I dragged a pair of Leak TL12's all the way from the UK when we emigrated to Canada and I ended up putting them in the trash!
If only there was internet back then.
If only there was internet back then.
Hi Nige, you do very much very good work.
In all your endeavours i have never seen you mention signal polarity switching.
I my world I regard signal/acoustic polarity as numero uno mission critical.
Dan.
In all your endeavours i have never seen you mention signal polarity switching.
I my world I regard signal/acoustic polarity as numero uno mission critical.
Dan.
Hi Dan. I played with this years ago. Simple way is speaker phase. Quad amps are inverting phase as are their preamps. That is very cunning as it gives more for less. Problem is people used Quad amps with other preamps. The Quad 303 is a nice easy input load whilst being inverting. That's win win. The problem with using reverse phase is for whatever reason the recording engineer didn't. It can be just one circuit in a mix that changes it. Possibly 90 % of music has no real phase element. The 10% has transients that need to be right. On a similar note EQ for records is seldom correct. A way of doing that is valid. The obvious one is 50/75 uS. Much is 50 uS. If one makes the EQ passive for 75 uS it is easy to make it 25 to 100 uS variable. That is worth doing. It also allows a cartridge to be tuned. Many MM sound better with high loading ( >100 K ) . The EQ can then be set to taste. Give it numbers 1 to 10 and sleeve notes.
If they were TL12 amps with ECC33 values ( red ) they are worth silly money. TL 12 + is a different animal. I like the 12 + more. I gave one away also.
If they were TL12 amps with ECC33 values ( red ) they are worth silly money. TL 12 + is a different animal. I like the 12 + more. I gave one away also.
Right Nige, for Ortodox Christians in Serbia, Greece, Romania, Bulgaria and Russia, today is the day.
I run all digital nowadays.
My feed is unbalanced into balanced input preamp via reversing switch.
I need to get around to fitting reversing switches to my active speakers in order to be able to define system individual stage throughput polarity.
I run Foobar as source with VST polarity switcher plugin.
If recorded polarity was uniform all of this would be unnecessary.
In practice, compilation cds are all over the shop.
Albums are problematic also....often the top 40 song is inverted wrt the others.
I am comfortable with flipping polarity every second track.
A PC software player (Foobar, Winamp etc) with settable track by track polarity meta info would be cool, way cool.
Dan.
My feed is unbalanced into balanced input preamp via reversing switch.
I need to get around to fitting reversing switches to my active speakers in order to be able to define system individual stage throughput polarity.
I run Foobar as source with VST polarity switcher plugin.
If recorded polarity was uniform all of this would be unnecessary.
In practice, compilation cds are all over the shop.
Albums are problematic also....often the top 40 song is inverted wrt the others.
I am comfortable with flipping polarity every second track.
A PC software player (Foobar, Winamp etc) with settable track by track polarity meta info would be cool, way cool.
Dan.
It would be fun to make a phase detector based on transients. Perhaps as crude as seeing which side has higher peak voltage. It might show something. My hunch is ears will find it quicker. Were it not for ears hi fi would be so easy.
If it was really cunning just LEDs with a dropper resistors to the speaker outputs ( LED's reversed biased to each other ). It might just show up as how bright on an LED's just showing some light. Our eyes are said to be sensetive to this. The old style tuning meters work this way. Match the LEDS in same bias to be sure they are a matched pair. Usually they are even when cheap. Triology valve amps used this for bias setting, others also.
BTW if the LEDs are in reverse bias to the same resistor they protect each other. That's where saving time and money really pays off. If the current is low I have reversed biased LED's by 24 V and had no problems. 5 V is usually advised. If you place 2 x reversed biased LED in series with 2 x 68K( 136 K ) > 0.6 watt they will work at 240 VAC .
If it was really cunning just LEDs with a dropper resistors to the speaker outputs ( LED's reversed biased to each other ). It might just show up as how bright on an LED's just showing some light. Our eyes are said to be sensetive to this. The old style tuning meters work this way. Match the LEDS in same bias to be sure they are a matched pair. Usually they are even when cheap. Triology valve amps used this for bias setting, others also.
BTW if the LEDs are in reverse bias to the same resistor they protect each other. That's where saving time and money really pays off. If the current is low I have reversed biased LED's by 24 V and had no problems. 5 V is usually advised. If you place 2 x reversed biased LED in series with 2 x 68K( 136 K ) > 0.6 watt they will work at 240 VAC .
Last edited:
I have thought about polarity detection and auto switching.
Maybe just peak wideband detection would work, maybe midband (vocals) filtering, or bass filtering is required.
Dan.
Maybe just peak wideband detection would work, maybe midband (vocals) filtering, or bass filtering is required.
Dan.
I must try the LED's . I have a hunch it might work. A bass pass filter should work. The vollume control could find the ideal level. Another way is set up a comparator. Advance the volume until in lights up. Change the phase via op amp and set up another comparator section. Another way is a VU meter showing positive and negative phase. It would need a diode gate. Make the signals mono and see which wins.
Polarity is another of those things I've never worried about - yes, I'm sure it would make a difference to the sound of various albums and tracks, switching it back and forth - but it's not something that concerns me; worrying about other considerations has been the valuable activity for getting the type of sound I chase ...
I tangled with a guy who showed a vactrol-based compressor-limiter on another forum years ago, and said I doubted his design worked. He had a cap in series with the LED in the optocoupler---and nothing else. Two of us conjectured that it wouldn't work once the cap was charged, but PRR pointed out, correctly, that the reverse breakdown of the LED provided the discharge path. The guy (who was uber-sensitive to criticism) turned this into a glowing (sorry) endorsement of his design. However, it's a risky business, as there is no guarantee that a given LED will have a well-determined breakdown voltage. I tried to get the Vactec folks to say what it was even typically---no response. I've measured at least as high as 30V on single isolated LEDs.
I've seen some signal asymmetry detectors published, from way back. Probably somewhere in files in storage.
Brad
The first thing to be sure of it the system you use has standard polarity. The Quad power amps are inverting phase. In a multi mic situation the absolute phase is due to placement. If a mixing desk with transformer input is wired up wrong all recordings will be adjusted to suit. Putting a mic in a drum from the back surely must invert the phase.
I think what Dan has said is a bit of fun. My worry is can I do it without making the signal path more complex. The better idea if me is two Neutrik speaker terminals on the speaker fronts for quick changes. They seem better than 4 mm sockets.
Going back to what I said before. 90% of music at a good guess will have no phase quality. 10 % will. The example of a bass drum might just be where the transient regardless of the recording engineers mistake ( ? ) should push forward. If I am right about bass letting you hear detail this will make a big difference. As Scott points out the engineer might have prefered it " wrong ". The pahse between mid and bass is always wrong. It can in the modern world be corrected with digital time delay. This makes a system sound more powerful and open using the same level of power. Even so it varries with frequency.
Dave Mate of SSL mixing desks made a speaker with very good squarewave response. It was a three way speaker using five drive units. The extra two were for error signal correction. Dave made a point. Humans are very good at processing signals with phase error. This processing causes fatigue. When I put this to Douglas Self he asked why we don't get visual fatigue ? For a start the brain gives over more processing power for images, if not driving a car would be impossible. Also I doubt we see much reflected light as image imformation. Dave said the best compromise is make the tweeter brake frequency as high as possible. 7 kHz being nice. On my recent speaker I got 6 kHz.
I think what Dan has said is a bit of fun. My worry is can I do it without making the signal path more complex. The better idea if me is two Neutrik speaker terminals on the speaker fronts for quick changes. They seem better than 4 mm sockets.
Going back to what I said before. 90% of music at a good guess will have no phase quality. 10 % will. The example of a bass drum might just be where the transient regardless of the recording engineers mistake ( ? ) should push forward. If I am right about bass letting you hear detail this will make a big difference. As Scott points out the engineer might have prefered it " wrong ". The pahse between mid and bass is always wrong. It can in the modern world be corrected with digital time delay. This makes a system sound more powerful and open using the same level of power. Even so it varries with frequency.
Dave Mate of SSL mixing desks made a speaker with very good squarewave response. It was a three way speaker using five drive units. The extra two were for error signal correction. Dave made a point. Humans are very good at processing signals with phase error. This processing causes fatigue. When I put this to Douglas Self he asked why we don't get visual fatigue ? For a start the brain gives over more processing power for images, if not driving a car would be impossible. Also I doubt we see much reflected light as image imformation. Dave said the best compromise is make the tweeter brake frequency as high as possible. 7 kHz being nice. On my recent speaker I got 6 kHz.
Indeed.
It is like the left-right positioning of the drums kit.
It always amazed-me that, in US, it was, most of the time, set according to the drummer point of view (Charleston at left), while in Europa, it is from the audience one. My first feeling was most of he american drummer were left handed ;-)
Last edited:
I will correct my own statement. We do see plenty of reflected light. In fact so much so as to be diffuse. I doubt we process it.
Whilst I accept the point about reverse biased LED's my idea protects them anyway. I would prefer a Schmidt action of the circuit. A VU chip will do that. A 10 K + 10 K op amp inverter to drive the out of phase VU meter. A diode in the signal path of each to ensure only that half cycle used. If using a 4 op amps chip a precission rectifier could be used to ensure a very low measuring threshold. The spare op amp an identical inverting buffer to match gain. TL074 would be OK or even LM324. The buffer could be filtered. A LM324 makes a very good comparator. The reason LM 324 survived is it is the tractor of chips. It's designer it is said was unhappy with it so suggested a low price to recover the devellopement costs. TL074/84 will often be for simple engineering the other choice. I have known LM324 beat TL074 on common mode problems . MC33079 beats them both and most other things.
Precision rectifier - Wikipedia, the free encyclopedia
Whilst I accept the point about reverse biased LED's my idea protects them anyway. I would prefer a Schmidt action of the circuit. A VU chip will do that. A 10 K + 10 K op amp inverter to drive the out of phase VU meter. A diode in the signal path of each to ensure only that half cycle used. If using a 4 op amps chip a precission rectifier could be used to ensure a very low measuring threshold. The spare op amp an identical inverting buffer to match gain. TL074 would be OK or even LM324. The buffer could be filtered. A LM324 makes a very good comparator. The reason LM 324 survived is it is the tractor of chips. It's designer it is said was unhappy with it so suggested a low price to recover the devellopement costs. TL074/84 will often be for simple engineering the other choice. I have known LM324 beat TL074 on common mode problems . MC33079 beats them both and most other things.
Precision rectifier - Wikipedia, the free encyclopedia
Last edited:
Nige, all this is fine for low current requirements, things like simple preamps which do not draw too much current. Once you go over the top, you either make several such small circuits for each block, or you need to revert to discrete regulators which are scaleable for larger currents.
For power amps, a corresponding regulator will in fact be almost of the same complexity as the power amp alone. OTOH, it will probably improve the sound, and not by a small measure. A bit like filling up the tank of a car made for say 90 octane fuel with 98 octane Super Plus fuel. If no wise guy electrinics stand in the way, your car will perk up and then some. 😀
For power amps, a corresponding regulator will in fact be almost of the same complexity as the power amp alone. OTOH, it will probably improve the sound, and not by a small measure. A bit like filling up the tank of a car made for say 90 octane fuel with 98 octane Super Plus fuel. If no wise guy electrinics stand in the way, your car will perk up and then some. 😀
We need 1 mA so it will be OK. It's how is screws up the signal path that bothers me. I think detect and change speaker cable is my option. If digital I dare say we can change phase without any great problems except cost. I don't want another op amp in the way. The text book valve or transistor phase splitter does it with least number of elements. It can be very low distortion. A transfomer can even have gain. Thus a bad thing becomes a good thing . A very good transformer can be close to - 80 dB . All that remains is to switch without loss. The Hypex module works equally well in either phase option. All you do is ground one side if running unbalanced in. How this is arranged is the input op amp has a gain of about 6. It uses two op amps and is balanced. This makes for a one resistor gain change to both sides, there is no DC blocking cap. It also is completely symetrical. It keeps the nasty power amp bit away from the input. This means the gain can be changed without screwing up the power amp stability. Changing the Hypex gain to about 400 mV works very well with a passive volume control for preamp .
Most 'artists' really don't have a clue technically...they just make the music.
Recording/mastering is where it all goes astray, with mics, recorders, tapes, outboard, desks etc with random polarity behaviour....Why the Fuss about Plus? - On XLR pin 2 hotness and absolute polarity.
Even harmonic distortion of mics, circuitry and loudspeakers implies asymmetric transfer function.
Impulsive sound is itself asymmetric in nature, and the ear readily detects this asymmetry.
According to the even harmonic behaviour of individual series connected stages, even harmonics can be exaggerated or diminished (cancelled, at least partially) at system acoustic output, or anywhere in the signal chain.
Each stage in my system has switchable input polarity in order to enable the subjectively most acceptable system distortion behaviour.
When this is done, recording acoustic polarity becomes perfectly revealed, and indeed mission critical to realistic reproduction.
It is further mission critical that polarity switching be done at the very source....to this end I use a Foobar polarity switching VST plugin.
On first inspection, reversed source material wav file polarity AND reversed speaker connection polarity ought to sound identical to normal polarity source wav file and normal polarity speaker connection....ime in practice the two conditions do NOT sound identical.
Dan.
View attachment Why the Fuss about Plus_ _ Recording Magazine -.pdf
Elementary Mistake....
Inverted acoustic polarity 'ain't never' going to sound correct, no matter how good the system is.
Dan.
Ime/imo acoustic polarity is the very first thing to get right 😉.
Inverted acoustic polarity 'ain't never' going to sound correct, no matter how good the system is.
Dan.
I made an interesting observation today, out in sunlight.
I had a car tyre spinning at such speed that the motion was a blur, and the tread pattern detail was not discernible.
However, when I blinked my eyelids as fast as possible, I was able to see the individual tread blocks perfectly, as if using a strobe light.
Dan.
- Status
- Not open for further replies.