You have it a bit wrong. You and I need the same answer. You do seems to be coming out with good suggestions. Some I think with a little thought are not fully hatched. We will see. I think the need for a sub woofer to do a test is slightly at odds with the evidence. I guess playing an LP via a filter to see if all the magic is lost would prove something. You would have to filter the CD also.
It's very strange to think liking vinyl is an error. I think it is strange I still do.
Do you realise if you have your wish what you get is a camp fire chat and not real debate. People with experiance will not join because the debate is sterile. Have you noticed how few experts do join. They are around in these forums. I don't claim to be one. I am time served only.
I think I will leave the debate now as the good I can do is ended. I hope I have made a window open to how the questions can be answered.
I suspect you are a younger man than me ( 60 )? Do you work nights?
The experts and the time served often look idiots when the simple truth is found. Usually it is simple. Who would guess MP3 would work. It works well enough.
We should have that pint sometime.
It's very strange to think liking vinyl is an error. I think it is strange I still do.
Do you realise if you have your wish what you get is a camp fire chat and not real debate. People with experiance will not join because the debate is sterile. Have you noticed how few experts do join. They are around in these forums. I don't claim to be one. I am time served only.
I think I will leave the debate now as the good I can do is ended. I hope I have made a window open to how the questions can be answered.
I suspect you are a younger man than me ( 60 )? Do you work nights?
The experts and the time served often look idiots when the simple truth is found. Usually it is simple. Who would guess MP3 would work. It works well enough.
We should have that pint sometime.
Last edited:
Interesting. I had already noted that small+sub worked less well with vinyl than digital, but never made the connection. I also don't prefer small+sub with any source for that matter. This is how I ended up with infeasible full range speakers.
Over the past few months I've restored 2 separate Leslie speaker cabinets that push lf sound around intentionally in various directions by the bucketful - they do sound fantastic and spacious - not for hifi BTW, for live rock music with a 1967 Hammond. I wonder if acoustically this effect is related? If you haven't heard a Leslie ever or for a while you might not know what I mean, but there is this amazing spaciousness and size to the sound..........? I think perhaps related.
Nigel: To be clear. All things being equal with a central sub you get
L (hpf): L+R (LPF): R(hpf) the L-R ambience is removed in the summing for the sub. In theory a simple test as to if the sub kills the ambience in theory. In practice getting this right might be tricky. You could also use an adjustable hafler setup with full range at the rear and mini monitors at the front to pull out L-R and see if the magic increases.
I would consider the two Douglases to be considered experts in their fields and I am just perching on shoulders.
Decade behind you, work days, like beer
L (hpf): L+R (LPF): R(hpf) the L-R ambience is removed in the summing for the sub. In theory a simple test as to if the sub kills the ambience in theory. In practice getting this right might be tricky. You could also use an adjustable hafler setup with full range at the rear and mini monitors at the front to pull out L-R and see if the magic increases.
I would consider the two Douglases to be considered experts in their fields and I am just perching on shoulders.
Decade behind you, work days, like beer
Last edited:
I thought you were working and writting. Good boss if so.
I have left the thread but like how it is going now. I will follow it.
Years ago I noticed single subs pulled voices to that centre position. Sometimes it was good as the voice needed to be centre stage. It was and I think is still against theory to question that. Much is wrong with theory. I would love a pair of the old Celestion subs for the SL6.
As for Douglas. I must have sold more books for him than a real agent. A lot anyway. I don't agree with all, but not anything worth saying.
I have left the thread but like how it is going now. I will follow it.
Years ago I noticed single subs pulled voices to that centre position. Sometimes it was good as the voice needed to be centre stage. It was and I think is still against theory to question that. Much is wrong with theory. I would love a pair of the old Celestion subs for the SL6.
As for Douglas. I must have sold more books for him than a real agent. A lot anyway. I don't agree with all, but not anything worth saying.
Maybe look upon it like bobsleigh? Good bobsleigh smoothly follows the locus of the course. Bad bobsleigh rattles about. Terrible bobsleigh leaves the course on the tight bends. Awful bobsleigh ends up in orbit somewhere.......
The reason vertical modulation is constrained in vinyl is rather like bobsleigh in that vertical G upwards is a big problem, nothing holds a bobsleigh down for pure vertical acceleration except 1G gravity. This is why bobsleigh courses don't have vertical dips (AFAIK!).............
Well, I'm glad it's not just me...
I tried feeding the input pins from specified impedances but I still couldn't get any sensible results out of it. I assume the schematic is wrong somehow, and not likely to be corrected after this effluxion of time.
The whole crossfeed business is more complicated than it looks. The Macaulay and Langvad crossfeed circuits work essentially on the same principle, which is less than obvious, and all credit to the designers. In each there are three paths which are summed/subtracted to create the output from each channel. Langvad does this more elegantly, using 4 opamps to Macaulay's 7. (possibly not a very fair comparison as Macaulays original circuit used discrete transistors, but nowadays using a single transistor as a summing amp isn't going to cut it)
Both, despite using second-order filters, have a roll-off for the antiphase signal of only 6dB/octave and a +2dB hump before the rolloff. This is inherent in the circuitry and not a matter of tweaking the gain of the three paths.
That can be overcome, but I am still working on my version.
Sorry, got interrupted there.
Another point about Macaulay and Langvad is that, as noted upthread, flatness of the frequency response for in-phase signals depends critically on component matching. This is fine for a one-off test box that can be tweaked as required (once you are aware of the issue) but looks unpromising for production. I suspect this is the true reason why this approach never seems to have gone commercial.
If you are happy with a 6dB/oct rolloff of the anti-phase signal then an elegant solution is crossfeed with two caps and a resistor, as described by Renardsen:
Phono Pre-amp Design
By exploiting the fact that signal can go through a resistor in both directions all the problems of accurate cancellation are avoided.
Another point about Macaulay and Langvad is that, as noted upthread, flatness of the frequency response for in-phase signals depends critically on component matching. This is fine for a one-off test box that can be tweaked as required (once you are aware of the issue) but looks unpromising for production. I suspect this is the true reason why this approach never seems to have gone commercial.
If you are happy with a 6dB/oct rolloff of the anti-phase signal then an elegant solution is crossfeed with two caps and a resistor, as described by Renardsen:
Phono Pre-amp Design
By exploiting the fact that signal can go through a resistor in both directions all the problems of accurate cancellation are avoided.
Who knows? monty78pig has not finished the design so it can hardly be analysed.
(The Macaulay circuit uses a summer to creat mono VLF but it is an uneconomic approach compared with Langvad)
I don't think we can take anything on trust. This is a rather non-intuitive area.
When I started simulating I had no idea that getting the in-phase response flat was going to be such an issue. But I did expect the 6dB/oct rolloff problem. Think about subtractive crossovers.
The Design of Active Crossovers by Douglas Self
Last edited:
I do have your book on my to buy list! I had the epiphany on the subtractive xover whilst trying to make sense of the Macaulay circuit.
I have not managed to simulate the M+S topology, so its a feeling nothing more at the moment. I fell off the simulation horse a while ago, so bear with my whilst I clear the rust and make old mistakes anew.
I have not managed to simulate the M+S topology, so its a feeling nothing more at the moment. I fell off the simulation horse a while ago, so bear with my whilst I clear the rust and make old mistakes anew.
I am carrying on the design work but I still have a way to go.
"Never show unfinished work" and all that.
I wonder if we can put together a spec for a Devinyliser?
Going back to that important B&K document, and looking at Figure 4, I'm thinking that the important band for anti-phase removal is from 5 to 13 Hz.
But how much attenuation there is enough?
Obviously the more anti-phase rolloff slope the better, but from what frequency should we start? 50 Hz? 100 Hz?
What about some sort of elliptical filter that puts one or more nulls into the anti-phase passband?
(Come to think of it, such a filter might work very well as a conventional rumble filter without any anti-phase crossfeed. Not everyone will like a null in the passband, but you can't hear phase. Especially not at 10 Hz)
Your thoughts would be much appreciated.
"Never show unfinished work" and all that.
I wonder if we can put together a spec for a Devinyliser?
Going back to that important B&K document, and looking at Figure 4, I'm thinking that the important band for anti-phase removal is from 5 to 13 Hz.
But how much attenuation there is enough?
Obviously the more anti-phase rolloff slope the better, but from what frequency should we start? 50 Hz? 100 Hz?
What about some sort of elliptical filter that puts one or more nulls into the anti-phase passband?
(Come to think of it, such a filter might work very well as a conventional rumble filter without any anti-phase crossfeed. Not everyone will like a null in the passband, but you can't hear phase. Especially not at 10 Hz)
Your thoughts would be much appreciated.
An experiment I did try. I have four subs in my living room. Two on the left, two on the right. I can run them either stereo or mono with a quick switch.
I have hundreds of ripped CDs and dozens of ripped vinyl albums on my server to test with. I've lived with both mono and stereo subs for months.
Honestly, I can't hear any difference. They're crossed over 24db/oct at 100 Hz.
I have hundreds of ripped CDs and dozens of ripped vinyl albums on my server to test with. I've lived with both mono and stereo subs for months.
Honestly, I can't hear any difference. They're crossed over 24db/oct at 100 Hz.
B&K suggests 5-13Hz, although by the time you remove it electrically the damage has already been done in FM of the signal. The measurement from the EMT suggests 50-150Hz is the critical point as this is audible. I would like to do some more analysis on some actual records before committing and a few tests which requires me to get off the internet and actually wield a soldering iron, but if I were testing with DSP I would try from 5-200Hz and try and find the point it gets obvious then back down a bit.
FM is a good point- no filter can undo that.
B&K suggests that the worst freqs for this are 3 - 4 Hz
Sure, so long as we can also put together its alter ego the 'Vinyliser' !!
I'm sure it is as billshurv posts, the action is way above this band, in the lf audioband. The range he suggests is a good ball park IMO.
The antidote for this is mostly cart/arm damping IME. Long story and OT though !
Do you have any ripped vinyl that you are able to analyze to see what the out of phase content is? or that you can pop onto dropbox for us to munge around with?
Just read about leslie speaker on the web. It sure does give ambience from samples I heard on the Internet. That is almost like varying levels and varying low frequencies of antiphase noise giving a feel of spaciousness and ambiance. Isn't it ? Thanks for sharing. That is nice innovation.
I like design of Mr. Morgan Jones (respect) tonearm. Where he says we need least damping for up and down motion. This might contribute to the vinyl sound with respect to the point we are discussing.
-----
Another point. I know below 10hz is very low. But what about harmonics we might be amplifying that in phono preamp stage. Isn't it ? We can design a filter to cut off these various low level frequencies and hear on the fly. Sorry a measly graphic designer here with very little electronics knowledge. Kindly bear with me.
Regards.
Seems to me a reasonable, if perhaps incomplete, theory. Griesinger has done lots of work on the audio illusion. This paper might be of interest because it describes some of the components of spaciousness illusion and their psychoacoustic function.
https://www.google.ca/url?sa=t&rct=...sg=AFQjCNHgzHS3r9a2CEg71Ndxb4aXHBy2VQ&cad=rja
https://www.google.ca/url?sa=t&rct=...sg=AFQjCNHgzHS3r9a2CEg71Ndxb4aXHBy2VQ&cad=rja
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.