of course this means you need to have adequate midbass up front to extend down to 60Hz as well, which is quite unlikely. So you will have to fiddle around and balance and compromise, but there are a lot of educated choices you can make versus just doing "what sounds good".

[dwnrodeo]

Nah, what sounds good to you isn't always what sounds right. Screw what your ears think until you learn what "sounds good" means first.

For example 6dB/oct is probably way too shallow of a slope for your subwoofer. It will be playing gobs of midbass with that setting, which makes it sound boomy and easy to localize. Rule of thumb is that the human ear can only localize sounds above 60Hz or so, so in general the lower the crossover frequency, the better (for imaging). Of course this means you need to have adequate midbass up front to extend down to 60Hz as well, which is quite unlikely. So you will have to fiddle around and balance and compromise, but there are a lot of educated choices you can make versus just doing "what sounds good".

Isn't what "sounds good" subjective to a degree? I mean, there's obviously stereo's that don't sound good, but when it comes time to fine tuning, I believe that your ears are a valuable tool. I do agree that you need to learn what "sounds good" first. I like to use a decent quality set of headphones or earbuds to let my ears learn what "sounds good". Car audio environments are terrible for "good sound". Drivers for the most part cannot be mounted on the same plane, at the same distance, or even on the same axis from the driver. This is going to cause phasing issues with your speakers. By experimenting and using differnet crossover slopes, you can improve these phasing issues, much like reversing the polarity on one driver to get it acoustically in phase with the other. There are more scientific ways to doing this, but they require much more work, time, and expensive equipment.

Begin with your highs only, turn your subwoofer off for the time being. For crossover slopes, I like to start high and go lower. Start at 24 db/oct and switch to 18db/oct. See which sounds better, (i.e. more midbass, better imaging, etc...) Try 12db/oct next. I would avoid using 6 db/oct unless you know your speakers frequency response curve. 6 db/oct crossovers are typically used where the speakers frequency response starts to fall off naturally to begin with. After you get your highs sounding good, turn your subwoofer on and keep the highs on. Start experimenting with the different crossover slopes. Also, I'm assuming your Alpine has an option to reverse the phase of your subwoofer. Try your 24db/oct crossover slope and then change the phase of your subwoofer. More bass? Keep that setting. Less bass? Go back to your original setting. Changing the phase of your subwoofer can allow you to keep your 24db/oct steeper crossover slope and allow the drivers to be acoustically in phase with each other. I do this because I'm using my subwoofer amp crossover which is fixed at 24db/oct.

Here's a little more information about crossovers from PG: http://download.phoenixphorum.com/TechSupport/
Click on X-Over Chart.

[stipud]

Isn't what "sounds good" subjective to a degree?

Yes and no. Theoretically the best sounding stereo is the one that accurately and faithfully reproduces the input signal precisely in the way that it was recorded (the output signal is identical to what the input signal is). So on paper there is no wiggle room at all, and there is such a thing as an objectively good sounding stereo. Since our transducer technology and the environment we use them in is far from ideal, there is definitely room for subjectivity in creating a better sounding system. However, one must make at least educated guesses about what sounds good, based on solid understanding of acoustics, use of measurement tools, etc.

For example while in smgreen's example a Dynaudio tweet can be crossed over at 6dB/oct, most cannot, since the drivers do not extend nearly as low as the dynaudios do. In order to estimate how to get something like the crossover slope correct, you would need a frequency response graph and full T/S parameters of the tweeter at a bare minimum, to see how low it extends without distortion, so you can plan your crossover frequency and slope accordingly. Similarly, getting a midrange to blend in properly, especially at a different slope than the tweeter, takes some work. For example, since the tweeter is rolling off at a lower rate, you may have to underlap the crossovers (i.e. setting the crossover frequency of the mid below that of the tweeter) in order to have accurate blending.

On top of all of that, where the driver is positioned, and how it is installed also create major changes in how a stereo sounds. Ideally there should only be two point sources for stereo sound, which provide the full range signal. However we don't have speakers that do this, so we have to compromise by mounting tweeter, midrange, midbass and subwoofer in different locations. Now we need to account for the phasing and time alignment changes, as well as how the various waves interact with each other at different angles, causing cancellation, etc.

We also need to account for the changes in the sound that the environment has created, or even the unnatural sounds the speakers themselves make. This requires EQing, which takes mastery of an RTA and a good understanding of acoustics to properly resolve.

While it is difficult to fully master everything, there are a lot of educated decisions you can make in order to put yourself further ahead. In my experience, when untrained or inexperienced people have simply done "what sounds good", it usually sounds absolutely terrible. Those of us with more experience can of course help to find more accurate starting points. But without actually being there in person with the correct measuring tools, we cannot tune his stereo for him. Of course none of us are nearly as talented as most acoustical engineers, and would be making a laughable attempt to someone in the know! But I do think we could come up with something better than a 6dB/oct slope @ 80Hz on a subwoofer?

[Phoenixcolt]

Thanks everyone, this has been helpful, sometimes just hearing how other people do things gives me enough insight to try something different or make a better decision on how I want to tune.

[dwnrodeo]

Yes and no. Theoretically the best sounding stereo is the one that accurately and faithfully reproduces the input signal precisely in the way that it was recorded (the output signal is identical to what the input signal is). So on paper there is no wiggle room at all, and there is such a thing as an objectively good sounding stereo.

Exactly! "Theoretically". When tuning our systems by RTA, we attempt to set our stereo's as flat as possible in an attempt to reproduce the input signal exactly as it was recorded, meaning if playing a 0db pink noise track, our stereos will produce each frequency at the same db level. However, I'm willing to bet that most people do not enjoy listening to their music like this. The way humans perceive sound is not quite the same as say an RTA meter. That's why we have A-weighting, B-weighting, and other filters to compare how humans perceive sound pressure levels and frequency (equal loudness). This is where using your ears comes in. Yes, you will need to be educated as to what sounds good and not everyone has the "golden ear" (myself included), but that's part of the learning and experimenting process.

Also, "the way it was recorded" isn't always the best for sound quality. The "way it was recorded" is subjected to the opinions of the people engineering it. Look at Metallica's last album as an example. Although, there's no amount of tuning available to correct a corrupt input signal. Clipped signal in = clipped signal out.

For example while in smgreen's example a Dynaudio tweet can be crossed over at 6dB/oct, most cannot, since the drivers do not extend nearly as low as the dynaudios do. In order to estimate how to get something like the crossover slope correct, you would need a frequency response graph and full T/S parameters of the tweeter at a bare minimum, to see how low it extends without distortion, so you can plan your crossover frequency and slope accordingly.

I am confused. You say the Dynaudio tweeter can be crossed over at 6db/oct and most cannot because they do not extend as low as the Dynaudio's. Why? If my tweeter has a natural roll-off at 5,000hz and the Dynaudio's have a natural rolloff at 1,200hz (I don't know the real frequency) why is a 6db/oct slope not appropriate? My understanding (and please correct me if I'm wrong) is that if a tweeter has a natural rolloff at a certain frequency, say it's 12db/oct at 5,000hz and we add a 6db/oct highpass crossover at 5,000hz, the driver will dispaly a frequency response that rolls off at 18db/oct at 5,000 hz. If another tweeter has a natural rolloff frequency of 12db/oct at 1,200hz and we add a 6db/oct highpass crossover at 1,200hz it will display a frequency response that rolls off at 18db/oct at 1,200hz. They will both essential display characteristics of rolling off at 18db/oct, but at different crossover points. Not only that, but they will only have a 90 degree phase shift because of the 6db/oct crossover slope instead of 270 degrees if an 18db/oct crossover slope was used at a point higher than the drivers natural rolloff frequency. Maybe Dynaudio chose to use a 6db/oct slope to prevent phasing issues with the midrange? IDK. The reason we use the higher order crossover slopes most of the time is because the crossover point we want to use does not always coincide with the drivers natural frequency rolloff point. Subwoofers for example rarely naturally rolloff at the crossover point we want to use (80hz or less most of the time). In order to prevent them from playing those higher frequencies, we give them a steeper crossover slope to keep more of the higher frequencies out.