|
|
|
|
|
|
BMW Garage | BMW Meets | Register | Today's Posts | Search |
|
BMW 3-Series (E90 E92) Forum
>
Base radio 2010 E90 measurements
|
|
12-24-2010, 11:00 AM | #23 |
New Member
0
Rep 8
Posts
Drives: 2011 328i, Formerly E46 M3
Join Date: Dec 2010
Location: SF East Bay, California
|
It would be curious to know how your setup (no re-coding) actually compared to the same setup, but with the head unit re-coded to HiFi. If you ever come across anyone with a re-coded setup, be sure to borrow there car for some measurements
Either way, a $500 certainly seems like the most economical (by far) of all the upgrade options out there, and considering that it is plug-and-play makes it a no-brainer. |
Appreciate
0
|
12-24-2010, 11:34 AM | #24 | |
Captain
157
Rep 716
Posts |
Quote:
|
|
Appreciate
0
|
12-24-2010, 12:01 PM | #25 | |
Lieutenant General
2295
Rep 13,001
Posts |
Quote:
The THA555 was designed to perform the same job as the Alpine kit but for any car with OEM HU and no OEM amp. |
|
Appreciate
0
|
12-24-2010, 12:03 PM | #26 |
Lieutenant General
2295
Rep 13,001
Posts |
Yes, it can do that. However, the Alpine OEM kit is not for HiFi systems (low level, balanced outputs) so it is unclear if that amp will accept those signals.
|
Appreciate
0
|
12-24-2010, 12:29 PM | #27 |
Captain
157
Rep 716
Posts |
The Alpine amplifier inputs (4 channels) are balanced, not single ended. The input impedance is 47 Ohms. Inside the amplifier, there are jumpers for high and low input levels. The default setting is high level.
|
Appreciate
0
|
12-24-2010, 01:13 PM | #28 |
Lieutenant General
2295
Rep 13,001
Posts |
Good to know, although using this kit for a HiFi signal is really a waste for the money, IMO.
|
Appreciate
0
|
12-24-2010, 02:50 PM | #29 |
New Member
0
Rep 8
Posts
Drives: 2011 328i, Formerly E46 M3
Join Date: Dec 2010
Location: SF East Bay, California
|
I stand corrected.
Technic, for the alpine kit, if not reprogramming to HiFi (low power balanced differential,) then is it just reprogramming to flat eq curve? Are these just settings available in the DCAN bus, or is there more to it than that? |
Appreciate
0
|
12-24-2010, 03:03 PM | #30 | |
Lieutenant General
2295
Rep 13,001
Posts |
Quote:
I think that the programming requirement of this new kit simplifies the install regardless of Business or Professional HU/iDrive and nothing else. |
|
Appreciate
0
|
12-25-2010, 12:04 AM | #31 | |
Brigadier General
289
Rep 3,201
Posts |
If you don't have a load on the passive crossover, the test results will have errors.
If you don't have the speaker load connected to the HU, your test will have errors (unclear whether or not you had such). The rear speakers, as have been published elsewhere, have an active high pass filter in the HU. Quote:
All in all, you seem to be verifying that the HU response is nasty without recoding or without external normalization. |
|
Appreciate
0
|
12-25-2010, 01:50 PM | #32 | |
Captain
157
Rep 716
Posts |
Quote:
The response of the HU was done with a 47 Ohms load (Alpine amp input impedance). BTW, audio amplifiers have low output impedance in order to get high damping factor (http://en.wikipedia.org/wiki/Damping_factor). This means that if you measure the frequency response of an audio amplifier with a load or without a load, there should not be large difference. Recoding to disable the equalization in the HU is a better option. I am looking at doing that. In the interim, I am seeking at getting acceptable results, a flatter response. I know I will not be able to get a flat electrical response with the parameters that I can tune: Bass, Treble and the switch on the Alpine amplifier. In post#20 second figure, I show the acoustical response of the drivers in the door after adjustment of the bass and treble. In the next figure, I show the response from the driver seat at ears level. As I said before, the peak and valley for acoustical measurements in a car are to be expected. You move the microphone by few inches and you get differents results. To get accurate measurements from speakers, the test has to done in a anechoic chamber (http://en.wikipedia.org/wiki/Anechoic_chamber), not a car. I do not have a figure with the settings that result in the best response because I take the results from many measurements, post-process and estimate the best setting after post-processing. I will try to take another acoustical measurement and post the results. |
|
Appreciate
0
|
12-25-2010, 03:55 PM | #33 | |
Brigadier General
289
Rep 3,201
Posts |
Quote:
Your method of switching between electrical and acoustic responses is a tricky one. If you have a target acoustic response, that's cool - but I don't think you do, and with lesser speakers, it's often better to suffer a response dip than to drive a marginal speaker into distortion in a particular range. Much of the poor sound quality you can find in this system is not going to be found with an acoustic response plot. Sure, there is missing bass (which is obviously a difference between the electrical and acoustic responses). But there is also bad-sounding bass and bad-sounding treble. The audio analysis won't tell you that you are listening to distortions of various kinds, it will tell you that there is energy present at those notes, but not its fidelity relative to the input signal. Unless you have an AP System One that you're going to wheel out on us (You could probably use the RightMark software to do some cool analysis). Response plots tell you if the speakers make noise at a given note, and if the interior absorbs or reflects it, at your mike. It doesn't tell you about bad-sounding bass or bad-sounding treble - distortions, whether they be steady-state or occurring at higher speaker excursions. If you've found what you think is a good acoustical response, posting the electrical response that got you that, and the settings, and then the acoustic response, and then your listening results, would be useful. But without the listening results, the response is unfortunately irrelevant. Attempting to correlate acoustic response measured in this manner to listening results, is a tricky business. When trying to upgrade this car's audio performance, you basically have a couple of paths. - Throw spaghetti at the wall, season as you can (i.e., bass and treble), and hope it sounds good. I think this is what you've been doing so far. Even if you have a flat electrical response, this is sort of what we do when we choose speakers, because your final acoustic response is impossible to predict without data from another similar installation. One of the reasons that listening rooms don't work for predicting driver response in-car. - Flatten the electrical preamp signal's response - really flatten it so that it's comparable to an aftermarket HU - and at that point, as above, your sound is dependent primarily on your speakers' performance and your crossover setup. This is what the CleanSweep is intended to do, and what the BitOne and MS-8 do with the input signal. It can be done manually with Audio Control EQs and with Zapco DSP amps or processor. Or it can be done with recoding. Then you're back to handling the difference between electrical and acoustic response once you've transduced all that energy. - Actually go beyond normalization to arrive at your own DSP optimized setup. This is what the BitOne allows you to do manually and the MS-8 does automatically, and what can be done manually with the Audio Control gear and with the Zapco DSP amps or processor. Your speakers are still a limiting factor, because with the DSP available today, you still can't make one speaker's electromagnetic tendencies seem identical to a totally different speaker, with just a few watts to use to push and pull that speaker. By the way, I just noticed that you are using white noise as a test signal. If you want to compare your test results to most, and you aren't adjusting for this in MatLab, I would suggest using pink noise, not white noise. (White noise is equal random energy at every frequency, and occurs with radios when no signal is found. Pink noise is equal energy at every octave, and corresponds to how humans hear). I would suggest using the WAV file on the Rockford Fosgate web site with pink noise. It's the longest track I've found, which makes this sort of testing easier to do, and it's already a WAV file, and it's free. |
|
Appreciate
0
|
12-25-2010, 05:26 PM | #34 |
Captain
157
Rep 716
Posts |
I have access to a Audio Precision 2700 but this is not what I used for the measurements. The AP is a rack type test equipment, not easy to bring to the car.
My goals are to get the electrical response and acoustical response of the HU and speakers. I have not seen this data elsewhere. I have seen various speculation about the purpose of the switch on the Alpine amplifier, I know now what this swich does. For the measurements, I am using white noise for some measurements and chirp "Sweep-Sine" for others (see: http://citeseerx.ist.psu.edu/viewdoc...=rep1&type=pdf). The white noise is flat across frequencies, my post-processing take care this into account. My test track is a 26 second and it is cyclic. This means that it can be loop forever without any discontinuities. With this noise, I make a CD with a track that is 10 min each. The sweep-sine technique has the same 6dB/Oct rolloff and is also cyclic. There are many papers on this technique on the Audio Engineering Society, see also: https://ccrma.stanford.edu/realsimple/imp_meas/ As for the sound quality, I am not saying this solution is much better than other solution. My goal was to improve the sound quality in my car. The result is a much improved sound quality over the base system. |
Appreciate
0
|
12-25-2010, 06:59 PM | #35 | |
Captain
157
Rep 716
Posts |
Quote:
|
|
Appreciate
0
|
12-25-2010, 07:32 PM | #36 |
Captain
157
Rep 716
Posts |
Here is a 2^20 length pink noise sequence PSD (generated using code : https://ccrma.stanford.edu/~jos/sasp...1_F_Noise.html)
|
Appreciate
0
|
12-26-2010, 09:27 PM | #37 |
Captain
157
Rep 716
Posts |
Here is the acoustical frequency response at the driver ears for the left and right channel. The response is measured from 100 to 20 KHz every 1/12 octave (92 data points for each channel).
At lower frequencies, the ripple in left channel response has a longer period than the right channel. This is from the sound reflections in the cabin and the left speakers that are closer to the driver than the right speakers. Note: I also measured the harmonic distortion of the amplifier+speakers. I is very low at the SPL where I made the measurements. |
Appreciate
0
|
12-27-2010, 01:23 AM | #38 |
Brigadier General
289
Rep 3,201
Posts |
Again, in my experience, data like this doesn't help if there isn't any correlation to listening in the car.
Sorry you didn't see where I'd published the electrical response before. I didn't publish the acoustical response because in my experience, it's not very helpful, for the reasons I mentioned before. I've tested that RF test track - don't remember seeing any rising response above 10K. You refer to the harmonic distortion level. Are you suggesting that you have a target threshold for harmonic distortion? Do you have any description of the odd order and even order distribution? Did you look at intermodulation products? Do you agree or disagree with Geddes' position on audible thresholds for harmonic distortion and intermodulation distortion? Don't link papers, I would like to know what you think. You can cite papers on this topic for any position you want to take - what's your position? What microphone are you using for the measurements? Any particular reason you aren't including information below 100 Hz? I'm glad your system sounds better. The base Stereo system sounds so heinously bad that you can improve it in many respects without too much analysis, but if you have access to the means, I totally understand doing measurements. My comments are intended to convey that I can't tell from your posts if you got to what some would call "high fidelity", or you just reduced the suckage. Either is fine, but it would be nice to know. I have had lots of people ask questions about the Alpine system. I've done a little research on it, and I did not see how on earth it could really be normalizing the signal. Your tests indicate that the Alpine system, with the bass and treble set to center, are not acceptable. Given that you obviously understand audio, I'm surprised that you didn't choose to address this problem in any number of other, more comprehensive ways. Again, I'm glad it's better - but from what you're saying - and not saying - it seems that you aren't that far away from getting even better results than you now have. By the way, is there any ground-loop or EMI RPM-related noise? Wondered how quiet that Alpine interface is. I'd turn on high beams and rear window defogger and any seat heaters, to load the alternator. |
Appreciate
0
|
12-27-2010, 10:56 AM | #39 |
Captain
157
Rep 716
Posts |
For the acoustical and some electrical measurements, I am using a chirp sine-sweep technique (100Hz to 20KHz sweep in 2.972 seconds). With this technique you can estimate the impulse response of the system under test. Distortion can also be estimated, they are separated in the time axis from the main impulse response (this is described in one paper that I linked). I can see distortion in some measurements (it is low), but I have try to quantify it more precisely. My software is not written for this at this time.
For the paper from Geddes, I have not seen it. I may take a look. The microphone is a "Bruel & Kjaer 2238 Mediator". It is calibrated every two years. The measurement setup is simple. It is a laptop (with a good sound card), a good microphone, a CD (test track) and the headphone output of a laptop. For the software, I modified the code from https://ccrma.stanford.edu/realsimpl...nt_Theory.html The software run in Matlab (expensive), it should also run in Octave (free). To do the electrical measurements, I made a simple pad to lower the level and a DC block (less than $1 of parts). The microphone that I use is expensive. An altenative can be found for less than $100 at http://www.cross-spectrum.com/ , http://www.parts-express.com/pe/show...-801&FTR=emm-6 Note: You need a phantom power for this mic. If anyone interested, I can share the software. I do not have any noise after the amplifier. I did the installation per instructions. In addition, I added 2 ferrite beads (left an right harness to/from the speakers/HU) to prevent noise. The part that I used is similar to http://www.parts-express.com/pe/show...number=110-450 The measurements are from 100Hz to 20KHz because this is the range that is mostly improve by the Alpine kit. I did not upgrade the woofer. For frequencies below 100Hz, in a car, you feel them more than hear them. I did make some measurement at lower frequencies. It is not flat at all, the woofer (and car?) is a tuned resonnator at 55Hz. This measurement is done with the bass=-2 and the switch on the Alpine amp at #1. I am now running the amplifier with the switch at position #2, the peak at 55Hz should be 4dB lower. To see how distortion show up in the time response, see "Impulse Responses" in Room Equalization Wizard (free software) http://www.hometheatershack.com/room...tml/index.html Last edited by montr; 12-28-2010 at 12:47 AM.. Reason: Adding link to "Room EQ Wizard" |
Appreciate
0
|
11-01-2011, 04:11 PM | #41 |
Registered
0
Rep 2
Posts |
Hi,
I have just had the alpine kit installed by a local car audio specialist. To cut a long story short there was no sound coming from front mid and lots of noise from the tweeter, so I look it back to him. It turned out that he had put the input into the midrange wire, the tweeter into the output and the midrange into the crossover input. The system seems to sound good now that it is correctly wired, but, what damage would this likely to have done to the tweeter and/or crossover? Just to put my mind at rest if anything else Kevin |
Appreciate
0
|
11-01-2011, 06:06 PM | #42 | |
Captain
157
Rep 716
Posts |
Quote:
Conclusion: Do not trust the label that are on the crossover. |
|
Appreciate
0
|
11-01-2011, 06:37 PM | #43 | |
Registered
0
Rep 2
Posts |
Quote:
Cant really justify buying 2 new tweeters just on a gut instinct though. |
|
Appreciate
0
|
11-13-2011, 07:47 PM | #44 | |
New Member
0
Rep 5
Posts |
Quote:
I am intending to upgrade my E90 Standard sound system using a Helix DB Five amp [4x60W at 4Ohm + 1x250W], front midbass from Individual System, DLS T25 tweeters, JBL Crossovers and other subwoofers underseat and Vibe Slick 40 in the back. I will also use a Helix AFC active frequency convertor [high-low] convertor. How should the recoding of the Professional HU be done? I have called BMW Dealership and they didn't know to answer my question regarding coding. Can anyone please tell me what should i ask them for? Eg: car recoded to hifi? Thank you very much! |
|
Appreciate
0
|
Bookmarks |
|
|