ON or OFF AXIS, What constitutes which?

[altoncustomtech 773]

I've been doing a lot of thinking and pondering how to get the sound stage and imaging in my Jimmy even better.  The thought of tossing the components for an 8 or 10" midbass and some full rangers in custom pods has crossed my mind many times, along with rebuilding the door installations with something that puts the aiming pointing right where it should and going with an active setup consisting of a 7" midbass and a large format tweeter.  The latter brought me to a though as to exactly what constitutes on or off axis.  Is on axis just directly aimed, or is there a few degrees of leeway there that can allow a person some wiggle room before it's considered off axis?  As it is It would be VERY hard to turn the drivers on the driver side door far enough to be directly aimed without what would seem like severe reflection and blocking issues from the rest of the door panel, legs, etc.  

 

Some clarification would be greatly appreciated so as to lend some much needed direction to my cacophony of thoughts going on in my head, lol.

[Impious 1,476]

Every driver operating outside of it's pistonic range is going to have some deviation in frequency response as you move from directly on-axis to any degree off-axis. The further off-axis you move, the larger the deviation is going to be. The question, however, is going to be whether the deviation in response is still within acceptable limits for your goals. There is going to be some range of degrees off-axis where there response will still be acceptable. However there are too many factors at play inside of an automobile outside of the direct radiated response from the loudspeaker itself to really be able to predict what that range will be with much accuracy. And that's before we consider the listening preferences and hearing abilities of yourself.

 

The short story is that they probably won't have to be perfectly on-axis. But how far off-axis you can move before response is no longer acceptable is something you are just going to have to test.

[altoncustomtech 773]

Damn.  I was afraid you were going to pretty much answer in that way.  I've gotten so much closer to what I really want with the tweaking and tuning I've done in the Jimmy thus far, but I've gotten a little more critical about it in the last few weeks and have come to the conclusion it's just not good enough.  I'm trying to figure out how to get even better staging and imaging.  As it is now the stereo width of the stage is wide and clear but it's still lower than it should be and the imaging is all over the place.  It really sucks when I'm listening to a kick ass guitar solo, for example, and can hear it moving up and down and some side to side as it shifts through all the notes being played.  I mean, I may be getting a bit too critical, but damn it I just know it can be better than it is.  Which is why I'm trying to figure out what my best options are.

[altoncustomtech 773]

On top of that, what specifications constitute what drivers will perform better off axis or on axis over other ones?  I mean, with SO many drivers to choose from out there how the hell can the average enthusiast begin to make a choice without understanding what to look for in a driver for a particular application.  I'd pay to take classes if there were any.

[Impious 1,476]

Off-axis response is almost entirely based on cone diameter.  Cone geometry is going to play a minor roll, but not enough to make a significant difference.  Cone diameter versus wavelength is pretty much what it boils down to.  There can be slightly better performance with a phase plug, but it depends on size/shape of the plug and such so it's hard to predict without knowing the specifics of the driver.

 

Problems with image movement is usually related to frequency response deviations between left and right drivers and uneven frequency response of the system as a whole. 

[altoncustomtech 773]

Off-axis response is almost entirely based on cone diameter.  Cone geometry is going to play a minor roll, but not enough to make a significant difference.  Cone diameter versus wavelength is pretty much what it boils down to.  There can be slightly better performance with a phase plug, but it depends on size/shape of the plug and such so it's hard to predict without knowing the specifics of the driver.

 

Problems with image movement is usually related to frequency response deviations between left and right drivers and uneven frequency response of the system as a whole. 

 

Hmm...  Well, now to figure out what the hell to do from this point.  Any pointers on what I need to go after to help improve the Bravox components for the time being?  I can't imagine a good reason why there would be deviations in the frequency response from one side to the other, aside from environmental effects from relections and such.  I swear to god the quest for near perfect staging and imaging is going to drive me insane, lol.

[Impious 1,476]

It's not a quest for the faint of heart But yes.....environment, reflections and differences due to varying listening axis are the large factors. For example, if speakers are in stock locations then you are sitting at a different listening axis between the left and right speakers. IIRC those Bravox use a low but shallow lowpass on the mid, so they are running into their beaming range so the top end of the mids response is going to be different. Reflections or absorption from seats/center consoles. Diffraction from the door panel. Differences in response due to reflections depending on the tweeters location. It's a big can of worms. Without knowing anything about your system it's hard to give any suggestions. An idea of install locations, the vehicle and tuning capabilities would be a good start.

[altoncustomtech 773]

It's pretty run of the mill, I've got a picture of how it looks below.  It's a 2001 GMC Jimmy and currently has a pair of the Bravox CS60K components powered by a Sundown 125.2.  Mids are in the stock locations mounted in very heavy duty MDF baffles which bring them out from behind the factory door panel, but unfortunately aim straight across to the other side.  Doors are deadned with Second Skin CLD and RaamAudio POI CCF.  I have SDS CCF and MLV to go on the doors, but I need to fix the hinges and work on attaching the door panel better before I put the MLV in.  The tweeters are mounted on the door panel a couple of inches above the mid.  Crossovers are mounted in the foot wells and the tweeters are set to the 0dB setting on the crossovers.  As for the tuning capabilities, I have everything the Clarion CZ702 has to offer which is 3 way active capability, DTA for the six separate channels, full high/low/passband crossover capability for all six channels, and a 5 channel PEQ that can be set for any frequency from 20hz to 20,000hz in very small steps (i.e. 20, 25, 31.5, 40, 50, 63, 70, 80, 90, 100hz, etc.) IIRC.

 

So, at this point it really does pretty decent.  I get midbass response that belies what most would expect from the 6.5" mids, the stage sounds wider than the truck but it ranges from the floor up to just level with the top of the dash.  Imaging is pretty much a mess though.  At times the instruments can be localized and sound believable, and at others it sounds as though they're everywhere at once, and on occasion when you can localize something on the stage it can sound as though it goes from higher up to down toward the floor as it shifts from a higher octave to a lower one, for example.  I would imagine these are problems that everyone experiences, but don't care enough about it to do anything to fix it.  I'm getting to where I want it better though.

 

 

 

 

 

 

[Impious 1,476]

Best place to start would be to grab another 2-channel amp or switch to a 4-channel and run the fronts active, IMO.  If you are having problems with stage height that would allow you to play with tweeter positioning while also optimizing the crossover for your install.  Unfortunately you're fairly limited on EQ bands & I presume you don't have the liberty of EQ'ing left and right independently, so you'll have to focus on the most troublesome areas once you button down the crossovers.

[///M5 2,833]

I'd isolate L from R even if it means just unplugging the R or L. Then do frequency sweeps listening for abnormalities and see how they differ from side to side. Then start removing them. The less drivers playing at a time the easier it is to hear things. Of course since they are passive you have to verify together as well.

[altoncustomtech 773]

I just happen to have a 4 channel JBL sitting in my office at work.  Unfortunately I don't have an easy place or way to hook it up and mount it.  As with the first set of components I had in the truck when I first started working on it, I tried moving the tweeters all over the place to get the best response and this set, like the last seemed to sound best in the same spot in the doors.  The difference is that I'm listening much more critically to it than I was then AND I now have a HU with much better crossovers, EQ's and DTA which has changed the entire paradigm by which I listen to it and decide what to do about it.  Ugh, ignorance truly is bliss.  You're also correct the HU does not have independent L&R EQ capability.  I can honestly say if it had more bands, plus seperate L&R capability it would be the ultimate unit.  Hell, give it that and the ability to adjust it separately for each of the six channels and the thing would be unstoppable.

 

I think though, if I were going to go through the trouble of doing all that, that I may as well begin planning on how I should go about going active with raw drivers or look at the big midbass and full ranger options.  I would imagine that even ran active the components will have more limitations to overcome than it would be worth spending the time for, am I thinking right on that?

[altoncustomtech 773]

I'd isolate L from R even if it means just unplugging the R or L. Then do frequency sweeps listening for abnormalities and see how they differ from side to side. Then start removing them. The less drivers playing at a time the easier it is to hear things. Of course since they are passive you have to verify together as well.

 

I started doing that once Sean, at your recommendation at that point in time, and got busy with so many other things I haven't had a chance to get back to it.  I do think I will begin it again and finish it this time so I can report on the findings.  Thanks for chiming in here.  I figured if I can get the two of you in on it I may have a chance at learning what I need to learn to gain the experience necessary to make this thing a real SQ machine.  I'm sure you understand, but I doubt you know how bad I really would love to get as close to perfection as possible.

[///M5 2,833]

Use what you have to first understand what you want better.<br /><br />This means listen to each driver independently. Maximize that. Then start adding pairs. ie, both tweets is one pair, the tweet & mid from one side is another, and so on.<br /><br />Understand what they do on music and on tones and then start to decide what you are missing and what'd you should fix. This will lead you down the right path of not wasting tons of dough.<br /><br />Of course, I guarantee you'll come to the same conclusion Brad did. You have the drivers, the processing, the amp so run what you have active first applying what you learn above and take it a step better. Then you'll actually be able to shop for drivers with a goal of improving something so the purpose will end up being one that will be not only an improvement but a good learning experience.

[///M5 2,833]

Alton, on 05 May 2013 - 21:59, said:

///M5, on 05 May 2013 - 21:54, said:

I'd isolate L from R even if it means just unplugging the R or L. Then do frequency sweeps listening for abnormalities and see how they differ from side to side. Then start removing them. The less drivers playing at a time the easier it is to hear things. Of course since they are passive you have to verify together as well.

I started doing that once Sean, at your recommendation at that point in time, and got busy with so many other things I haven't had a chance to get back to it. I do think I will begin it again and finish it this time so I can report on the findings. Thanks for chiming in here. I figured if I can get the two of you in on it I may have a chance at learning what I need to learn to gain the experience necessary to make this thing a real SQ machine. I'm sure you understand, but I doubt you know how bad I really would love to get as close to perfection as possible.

Just to beat a deadhorse. Perfection is in the ear of the beholder. That means you should use your ears to decide what you do and don't like about the drivers you have. Then you can easily cite what you want to improve. Every driver I've owned has done something better or worse than another. Imagine that, compromise in audio. Compound that in a car and listening for what YOU want is critical. Also realize your first time out there is a ton that can be improved so even after you upgrade once you'll be right back into the ugh, upgrade stage. This is also a good and bad thing. Your ignorance is about to become even less. Means your critical listening will become more demanding. At the same time, you'll be 1000% behind the follow the sheep mentality that destroys the rest of the car stereo forums on the net.

As you play, think about how ridiculous of a reality the above is. Alone transients here make up for how many super tweeter front stages? What you have makes that sound silly, but that'll only continue.

[///M5 2,833]

For instance, you'll notice if you search through all the setups I've recommended that I never once offered up which was by far my all time favorite combo that includes a 7" and a small format tweeter. It's never matched what anyone on here would want and it was hell to minimize the processing. Neither which are good for a, "I don't like my comp set and want to step up" situation.

[ricksi30 923]

For instance, you'll notice if you search through all the setups I've recommended that I never once offered up which was by far my all time favorite combo that includes a 7" and a small format tweeter. It's never matched what anyone on here would want and it was hell to minimize the processing. Neither which are good for a, "I don't like my comp set and want to step up" situation.

Have you ever said what that setup is? Just curious...

 

 

And thanks everyone this thread is very informational.

[altoncustomtech 773]

Thank you for the help and honestly I did mean only perfection to me, I KNOW there's no such thing as real world perfection for everyone's ears, I've learned that much during my short time here, LOL.  If it were as easy as just saying go here, buy this, install it this way and be perfectly happy this hobby would be quite boring.  I have a ton to learn, I just wish I had more time to learn it and resources to learn it from.  In a hobby where almost everyone is just looking to get louder than their friends or previous setup finding good, useful information on the how's, why's, do's and don't of response issues, driver choice & placement, adjusting EQ's/crossovers/DTA, and everything else that goes into tuning a system for SQ over SPL is just tougher than hell.  I'm CERTAIN it's out there, and with yours and Brad's help I get closer every time someone posts questions on some aspect of the subject and I read, re-read, and often times save the answers you guys give because it's priceless information.  You're right though, my listening has become considerably more critical than it used to be.  I've started to become disgusted by the god awful quality of all my digital audio as well, it's REALLY not helping the issues I'm hearing with the components.  

 

 

Now to get to the critical listening of everything as individuals, pairs and as a whole to get the full picture.  Would you also recommend, when starting out with each driver/pair/set to start with the EQ flat and the DTA at 0 and adjust for each driver/pair/set as I go?  I picked up the entire set of the Autosound 2000 test CD's from Sheffield Lab which I found one hell of a deal on.  Are these going to be of any real help or are they mostly hype?  They have a ton of sweeps, tones, Pink noise, which should be helpful in some aspect even if the rest of the specialty music tracks and stuff are useless.  Either way, between the equipment I've got already and you guys' guiding I'm certain I can make my ears happy in due time.

[///M5 2,833]

Test discs are insightful in the beginning.

Let me state this a bit different.

To me this isn't a hobby, a sport, or something I even want to have to do. I just love music and hate it when the stereo fucks up the music.

That being said, the best (and worst if you want to be apathetic about speaker choices) advice I can give is to focus on teaching your ear. It's time to learn. Not to get louder or sound better, bu to understand WHY what sounds like how it does and then decide if you like it or not.

I took 14 years of piano lessons, 12 years of classical guitar, 4 years of upright bass, 3 years of voice lessons, and have played in countless bands of all genre's. One thing in common with ALL the learning. A major focus was on training your ear. The other of course is concentration which is a priceless skill for anyone and exactly why ALL children should be required to take music lessons. Needless to say those skills help more than anything in determining tones.

I also ran a soundboard 1-2x per week as a job for 3 years in an acoustically super live venue. Absolutely horrid place for sound, but damn does it teach you how to listen for anomalies and fix them.

Focus on your ear, the rest will then come easy.

[Impious 1,476]

Now to get to the critical listening of everything as individuals, pairs and as a whole to get the full picture.  Would you also recommend, when starting out with each driver/pair/set to start with the EQ flat and the DTA at 0 and adjust for each driver/pair/set as I go?

When you start listening to the system as a whole, you'll have to set the time alignment (and phase, don't forget to play with phase) before you start messing with the EQ.  TA and phase are going to affect the frequency response since it is changing the interaction of the sound waves at the listening position.  You don't want to get the EQ set to where you like it, hit the TA and have the EQ go out the window since you changed the constructive/destructive interference occurring at the listening position.  Plus it ALWAYS kills my concentration when the time alignment is off screwing everything up from the start.

 

Test tones/sweeps can be helpful for identifying trouble areas like Sean mentioned.  And once you have an idea of what the audible tone of a certain frequency sounds like, it can point you in the direction of more easily identifying trouble areas in the music.  I don't have the A2K test discs, but I do like the IASCA disc because the manual gives some instruction as to what you should be listening for in each track. 

 

It also helps to just spend some time playing with the EQ.  Run through the center frequencies with music playing, increasing/decreasing the amplitude adjustment so you can hear how changes at certain frequencies affects the sound.

[///M5 2,833]

Thanks Brad, I left out step one which is T/A. Of course, I cheat for that and actually measure time delay as it is pretty much the only non-subjective quantity in the setup. Really need to do this before listening in pairs of any sort.

[altoncustomtech 773]

Well, I've looked at a few things and tried some listening to the individual drivers and I'm convinced I need to take them active to do any real good.  Mounting the plus sized JBL 4 channel is going to be a PITA with the way I currently have the amp rack built and configured so I think I'm just going to pick up a little 2 channel amp that will be easy to hide for the tweeters.  I'm looking at getting a PPI S200.2 to run the tweeters with as it's pretty small and priced right, or perhaps a Polk Audio PA D2000.2 or some similar little class D.  I think 70 watts a channel will be more than enough though for tweeters and then I can do more with ensuring placements, phasing and the like are in good shape and begin to zero in on the actual problem areas.

 

Just in case you're wondering I came to that conclusion after running the mids without the passive crossovers and getting a MUCH better overall experience from them.  For testingI had them low crossed at 2.5khz and they sounded 10x smoother and tamer than on the passive crossovers along with a seemingly fair improvement on the bass (still questioning that).  I'm sensing the need to ditch the passive crossovers though.

[altoncustomtech 773]

This probably sounds like a dumb question, but what's the best ORDER in which to make the DSP adjustments when running active?  Something like DTA, crossovers, then EQ?  Any hints, tips or tricks on what to listen for to identify anomalies that my uneducated and limited experience may not know is an issue?

 

Believe me I KNOW it takes experience to do this, and it takes time to learn it, but I'm a very strong analytical personality type which makes me want to get things done just right a lot and I'll study it to death to do so.  Makes it really easy to miss the forest for all the trees.

[///M5 2,833]

The order is all of the above or any permutation you can imagine. Have to train your ear first.

[altoncustomtech 773]

One step at a time, I know.  

 

Per the post earlier today, 60 or 70 watts of power should be plenty to run those tweeters active, should it not?  I'm fairly certain it is, but some verification before I throw a little money at one would ease my questioning tendencies to over analyze every damn thing, lol.

[///M5 2,833]

10w will be enough.

 

Start with every driver independently.  I'd pick the passenger mid first.  Start with "normal" listening levels.  Eventually you'll have to test at higher levels to see if you need to compromise on the crossover to protect them, but first start where that won't be an issue.

 

Tweeters are confusing at first.  Beware the level.