• 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!

Minimum Phase vs Linear Phase

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
Whilst I probably agree with what John is saying here, again it has to be put into context. How audible is an effect of moving the phase of say high frequency harmonics of a percussive sound? How much audible impact is there of changing the phase of a harmonic at 20 kHz that might be -70dB down?
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
In the DAC, minimum phase by default, of course.

But if you make resampler from a computer, depending on the recording, it may sound better in another way, something I have verified many times. Usually I prefer Intermediate, if the original sound does not convince me. If after the process still does not convince me, I delete the files.

I can differentiate between the Pass Band choiced. Of course, I talk about very good recordings with high DR.

View attachment 32296

Come back when you have performed some blind tests that verify your subjective opinion. ;)
 

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,161
Location
Tarragona (Spain)
If the recording is well made, the minimum phase is the right choice. But unfortunately there are and there were many bunglers and with a fairly resolutive system their bad arts or incompetence come to light when you listen to the recording.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
If the recording is well made, the minimum phase is the right choice. But unfortunately there are and there were many bunglers and with a fairly resolutive system their bad arts or incompetence come to light when you listen to the recording.
On what basis is it the right choice. Please explain.
 
Last edited:

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,008
Likes
3,999
Location
Amsterdam, The Netherlands
Just as a reminder, here is the classic illustration of the Gibbs phenomenon. The animation builds up a square wave from sine wave components, and you see the "ripple" that is simply a result of not using an infinite number of harmonics - so not having infinite bandwidth. It is not "ringing" as such.

SquareWave.gif
 

JanRSmit

Member
Joined
Jan 21, 2019
Messages
54
Likes
21
Whilst I probably agree with what John is saying here, again it has to be put into context. How audible is an effect of moving the phase of say high frequency harmonics of a percussive sound? How much audible impact is there of changing the phase of a harmonic at 20 kHz that might be -70dB down?
Yes, would like to know that as well. In our thinking and research we use phase, however our hearing system acts on time differences. The outcome may well be that human hearing is much more sensitive to “phase” in low frequencies.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
Yes, would like to know that as well. In our thinking and research we use phase, however our hearing system acts on time differences. The outcome may well be that human hearing is much more sensitive to “phase” in low frequencies.
At low frequencies dacs and amplifiers are unlikely to have any issues with phase. It's only at their response top end where phase shifts are likely to occur.
 

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,161
Location
Tarragona (Spain)
On what basis is it the right choice. Please explain.

I do not understand you. I have read bass and not basis :)

One example of my use of resampler with Intermediate is when I made a rip from CD made in 80 - mid 90. Usually the commercial music sounds bad or very bad, probably because it has a lot of jitter inside. Sometimes with a new SSRC resampler Intermediate you can cover up the jitter and make listening much more bearable, but sometimes not so.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,008
Likes
3,999
Location
Amsterdam, The Netherlands
I do not understand you.

And I don't understand you. :)

One example of my use of resampler with Intermediate is when I made a rip from CD made in 80 - mid 90. Usually the commercial music sounds bad or very bad, probably because it has a lot of jitter inside.

How can music have "jitter inside"? A ripped file is a ripped file - if there was enough jitter in the ripping CD drive to cause bit errors, there are incorrect bits in the file, otherwise the bits are the same as in the original. You can't store jitter in a file.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
I do not understand you. I have read bass and not basis :)

One example of my use of resampler with Intermediate is when I made a rip from CD made in 80 - mid 90. Usually the commercial music sounds bad or very bad, probably because it has a lot of jitter inside. Sometimes with a new SSRC resampler Intermediate you can cover up the jitter and make listening much more bearable, but sometimes not so.
You have lost me I'm afraid.

Linear phase (in a well made recording, i.e. One that is properly filtered and bandwidth limited at half sample rate) is the right choice.
 

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,161
Location
Tarragona (Spain)
https://www.google.com/search?&q=minimum+phase+dac+why

-> The Linear vs. Minimum Phase Upsampling Filters Test [Part I]: RESULTS

http://archimago.blogspot.com/2015/07/the-linear-vs-minimum-phase-upsampling.html

I usually listen to recordings with acoustic (more) or electrified (less) instruments and almost nothing with synthetic instrumentation and voices manipulated with autotune as with current commercial music. With the first, and the second type of recordings, you usually have memory of how they sound live, and their tonality. With these recordings the best sound is with minimal phase.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,008
Likes
3,999
Location
Amsterdam, The Netherlands
Linear phase (in a well made recording, i.e. One that is properly filtered and bandwidth limited at half sample rate) is the right choice.

I agree.
 

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,161
Location
Tarragona (Spain)
In those years many things were done wrong when digitizing. One of them, out of ignorance at that time, despise the importance of the jitter. Over the years it was verified that we are more sensitive than previously believed.

I guess that came from the preference of vinyl, without digital processing. Now digital technology is much better but the products created are horrible.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,008
Likes
3,999
Location
Amsterdam, The Netherlands
The Linear vs. Minimum Phase Upsampling Filters Test [Part I]: RESULTS

Did you actually read the "conclusions" part?

"If we can't find clear preferences with the amount of pre-ringing in the 99% steep filter as compared to minimum phase with no pre-ringing, then what are the chances that the comparatively small amount of pre-ringing with the default 95% SoX linear filter is of any concern?"
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,008
Likes
3,999
Location
Amsterdam, The Netherlands
In those years many things were done wrong when digitizing. One of them, out of ignorance at that time, despise the importance of the jitter. Over the years it was verified that we are more sensitive than previously believed.

You make a lot of statements and claims, but I see very little logic or rationale - not to mention evidence....
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Archimago had the following conclusion in part III:

As we know last week, statistically, indeed there was a preference for minimum phase filtering, but not strongly significant based on statistical data.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,008
Likes
3,999
Location
Amsterdam, The Netherlands
Archimago had the following conclusion in part III:

As we know last week, statistically, indeed there was a preference for minimum phase filtering, but not strongly significant based on statistical data.

Yes, there was only one sample where the preference was actually statistically significant, and as pointed out earlier, the test was artificially exaggerated.
 

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,161
Location
Tarragona (Spain)
I guess it will have a lot to do with the kind of music used for the test. As I said, with acoustic instrumentation and voices without autotune, the difference is very evident to me.

For the rest, I have no confidence in the tests performed from the Internet. Blind tests must be in the same room and with a good stereo that does not have bottlenecks that prevent appreciating differences.

And, of course, without using foobar2000 as a software player, with or without the ABX plugin. If the source is a computer, it must be optimized to play multimedia.

- End off topic -
 

simbloke

Senior Member
Forum Donor
Joined
May 30, 2019
Messages
354
Likes
584
Location
North Wales, UK
A few years back I was using a DIY WM8741 based DAC with an Arduino based controller.

The DAC had 5 (IIRC) different digital filters and I was never 100% sure my code was properly changing the filter as I could never hear the difference.

I think I was doing the right thing as all the other programming of the DAC worked fine. Now with my bought DAC I don't worry at all that there's no way to select filters!
 
Top Bottom