• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Understanding Phase

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
I agree with this position - but the interesting thing is that if the direct sound is of high enough quality then purely reflected energy has sufficient integrity to allow the musical content to remain intact, subjectively. The current system I'm playing with, tweaking, faces completely away from me when sitting down at the laptop - the only direct sound are the reflex ports at the back of the speakers; yet, the bounce from the walls, and glass that see the direct sound passes on the message with full coherency ... it's fascinating how the brain can reconstruct the sound event if enough untainted information is available.
Exactly. Even though the room seems to "mess up the phase" if we look at a measurement, in fact it has merely contributed delayed reflections in a completely coherent manner relating to its shape, dimensions and materials. We could, if we wanted to, work back to the direct sound by deconvolution. Our brains seem able to do this without effort.
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
[...]
First part does a very nice job explaining the odd displays of phase you see in programs like REW:


Part 2 starts somewhat repetitive but then gets interesting half way through as he analyzes the response of a real speaker (Meyer Sound?):

[...]

Hello,

I'm unable to wathc the movies, it says that "video is unavailable". Are the same movies as below, please?

Thank you!
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
And also, how important is phase accuracy to getting competent sound? For myself, I have never, ever worried about phase - it can do what it likes in a system, ;) - and such has never helped or hindered me getting the quality I was after ...

Hi,

Have you ever reverted + and - from a bass speaker? Have you seen that coil/diaphragm is moving vice-versa? You can also test this with a 1.5V battery and you'll see it easily.

Here's good example from my bedroom: after placing my subwoofer with backside to my chair I realised that bass can be heard with much more weight in the other half of the room than I intended to, hence I switched 180-degrees phase-shift switch from the sub to correct this. Now bass sounds bigger and with great weight, like it should be.

Here's a 20 kHz sine-wave on a headphones amplifier:
Sine_20kHz_before_LPF_mod.png vs. Sine_20kHz_after_LPF_mod.png
Left - original circuit, right- LowPassFilter modified by DIY.

Same as above, but with a 100 kHz test signal:
Sine_100kHz_before_LPF_mod.png vs. Sine_100kHz_after_LPF_mod.png
Left - original circuit, right- LowPassFilter modified by DIY.

You can see on the above pics that phase is visible shifted. Blue is the original audio signal, red is what the amplifier gets through the headphone-out 6.3mm jack plug. Is any sound difference after modifying the LPF? I say YES, higher treble sounds better now and more detailed...it's hard to explain, but after all, the phase shift difference is audible, at least in my case.

Let' continue with some square signals too.

Here's a 20 kHz square test signal:
Square_20kHzGain_0dB.png vs. Square_20kHz_after_LPF_mod.png
Left - original circuit, right- LowPassFilter modified by DIY.

Here's a 20 Hz square test signal:
Square_20Hz_before_High_Pass_mod.png vs. Square_20Hz_after_cap.png
Left - sound gets through the HighPassFilter filter of the amplifier, right - audio signal was applied after the 2.2uF/250V input caps (input caps are usually there to prevent DC-voltage inject from DACs or other audio sources).

Low-bass has more kick and weight when the input caps are bypassed. Not recommended to inject audio signal without input caps connected, but probably doubling their capacitance resolves the phase shift. However, 4.7uF caps costs more than 2.2uF, so this is why most manufacturers consider not to spend lot of money on this.

L.E.: The 100 kHz sine-wave phase shift also shows us the speed of the amplifier itself. With LPF removed signal is almost perfect with almost no phase-shift.
 
Last edited:

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,274
Location
Gold Coast, Queensland, Australia
hence I switched 180-degrees phase-shift switch from the sub to correct this

It is a polarity switch on your subwoofer, not a phase switch. Playing with a battery and your woofers is about polarity- not phase. Phase relates to timing of a waveform. It doesn't help that manufacturers of subwoofers perpetuate the confusion by labeling their little toggles 'phase/180' or the like.

That said, the physical placement of the subwoofer/s, with respect to the main woofers influences phase coherence. But, as we are speaking of low frequencies that are repetitive in nature (in most cases) and do not contain much in the way of positional information orspatial cues, a simple polarity switch can help in some cases.

Some of fully phase-coherent guys here with full blown DSP on everything would be able to add more no doubt...
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
Many thanks for clearing up the difference between phase and polarity. Just found some good readings about it too:
http://thehub.musiciansfriend.com/tech-tips/polarity-versus-phase
https://churchtecharts.org/home/2014/8/10/whats-the-difference-polarity-vs-phase

However, not sure about the back-switch of the subs and manufacturers confusion. I always thought it's a phase shift adjust and not the polarity, especially that expensive subs are not using only 0 and 180 degrees, but more intermediate adjustments too. Please have a look over here
(start watching from 4':40"). Also SVS guys say that their subs are phase-adjustable too: https://www.svsound.com/blogs/glossary/phase and their internal DSP is taking care of gradually phase-adjust: https://www.svsound.com/blogs/svs/75346755-understanding-dsp ("true variable Phase control is available to achieve optimal phase integration between the speakers and subwoofer"). Not owing any SVS subs and not trying to sell something here, but I do hope that what I'm adjusting to my Mackie sub is the phase and not the polarity.

However, one thing's for sure: after I was adjusting my sub's back-switch from 0 to 180 degrees my bass is in sync with my active monitors.
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
In some earlier comments, I suggested that phase shifts are not delays. Here is a paper explaining why that is the case.
https://inst.eecs.berkeley.edu/~ee123/sp14/NegativeGroupDelay.pdf
Basically, if you hold that a phase lag *is* a delay, you have to explain why a phase lead is not time travel into the future.

"Group delay" is merely the resemblance of a phase response against frequency to the phase shifts you would get with a real delay - including a negative delay. It is not an actual delay.

This may help explain to some people why, in the real world of transients as opposed to steady state sine waves, time alignment of drivers needs actual delays and cannot be performed by the crossover filter (which can only apply phase shifts).
 

Wombat

Master Contributor
Joined
Nov 5, 2017
Messages
6,722
Likes
6,459
Location
Australia
In some earlier comments, I suggested that phase shifts are not delays. Here is a paper explaining why that is the case.
https://inst.eecs.berkeley.edu/~ee123/sp14/NegativeGroupDelay.pdf
Basically, if you hold that a phase lag *is* a delay, you have to explain why a phase lead is not time travel into the future.

"Group delay" is merely the resemblance of a phase response against frequency to the phase shifts you would get with a real delay - including a negative delay. It is not an actual delay.

This may help explain to some people why, in the real world of transients as opposed to steady state sine waves, time alignment of drivers needs actual delays and cannot be performed by the crossover filter (which can only apply phase shifts).



Three cars are travelling in-line on a road. Wrt to the middle car, one is advanced(leading) and the other delayed(lagging) on the same road. They are all in the same time reference frame(not past, present and future), there is just a positional difference.

Same with phasors - angular displacement wrt reference phasor.
 
Last edited:

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
Nice reading Cosmik, thanks.

However, if you're looking to one of my few posts above you'll see that for 20 kHz sine-waves the amplifier is lagging and not advancing, because scope is analyzing the signal from right to left so output signal is time-delayed and not time-advancing.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,834
Likes
16,496
Location
Monument, CO
"Group delay" is merely the resemblance of a phase response against frequency to the phase shifts you would get with a real delay - including a negative delay. It is not an actual delay.

Group delay is defined as the negative of the derivative of phase in frequency: GD = -dP/dF where GD = group delay, dP = difference in phase, and dF = difference in frequency. Linear phase (over frequency) leads to constant group delay, a desirable outcome to preserve pulse integrity (in time) since all frequencies are delayed equally. True time delay delays every part of the signal equally regardless of frequency. Linear phase filters and true time delays have application in many radar, lidar, and communication systems.
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
Given the name of this thread and also of the forum the best shot I got is the one posted half page above: https://audiosciencereview.com/forum/index.php?threads/understanding-phase.757/page-2#post-69737. It's exactly what's happening inside an amplifier, nothing else, so I believe this is related to phase issues.

The square-waves with 20Hz bass from my pics are showing an imperfect response of the amplifier caused by the high-pass filter; bypassing that filter shows the perfect square-wave; there's a phase-advance showing a bass roll-off actually (deficiency in low frequency, based on http://users.tpg.com.au/users/ldbutler/Waveforms.htm and http://www.r-type.org/articles/art-125.htm).
Square-waves with 20Hz shows deficiency with trebles (based on http://users.tpg.com.au/users/ldbutler/Waveforms.htm and http://www.r-type.org/articles/art-125.htm)
The sine-waves with 20kHz trebles looks like time-delay to me (phase-shift), but please correct me if I'm wrong...it's just what I see with my scope and this is only affecting trebles from my amplifier and not the midrange or other freqs. More details here: http://etc.unitbv.ro/~olteanu/Tehnici de masurare in tc/Phase Measurement.pdf, especially fig. 41.2 with 2 signals having the same amplitude, but small phase difference.

Found a good read over here: https://www.uaudio.com/blog/understanding-audio-phase/ where it states that most common phase issues are on low-end freqs, so subwoofers seems to be more affected.
Also, phase issues from mics being at different distances from the singer/instrument:
.
And again phase issue with mix: https://www.youtube.com/watch?v=kneNsn65EBg and http://behindthespeakers.com/fix-phase-cancellation/.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL
However, one thing's for sure: after I was adjusting my sub's back-switch from 0 to 180 degrees my bass is in sync with my active monitors.

Maybe your monitors invert the signal.
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
Not at all, I already explained this few posts above: my sub was vice-versa mounted in the room, it was not facing the monitors, it was facing the back-wall. Adjusting the phase with 180 degrees solved the problem.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL
Last edited:

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
Ya, I know that...it's impossible to do it with speakers, but this has nothing to do with phase inside an electronic amplifier.
However, Tyll's measurements prove that some planar headphones can achieve almost perfect square-wave shape, so less air between ear and driver = better results. :)
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL
However, Tyll's measurements prove that some planar headphones can achieve almost perfect square-wave shape, so less air between ear and driver = better results.

I'll assume (for now) that's frequency dependent...

A sweep tone will reveal some reasonable squares at cherry-picked frequencies:

upload_2018-3-13_17-34-3.png
 
  • Like
Reactions: trl

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,274
Location
Gold Coast, Queensland, Australia
Now, try to get that perfect electrical square through the air to your ear

Or the microphone...

Square wave testing is phenomenally useful for electronics, but pointless for transducers (speakers and mics) IMO.

Impulse and toneburst testing however, is a whole other story and very useful for speaker testing.
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
I'll assume (for now) that's frequency dependent...

A sweep tone will reveal some reasonable squares at cherry-picked frequencies:

Looks good actually, how did you measured it? What mic and speakers, please?
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL
Looks good actually, how did you measured it? What mic and speakers, please?

UMIK-1 and MartinLogan reQuest, both speakers active at 10 feet.

A few Hz up or downscale from that it goes wild again, cyclical.
 
Last edited:
  • Like
Reactions: trl

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL
Impulse and toneburst testing however, is a whole other story and very useful for speaker testing.

I was curious about the usefulness/accuracy of Impulse and Step Response, when calculated from a swept sine tone and not from an actual impulse or step signal.

A little testing showed me the calculated-from-swept-sine values matched recorded audio waveforms for actual impulse and step signals, much to my surprise.

https://audiosciencereview.com/forum/index.php?threads/impulse-response.1765/
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,274
Location
Gold Coast, Queensland, Australia
I was curious about the usefulness/accuracy of Impulse and Step Response, when calculated from a swept sine tone and not from an actual impulse or step signal.

Toneburst testing allows testing the upper limits of power handling of speakers without having to burn them out, or go deaf, which is always a good thing. :)

It also allows you to more effectively remove the room effects from plots, which is much more difficult with a continuous/repeating or swept sine.

I have the design for a tone burst generator I built many years ago I will dig out and scan.
 
Last edited:
Top Bottom