DiyMobileAudio.com Car Stereo Forum banner

Status
Not open for further replies.
61 - 80 of 418 Posts

·
Registered
Joined
·
8,967 Posts
I would compare them full range to compare the extremes of their operation and then again bandpassed but without tweeters or midbasses.
 

·
Registered
Joined
·
3,032 Posts
I agree that the mids should be tested outside the car, alone, to get all of the essential base data & detailed comparison. However, I think testing inside the car will also be important to get a "real world" response & comparison.
 

·
Registered
Joined
·
3,498 Posts
Discussion Starter #63
Okay...so it is looking like the consensus is stand alone testing of the midrange. I will try to test both in car and out of car (will let you guys know more one I get all the drivers in my possession).

Do you guys want them tested in a sealed enclosure in the car or free air? Same question for out if the car, in an enclosure or free air? We really need to think more about not what the speaker would preform better in, but what would 95% do in an install with these speakers. I went to the extreme and have huge pods on my dash...but I know most people probably wont install them in that fashion.

1 idea is open baffle outside the car and enclosure inside the car...what do you think about this?
 

·
Moderator
Joined
·
18,499 Posts
^^^I was going to say the exact same thing and he's absolutely correct. Testing as a system is absolutely the wrong way to do this.
x3, then.

When I saw that you were going to test different drivers with the dyns the first thing I thought of was the tonality wreckage that was to occur. ;)


I'd test them as mids, and solely that.

IMO these are a few things you want to do at a minimum, all outside the car:
Free Air Tests:
  • Get measurements in free-air conditions with the woofer tester 3.
    You might want to test them with the drivers laying on their side so you don’t block the vent hole. I believe this can skew your results, but can’t recall 100%. Or, you could simply test it first. Regardless, make sure you test all the drivers in the SAME conditions and position.
  • Test the drivers playing free air, mounted only to a simple baffle.
    Nothing too large so that you won’t skew results from driver to driver. There’s a point at which the baffle size is large enough to block back waves and it’s dependednt on bandwidth. I’ve seen it talked about specifically for open baffle subwoofers in home audio. To keep things apples to apples, build a baffle simply to mount the drivers to. Then play free air. Test the lower limits of the driver playing free air. Note breakup at lower and higher end as you move the crossover points.
Enclosure Tests:
  • With your woofer tester results, model up nominal box sizes. Do this for every driver and save the results.
  • Build the largest enclosure you need to build (whatever driver requires the largest enclosure).
    Now that you have the largest enclosure size and all baffles made, you can simply get the nominal size for each subsequent driver test simply by adding a piece of wood, brick, sand, etc to the enclosure. This will allow you to quickly and easily achieve the nominal enclosure volume and continue right on with your test.
  • Make a separate baffle for each driver. Use t-nuts/screws for mounting so that you can swap baffles quickly without marring the mdf that the baffles are mounted to.
  • Note the low end and the top end. I wouldn’t expect top end to change, but it’s possible the enclosure could cause a roll off sooner than in free-air.
  • Use your woofer tester in each nominal enclosure to do an impedance sweep and spit out the new t/s parameters. This is important.
  • Also, if you want to, you can try the different drivers in different sized enclosures by using your bricks/scrap wood/etc. As long as your box size is not any smaller than the largest required volume, you should have plenty of elbow room to do this testing. As you do this, make sure again to use the woofer tester for impedance plots and t/s parameters.
Might want to consider taking some measurements at different angles, too. I did that with the h-audio souls. Here’s the thread:
http://www.diymobileaudio.com/forum/member-product-reviews/66611-h-audio-soul-midrange-x2-wideband-tweeter-review.html



Make sure to take good notes. You might even want to get a dry-erase board on hand so you can write down what each test is as you’re running the woofer tester sweeps.
I suggest that when you take pictures of the group, simply use a sheet of paper and write the driver on that and put it next to the driver. It’ll save you time when you’re typing up the report. ;)
I recommend doing this anytime you can, for any test you may have, just so you can remember more easily later what the picture is of (ie: picture of x driver in xft^3 enclosure). Assign the tests a number so when you write up your posts and try to search through it, it’ll be easier to find. Ie: Test #4 = Scan 12m in 1ct^3 enclosure. Your heading for listening test would be: #4. Just easier overall to find, because your post(s) are going to be pretty long I have a feeling.

For in car testing, I’d go with whatever makes the most sense for you. Honestly, I’d try putting them in free-air and not worry about an enclosure. You’re going to get the general idea of what the enclosure does when you test it outside the car. Plus, we know that car’s geometry will vary, as will aiming. If you put it in the car and try to do enclosures, etc you’re talking about some SERIOUS headaches. Especially if you start trying to do on/off axis, etc.
For simplicity, I say free-air in the car, maybe on axis and 90* off axis… no in betweens. Just to get a very general idea.

That’s all I can think of right now. Hope that helps.
Good luck!
 

·
Registered
Joined
·
3,498 Posts
Discussion Starter #65
x3, then.

When I saw that you were going to test different drivers with the dyns the first thing I thought of was the tonality wreckage that was to occur. ;)


I'd test them as mids, and solely that.

IMO these are a few things you want to do at a minimum, all outside the car:
Free Air Tests:
  • Get measurements in free-air conditions with the woofer tester 3.
    You might want to test them with the drivers laying on their side so you don’t block the vent hole. I believe this can skew your results, but can’t recall 100%. Or, you could simply test it first. Regardless, make sure you test all the drivers in the SAME conditions and position.
  • Test the drivers playing free air, mounted only to a simple baffle.
    Nothing too large so that you won’t skew results from driver to driver. There’s a point at which the baffle size is large enough to block back waves and it’s dependednt on bandwidth. I’ve seen it talked about specifically for open baffle subwoofers in home audio. To keep things apples to apples, build a baffle simply to mount the drivers to. Then play free air. Test the lower limits of the driver playing free air. Note breakup at lower and higher end as you move the crossover points.
Enclosure Tests:
  • With your woofer tester results, model up nominal box sizes. Do this for every driver and save the results.
  • Build the largest enclosure you need to build (whatever driver requires the largest enclosure).
    Now that you have the largest enclosure size and all baffles made, you can simply get the nominal size for each subsequent driver test simply by adding a piece of wood, brick, sand, etc to the enclosure. This will allow you to quickly and easily achieve the nominal enclosure volume and continue right on with your test.
  • Make a separate baffle for each driver. Use t-nuts/screws for mounting so that you can swap baffles quickly without marring the mdf that the baffles are mounted to.
  • Note the low end and the top end. I wouldn’t expect top end to change, but it’s possible the enclosure could cause a roll off sooner than in free-air.
  • Use your woofer tester in each nominal enclosure to do an impedance sweep and spit out the new t/s parameters. This is important.
  • Also, if you want to, you can try the different drivers in different sized enclosures by using your bricks/scrap wood/etc. As long as your box size is not any smaller than the largest required volume, you should have plenty of elbow room to do this testing. As you do this, make sure again to use the woofer tester for impedance plots and t/s parameters.
Might want to consider taking some measurements at different angles, too. I did that with the h-audio souls. Here’s the thread:
http://www.diymobileaudio.com/forum/member-product-reviews/66611-h-audio-soul-midrange-x2-wideband-tweeter-review.html



Make sure to take good notes. You might even want to get a dry-erase board on hand so you can write down what each test is as you’re running the woofer tester sweeps.
I suggest that when you take pictures of the group, simply use a sheet of paper and write the driver on that and put it next to the driver. It’ll save you time when you’re typing up the report. ;)
I recommend doing this anytime you can, for any test you may have, just so you can remember more easily later what the picture is of (ie: picture of x driver in xft^3 enclosure). Assign the tests a number so when you write up your posts and try to search through it, it’ll be easier to find. Ie: Test #4 = Scan 12m in 1ct^3 enclosure. Your heading for listening test would be: #4. Just easier overall to find, because your post(s) are going to be pretty long I have a feeling.

For in car testing, I’d go with whatever makes the most sense for you. Honestly, I’d try putting them in free-air and not worry about an enclosure. You’re going to get the general idea of what the enclosure does when you test it outside the car. Plus, we know that car’s geometry will vary, as will aiming. If you put it in the car and try to do enclosures, etc you’re talking about some SERIOUS headaches. Especially if you start trying to do on/off axis, etc.
For simplicity, I say free-air in the car, maybe on axis and off axis… no in betweens. Just to get a very general idea.

That’s all I can think of right now. Hope that helps.
Good luck!
I have talked to several of you guys in email and on the phone the last couple of days and it seems like everything is becoming more clear on this whole process, so thanks for all your help!

What is your opinion of in car vs out of car testing? Should I do the sealed enclosure in car?
 

·
Moderator
Joined
·
18,499 Posts
^ check the post... I did a ninja edit. ;)


I'm sorry I didn't get a chance to call you last night.
Plus, could you imagine me trying to tell you all of that? I would’ve confused both of us, and you would have thought I was raving mad! Lol!!!
 

·
Registered
Joined
·
8,967 Posts
I agree with fish, I think they should at least me measured in and out of the car. A perfect reason is the last tweeter test. The L1s didn't sound very good out of the car but in the car they really woke up. More of a "just in case" kind of thing but ya never know.
 

·
Registered
Joined
·
3,498 Posts
Discussion Starter #68
^ check the post... I did a ninja edit. ;)


I'm sorry I didn't get a chance to call you last night.
Plus, could you imagine me trying to tell you all of that? I would’ve confused both of us, and you would have thought I was raving mad! Lol!!!
Crazy Ninja...thanks!
 

·
Registered
Joined
·
998 Posts
You already know my stance on this, but I'll state it here so others know why I'm on the "stand alone" side.

One of the main things we try to do when designing a system is timbre/tonal matching between drivers. There will be several different types of midranges that will be tested. Some will mate tonally well with the Dyn woofers and tweeters, and some others won't. My concern is that the test would naturally become a test on how well the particular midrange sounds with the Dyn speakers.

It would be EXTREMELY hard to not let that affect your judgment.

So, I say test them stand alone and let them be listened to on their own merits.
Not trying to be contrary, but this is honestly something I've never understood :(

What does timbre/tonal matching between drivers playing different frequency ranges really mean? I mean, what are the underlying quantitative principles? Obviously, can't be frequency response. Is it distortion? Even that doesn't make a lot of sense, over different frequency bands.

In the objective camp, where we strive for accuracy in reproduction, I just don't get what quantitative characteristics of the drivers we're trying to "match" :(

How do you "tonally match" drivers that are, by definition, playing separate and different "tones"? Know what I mean?

Again, not trying to be argumentative. Just looking for objective clarity.
 

·
Moderator
Joined
·
18,499 Posts
to me it's rather simple.

You may have one driver that, by design, sounds a bit ‘dry’.
Some drivers sound more warm. Some sound more neutral, etc. I think we can all agree that different drivers often have their own sound characteristics.

The problem I see when doing these kind of tests is that, for example, you have a system that consists of drivers that have a more warm sound to them. You throw in a mid that doesn’t have that warmth and is rather a dry sound and you may not like what you hear relative (this is key) to what the rest of the system sounds like.
Does this mean you might like it by itself? Possibly… There’s the chance for it to happen, for sure.

I’ve personally experienced this, so I’m speaking from my own experience. Some may be better at controlling their brain enough to say, ‘no, what I’m hearing is the contrast between drivers, not a bad sounding driver itself’. I have a hard time doing this.

The problem with this, and this test, is that there’s one guy looking to evaluate a bunch of drivers. If he were looking to do this for himself only, and solely to find what drivers he wants to keep then it makes ABSOLUTE sense to test them with the other drivers he likes. Assuming, of course, that he’s keeping the tweeter and midbass and trying to pick a midrange. In this case I think you’d be silly not to test the midrange with the rest of your system.

But, again, for the purpose of this test, where one person is trying to give a characteristic of individual drivers, I think it’s best if you remove all other potential inhibitors. At least then you’ve narrowed everything down to one driver and have no potential for coloration from other drivers.

This is all my opinion... no special degree from me. ;)


Hope that helps. :/
 

·
Registered
Joined
·
8,967 Posts
Okay, so you guys want me to do open baffle and sealed out of car and which one inside the car, sealed or free-air?
That's hard to say. I now objectively any results will have to be taken with a grain of salt if they're all in the same enclosure and that that will make the results more heavily skewed heavily towards performance in your car. But by the same token, most people running a 3-way front stage are running pods that are built to fit, not necessarily what performs best.
 

·
Moderator
Joined
·
18,499 Posts
But by the same token, most people running a 3-way front stage are running pods that are built to fit, not necessarily what performs best.
Exactly.

Which is why I suggest building the largest enclosure and using stuff to fill it in to change the enclosure size to various sizes and test at those sizes. You can get a general idea of how a driver will perform in the car, for whatever enclosure size, based on how many variations in size are used outside of the car for each driver.

So, with the above quoted, I see no reason to try to do enclosures in the car. It’s more of a PITA for you, and really doesn’t tell me anything more than an enclosure outside of the car + what we already know about cabin gain and how geometry of nearby things effect the driver (standing waves, etc).

Plus, you can’t account for pillar or kick mount… that’s just too much work on you.
Again, we can draw conclusions from your tests outside of the car with various sized enclosures.
One thing you can do to make this REALLY easy is to pick 3 sizes… maybe 1ft^3, 0.5ft^3, & 0.25ft^3. Then measure each driver with the woofer tester in each of these 3 configuarations. Then show the results on top of each other. We can draw a lot of conclusions from the impedance plot and t/s parameters you get from these three. And it’s much more useful to me than you testing them in the car like this. That’s solid information that will not change. Conversely, anything you do in the car is subject to change.
 

·
Registered
Joined
·
998 Posts
to me it's rather simple.

You may have one driver that, by design, sounds a bit ‘dry’.
Some drivers sound more warm. Some sound more neutral, etc. I think we can all agree that different drivers often have their own sound characteristics.

The problem I see when doing these kind of tests is that, for example, you have a system that consists of drivers that have a more warm sound to them. You throw in a mid that doesn’t have that warmth and is rather a dry sound and you may not like what you hear relative (this is key) to what the rest of the system sounds like.
Does this mean you might like it by itself? Possibly… There’s the chance for it to happen, for sure.

I’ve personally experienced this, so I’m speaking from my own experience. Some may be better at controlling their brain enough to say, ‘no, what I’m hearing is the contrast between drivers, not a bad sounding driver itself’. I have a hard time doing this.

The problem with this, and this test, is that there’s one guy looking to evaluate a bunch of drivers. If he were looking to do this for himself only, and solely to find what drivers he wants to keep then it makes ABSOLUTE sense to test them with the other drivers he likes. Assuming, of course, that he’s keeping the tweeter and midbass and trying to pick a midrange. In this case I think you’d be silly not to test the midrange with the rest of your system.

But, again, for the purpose of this test, where one person is trying to give a characteristic of individual drivers, I think it’s best if you remove all other potential inhibitors. At least then you’ve narrowed everything down to one driver and have no potential for coloration from other drivers.

This is all my opinion... no special degree from me. ;)


Hope that helps. :/
doesn't work for me :(

"dry" or "warm" are terms that must be related to frequency response, distortion, energy storage, etc. in order to answer my question : what are the objective characteristics that we're trying to "match" ... for drivers playing different frequency ranges? And when you do express it in these quantitative terms, i always arrive at the same conclusion : it doesn't matter.

I can elaborate : let's say you're trying to match left & right midrange drivers. Obviously, we can strive for similar frequency response and distortion signatures here, because they are playing the same frequencies. I just can't mentally extend that argument to a midrange and midbass on the same channel.

I humbly submit that "matching" drivers playing different frequency ranges, according to "tonality" or "timbre", or subjective adjectives like "warm" or "dry", is one of those audio myths that has no real basis in reality :p

I'm sure this won't be a popular conclusion :( But before anyone jumps in with an anticipated counter-point ... yes, i understand that a saxophone has a very different timbre than a clarinet. Objectively, we can describe this by looking at the harmonic signature of each instrument. HOWEVER ... we must not confuse the Mona Lisa with the glass through which we are observing her ;) I'm not looking for my reproduction system to magically "create" those rich harmonic overtones that define each instrument. I expect my audio system to reproduce those rich harmonic textures, captured in the recording, with accuracy.
 

·
Registered
Joined
·
40 Posts
You already know my stance on this, but I'll state it here so others know why I'm on the "stand alone" side.

One of the main things we try to do when designing a system is timbre/tonal matching between drivers. There will be several different types of midranges that will be tested. Some will mate tonally well with the Dyn woofers and tweeters, and some others won't. My concern is that the test would naturally become a test on how well the particular midrange sounds with the Dyn speakers.

It would be EXTREMELY hard to not let that affect your judgment.

So, I say test them stand alone and let them be listened to on their own merits.
x4:)
 

·
Moderator
Joined
·
18,499 Posts
I understand… the sound characteristics I’m talking about are related to FR and the harmonics. I understand.
I just don’t know how one could show this without a klippel. The only way I know how to evaluate this is by subjectively testing the drivers. RTA could and will show FR, though... hmph...

If this was purely a technical test, then I’d agree 100% with you. But, given the following:
1. Niebur’s going to do listening tests
2. We don’t have no stinkin fancy equipment here, feller. :mad:

I don’t see really see how to avoid any ‘mind coloration’ as I will call it, in conjunction with what I’ve said above as rationale.
 

·
Registered
Joined
·
998 Posts
I understand… the sound characteristics I’m talking about are related to FR and the harmonics. I understand.
I just don’t know how one could show this without a klippel. The only way I know how to evaluate this is by subjectively testing the drivers. RTA could and will show FR, though... hmph...

If this was purely a technical test, then I’d agree 100% with you. But, given the following:
1. Niebur’s going to do listening tests
2. We don’t have no stinkin fancy equipment here, feller. :mad:

I don’t see really see how to avoid any ‘mind coloration’ as I will call it, in conjunction with what I’ve said above as rationale.
Klippel won't help you. It can do a great job of quantifying the large signal performance of drivers, but that won't help the cause of the illogical.

There's just no meaning behind timbre & tonal matching of drivers playing different frequency ranges.

How will I match tonality (frequency response) and timbre (harmonic signature) of an 93 Hz sinewave, with a 347 Hz sinewave? You have ALL the tools imaginable at your disposal ... how will you do it?

Furthermore, i humbly submit that for those who have heard "driver mismatch", what they've really heard is a poorly designed crossover between the drivers ;)

Probably subject for another thread, but it does relate to whether or not a single-brand midbass can be used for a midrange comparison test.
 

·
Moderator
Joined
·
18,499 Posts
Fair enough. I see your point.
I’ll just kindly disagree and let this thread continue on. We could keep it up, but you’ve shown your points and I’ve talked about mine. Really nothing left for either of us to say. At least not in this thread.

:)
 

·
Registered
Joined
·
2,298 Posts
Not trying to be contrary, but this is honestly something I've never understood :(

What does timbre/tonal matching between drivers playing different frequency ranges really mean? I mean, what are the underlying quantitative principles? Obviously, can't be frequency response. Is it distortion? Even that doesn't make a lot of sense, over different frequency bands.

In the objective camp, where we strive for accuracy in reproduction, I just don't get what quantitative characteristics of the drivers we're trying to "match" :(

How do you "tonally match" drivers that are, by definition, playing separate and different "tones"? Know what I mean?

Again, not trying to be argumentative. Just looking for objective clarity.
A well damped speaker like the Dyns are going to sound very different than say a metal coned driver with a thin metal cone. I understand what you're saying about different pass bands, but you will still have summing of the two drivers at crossover points.

It's about avoiding confusing the senses around the top and bottom end of the the midrange pass band. Not a declaration of "you can't run a poly cone and metal cone together".
 
61 - 80 of 418 Posts
Status
Not open for further replies.
Top