[glynnw]

Any progress on getting balance to work without increasing overall volume?

[kernelbob]

I prefer the current arrangement for balance control in the Tortugas.  Whether the adjustment is made to one channel or the other (as currently implemented) or if both channels are adjusted at the same time (one channel raised and one lowered), the changes will use the existing LDR attenuators.

As it stands now, this allows one channel to be changed by one attenuation step.  If a "no net volume change" protocol is used, then the minimum balance adjustment is two attenuation steps, either two steps left or two steps right.  This is because in order to maintain the same overall level, you have to increase the level of one channel and decrease the other by the same amount.

The result of that would be a more coarse balance control, which is exactly what I don't want.  The typical use of a balance control is to compensate for a slight difference in the gain of a left versus right amp's gain, the impact of an asymmetric room, a slight imbalance in a recording, differently aging tubes, etc.  The point is that L/R differences addressed by a balance control are typically slight and subtle.  Even the small change of one channel's level by one attenuation step may well be more of a change than desired.  Doubling the minimum L/R shift amount not what I would want to have to live with.

Best,
Robert

[glynnw]

I am hoping Morten can avoid this problem by some reprogramming of the step size magic.  If not, I agree with you.

[tortugaranger]

I prefer the current arrangement for balance control in the Tortugas.  Whether the adjustment is made to one channel or the other (as currently implemented) or if both channels are adjusted at the same time (one channel raised and one lowered), the changes will use the existing LDR attenuators.

As it stands now, this allows one channel to be changed by one attenuation step.  If a "no net volume change" protocol is used, then the minimum balance adjustment is two attenuation steps, either two steps left or two steps right.  This is because in order to maintain the same overall level, you have to increase the level of one channel and decrease the other by the same amount.

The result of that would be a more coarse balance control, which is exactly what I don't want.  The typical use of a balance control is to compensate for a slight difference in the gain of a left versus right amp's gain, the impact of an asymmetric room, a slight imbalance in a recording, differently aging tubes, etc.  The point is that L/R differences addressed by a balance control are typically slight and subtle.  Even the small change of one channel's level by one attenuation step may well be more of a change than desired.  Doubling the minimum L/R shift amount not what I would want to have to live with.

Best,
Robert

I've not focused on this of late but I think there's a middle ground approach that while not perfectly "zero net volume change" on every step would be close enough compared to current approach.

Say you want to shift balance to the left and make several steps in that direction for whatever reason:
step 1:  +1 to left channel (not quite a dB increase and when averaged with right perhaps no more than 0.4 dB total increase)
step 2:  -1 to right channel (now there's a 2 step overall differential but net volume change should be zero)
step 3:  +1 to left again (no change to right)
step 4:  - 1 to right (4 step difference but no net increase in volume)
etc.

Should work ok. Right? 

[glynnw]

That should work, but is a bit inelegant.  I was hoping for a simple 1 touch solution that would work like some of the balance controls that use paired reversed potentiometers to keep the level constant.  I may be an outlier, for I use the balance feature constantly, especially when listening to collections that were recorded 1 song at a time by different engineers. But Morten, I remain a true believer and love my Tortuga volume control as it is.

[tortugaranger]

That should work, but is a bit inelegant.  I was hoping for a simple 1 touch solution that would work like some of the balance controls that use paired reversed potentiometers to keep the level constant.  I may be an outlier, for I use the balance feature constantly, especially when listening to collections that were recorded 1 song at a time by different engineers. But Morten, I remain a true believer and love my Tortuga volume control as it is.

I'm thinking you may have misunderstood what I was describing. It is simple 1 touch. It works exactly how balance adjust works now. Press left to shift left. Press right to shift right.

The difference would be what happens internally which is what I was describing with those steps. It would alternately increase one side and then decrease the other to try and maintain a constant overall volume. As far as the user is concerned all you're doing is pressing either left or right button on the remote.

[glynnw]

Leave it as it is.  I believe what you describe would give a 2 or 4 step difference between channels, which is too coarse an adjustment.

I would be looking for something like this:  1 touch to left raises it by 1/2 step and lowers right by 1/2 step.  This gives a 1 step increase to the left with same overall volume.

Seriously, Morten, this is not a big deal to me.  Work on this when you have accomplished all your other projects (as if that will ever happen   ).

[kernelbob]

Leave it as it is.  I believe what you describe would give a 2 or 4 step difference between channels, which is too coarse an adjustment.

I would be looking for something like this:  1 touch to left raises it by 1/2 step and lowers right by 1/2 step.  This gives a 1 step increase to the left with same overall volume.

Seriously, Morten, this is not a big deal to me.  Work on this when you have accomplished all your other projects (as if that will ever happen   ).

I think you're missing a key point.  The attenuation steps are fixed at whatever number of increments that Morten has implemented in his software managed attenuation curve.  There is no "1/2 step" position in between the response curve steps.  If there were more steps, that would be great, but we would be back to square one of my point.  That is: whatever the number of attenuation steps are available, independent L/R adjustment to fine tune balance allows smaller, more subtle L/R shifts.  Adding the requirement to increase one channel simultaneously with the other channel, effectively throws away the benefit of implementing a more fine grained attenuation curve as it relates to allowing correspondingly smaller balance steps.

The LDR "steps" are actually discrete calibration points managed and maintained by the Tortugas software-- not continuous.  You can't just move the L/R balance an arbitrary smidgen as you can with the ganged potentiometers.  On the plus side, getting rid of all potentiometers and switches in the signal path is a key as to why the Tortuga sounds as great as it does.

One possibility is that Morten could add another configuration option (submenu #6?).  That option would allow the user to choose whether they want the L/R balance control to change the volume one channel only at a time as is now done.  The other choice would increase one channel's level a step and decrease the other channel's level a step, effectively shifting the L/R balance two steps, but increasing the coarseness of the shift.

For me, that isn't a priority, but then we all have different priorities.  For example, mine was the ability to switch absolute phase from the remote which Morten was able to implement in the LDRxB-V2 elegantly and without adding any additional sonically degrading complexity to the signal path (thanks again Morten).

Best,
Robert.

[kernelbob]

One last detail.  The attenuation step sizes in the Tortugas vary with volume level.  For example, near the maximum level (least attenuation) the step sizes are small than those as you approach the midipoint of the attenuation curve.  So, if you simultaneously lower one channel one step and raise the other channel one step, the combined volume level may not be exactly the same as before the change.  Similarly, if you lower just one channel near the top of the attenuation curve, the level change in that channel may not be the same dB as if you lower that same channel when listening nearer the midpoint of the attenuation curve.

[tortugaranger]

Leave it as it is.  I believe what you describe would give a 2 or 4 step difference between channels, which is too coarse an adjustment.

Sometimes I think kernelbob understands our preamp controller better than I do. Seriously, he's provided insights into how it does work and might work that have been very helpful.

But I've clearly failed to properly describe a potential fix to glynnw's original valid complaint re the channel balance algorithm (glynn you are incorrect in your interpretation).

Current Channel Balance Scheme
a) each balance adjust step increases that channel's volume by roughly 0.8 dB relative to the other channel which remains fixed
b) there are 9 total steps of adjustment available so a maximum of ~7.2 dB shift
c) although the channel balance does indeed shift, the total volume goes up as well because one side goes  up 7.2 dB (after 9 steps) and the other remains fixed with 0 dB change. Call it 3.6 dB average increase in volume.

Proposed Channel Balance Scheme
a) each balance adjust step will alternate as follows (let's assume we want to shift balance to the right channel):
    i) each odd step (1,3,5 etc) will INCREASE right channel volume by roughly 0.8 dB relative to the left channel which remains fixed. Avg volume increases by 0.4 dB.
   ii) each even step (2,4,6 etc) will DECREASE left channel volume by ~ 0.8 dB while the right channel remains fixed. Avg volume increase is now back to 0 dB.
b) there are 9 total steps of adjustment available (by design) so a maximum of ~7.2 dB shift
c) the volume remains almost constant alternating between +0.4 dB and 0 dB.
d) the only way to increase the granularity further is by increasing the overall number of attenuation steps from 70 to some higher number. let's say it was increased to 120 steps. given that the attenuation range remains fixed at 60 dB overall, each step size would then be 0.5 dB.

Just to be crystal clear here, in either scheme all the user knows and does is press the right (or left) button on the remote to shift the channel balance to the right (or left). The rest is done by the software under the hood.

[glynnw]

DOH!!  Now I get it.  Yeah, that would be good.

As an aside, I really enjoy the diversity of hearing we all have.  While I am seriously concerned about balance (proper balance can affect the tone), I never hear any change when switching the polarity.  Good luck to you, Morten, at keeping us all happy.

[kernelbob]

Excellent idea Morten!  I'm curious what the behavior would be when the overall level of both channels starts out close to the maximum.  As successive balance clicks cause one channel to reach the maximum level, I assume that successive balance clicks would simply continue to reduce the channel being attenuated followed by the next click having no affect on the channel that's already at maximum volume, then the next click reducing the attenuating channel, etc..  That would probably avoid an even more complicated software programming challenge, but I'm not the programmer (lucky me!).

[kernelbob]

glynnw,

Having the ability to switch absolute phase/polarity (some folks insist on one terminology or the other) from the listening position makes the effect much more obvious.  It's also important that the alternate settings have similar signal paths.  When the "inverted" setting of a component requires that the signal go through an additional switch or other amplification stage, the simpler path invariably sounds better.

Robert

[tortugaranger]

Excellent idea Morten!  I'm curious what the behavior would be when the overall level of both channels starts out close to the maximum.  As successive balance clicks cause one channel to reach the maximum level, I assume that successive balance clicks would simply continue to reduce the channel being attenuated followed by the next click having no affect on the channel that's already at maximum volume, then the next click reducing the attenuating channel, etc..  That would probably avoid an even more complicated software programming challenge, but I'm not the programmer (lucky me!).

Yes, when the volume is at either end of its range (a few steps above step 0 or below step 70) the channel balance feature runs out of differential headroom. For example if you're at step 70 (max volume) the only way to adjust channel balance is to reduce one side with respect to the other. So one side stays at 70 and the other might be at say 67. No real way to maintain neutral volume level when hard up against either limit. I honestly don't see this as a concern since it's fairly rare that any one will operate at either end of the attenuation range (0-5 or 65-70) AND also need channel balance adjustment.

[craig sawyers]

For me, phase reversal is more relevant to me than balance (I've never used a balance control. Ever. On any preamp I've owned over 40-odd years)

Craig

[Randy]

DOH!!  Now I get it.  Yeah, that would be good.

As an aside, I really enjoy the diversity of hearing we all have.  While I am seriously concerned about balance (proper balance can affect the tone), I never hear any change when switching the polarity.  Good luck to you, Morten, at keeping us all happy.

I had a CD player from Mark Levinson that had a polarity switch that could be done via the remote. I don't think I ever could really hear any difference between polarities, but sometimes you can talk yourself into hearing anything you want.

[kernelbob]

I had a CD player from Mark Levinson that had a polarity switch that could be done via the remote. I don't think I ever could really hear any difference between polarities, but sometimes you can talk yourself into hearing anything you want.

Your post reads to me that you don't believe the difference in absolute polarity is audible.  Well, every time I've demonstrated it to even total non-audiophiles, they can hear the differences and hear the same kinds of differences in bass power and extension, improved tonal balance up through the midrange, and focus.  Now, you do need to start with a system that's capable of good resolution and speed, especially in the bass.

[glynnw]

While I have never heard a change in sound due to polarity, I wouldn't mind readdressing the question.  I am positive many people are noticing polarity as well as other aspects of sound that bypass my brain.   I probably should simply be happy that the set polarity of my system works well for me, but if I do begin to notice polarity differences, it will give me another area to worry about, in true audiophile fashion.  Is there a simple device I can insert in my system that will give me remote control of polarity?  I am using a Tortuga non-balanced control which feeds 2 things -1.  a DSpeaker controlled dual sub system and -2.  a tube buffer feeding a 2 channel amp. I think with a little soldering I can come up with something but if a simple inexpensive small box solution exists I'd be interested.

[Elizabeth]

I can hear polarity a lot easier with dipoles!
With a positive polarity the presentation is 'forward' and slightly in front of the speakers.

With a negative polarity, the music appears to be coming from behind and around the speakers.
Easy as pie... With dipoles.

Also bass is slightly more coherent with a positive polarity on dipoles like my Magnepan 3.6

With box speakers I find it very iffy to know polarity.

[kernelbob]

I'm attaching a post from another Tortuga thread where I described the sonic qualities that I hear in my system-- Von Schweikert VR100XS speakers biamped from a Tortuga LDRxB-V2 with the balanced option and a Tortuga LDR1B ahead of the bass amp for level matching.

Following up with the differences between the alternate polarity settings that I hear in my system.

There's no way of knowing whether the "normal" or the "inverted" setting in a playback system corresponds to the original performance's polarity.  There are many components in the recording and reproduction chains.  So, I'll refer to the differences that I hear between the alternate settings as Left and Right.  The easiest and fastest way to identify which is which in your system for any particular recording is the switch the polarity setting using your remote from the listening postion.  I consistently identify a set of characteristics of the alternate phase/polarity settings that span recordings, labels, and genre as follows.

1) Bass.  in the Right setting, the bass has better weight.  It also extends more deeply-- this is more evident on some labels than others since some labels just don't have much really deep bass.  Hall ambiance is well defined and more distinct.  The bass is tighter and clearer.  Pianos have more weight and power.  Orchestral recordings are especially easy to distinguish the differences between Left and Right.

2) Lower midrange.  Bass harmonics are richer.  The tenor voice of a piano has a better balance with the Right setting and can even go missing with the Left setting.  Transients are more integrated.  Check out the sound of plucked bass in jazz recordings.

3) Midrange to upper midrange.  There's an odd quality that appears in the Left setting.  The middle to upper midrange starts to sound a bit to forward relative to the rest of the frequency range.  Also in the Left setting, the upper midrange starts to take on a degree of hardness.  Both of these qualities are corrected in the Right setting.

4) Treble.  With the Right setting, all the improvements in the bass to upper midrange are combined with a more balanced treble.  Much of the treble information are the harmonics of fundamentals in the bass and midrange.  So getting the bass and midrange right allows the treble to sit comfortably on top.

One point to keep in mind.  Our hearing isn't symmetrical when it comes to the compression and rarefaction components of a sound wave.  It's actually a bit more sensitive to rarefaction.

Best,
Robert