• 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!

Phase shift

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
YES

-> At 20 kHz < 10º But it can be better.

Class AB amplifier?
Yes, class AB. It can be better very easily if you extend amplitude frequency response. It is very easy to do, however at the expense of RFI immunity and stability. Phase response is directly derived from amplitude response and as isolated it has no effect on audibility. I hope you will once realize this fact.
 

wisechoice

Active Member
Joined
Jun 22, 2021
Messages
155
Likes
141
As this thread is in the amplifiers section and not the room correction section, I hesitate to post this here. However, it may provide some interesting data for discussion. I myself have been trying to understand phase, and I'm finally just beginning to do so.

I have just added fibreglass acoustic panels to my small office, which has a pair of Neumann KH120s on the desk. And I've taken measurements with and without the panels, as well as a couple of recordings to listen for comparison with headphones (I'm using SRH-1840s which measure extremely smooth & neutral).

At first I smoothed the measurements at 1/12th octave. With smoothing, there isn't a huge difference between the before (red) and after (green) in the frequency response [EDIT: these measurements are for the left speaker only]. The difference in unwrapped phase is more dramatic, and it occurs above the transition band, where room reflections and loudspeaker variation take over from room modes.

I suspect this difference in phase is a result of the reduction in reflections and reverberation in the room, which are audible both in direct listening and in the recordings.

Before:
Screen Shot 2022-07-14 at 4.36.27 AM.png


After:
Screen Shot 2022-07-14 at 4.36.52 AM.png


The reduction in resonances and reverberations is more noticeable in the un-smoothed frequency plots, overlaid here. However, I would still argue that the difference is a little easier to discern in the phase shift plotted above. In any case, a difference is there, albeit differently represented, in both frequency and phase plots.

Screen Shot 2022-07-14 at 4.55.53 AM.png


Whether this has any relevance at all to the discussion is for others more knowledgeable to decide.

Here are the recordings for comparison (performed with a single Sennheiser ME 64 into a Scarlett 2i4):

Note that I may have also engaged the bass tone control on the speakers between recordings, which was a mistake (although it sounds better), and which explains the difference in the boomy quality of the bass in the "before" recording. However, if you focus on the vocals, you can clearly hear a difference.

[EDIT: the tracks were each normalized to a peak of -3db in Audacity -- and because of the mistake I mention above, and the bass frequencies therefore being louder than the other frequencies, you may need to turn up the volume on the track without treatment to sufficiently hear the difference in the vocals. That said, the better balance in the other track should allow you to turn that one up as loud as you like, without too much discomfort. Not a perfect listening test, unfortunately, but it should be sufficient for a demonstration if played back on a high enough quality system. I can hear the difference clearly on both headphones and speakers.]

Also note that phase shifts are more audible when stereo tracks are summed into mono. So what is audible in a mono recording vis a vis phase may not be audible in person. In other words, a difference is audible in person, but it may not be phase, even though the phase difference may be audible in the recording (I think it is).

And here is the original track:
 
Last edited:

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,191
Location
Northern Virginia, USA
Top Bottom