Good to know I'm not imagining the difference.
The solo is straight from the original mix which IIRC was originally an mp3.
It is true that "up" isn't always better - I agree it just depends on what you're looking for.
The most night and day difference I ever found was maybe 30 years ago using a function generator to produce a single-cycle 100 Hz sinewave with a slow repeat rate. That was my kick drum simulation.
I can't recall if I used a full single-cycle or a half-wave burst.
I was listening in my old shop which had a hard wall behind me. The room maybe 10x12'.
I focused on the incident wave and first-order slap echo.
If the leading edge (or first half-cycle) was positive, producing pressure, the incident wave and echo seemed to be fuller.
Reversed polarity, with the first half-cycle producing rarefaction, sounded sucked-out. The slap echo also sounded sucked out.
If there was some kind of acoustic null it would seem to occur regardless of the stimulus polarity.
I did conclude that in this case "out" is better than "in" when it comes to cone movement on a pulse-like waveform.
The difference is striking ... after hearing the inverted version I just did not like the mix on the original at all (although that may be because it was from an mp3). I like to listen around 84dB then a bit louder and it was almost unpleasant here when louder. In comparison the balance and all elements are more pleasant in the inverted file but may be only for that section of the track?
LOL that's devotion for you, it's been 30 years since I did anything like that ... back in the day I used the phase switch on channels all the time but only in e.g. multiple mic situations or mic/DI and just chose the best direction for the context of the mix ... sometimes the out of phase thinner sound fitted in better ... 🙂
Just to throw it out there:
Classic Tannoy and JBL drivers move inwards when faced with a positive voltage (My Little Reds certainly do).
Both companies claimed that this is to preserve the absolute phase.
Classic Tannoy and JBL drivers move inwards when faced with a positive voltage (My Little Reds certainly do).
Both companies claimed that this is to preserve the absolute phase.
Handy for adjusting phase and widely used in recording studios because it sounds good too: https://www.littlelabs.com/products/ibp
Its commonly used to adjust phase when multiple mics are used on one source.
Apparently a key feature is its a Wein bridge all pass filter. The bridge topology may tend to have more favorable phase characteristics in terms of human perceptual preference.
Its commonly used to adjust phase when multiple mics are used on one source.
Apparently a key feature is its a Wein bridge all pass filter. The bridge topology may tend to have more favorable phase characteristics in terms of human perceptual preference.
The IBP doesn't adjust phase, it's an allpass filter, it compensate on a given freq.
Indeed used in case of multi mic, but you can achieve even better results by shifting ( offset) one mic over another ( on the timeline in the DAW- re alignement of mics).
Ibp are handy and faster to setup but in no way they 'adjust' phase, at least not over a small freq area.
Indeed used in case of multi mic, but you can achieve even better results by shifting ( offset) one mic over another ( on the timeline in the DAW- re alignement of mics).
Ibp are handy and faster to setup but in no way they 'adjust' phase, at least not over a small freq area.
You can also re-align waves in daw but that won't solve all the vintage classic recordings ... any with multiple mic recordings will have all sorts of phase anomalies in different parts of the track .... here's a chart showing prevalent phase for record companies 🙂
WRT using a sawtooth: Thanks for reviving a few almost-dead brain cells. I recall a construction article for a polarity meter in either Audio or Radio Electronics sometime during the 70s that used a sawtooth as stimulus.
I'll revisit my hard copy "polarity" file to see if I have it.
There might be an Eric Johnson cite in there as well.
Charles Darwin: Yes a classic JBL definitely would go "in" with a battery connected to the +terminal of the driver.
A couple of simple tools we all have:
The best reference to validate a measurement mic's absolute polarity is your own voice.
Connecting a battery and watching the woofer to validate mechanical polarity is also worth doing.
There might be other polarity surprises in the chain as well.
Slightly OT but on the subject of all pass filters and "phase" we have the "90° Filter" useful for making better mono.
For the record you'd never use it for stereo playback as its an image-destroyer but it does, in many cases, make better-sounding mono for AM and BG music by phase shifting L and R 90° relative to each other before summation into a single channel. https://proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=1055 (Schematic on page 2.)
Conventional L+R, when folded down to mono, eliminates all L-R difference information.
90° filtered L+R, we'll call it "I+Q," fold down L+R and L-R to mono with equal weight.
To my ear the I+Q version of "Paperback Rider" (for example) sounds better and in the experiments I tried the I+Q version has the same or similar spectral and instrumental balance and "life" as the stereo version. https://proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=1055&start=20
For those that are curious the 90° filter is worth a listen.
The one studio application for it (besides the rare stereo to mono master) is taming out-of-polarity synth patches during tracking which are quite common and bothersome.
Sometimes it can be quite difficult to get mono that folds down properly out of a synth without simply reverting to using only a single channel's output.
Murphy's law usually makes the problem track a bass line.
An ME over at GearSpace once said:
"Stereo bass is interesting: Mono bass is powerful."
I second that motion.
I'll revisit my hard copy "polarity" file to see if I have it.
There might be an Eric Johnson cite in there as well.
Charles Darwin: Yes a classic JBL definitely would go "in" with a battery connected to the +terminal of the driver.
A couple of simple tools we all have:
The best reference to validate a measurement mic's absolute polarity is your own voice.
Connecting a battery and watching the woofer to validate mechanical polarity is also worth doing.
There might be other polarity surprises in the chain as well.
Slightly OT but on the subject of all pass filters and "phase" we have the "90° Filter" useful for making better mono.
For the record you'd never use it for stereo playback as its an image-destroyer but it does, in many cases, make better-sounding mono for AM and BG music by phase shifting L and R 90° relative to each other before summation into a single channel. https://proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=1055 (Schematic on page 2.)
Conventional L+R, when folded down to mono, eliminates all L-R difference information.
90° filtered L+R, we'll call it "I+Q," fold down L+R and L-R to mono with equal weight.
To my ear the I+Q version of "Paperback Rider" (for example) sounds better and in the experiments I tried the I+Q version has the same or similar spectral and instrumental balance and "life" as the stereo version. https://proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=1055&start=20
For those that are curious the 90° filter is worth a listen.
The one studio application for it (besides the rare stereo to mono master) is taming out-of-polarity synth patches during tracking which are quite common and bothersome.
Sometimes it can be quite difficult to get mono that folds down properly out of a synth without simply reverting to using only a single channel's output.
Murphy's law usually makes the problem track a bass line.
An ME over at GearSpace once said:
"Stereo bass is interesting: Mono bass is powerful."
I second that motion.
In the end it makes no difference.
If you are concerned about absolute polarity, you have to realize that most recordings pay little attention to it, and those that do are randomly one way and the other. One has to have a switch and somehow note the correct polarity for that soon (if any)
dave
"Paperback Writer" ... surely you've not been hearing it wrong all these years?
You mean Hendrix didn't sing "excuse me while I kiss this guy?" The 90° filter cleared that up. LOL,
At least I got Paperback Writer correct in the post: https://proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=1055&start=20#p16555
It's approaching 104° today: The mind is going...
In the meantime I found the article "Build an Audio Phase Detector." https://proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=1055&start=20#p16555
FWIW I found that many modern recordings do get it right at the file level. Probably because they're edited on a DAW and the asymmetry becomes visually obvious.
Update: I just opened the original Renee Olstead file and the absolute polarity was positive heavy or "up." I've edited at least 5000 or so files and its rare to find one overall that has negative asymmetry which I invert. Maybe less than 10 times. Problem is most modern music is hyper-compressed and there's no asymmetry or DR left.
At least I got Paperback Writer correct in the post: https://proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=1055&start=20#p16555
It's approaching 104° today: The mind is going...
In the meantime I found the article "Build an Audio Phase Detector." https://proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=1055&start=20#p16555
FWIW I found that many modern recordings do get it right at the file level. Probably because they're edited on a DAW and the asymmetry becomes visually obvious.
Update: I just opened the original Renee Olstead file and the absolute polarity was positive heavy or "up." I've edited at least 5000 or so files and its rare to find one overall that has negative asymmetry which I invert. Maybe less than 10 times. Problem is most modern music is hyper-compressed and there's no asymmetry or DR left.
Last edited:
I know ... I was just "pulling your leg" ... I'm Scottish it's our sense of humour ... it's 25c here today (boiling) ... 🙂
Those are interesting files too ... there are elements I like about both but A is my favourite here ... which was the original?
My name is Kirkwood which makes me Scottish, I'm a serial bad punster, I write parody lyrics for popular songs and I've done some hideous mash-ups. They pop into my head and various random and "inopportune" times.
My best work may be a song about breakfast called "Bacon I'm Going to Eat You" performed to "Babe I'm Going to Leave You." When my daughter auditioned it for her band they couldn't stop laughing. My worst work may be a mash-up of Bali Hai and Ann Wilson's version of Immigrant Song. They're definitely the same song ya know. It's a sickness but somehow I cope with it.
Got the link wrong for the "Phase Detector." (Kitchen really meant polarity.)
https://proaudiodesignforum.com/ima...tector_Charles_Kitchen_Audio_January_1978.pdf
Spoiler Alert: Paperback Writer Cut A is the I+Q version.
The input levels were equal in both cuts.
The output levels were not normalized to be equal.
The hotter "A" version is the result of the added L-R which, in that cut, is a large amount.
Cut A, I+Q, has a peak level of about -1 dBFS and a total RMS power of about -16 dBFS.
Cut B, L+R, has a peak level around -3.3 dBFS and a total RMS power of about -17.3 dBFS.
https://proaudiodesignforum.com/ima...tector_Charles_Kitchen_Audio_January_1978.pdf
Spoiler Alert: Paperback Writer Cut A is the I+Q version.
The input levels were equal in both cuts.
The output levels were not normalized to be equal.
The hotter "A" version is the result of the added L-R which, in that cut, is a large amount.
Cut A, I+Q, has a peak level of about -1 dBFS and a total RMS power of about -16 dBFS.
Cut B, L+R, has a peak level around -3.3 dBFS and a total RMS power of about -17.3 dBFS.
Brilliant! ... I've been there too, I was a pro-guitarist for many years alongside the sound engineering, one of the most lucrative bands was a holiday camp comedy spoof rock band with many stupid lyrics (& costume changes). Also did a whole album with Gerry Rafferty's nephew and other Rafferty family members of ritually humiliating songs about football and working class life in Scotland to famous TV theme tunes ... LOL, he got a TV appearance off of that on a popular football show here .... Oh yes we like parody lyrics here! ... 🙂
Regarding loudspeaker polarity: many multiway loudspeakers work (more or less) as all-pass filters, and it's not unusual for the polarity to be swapped above a cross-over frequency compared to below it.
Typical example: two-way loudspeaker with a second-order Linkwitz-Riley cross-over. If the acoustic centres of the woofer and tweeter are well-aligned and the driver phase shifts around cross-over are not too bad or are compensated for, then the tweeter needs to be connected with reversed polarity.
Hence, no matter how you connect the amplifier to the speaker, either the tweeter or the woofer will work with the wrong polarity.
Typical example: two-way loudspeaker with a second-order Linkwitz-Riley cross-over. If the acoustic centres of the woofer and tweeter are well-aligned and the driver phase shifts around cross-over are not too bad or are compensated for, then the tweeter needs to be connected with reversed polarity.
Hence, no matter how you connect the amplifier to the speaker, either the tweeter or the woofer will work with the wrong polarity.
I forget where the transition region is in human hearing from temporal/phase sensitivity (and polarity?) to spectral.
Isn't it somewhere in the lower midrange to "mid-ish" midrange?
Just wondering about some things...
Over a span of midband frequencies, on just one frequency? Doesn't IBP actually has two all-pass filters one for LF and one for HF so as to extend useful bandwidth?
Regading sliding tracks in a daw, what about for people who prefer the sound and can afford the expense of recording to tape? How does one know in advance, say, for example, if a mic'ed up guitar cab is going to sit in the mix when tracking? Isnt' the idea to record it so that it sounds like a mix going in, without a lot of post production work to fix it later?
Thanks in advance for your insights 🙂