High End Tone Control

[ThorstenL]

Hi,

Originally Posted by PierreQuiRoule

For width expansion only, one could do without the blend part and add a pot in the cross talk cancelation circuit. Clean and elegant.

Yes, indeed.

The typical "Ghetto-Blaster" in the 80's had a switch for "mono/stereo/wide", mono and stereo being obvious, wide simply connected a resistor and capacitor between the inverting inputs of the output chip-amp's. It was usually overcooked, so I changed the resistors and capacitors in the various models I owned.

Originally Posted by PierreQuiRoule

As far as I can tell, though, the width expansion function requires a feedback loop or an active stage with differential inputs. Feedback and differential amplifiers are quite acceptable when done right. Yet I remain attracted by the daring german eq circuit which has zero feedback. It is so simple.

Actually, the "ZenEQ" could have a simple blend control added into the inverted output circuit (drains) while the crosstalk cancellation could be applied to the source inputs.

Ciao T

[PierreQuiRoule]

Originally Posted by ThorstenL

Actually, the "ZenEQ" could have a simple blend control added into the inverted output circuit (drains) while the crosstalk cancellation could be applied to the source inputs.

Yes, I see it now. This is applying the polarity-reversed blend signal back to the inputs through a resistor network. Filter to taste. Easy to add by those who want it. Others can leave it out.

[sreten]

Hmmm ......

I'm really not spoiling for a invective argument.

No issues with post #36 being a good approach.

Looking at L+R signals as M+S is mathematically valid whether
or not you actually physically create the M+S channels. Any
manipulation of width is easier to understand using the M+S
concepts of the L+R signals, and using M+S concepts it is
still easier to understand the mono constraints (for radio)
that have been and still are applied to nearly all recordings.

I'm not trying to say what I'm discussing is the way it should
be done, but that its food for thought, and it really doesn't
need to be denigrated with alleged OTT limitations.



This is a circuit that does create M+S channels, very 70's.
(Rear is S or difference, Centre is M or sum).

Shows you can combine buffering and sum and difference
with a small noise penalty (i'm assuming you'd use ~ 10K
rather than 100K), and gain is easy to include. If its not
obvious the same circuit converts M+S back to L+R,
depending on the EQ topology may need some gain.

1 "op-amp / gain stage" per channel is not doable, but 3 is,
or 2 if your talking fully passive EQ, can't have 1 for that.
Which is nothing like the 10 its been I'm allegedly suggesting.

rgds, sreten.

As there is a separate thread about the actual pre-amp to
be built I assume this thread is about the options, why or
why not they were chosen, but that its not intended to
make out the approach taken negates other approaches.
Their validity isn't something the point is to slag off.

[ThorstenL]

Hi,

Originally Posted by sreten

This is a circuit that does create M+S channels, very 70's.

...

Shows you can combine buffering and sum and difference
with a small noise penalty (i'm assuming you'd use ~ 10K
rather than 100K), and gain is easy to include. If its not
obvious the same circuit converts M+S back to L+R,
depending on the EQ topology may need some gain.

Yes.

Now, in order to work reasonably independent from source impedance and to not load the inputs heavily we either incur a heavy noise penalty by using high value resistors (compared to "brutally simple" # 36) or we need to add a buffer to be able to use low value resistors.

Equally, using the output M/S decoder for gain may be possible, but it may compromise operation, especially is limited to standard component values.

So, while this is an option for a tone control (as in it will pass a signal from in to out and allow control of tone), I find that it adds much complexity without any compelling advantages # for a high quality design. So as an option I would probably, personally place it VERY LOW on the list of options.

Originally Posted by sreten

As there is a separate thread about the actual pre-amp to be built I assume this thread is about the options, why or
why not they were chosen, but that its not intended to
make out the approach taken negates other approaches.

The other thread is about "lets make a Cello Palette functional clone".

This one is called "High End Tone Control" and aims to discuss and illustrate approaches to providing tone controls for situations where there are needed and desired, WITHOUT introducing designs that are subject to the usual, common and qite accurate criticism of tone-controls, namely that "they mess up the sound".

I would argue that the circuit in post # 36 answers this requirement excellently, if implemented with high quality passive components. It forms a complete line-stage with gain, with very decent noise and distortion performance if the active IC (or discrete circuit) is chosen with care.

In "defeat" position not even a switch contact is added to the signal circuit, in operational condition nothing but two switch contacts and the actual filter circuits are added and then in parallel with the original feedback loop, so again, their influence, especially when set at small amounts of boost or cut, is reduced.

If the filter circuits are arranged using stepped controls (which for highest quality are / may be required) and high grade film and film/foil capacitors the additional sonic footprint over the straight line stage is minimised.

With all due respect, I cannot remotely conceive any situation where the A/B to M/S -> Tone -> M/S to A/B system excluding additional circuitry can offer similar results and again, with respect, the resultant sonic impact would be unlikely to be as small as the minimalist scheme I proposed.

There may be very good reasons to employ such a M/S tone control system anyway#, however I cannot conceive it being able offer what is commonly understood as "High End Performance".

Ciao T

# The M/S tone control may be applicable in a situation where the fact that poor channel tracking does not lead to L/R image shift, but rather to an image width shift is appreciated.

Such a case may be one where even credible tracking performance linear potentiometers (costing around 50 Cent US) and 10% tolerance film capacitors are by far too expensive and instead it would be considered that adding some extra ultra-cheap (well below 5532 grade) is a good trade-off for allowing to use the cheapest dreck for tone controls and where the ability to then have a free "wide stereo" option is a boon, as is the availability of a mono signal for a sub etc...

That is I can see use in Lo-Fi or No-Fi ultra cheap systems, where the benefits may outweigh the problems introduced. Even at Nad/Marantz/Onkyo Mid-Fi levels of quality I cannot perceive any benefits.

The other reason I can conceive is to find a way to justify the use of even more Op-Amp's that D. Self uses in his 2012 Preamp in order to badly do the same the job that a single one can do well...

[ThorstenL]

Hi,

Originally Posted by sreten

This is a circuit that does create M+S channels, very 70's.

...

Shows you can combine buffering and sum and difference
with a small noise penalty (i'm assuming you'd use ~ 10K
rather than 100K), and gain is easy to include. If its not
obvious the same circuit converts M+S back to L+R,
depending on the EQ topology may need some gain.

Yes.

Now, in order to work reasonably independent from source impedance and to not load the inputs heavily we either incur a heavy noise penalty by using high value resistors (compared to "brutally simple" # 36) or we need to add a buffer to be able to use low value resistors.

Equally, using the output M/S decoder for gain may be possible, but it may compromise operation, especially is limited to standard component values.

So, while this is an option for a tone control (as in it will pass a signal from in to out and allow control of tone), I find that it adds much complexity without any compelling advantages # for a high quality design. So as an option I would probably, personally place it VERY LOW on the list of options.

Originally Posted by sreten

As there is a separate thread about the actual pre-amp to be built I assume this thread is about the options, why or
why not they were chosen, but that its not intended to
make out the approach taken negates other approaches.

The other thread is about "lets make a Cello Palette functional clone".

This one is called "High End Tone Control" and aims to discuss and illustrate approaches to providing tone controls for situations where there are needed and desired, WITHOUT introducing designs that are subject to the usual, common and qite accurate criticism of tone-controls, namely that "they mess up the sound".

I would argue that the circuit in post # 36 answers this requirement excellently, if implemented with high quality passive components. It forms a complete line-stage with gain, with very decent noise and distortion performance if the active IC (or discrete circuit) is chosen with care.

In "defeat" position not even a switch contact is added to the signal circuit, in operational condition nothing but two switch contacts and the actual filter circuits are added and then in parallel with the original feedback loop, so again, their influence, especially when set at small amounts of boost or cut, is reduced.

If the filter circuits are arranged using stepped controls (which for highest quality are / may be required) and high grade film and film/foil capacitors the additional sonic footprint over the straight line stage is minimised.

With all due respect, I cannot remotely conceive any situation where the A/B to M/S -> Tone -> M/S to A/B system excluding additional circuitry can offer similar results and again, with respect, the resultant sonic impact would be unlikely to be as small as the minimalist scheme I proposed.

There may be very good reasons to employ such a M/S tone control system anyway#, however I cannot conceive it being able offer what is commonly understood as "High End Performance".

Ciao T

# The M/S tone control may be applicable in a situation where the fact that poor channel tracking does not lead to L/R image shift, but rather to an image width shift is appreciated.

Such a case may be one where even credible tracking performance linear potentiometers (costing around 50 Cent US) and 10% tolerance film capacitors are by far too expensive and instead it would be considered that adding some extra ultra-cheap (well below 5532 grade) is a good trade-off for allowing to use the cheapest dreck for tone controls and where the ability to then have a free "wide stereo" option is a boon, as is the availability of a mono signal for a sub etc...

That is I can see use in Lo-Fi or No-Fi ultra cheap systems, where the benefits may outweigh the problems introduced. Even at Nad/Marantz/Onkyo Mid-Fi levels of quality I cannot perceive any benefits.

The other reason I can conceive is to find a way to justify the use of even more Op-Amp's that D. Self uses in his 2012 Preamp in order to badly do the same the job that a single one can do well...