Upgradin' door speakers... (or: how to fix the crappy Bose system)
#16
Jon8,
It is refreshing to read a knowledgeable and detailed assessment of the Bose acoustic characteristics in our P cars.
Your analysis makes perfect sense and aligns with what I suspected was going on acoustically in the cabin; in my prior life I was an audio engineer at Bell Labs, so I follow your analysis and agree with your conclusions. I am glad to see someone talking about, and explaining, decibels, band pass filters, enclosures, transducers, etc.
I look forward to reading about the solution you choose.
Rep points for you!
Cheers.
It is refreshing to read a knowledgeable and detailed assessment of the Bose acoustic characteristics in our P cars.
Your analysis makes perfect sense and aligns with what I suspected was going on acoustically in the cabin; in my prior life I was an audio engineer at Bell Labs, so I follow your analysis and agree with your conclusions. I am glad to see someone talking about, and explaining, decibels, band pass filters, enclosures, transducers, etc.
I look forward to reading about the solution you choose.
Rep points for you!
Cheers.
#17
Me too
As an audio enthusiast, I too am looking for a reasonably priced upgrade to the audio quality in my 997.2 cab. Thanks for taking the measurements that back up what we are all not hearing. Looking forward to the progression of this thread!
tim
tim
#21
A true but sad commentary on Bose.
Reus audio (www.reusaudio.com) keeps the head unit but replaces everything else. It appears they do extensive sound testing for each application. They are not cheap. But you're still stuck with the less the state of the art head/nav/bluetooth/ipod interface unit.
Reus audio (www.reusaudio.com) keeps the head unit but replaces everything else. It appears they do extensive sound testing for each application. They are not cheap. But you're still stuck with the less the state of the art head/nav/bluetooth/ipod interface unit.
#24
Ok i've spent an hours in my garage doing sound tests on the car tonight.
I used pro equipment: industrial 1/3 octave RTA with 2 different calibrated mic (one for SPL and the other for frequency response). I could have use higher resolution RTA (such 1/24 oct) but the problems are big enough, no need for a microscope, trust me...
Quick diagnosis: Ugly.
I'll be honest: i wasnt expecting the kind of flat frequency response i have in my home. Neither i want a concert hall in my P car. All i need is something decent. I'll let the car audio aficionados out there put 5k or 10k or gears in that (noisy) car...
Personally i would rather listen to the flat 6 and aftermarket exhaust sound, for most of the time i'm enjoying this car.
Being that said, i still want to listen music (weekdays traffic, highway long trip, etc..)
Ok here's the good news: There is plenty of power (for OE system). I've hit 98 db on a pink noise (no bass peak here) without pushing too far the ****... I'm pretty sure it's possible to work something out with the actual power amp.
That's pretty much it for the good news.
Subwoofer: i don't know much about it, but it looks like a tiny over-exploited 6-inches transducer with bandpass (4th or 6th order) config enclosure.
Oups sorry guys, i'll translate: it's like a 3-cylinder motor with a small turbo pushed with the max Psi possible..
Subwoofer for dummies: a bandpass enclosure is always restrained to certain frequencies. In that case, i guessed 40-50hz and i was right, i measured an hefty peak at 40-50hz. Mostly shifted to 50hz. And that makes sense: Bose (and Porsche) probably needed to impress the customers and 40-50hz is very good frequency range to do so. It sounds ''powerful'' while not space (enclosure requirements) and power consuming like 20-35hz subwoofer can be, but not ''thin sounding'' like a peaky 60-70hz sub could be.
So, in a marketing point of view, that was smart.
Problem is (regarding overall bass response):
While the peak in the 40-50hz range is huge (and unavoidable -read below-) there is a major lack of midbass (low-midrange) especially in the ''Barry White & snappy kick drum'' frequencies 125-315hz, which makes the contrast horrible.
We're not talking about snobish audiophile tiny flaws here, we're talking about a +/- differencial of more than 20-30db (every 3db step is double when measured and every 10db is double to the human ear perception).
Ok what all that means ?
It means the speaker's overall design/system is basically defective. The conception is defective.
But, hey!, we have Treble/Bass adjustement, right ? We will be good.
Wrong. The peaky 40-50hz STAYS even at -9 bass correction while making the midbass even worst. You want more 100-300hz ? The bass correction barely helps and makes the peaky sub sounds horrible.
I've measured the system FLAT (0 bass 0 treble) and without surround CDS. besides the subwoofer peak and midbass dip problem, the highs seriously falls after 8khz (a good sound system should cover at least 18-20khz). The treble correction helps a little, while giving a boost in the 8khz-15khz range. But the sonic caracteristics of the med-high transducers are obviously not comfortable with too much correction (harsh sound) and it makes the lack of midbass even more obvious (and we absolutely don't want that).
I'm sorry, i'm not done with the bad news.
To get things even more complicated, the frequencies response balance change... with the volume.
That's normal with active amplified speaker system (and even on passive) but that much ? It's unacceptable, to say the least. I don't know how many amp's channels there is, but it's badly balanced. The overall speaker's sensitivity/outputs are mismatched (but it makes me think the system was built with the convertibles in mind -a lot of bass...tuned for high volume-)
What that change in your life ? That means if you listen it very loud, the peaky sub won't follow the now-very-loud-mid-highs. And if you listen at low or moderate levels, the sub will be too loud or the mid-highs won't be loud enough (depends of your point of view..)
Problem is: it makes the Treble/Bass adjustements very difficult, and only good for certain SPL (sound pressure level).
Next week guys, we will add another complication to all that mess: we will measure the in-cabin noise while running the car! I'm already expecting a very conflictual low-mid noise pollution against an already anemic low-mid sound reproduction system!
I'm motivated to find a solution, so i'll work hard on that in the next weeks...
I used pro equipment: industrial 1/3 octave RTA with 2 different calibrated mic (one for SPL and the other for frequency response). I could have use higher resolution RTA (such 1/24 oct) but the problems are big enough, no need for a microscope, trust me...
Quick diagnosis: Ugly.
I'll be honest: i wasnt expecting the kind of flat frequency response i have in my home. Neither i want a concert hall in my P car. All i need is something decent. I'll let the car audio aficionados out there put 5k or 10k or gears in that (noisy) car...
Personally i would rather listen to the flat 6 and aftermarket exhaust sound, for most of the time i'm enjoying this car.
Being that said, i still want to listen music (weekdays traffic, highway long trip, etc..)
Ok here's the good news: There is plenty of power (for OE system). I've hit 98 db on a pink noise (no bass peak here) without pushing too far the ****... I'm pretty sure it's possible to work something out with the actual power amp.
That's pretty much it for the good news.
Subwoofer: i don't know much about it, but it looks like a tiny over-exploited 6-inches transducer with bandpass (4th or 6th order) config enclosure.
Oups sorry guys, i'll translate: it's like a 3-cylinder motor with a small turbo pushed with the max Psi possible..
Subwoofer for dummies: a bandpass enclosure is always restrained to certain frequencies. In that case, i guessed 40-50hz and i was right, i measured an hefty peak at 40-50hz. Mostly shifted to 50hz. And that makes sense: Bose (and Porsche) probably needed to impress the customers and 40-50hz is very good frequency range to do so. It sounds ''powerful'' while not space (enclosure requirements) and power consuming like 20-35hz subwoofer can be, but not ''thin sounding'' like a peaky 60-70hz sub could be.
So, in a marketing point of view, that was smart.
Problem is (regarding overall bass response):
While the peak in the 40-50hz range is huge (and unavoidable -read below-) there is a major lack of midbass (low-midrange) especially in the ''Barry White & snappy kick drum'' frequencies 125-315hz, which makes the contrast horrible.
We're not talking about snobish audiophile tiny flaws here, we're talking about a +/- differencial of more than 20-30db (every 3db step is double when measured and every 10db is double to the human ear perception).
Ok what all that means ?
It means the speaker's overall design/system is basically defective. The conception is defective.
But, hey!, we have Treble/Bass adjustement, right ? We will be good.
Wrong. The peaky 40-50hz STAYS even at -9 bass correction while making the midbass even worst. You want more 100-300hz ? The bass correction barely helps and makes the peaky sub sounds horrible.
I've measured the system FLAT (0 bass 0 treble) and without surround CDS. besides the subwoofer peak and midbass dip problem, the highs seriously falls after 8khz (a good sound system should cover at least 18-20khz). The treble correction helps a little, while giving a boost in the 8khz-15khz range. But the sonic caracteristics of the med-high transducers are obviously not comfortable with too much correction (harsh sound) and it makes the lack of midbass even more obvious (and we absolutely don't want that).
I'm sorry, i'm not done with the bad news.
To get things even more complicated, the frequencies response balance change... with the volume.
That's normal with active amplified speaker system (and even on passive) but that much ? It's unacceptable, to say the least. I don't know how many amp's channels there is, but it's badly balanced. The overall speaker's sensitivity/outputs are mismatched (but it makes me think the system was built with the convertibles in mind -a lot of bass...tuned for high volume-)
What that change in your life ? That means if you listen it very loud, the peaky sub won't follow the now-very-loud-mid-highs. And if you listen at low or moderate levels, the sub will be too loud or the mid-highs won't be loud enough (depends of your point of view..)
Problem is: it makes the Treble/Bass adjustements very difficult, and only good for certain SPL (sound pressure level).
Next week guys, we will add another complication to all that mess: we will measure the in-cabin noise while running the car! I'm already expecting a very conflictual low-mid noise pollution against an already anemic low-mid sound reproduction system!
I'm motivated to find a solution, so i'll work hard on that in the next weeks...
Every piece of your audio system plays a part in its overall sound. How do you know that the head unit is putting out a flat audio image/signal across the range? You dont. Maybe the speakers are perfect and the signal going to them is whats flawed...
The PCM is crap, plain and simple. The output signal it generates is crap. All you do to change the amps, speakers, x overs, whatever will not change the fact that signal being output is crap...
Anyway, good luck with your testing. I will say it again. Its a small and very noisy vehicle-interior. Whatever you do will not change that. Pump many $1,000.00 in it with upgrades and you will still have a noisy and small interior. You will never have a proper sound stage, you will never have a chance for the sound waves to do what they need to do before they hit your ears.
If you plan on sitting in your garage with the engine turned off and you sit between the 2 seats, you may achieve some level of decent sound....beyond that, give up. Its a total waste of time and energy. The car was not meant for audiophiles...
Good luck
Steve ( a former sound and recording engineer).
#25
Question is which processor to choose. With speakers I am pretty much set for Focal, but I still have no decision on how to build audio tract. Will probably wait until somebody else will find best solution as to copy is simpler than to try and fail by yourself.
#26
Motus labs
hey guys, check out the motus labs agw+. This thing ties into the most network and basically bypasses everything except the controls of the factory head unit. It allows new amplifier connection and also most importantly true sound processing with a laptop......this is what i am going to use to run dynaudio esotars in mine....
#27
All your testing is great and wonderful....BUT, your tests are completely flawed if you are planning on keeping the head unit.
Every piece of your audio system plays a part in its overall sound. How do you know that the head unit is putting out a flat audio image/signal across the range? You dont. Maybe the speakers are perfect and the signal going to them is whats flawed...
The PCM is crap, plain and simple. The output signal it generates is crap. All you do to change the amps, speakers, x overs, whatever will not change the fact that signal being output is crap...
Anyway, good luck with your testing. I will say it again. Its a small and very noisy vehicle-interior. Whatever you do will not change that. Pump many $1,000.00 in it with upgrades and you will still have a noisy and small interior. You will never have a proper sound stage, you will never have a chance for the sound waves to do what they need to do before they hit your ears.
If you plan on sitting in your garage with the engine turned off and you sit between the 2 seats, you may achieve some level of decent sound....beyond that, give up. Its a total waste of time and energy. The car was not meant for audiophiles...
Good luck
Steve ( a former sound and recording engineer).
Every piece of your audio system plays a part in its overall sound. How do you know that the head unit is putting out a flat audio image/signal across the range? You dont. Maybe the speakers are perfect and the signal going to them is whats flawed...
The PCM is crap, plain and simple. The output signal it generates is crap. All you do to change the amps, speakers, x overs, whatever will not change the fact that signal being output is crap...
Anyway, good luck with your testing. I will say it again. Its a small and very noisy vehicle-interior. Whatever you do will not change that. Pump many $1,000.00 in it with upgrades and you will still have a noisy and small interior. You will never have a proper sound stage, you will never have a chance for the sound waves to do what they need to do before they hit your ears.
If you plan on sitting in your garage with the engine turned off and you sit between the 2 seats, you may achieve some level of decent sound....beyond that, give up. Its a total waste of time and energy. The car was not meant for audiophiles...
Good luck
Steve ( a former sound and recording engineer).
It is not quiet in my cabin!
#28
I 2 will follow this post. My CD got jammed in the factory PCM and they special ordered me a new one which has just come in. I am thinking about selling the brand new unit, and using the money to invest in the Kenwood or Pioneer 2 DIN Nav units. Previously, I heard that just changing the head unit would improve the sound. So that will be my first step and if that doesn't work I think I will go the interative steps of adding amp, and then speakers in that order.
Last edited by anthinino; 04-10-2010 at 09:11 PM.
#29
"How do you know that the head unit is putting out a flat audio image/signal across the range? You dont. Maybe the speakers are perfect and the signal going to them is whats flawed...
The PCM is crap, plain and simple. The output signal it generates is crap. All you do to change the amps, speakers, x overs, whatever will not change the fact that signal being output is crap..."
Socalsteve...I follow you on your first statement. The sound that was measured includes all the issues from head unit though transmission, through amplification through speakers, etc. This makes perfect sense to me.
You then make a bit of a leap and say the PCM output is crap. What are you basing that on? Have analysed the output? Have you changed out the unit and seen improvement? I am looking to understand what your basis is.
From my experience with Car audio (medium to high level of experience including about a dozen of fairly compicated installs), when looking at two systems that have CD head-unit/separate amplification/speakers (all factory)...the weakest link is invariably the speakers. Next weakest is the amplification, then I would be concered about the head unit. It still may be a good idea to change them all out...but to save dollars or avoid changing look of car, I would consider working my way from speaker to amps to head unit.
Factory speakers are typically very low quality. Paper cone/surround with very limited throw, thin stamped cage, small magnet/voice coil. Bose is especially notorious for poor quality drivers.
Anyhow...I don't doubt at all that the PCM output is poor, just looking for what evidence exists...
The PCM is crap, plain and simple. The output signal it generates is crap. All you do to change the amps, speakers, x overs, whatever will not change the fact that signal being output is crap..."
Socalsteve...I follow you on your first statement. The sound that was measured includes all the issues from head unit though transmission, through amplification through speakers, etc. This makes perfect sense to me.
You then make a bit of a leap and say the PCM output is crap. What are you basing that on? Have analysed the output? Have you changed out the unit and seen improvement? I am looking to understand what your basis is.
From my experience with Car audio (medium to high level of experience including about a dozen of fairly compicated installs), when looking at two systems that have CD head-unit/separate amplification/speakers (all factory)...the weakest link is invariably the speakers. Next weakest is the amplification, then I would be concered about the head unit. It still may be a good idea to change them all out...but to save dollars or avoid changing look of car, I would consider working my way from speaker to amps to head unit.
Factory speakers are typically very low quality. Paper cone/surround with very limited throw, thin stamped cage, small magnet/voice coil. Bose is especially notorious for poor quality drivers.
Anyhow...I don't doubt at all that the PCM output is poor, just looking for what evidence exists...
#30
"How do you know that the head unit is putting out a flat audio image/signal across the range? You dont. Maybe the speakers are perfect and the signal going to them is whats flawed...
The PCM is crap, plain and simple. The output signal it generates is crap. All you do to change the amps, speakers, x overs, whatever will not change the fact that signal being output is crap..."
Socalsteve...I follow you on your first statement. The sound that was measured includes all the issues from head unit though transmission, through amplification through speakers, etc. This makes perfect sense to me.
You then make a bit of a leap and say the PCM output is crap. What are you basing that on? Have analysed the output? Have you changed out the unit and seen improvement? I am looking to understand what your basis is.
From my experience with Car audio (medium to high level of experience including about a dozen of fairly compicated installs), when looking at two systems that have CD head-unit/separate amplification/speakers (all factory)...the weakest link is invariably the speakers. Next weakest is the amplification, then I would be concered about the head unit. It still may be a good idea to change them all out...but to save dollars or avoid changing look of car, I would consider working my way from speaker to amps to head unit.
Factory speakers are typically very low quality. Paper cone/surround with very limited throw, thin stamped cage, small magnet/voice coil. Bose is especially notorious for poor quality drivers.
Anyhow...I don't doubt at all that the PCM output is poor, just looking for what evidence exists...
The PCM is crap, plain and simple. The output signal it generates is crap. All you do to change the amps, speakers, x overs, whatever will not change the fact that signal being output is crap..."
Socalsteve...I follow you on your first statement. The sound that was measured includes all the issues from head unit though transmission, through amplification through speakers, etc. This makes perfect sense to me.
You then make a bit of a leap and say the PCM output is crap. What are you basing that on? Have analysed the output? Have you changed out the unit and seen improvement? I am looking to understand what your basis is.
From my experience with Car audio (medium to high level of experience including about a dozen of fairly compicated installs), when looking at two systems that have CD head-unit/separate amplification/speakers (all factory)...the weakest link is invariably the speakers. Next weakest is the amplification, then I would be concered about the head unit. It still may be a good idea to change them all out...but to save dollars or avoid changing look of car, I would consider working my way from speaker to amps to head unit.
Factory speakers are typically very low quality. Paper cone/surround with very limited throw, thin stamped cage, small magnet/voice coil. Bose is especially notorious for poor quality drivers.
Anyhow...I don't doubt at all that the PCM output is poor, just looking for what evidence exists...
I was PLEASANTLY surprised at how much better sounding the system was after bypassing the PCM and Bose main amp. Much crisper, clearer and all around better sound quality. Its not audiophile by any means, but for a very loud sports car, its good enough. The PCM-Bose amp-Bose speakers (even with a high quality CD) sounded like crap. I mean, really bad. The modification improved the sounds by leaps and bounds.
So, yeah, I have tested this out in the real world, with my real ears. Ears that were trained to pick up nuances in sound, music and sound quality. Again, its not even close to audiophile, but its very acceptable given the circumstances of how loud and how small the interior of the car is.
You can test until the cows come home and throw as much pink noise, white noise (its all still just noise) as you want into it. Until you listen to a Steely Dan album (or whatever else you are into) and hear the sound quality with real music, its all just noise and tests.
Just sayin'
Steve
Last edited by socalsteve; 04-10-2010 at 11:57 PM.