Rerurn to Romy the Cat's Site

Horn-Loaded Speakers
Topic: About the tweeters phase alignment.

Page 1 of 1 (4 items)


Posted by Romy the Cat on 09-29-2006

Tweeters Alignment - what is a bitchy subject! I was mildly complaining about it for years and during the last few weeks I complain about my own tweeters - primary because the very same issues of the imperfect alignment. This there is a derivation of the thread: An amplifier for Tweeters and I will start here when I finished in there....

It is well known that HF channels should be time aligned however it is not really know to majority of people to which degree of precision that alignment should be done, what variables are involved, what negative consequences Sound has if the tweeters are unaligned (misaligned? disaligned?... go figure). Also it is not clear for public how not properly aligned tweeters negatively affect listening experiences. I will slowly collect form my mind the bookmarks of my experiences on the subject and will be posting them in this thread.

As few words about the perspective beneficiaries of this thread. If you use 2-way monitors then you might skip reading further. My definition of tweeter is something that starts to operate after a MF driver. It is irrelevant what cut off your MF driver has but you should have a fully functional MF channel. In my world of compression drivers, when the drivers run up to 8kHz-13kHz the tweeter are something the come, AFTER the MF drivers. If your tweeter kicks in at 3 kHz-4kHz then skip reading further – you have no MF driver and this thread is not for you.  My experience indicates that ~90% of people out there among those who used stand alone tweeters have their tweeters unaligned. I practically never head any HF capable horn installation with properly sounding HF.  As I said: people who use box speakers has some baffle reference but the horn folks have mostly no geometrical reference points as their damn horns are usually very big and there is no space for tweeters reference baffle.

I am sure you heard a lot of horn installation and you have noticed that most of the compression driver, although in their papers they go up to 20.000Hz but in realty they all die much sooner. Some people do not extend their MF compression drivers (if their specific MF does have good HF and the room allows it) and listen then systems without complimentary tweeters. I do it sometime as well but only in the case of the very specific recordings. Generally people and I agree with them do use some kind of tweeter that kick after the regular MF compression driver. So, if you did hear a lot of horn installations then you know that MF and tweeters combined very really sound “properly”.

It would be nice to blame tweeters phase misalignment for everything but it would not be correct: there are a lot of reasons why tweeters and MF did not sound like one entity. However, time misalignment does play a huge role in there. I know that most of the people out there do not play serious music that requires a perfect time alignment. It mostly happens because of ignorance, not only musicals but audio ignorance. Let me to assure you that when one would learn how to detect in reproduced sound the evidences of tweeters misalignment then it would be much more difficult for this person to listen a tweeters misaligned playback.

It would not be difficult to describe a contribution of misaligned tweeters but I would like do not do it as explicit “education” but rather offer you’re a change to rediscover the subject for yourself.  Let start for instance from a suggestion that the well know to anybody artificial effects of reproduced sound: narrow sounding sibilants, fast sounding shibilants, loud sounding fricatives are not due to the “wind form singer’s mouth blow into the microphone grill”. In 99% of all occasions those artificial effects because the phase anomalies and the HF transducers are primary guilty parts. Even more – even if the recording made badle and the recording has “taught sounding “TH” and “SS” then a “properly aligned” tweeter might greatly “fix” the situation.

The next post I will outline some techniques how you might educate yourself about the subject of tweeters phase misalignment and what methods you might use to resolve the problems

Rgs,
Romy the Cat

Posted by Romy the Cat on 10-02-2006

Ok, none it is time to dive into the practical process of tweeters alignment. Once again I presume that we are taking about monopole tweeters that use AFTER your midranges compression drivers (read my previous post). Some people use term “supertweeters” but I tend do not use that name and that concept.

I have my reasons do believe in the ability to align tweeters using methods of objective control, at least I never was successful in doing it.  You might rune a high resolution RTA and move tweeters until you get max (if they are in-phase) amplitude at crossover point. You might run 2 microphones into dual trace oscilloscope (or one trace + memory) and perfectly synchronies two sinusoids. You might run impulse response synchronization and have your tweeters perfectly aligned. However all those methods of alignment contradict the auditable experience that I usually get from tweeters and therefore I discard the methods of objective tweeter alignment. Furthermore the ideal setting for the tweeters that I get from objective methods is always deterrent then the one that I get from the methods of subjective alignment. So, my methodology of tweeters alignment is a sequence of subjective actions, shaped into a methodological pattern- or something that I call “the real objective methods”.

I few word before I go further. Over the course of the 6 years that I’m working on my Macondo I had approximately two dozen of tweeters and in one way or another I my further writing is a collective experience hot to make my tweeters to sound “right”. One exception though. I never tried to stabilize the tweeter’s impedance and to write the “phase neutral crossover” against it. All my experiments were around the regular 6, 12, 18, 24 db per octave, none series, crossovers. I do not know if with subtraction crossovers satiation would be different and I have no experience with “phase neutral crossover” (in fact for HF I believe that they are bogus). Also, and it is very important to understand: that the method below juts synchronization of the HF and MF alignment and it is not responsible for the tonal attributes of HF and MF drivers.

So, below is my “Real-Objective Tweeters Alignment Survival Guide”

1) Mount your tweeter in the position where it’s diaphragm would be roughly at the same vertical and horizontal plane + 2 inches back relative to the diaphragm of your MF driver. Use juts one right of left channels. All further alignment should be done for each channel separately.

2) Regardless the time or order of the crossover you use set your tweeter in phase with your MF driver. There is a LOT of reasons or school of thought suggesting to invert the tweeter with second order as it has 180 degree phase shift and there are many other reasons. Nevertheless, do not invert tweeters under any circumstances. Tweeter and MF driver much be in-phase. (Be advised that mic of your phrase tester might not pick HF well, use alternative methods to synchronize the polarity of your MF and HF)

3) Set up a correct listening volume of your tweeter relative to MF driver. It should not be exact but it should be from exact to +2db. It is preferably to have your tweeter at this point slightly “hotter” then necessary. It is important do not set the tweeter too soft at this stage. A general perception “a little too bright” but not “aggressively bright” works well at moment.

4) Find a good quality, fill range recording of an opera with strong soprano and tenor. The opera must be recorded during 1950s, preferably before 1955 (less HF screwing global feedback was used at that time)

5) Play in a loop (do not convert it in WAV file) the soprano/tenor recording and pay attention that the highest notes of the singers have a certain “clippings” at “TH” and “S” sounds. Monitor it from the listening position.

6) Begin very gradually to move your tweeter forward, approximately 1/8” per move. At certain point the “clippings” will be minimized or disappeared. Make those locations. Continue to move the tweeter forwards at the distaste of the MF’s diaphragms, minus 2 inches.

7) Now you covered with your tweeter 4 inches and you have 2-3 spots what you feel that the “clippings” were minimized.

8) Set your tweeters in each location and find a better among those marked locations. Still do not loose the markings of the other marked locations – you will return to them.

9) Set your tweeter in the location-winner and while you listen ask someone to very gently angle the tweeter to 3-5 degrees to different directions.

10) If a very minor deviation of your tweeter form being aligned to it’s axe produce a very slight improvement in reproduction of fast sibilants or shibilants then keep moving your drivers. Your goal is to find a location where the driver will not demonstrate any improvements being angle and where it will be sitting perfectly on the axe of the MF driver.

11) After you feel that you found the correct location and the shibilants are the best you can get try further move of your driver but now the increments of the movement should be 1/32 inch or even less. At this point you should pay attention to the raise of dynamic in the singers singing and the decays of the singer voice into the background nose. The raise of dynamic (usually with frequencies doing up in sopranos) should be without any micro events (unless it justified musically) and it should be very gradual and constant. If you have little “thorns” in the dynamic raise then your have timing anomalies and be further movement of the tweeters you should be able to resolve it. The same is with voices decay to “dark”. It is very rare when voice can go to “dark” smoothly and usually at certain point you will feel a “step”. However, the softer this step would be the better time alignment of your tweeter is.

12) Now you are within 1/8”- 1/16” from the correct alignment. Take a large chorus peace recorded not “ambianic” but with more or less up-closed microphones. Play it at different volume levels. Pay attention that sometime some singers “jump” out from the chorus by having that “clippings”   Move your tweeter more in order to minimize the previously mentions effect of sibilant consonant “clipping”.  When the choirs hits the aerodynamicy defended letters/sound they should sound as not problematic as strident vowels. At this point the adjustment of the tweeter positioning will be very-very minor.

13) After you reach the point of perfection in the paragraphs above then begin to walk across your room listening the same things that you were listening in the paragraph #12. If your were able to detect with your “side hearing” any HF artifact of chorus reproduction when you are at 30%, 60% or 90% from your tweeter axes then you most likely have chosen a wring marked point at the step#10. Return to different marked location and start everything again.

14) After you succeed. Set the tweeter volume back to the normal level, or preferably .5dB less.

15) Listen an act or two of the opera and confirmed that your have no HF artifacts.

16) Do again the paragraph # 15 at the day with different electricity. Do not do the specific intentional listening … do something else and let the music just to play on background loud. If you pick with your “side hearing” any HF anomalies then … you have them and then your should look at them further.

17) Listen the large string groups, preferably of the orchestras that have bold and aggressive string sections. Fine adjust the volume of the tweeters…

18) Make the final location of the tweeter very precisely and do it with another channels of you’re using 2 channels Stereo. If you use 5 channels then do not worry about the tweeter alignment but kill yourself – you deserve it.

You should be all set from here. The tweeter alignment might be a pain in ass ceremony but it is a mandatory if you use tweeters. If your tweeters are crossed over 12K then you should not hear any changes ion sound when you turn you tweeters on an off. Of you by turning on the tweeters have auditable “tonal event” then only two conclusions are possible:

A) The amplitude of your tweeters set too high.
B) Your tweeters are garbage in relation to your MF channels and it is better do not use them

Rgs,
Romy the Cat

Posted by stuck.wilson on 10-02-2006
now THAT is the kind of instruction in critical listening that gives you some real guideposts for making your system sound right!  romy, i believe you have honestly set a new standard...

 i have a 2 way system, so i can't use the exact instruction-- but the PROCESS you outlined, and the fact that it's clearly delineated and specific goal-oriented.. that in itself is the kind of thing that could cure a lot of ills in the audio world!  if more folks focused on that sort of information sharing-- there'd be a lot less bad sounding systems-- especially if coupled with the distribution of symphony tickets...

it's funny that this post corresponds exactly to my own search for good, real information on time alignment-- even with a simple 2 way system-- guess the stars.. ahem.. 'align'.. sometimes...

thanks!

d.

Posted by Romy the Cat on 10-07-2006

A German company Avatar Acoustics makes loudspeakers with adjustable positioning of their ribbon-tweeter. This is very much with the context of this thread although I’m questioning a lot of things in those loudspeakers (pretty much everything beside the DESIRE to have adjustable tweeters). The Avatar Acoustics suggest that they did it for “time Alignment for the listener's position”. Since they clamed that they implemented the phase-linear crossover their desire to fine-tune the tweeter’s distance looks like a noble task, at least at surface. However, if to look deeper then for the given design it is PERFECTLY POSSIBLE to set the correct location of the tweeter and do not move it. The “perfect” poisoning of the tweeter might very slightly fluctuates depends of the elevation of the listener’s position but in real world the delta of the position would be within a fraction of an inch but not within a whole 12 inches as the Avatar Acoustics suggests (even considering the low crossover point). Unless the Avatar people propose to move the tweeter console in the real time when a person is walking around the room?!

Anyhow, I do not feel that the time adjustability was is really necessary in the GIVEN design (12 inches!!!). It is like a tonearm with adjustable azimuth of the tonearm wand: completely unnecessary and crates more problems then helps (unless you are wiling to play defective cartridges).

The caT

Page 1 of 1 (4 items)