About the importance of Phase adjustment !!!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello fullranger addicts !!!


I've just finished a fullrange system set-up (my home theater front satellites), and I've been amazed by a little thing called "phase adjustment".

I used the THX optimizer found in a movie (eg:Star Wars) and launched the different optimization programs.

The phase adjustment sends a mono tone (seems to be some kind of pink noise) on two opposite speakers, in my case, the front sats.

Out-of-phase, the sound is difficult to locate, and you separately hear the two speakers.

In phase, the tone should be heard precisely centered between the speakers, like if coming from a physical center speaker.


It is the most obvious adjustment I've ever heard ! My speakers where definetely out-of-phase !

The solution to this, according to the THX optimizer, was to reverse the speakers connexions. I did it.

I stopped the movie and launched a CD.

WOWww ! :bigeyes:

Immediately, the music was brought to another dimension. The singer's voice was well-located at the center of the soundstage, and the stereo separation was more obvious. More than this, the highs got immediately sharper, the percussions were light, aerial, brilliant, fast, and runned from left to right with ease.

I litterally rediscovered my speakers ! Before, I found them flat, hollow and I thought they had not a good imaging for fullrangers. I just found them good enough for the price.




Surprised by this, I asked myself : why do I have to reverse the speaker's connexions ? it must mean the speakers arent' synchronized due to their position !

I moved one speaker slowly with the THX movie running, until the tone sounded correctly with the connexion plugged normally. It was just about centimeters !

Then I asked myself : if it is a phase problem, it is something progressive, and must be set precisely.

I found the pink noise tone was not accurate enough for this smooth adjustment.

I launched again my CD, and adjusted the speaker's position with patience. I used the singer's voice as reference.

Half an hour after, I got the perfect position. Now, I could hear everything with a positionning precision I never heard before. I could hear until the singer taking her breath behind the instruments, until her saliva between her teeth (I'm not laughing lol !!), and I really believed I had a physical center speaker.

The difference between two speaker positions was audible +/- 5mm !!

It was really worth a try, my speakers are now transfigured !

I can only recommend you doing this yourself, it is very well spent time.


And now, I understand the superiority of a fullrange driver on every other concept : one source, one wave, no phase decay possible.

And I now know what imaging means !
 
Stereo speakers that are out of phase which each other sound like ca-ca. And have no low end.

I have an old test record that has a phase test, makes the difference quite obvious.

I don't know about absolute phase. Apparently different CDs even have different absolute phase directions!
 
Yes, some CDs have a reversed absolute phase, but in this case, both channels are reversed, and they are still synchronized.

The CD I use is a compilation of musics and test sounds I burnt myself; with samples of the hardest music material I listen to.

I think there are not so many people who know about phase adjustment, and they would really appreciate the difference I think.

Concerning the out-of-phase sound, I rather found the speakers had no high end, though the low-end suffered also. At the beginning I thought it was a problem with the brand new DVD player (a Pioneer DV585, very good BTW) !

That's why I posted this :smash:
 
youyoung21147 said:
Yes, some CDs have a reversed absolute phase, but in this case, both channels are reversed, and they are still synchronized.

The CD I use is a compilation of musics and test sounds I burnt myself; with samples of the hardest music material I listen to.

I think there are not so many people who know about phase adjustment, and they would really appreciate the difference I think.

Concerning the out-of-phase sound, I rather found the speakers had no high end, though the low-end suffered also. At the beginning I thought it was a problem with the brand new DVD player (a Pioneer DV585, very good BTW) !

That's why I posted this :smash:

It sounds like you've been baptized. :D
When you get to the point where you hear the difference between the absolute phase, you've reached new highs.:hot:
 
youyoung21147 said:
Yes, very interesting article.

Maybe one day I'll try reversing my speakers :D

But I'm a little sceptical about their metal orientation in cables, as far as sound is AC signal.


I'd be very interested to have a phase adjuster on my amp/preamp. The best thing would be a progressive adjustment for each channel, like on the best sub modules :smash:

Sound is AC, but not steady AC, nor symetric AC. I think one would definitely tell the difference between a suck on the ear or blow in the ear.
:yes:
 
Sure it is possible to feel the difference between a peak and a dip, but it becomes harder when you have to care about several pressures at the same time, like in music.

I have maybe never heard an "absolute phased" recording, and sometimes the listening room itself can reverse the pressure at a special point, which could be the listener !

One day I have time, i will surely test this, I find this to be a quite subtle issue in music reproduction ! It's moreover quite easy to test : reverse the speakers wires. I think the " tone in - tone out" swith on my amplifier can do the trick also.

A good thing to test during my Sunday :smash:
 
A friend of mine came to hear my system this afternoon, and I didn't forget to test absolute phase with him !

So I have two conclusions to draw :

- Yes, absolute phase is audible and is even obvious on some music material

- the absolute phase is sometimes reversed on some CDs


More details about the test protocole :

we tried different CDs with music material that I thought having obvious "phase testers", such as piano and bass percussions.

During the track, I reversed the phase on the receiver (thanks it has this adjustment :) ) to hear the difference, if there was one. Then, my friend and I gave his opinion to each other about the sound : we always thought the same thing about the sound.

As I said before, it is more obvious with percussion instruments, like piano and drums. Didn't have time to test guitars.

When the phase is reversed, you miss all the impulsions and feel a lack of sound instead of it. It's maybe one of the reasons why my mother doesn't like recorded piano. :rolleyes:

And according to this, I found some tracks of my favorite CDs that had a reversed phase !


I think my speakers are well-adjusted because most of my recordings sounded better with the phase set normally.

It was a really interesting test, what I find to be well-spent leisure time :smash: ;)
 
I assume you wired the correct terminals on the drivers to the correct binding posts inthe first place? (easily done -I've made that mistake myself in the past. Grab the Stereophile Test CDs if you can -good value, and they have lots of excellent test tracks for this sort of thing.)

Phase is certainly audible -much more so than many people realise. The use of recording elements and mixing them in different phase was partially how the original Matrixed Quadraphonic syestms like SQ and QS / RM worked. And they did work quite well you know; I've been a big fan of Quad for years, and it's a much maligned topic in audio. Visit our little home (for surround, and especially Quad-fiends like myself) at www.quadraphonicquad.com -lots of information archived there; I was one of the first 100 members I believe, or close to that at any rate!

For historical interest, SQ equations (SQ was far and away the most popular, coming, as it did, from Sony / CBS) ran like this:
Lt=Lf+0Rf-jO.707Lb+0.707Rb
Rt=0Lf+Rf-0.707Lb+j0.707Rb

Where Lt and Rt are the total left and right values, Lf and Rf and the front left and right values, Lb and Rb and the rear left and right values, and j indicates the rear channel components are shifted in phase by +90 degrees in one channel, and -90 degrees in another.

SQ was probably the best of the matrixed formats for, as you can see from the equations, the front stereo image is preserved intact, there being no cross-talk. The analogue decoder the signal was fed into performend sum and distance operations to re-build the 4 channels of information. Note however that the rear channels did contain a lot of information from other components, so separation here was a pretty minimal 3db until gain-riding logic decoders emerged, and, ultimately, the Fosgate Tate II 101a
decoder, which altered the matrix coeficients in real-time to provide a clearer lock onto channel information.

Anyone with a copy of Adobe Audition can actually recreate this decoding (actually re-encoding) and listen via 2 stereo amps and 4 speakers. The decoders frequently worked well on ordinary stereo recordings, which have a lot of out of phase information folded into them, and many Cds are still available with the matrix quad mixes still on them. The box set of Mike Oldfield's first 3 albums for example, preserves the SQ Quad mix of Tubular Bells (the some as on the SACD). Have a look on the above website for more information on doing this: it works very well.

OK, so all of this is a little extreme (but quite interesting if you like that kind of thing.) There's something else you can do too if you get into fiddleing about with phase; namely the Hafler circuit -no amps or anything other than a pair of speakers and some wire required.

Stick two extra speakers behind you. Wire their negative terminals together, the + on the left rear speaker to the left amp + terminal, the + on the right rear speaker to the right amp + terminal. The rear speakers will produce everything in the R channel, and everything in the L channel, but all the mono information, i.e. phantom centre L+R is put out of phase with itself and cancels out. Good for ambiance extraction. You can also try reversing connections to play still more with the phase of the rears, wiring the speaker + terminals together and the - terminals to the + amps terminals. All good, and quite instructive fun. Ultimate cheap version? Set up the speakers as per the first version mentioned above. Bridge the - terminals on the amp with a wire that leads to the wire connecting the - terminals on the rear speakers. That'll give a very basic varient of the SQ setup mentioned above, or at least a similar blend of channels. Just for heavens sake be careful: the bridging of their - terminals might case a few amps some headaches!

Brian Eno used to advocate a 3 speaker setup -identical to the firstr version above, just using a single rear, its terminals fed to the two + terminals on the amplifier. He even provided a diagram on the rear of the first pressing of Ambient 4: On Land.

OK, enough of all that now!
Best
Scott
 
Nuuk said:
I can clearly hear the difference on my open baffles and I wondered if there is any reason why dipole speakers would display the phenomenon more clearly than boxed speakers. :att'n:


Maybe because those speakers are more sensitive to room boundaries and reflections in the highs, because they diffuse the highs on 360° contrary to normal speakers which beam at high frequencies.
The waves reflecting on the wall add more obviously to the speaker's direct radiation, and the phase/delay errors induced become stronger and more audible.

When moving your dipole speakers, you move with them the phase/delay of the reflected waves. As those are more audible than on normal speakers, you hear more differences.


Well that's just my theory but maybe a more "scientifical" guy will tell me I'm completely wrong :D
 
bigwill said:
I really doubt inverting the signal is audible. Stop imagining things! :D

According to the first post, one speaker out of phase from the other is the most audible.

Positioning speakers for optimum position to minimize early reflection and equal distance to the listening postion is the second most audible.

Absolute phase is the most difficult, and most cannot be detected for the following reasons:
1. Speaker driver has too much sonic signature of it's own.
2. XO is not designed to be transient perfect.
3. The listener has little experience with unamplified live performances.
4. Health condition of the listener.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.