[mourip]

For instant channel balance and phase checks (hugely obvious & easy), I use this whenever I tear down & reassemble. Have it ripped to iMac as well.

Great tool. I have owned one for probably 25 years. I am always using the L/R track for when I screw up my interconnects!

For balance I pretty much have to use my ears as it can vary so much depending on my sinus congestion. TMI?

 

[johnjen]

Are there 2 types of listeners?

It could be said that there might be 2 types of listeners, those who can hear INTO the music, vs those who listen TO the music.
This might seem like a rather subtle distinction, one that may not seem all that important.
But it could go a long way in helping to explain THE major dichotomy within the world of audio, the primary split between the Objective-ists vs Subjective-ists.

This might help to explain, at least in part, why some insist that upgrades such as cables, fuses, etc, make no difference and so define them as “snake oil” or worse etc. And perhaps they haven’t yet learned how, nor what to listen for, and/or perhaps they simply can’t hear these subtle cues, and so for them these beliefs become ‘fact’.
As such, based upon their experience, they ‘Know’ and so proclaim, as fact, albeit often indirectly, that these very same subtle cues can’t be real.
Which is probably a true statement for them, but by the rules of logic, simply can't apply universally.
And of course a variety of supporting rationalizations and arguments that call on authority can and have been employed as well.

But our appreciation of music still boils down to all of our experience and the degree of Calibration that the individual has achieved. This is a form of learning, but not by the analytic portion of our minds, rather it involves the auditory processing portion which is directly linked (like our vision and smell etc.) to our awareness of what occurs in the moment.
IOW what we hear is live and direct into our awareness.

And this points directly at the differences between listening TO the music vs. INTO our music.
For myself sometimes when the music is playing in the background, I’m not even listening TO the music, but using it as a source for a pleasant distraction in the background while I do other things.
And sometimes it will grab my attention, forcing me to pay attention and listen TO the music.
And if by paying closer attention I am drawn INTO the music, well then the fun really begins.
Being drawn INTO the music is a learned trait which increases the depth, dimensionality of the experience, and so the meaning of the experience is enhanced.
But most importantly, it engages my ability to fully immerse myself INTO the music, experience the music as if it surrounds me, as it grabs me and sucks me INTO the performance and transports me there INTO the successively unfolding now moments, which IS the music.

I call this ‘connection’ my CNST (Central Nervous System Tap) where I get fully involved and engaged with and INTO the music itself.
As these experiences continue they perform a Calibration which allows me to know what is possible, what aspects of these experiences are worthy of further exploration AND they become compelling enough that I want to seek them out.
This also provides the motivation to tweak my playback system to determine if I can enhance this experience, a trend I refer to as striving towards REALNESS.

Calibration is based upon the process of listening INTO the music and gaining a greater degree of experience of SQ than what the listener has heretofore heard before.
Which translates to hearing more of what is possible, which leads to knowing WHAT you're hearing, which then feeds back to WHAT to listen for in the first place.
This is a form of a feedback loop which refines, delineates and helps us focus our ability to hear even further INTO the music.

And in the end, audio is an experiential based endeavor, meaning we decide if the SQ is acceptable, or not, if the changes made were an improvement, or not, solely based upon our experience of what we desire and hear. Gaining this experience has little to do with measurements, other than recognizing them as broad cursory strengths/limitations to the systems overall performance envelope.

But that isn’t to say that we can’t pay particular attention to aspects of this ‘live stream’ of music and learn to recognize changes in particular aspects of a ‘voice’ as I call them.
A ‘voice’ is any source of sound in a piece of music, be it a saxophone, drums (in a set), the human voice, a piano etc.

Each of these has a unique set of frequencies and ALL of their natural resultant harmonics, along with various room/effects.
And let’s not ignore the musicians gift at exploring and expressing all of what each ‘voice’ can truly provide, as this is where the ‘good stuff’ resides, and where much of the meaning we attain from listening INTO the music originates.

We as audiophools can and do vary these harmonic relationships by various means and methods and can choose to do so in such a way that the music can now express ‘more’ of what it inherently is.
AKA There’s more there, there.

This is where tweaking and refinements kick in, like eliminating ALL SMPS’s from the household electrical load (if possible) etc.
And for me, buying the gear is but the first step in integrating and optimizing the entire system, as there are many additional refinements that I use to help, which many would probably simply dismiss out of hand.

And that's great, for those who have found their version of musical enjoyment, but as such they will never hear what is truly possible out of a truly well refined and dialed in system, and perhaps they simply can't hear these sorts of audible subtleties, which is a great way to avoid such wallet draining circumstances in the first place.

And it certainly precludes Audio Nervosa from even gaining a foot hold.

To wit here is a revealing talk given the the CTO of ESS (DAC chip maker fame) skip to ≈ 19 minutes for the start of the directly meaningful stuff, and the really pointed subject matter begins at 25:30, (But the entire talk is interesting).


Yes all this time spent tweaking is significant and involved and complex and involves the need to merge multiple technical capabilities all at the same time. IOW it’s a complicated mix of different kinds of technical know-how that the DIY’r and tweaker brings together to impact the operation of our systems. This process of experimenting, learning and discovery can take many years, but then this is a hobby and the time spent is ours to freely give.

And the resulting changes and improvements are for us to appreciate.
These changes, that allow more of the music thru the system which is then presented to us to hear, are why we pursue this course of action.
Because when we hear into the music we are rewarded with a degree of added musical enjoyment and no amount of nay saying can ever negate nor change these experiences.

And in the past the opinion has been proffered that how could I, as a lone experimenter without the resources of the factory behind me, improve upon their cumulative expertise in the 1st place…?
It’s a kin to the ‘so you think you know more than the designers, better than the manufacturer, better than all the really smart guys. Oh the (at least implied) arrogance’, argument.

Except that the manufacturer has to pay attention to many significant details, like having to meet federal and safety regulations, having to meet the marketing depts expectations, and the accounting department, let alone make a profit, and a whole host of other aspects to running a business, all of which we as DIY’rs can and do, more or less, ignore. Not to mention practicality and user friendliness and the demand for an aesthetic standard and being willing to have wires poking out of places they shouldn’t be etc. And to a certain degree the argument is correct, we don’t have all of the resources available, but we do know a few tricks the factory rarely employes for numerous reasons.

And we audiophools, being the step children of the audio world, are willing and able to explore and push those envelopes just to see what we find and learn for ourselves what does and doesn’t provide us with more ability to hear INTO the music and delight in what we find there.

JJ

 

[gefski]

Once we get INTO well recorded music played on equipment that goes after that REALNESS, we can't "unhear" those cues, Truth of timbre, texture, touch, graceful and natural dynamic flow, real instruments in space, the perception of hall sound even in "silence", are just a few things that matter.

And the amazing acuity of human hearing, even for those of us with "old" ears, allows us to pretty consistently achieve genuine improvements in our systems over the years.

 

[johnjen]

Yeah gefski, I agree, once those subtle nuances and acoustic aspects are heard and recognized you can't un-hear them.

And 'old ears' have the experience to know WHAT to listen to and for.
Besides as you have pointed out, it's not about frequency response or being able to hear all of that last octave in the top end.
It is about being able to resolve and then integrate the whole of the midrange (where the fundamentals are generated) and so enhance our ability to hear INTO the music, that is, at least for me, where the meaning and 'power' of music is found.

JJ

 

[johnjen]

So the system has settled down and stabilized enough so I decided to try to hear the differences between up sampling vs down sampling via my AOIP feeding my Jggy dac.

I have been using the 44.1 sample rate as the standard and if needed, down sampling anything required to that setting for some time now.

So I grabbed a few 192 and 176 files and ran them 'native', then down sampled them to 88.2 and again down to 44.1.

And the lower I down sampled the 2 high rate files, the better they sounded.
Which might sound a bit counter intuitive on the surface.

But the Jggy was built to maximize 44.1 sample rate files as it's priority, so feeding it these type of files would optimize it's operation.
Which is what I'm getting.

Now to be sure the differences are very small, like at the threshold of perceptibility or a bit above, but if I can tweak my SQ just by shifting my sample rate, that is like a free upgrade or mod just by adjusting my s/w settings.

JJ

ps I'm working on a 2-3 part paper on power delivery with color photographs and readouts and numbers and stuff.
All I'm missing are a few power measurements and some additional comparisons etc.

 

[gefski]

So the system has settled down and stabilized enough so I decided to try to hear the differences between up sampling vs down sampling via my AOIP feeding my Jggy dac.

I have been using the 44.1 sample rate as the standard and if needed, down sampling anything required to that setting for some time now.

So I grabbed a few 192 and 176 files and ran them 'native', then down sampled them to 88.2 and again down to 44.1.

And the lower I down sampled the 2 high rate files, the better they sounded.
Which might sound a bit counter intuitive on the surface.

But the Jggy was built to maximize 44.1 sample rate files as it's priority, so feeding it these type of files would optimize it's operation.
Which is what I'm getting.

Now to be sure the differences are very small, like at the threshold of perceptibility or a bit above, but if I can tweak my SQ just by shifting my sample rate, that is like a free upgrade or mod just by adjusting my s/w settings.

JJ

ps I'm working on a 2-3 part paper on power delivery with color photographs and readouts and numbers and stuff.
All I'm missing are a few power measurements and some additional comparisons etc.

From Mike Moffat May 2017
"The megaburrito filter is unique in the D to A world. It is designed for the reproduction of redbook material above all else"

I've been aware of this approach since before I got my Yggy in 2015, and remain thrilled with my 99% redbook rips. However, it does seem funny that "messing with" higher bit rate files as you describe doesn't seem to damage them. It would seem logical to process everything natively? Interesting.

 

[NorCal]

Probably a dumb question, with an obvious answer, but this is something I've wondered about.
If one was designing a power amp that would need a 1:1 transformer, so that the pre-rectified AC voltage is equal to the line voltage, why use a transformer at all? Seems like this would be a major savings in costs, and the weight of the amplifier. [I realize that some amps need to step up the voltages, but if not ... ].

 

[Golfnutz]

So the system has settled down and stabilized enough so I decided to try to hear the differences between up sampling vs down sampling via my AOIP feeding my Jggy dac.

I have been using the 44.1 sample rate as the standard and if needed, down sampling anything required to that setting for some time now.

So I grabbed a few 192 and 176 files and ran them 'native', then down sampled them to 88.2 and again down to 44.1.

And the lower I down sampled the 2 high rate files, the better they sounded.
Which might sound a bit counter intuitive on the surface.

But the Jggy was built to maximize 44.1 sample rate files as it's priority, so feeding it these type of files would optimize it's operation.
Which is what I'm getting.

Now to be sure the differences are very small, like at the threshold of perceptibility or a bit above, but if I can tweak my SQ just by shifting my sample rate, that is like a free upgrade or mod just by adjusting my s/w settings.

JJ

ps I'm working on a 2-3 part paper on power delivery with color photographs and readouts and numbers and stuff.
All I'm missing are a few power measurements and some additional comparisons etc.

Something I noticed with my system is as the sample rate goes down, the latency goes up.

With 44.1/16 my latency is 1.8msec, and using a 176.4/24 sample rate, it drops down into 900-999usec. IOW, latency has almost doubled when using 44.1/16.

It would be very difficult for anyone to sit down and tell me what sample rate they were listening to at any point in time, with say a 15 minute break in between As far as I can tell it would be a complete guess. You might notice some differences doing an A/B within seconds to a couple of minutes. But walking in cold to a room and listening to music, I doubt anyone would be able to duplicate the results repeatedly. I also doubt latency has any affect other than making sure the music is flowing without dropouts.

JJ, I agree with your comments on the 'Rednet' thread, that focusing on the latency isn't going/shouldn't improve anything with a simple two channel system - only in a network system where there's so much going on that the RN Pci/e card is a requirement.

 

[johnjen]

I agree, latency in the amounts we are seeing, in our very simple networks, seems to be a non issue, especially with readings in the ≈1ms ± range.

And I also agree that these differences, in just sample rates for instance, taken singularly, are rather slight and not night and day changes.
However…
All of these various tweaks when taken together, as the sum of all of them, results in rather obvious and very welcome and enticing changes for the better.

And my sub 1ms reading (997µs) is at 44.1ms and when I switch to 192KB files I get slightly lower (970's to 980's µs) readings.
So yes the highest sample rates do have lower latency readings, but not by all that much.
But it does make my imagination step up and try and figure what is happening all the same.

JJ

 

[johnjen]

snip
I've been aware of this approach since before I got my Yggy in 2015, and remain thrilled with my 99% redbook rips. However, it does seem funny that "messing with" higher bit rate files as you describe doesn't seem to damage them. It would seem logical to process everything natively? Interesting.

Messing with higher bit rate files has always struck me as a potential problem, and the best I could hope for was that there wasn't any degradation to the SQ.

This 'concern' was effectively 'put to rest' with the Jggy since it LIKES the lowest bit rate files more so than the higher bit rate files.
And granted not in every instance but the preponderance of those high SR files are not adversely affected by down sampling them to 44.1.
And the parentage of those high SR files is always a question, and often with uncertain or questionable answers.

This makes it much simpler to NOT have to fuss with switching sample rates on the fly while using my AOIP system (which is a bit of a PIA).
Instead I perform the SR down sampling in JRiver, and they do seem to handle all of this in DSP very well.

And I figure its ALWAYS better to discard data than to create data all based upon a mathematical model, especially when you can divide by zero and not collapse the known universe, doing so.

JJ

 

[johnjen]

Probably a dumb question, with an obvious answer, but this is something I've wondered about.
If one was designing a power amp that would need a 1:1 transformer, so that the pre-rectified AC voltage is equal to the line voltage, why use a transformer at all? Seems like this would be a major savings in costs, and the weight of the amplifier. [I realize that some amps need to step up the voltages, but if not ... ].

Very few amps use 60 volt rails as the basis for their design.
And transformers can act as a buffer of sorts between the ac source and the diode bridge, which is a good thing as it tends to keep the diode bridge from letting ALL of their smoke out and perhaps crispy crittering the board itself.

And this buffering can also limit the very fast high voltage spikes that can sometimes 'sneak' in, which would do bad things to those very same diodes.

Just a thought or 2.

JJ

 

[johnjen]

snip
JJ, I agree with your comments on the 'Rednet' thread, that focusing on the latency isn't going/shouldn't improve anything with a simple two channel system - only in a network system where there's so much going on that the RN Pci/e card is a requirement.

I really wonder about the reports of better SQ from the Rednet PCIe ethernet card.
But not enough to want to spend 1K$ to find out.

And really Rednet doesn't have SQ as the highest of priorities in the 1st place, which sorta makes me wonder what would happen IF they decided to optimized their h/w and s/w with that goal in mind.

It kinda makes me go hmmmmmmmmmm…

JJ
ps I just received an ac power meter and I made myself a pair of dummy loads to run my Mojo amp all the way up to clipping to monitor various operational parameters (peak current, current flow duration, power used, etc). Thus far the results I'm seeing are insightful and providing much food for thought.

 

[Golfnutz]

When Focusrite Customer Support receives questions about latency issues, including RN PCIE/R cards, I would imagine this is what their mind is seeing (pic below). Not some guy sitting at home in front of his 2 channel system listening to Pink Floyd being feed from a PC with a Rednet device.
Source - http://www.yamahaproaudio.com/europ...g/audio_quality/chapter5/05_absolute_latency/

Results of audio latency problems - Dropouts, clicks, and pops.
Source: http://www.resplendence.com/latencymon

Nothing I've found on the internet indicates the better the latency, the better the source will sound, or vise versa.

What do you think the answer would be if you asked a recording engineer what good latency sounds like?
I bet it would have nothing to do with the stuff we perceive as good audio (i.e. better soundstage, bass, imaging, etc.).

I'm not sure I agree with your priorities about SQ. Have you compared your computer source + Rednet (without any Mutec external devices) with a really good CD transport system?
If I want a point of reference, sometimes I use my Teac VRDS 25 as a transport only to my DAC, and it's amazing how close it sounds to my PC + D16 at 44.1/16.
How much different would you expect the SQ to be using the very best PC + AOIP system feeding a DAC, than the very best CD transport available today feeding the same DAC?

The other thing I forgot to mention is there may be more data coming from the source than we already know. By simply changing the amp, you may experience better resolution, details, imaging, bass, etc. that we didn't know already existed. Same for speakers or DAC's. IOW, all the focus shouldn't just be on the source.

 

[johnjen]

Yeah latency for us in our setups is a non issue as far as I can tell.
Our use of the Dante network is so simple and we are such a minor blip on their radar that I don't expect that SQ would ever be a priority for them, at least until they decided to add us as a market for them to address.

As for trying a CD player, I have a McCormick UDP-1 transport which I have used in a comparison with my (at the time) tweako USB feed.

The difficulty for me in using a cd player straight into my dac is, it bypasses my entire DSP setup in JRiver which is an integral part of my setup.
For example I have adjusted the channel balance by 1.5dB in the DSP stack, which alone makes a substantial difference in the very bottom end let alone the stereo image using speakers.
And that is but one of 6 other adjustments I am using to dial in my setup.

Even so while the CD player was nice and all, by itself it didn't measure up then and now with the substantial improvement that my tweaked AOIP setup provides, the gap is even greater.

As for Focusrite aiming at SQ as a priority, they are aimed at the commercial/professional audio market, which is in full evidence by their choice of the power supplies they use, (SMPS) vs LPS's.
They are WAY less expensive and create far less heat etc.
But are not optimal for SQ, not by a long shot.
This is but one indication of where we as audiophools would (and do) make modifications to improve the SQ of their gear with satisfying results.

I see the changing of gear, like in your example the amp, as a way of reducing or (hopefully) eliminating CP's (Choke Points), which in turn allows more of the original signal to pass thru the entire system and then be delivered to our ears.
But if the source alters the signal to begin with, there is very little that can be done downstream to 'correct' those alterations.
The same can be said for any link in the chain of course, but the signal source (and any transducers) are usually the most critical in this regard.

And any time a signal is transformed from one form to another (digital to analog, electrical to acoustic etc.) that is potentially where the greatest gains in terms of removing CP's is also to be found.
Assuming of course that these alterations can be identified and addressed directly, which isn't always the case.

JJ

 

[johnjen]

So my latest tweaks involve dithering and the new Dekoni ear pads for the HD800.

Another thread was talking about the use of dithering as a way of randomizing the noise generated by dacs when they choose the value for the LSB (Least Significant Bit).
Essentially this a mathematical 'trick' to help reduce noise created when reconstructing the analog waveform.
So I figured I'd play around with the 3 settings in Media Center that control the 2 types of dithering that can be added, or not (as in turned off).
And after fussing with the controls I settled upon the option of not adding any dithering as giving me the 'better' SQ.

This tends to run counter to the 'accepted' norm of using dithering to 'hide' patterns that can heard as added noise.
Except that what I hear after turning off dithering, is added detail and spatial cues that I deem as 'better'.

And granted these changes aren't big, indeed they are close to the ToP (Threshold of Perceptability) and as such many may not choose this option.
But for me it does add just that much more REALNESS.

And my next tweak is, when massdrop offered them, I ordered a pair of the Dekoni fenestrated sheepskin earpads for my 800's, which arrived yesterday, so I yanked my nearly new stock pads off and installed the new Dekoni's and have been listening to my 800s using them ever since.
And one mod that many seem to make is to cut out the inner dust barrier from these pads.
I'll make that mod after I acclimatize to the changes these pads make in stock form, and then swap back to stock pads to get a better understanding of the changes.

Thus far what I am hearing is a 'push back' as in the soundstage has receded away from me a bit.
Also the bass has a softer leading edge to it and it seems like I'm missing some details, most likely small signal aspects of the acoustic presentation.

This test, going back to stock after making a change, is needed in this case especially, due to the subtle nature of the changes these pads make.
And this test will help determine what is 'better' for me in my system.

But I can see that for some these pads may help tame aspects that 'stick out' and so this is may be a way of 'rounding off the sharp edges', so to speak.

And in my case my 800's are by no means stock.

Here are 3 examples of a nearly stock 800, my 'B' set of 800's and my primary 'A' set of 800's
These 3 examples are showing, with a greatly expanded in the time domain, leading edge of their respective square wave responses.
The time domain (the horizontal axis) has been expanded to near the resolution of the bit depth of the adc used to capture the analog signal from the microphone.

First up is the 20Hz square wave used to drive the amp.



Next I'll show all three graphs in close proximity, then explain what your looking at.





The top graph is my majorly modded Gen-4 800's.
These are hardwired and are the 4th generation of a series of tweaks I've been working on for a while now.
They use the SAA balanced cable with a rhodium 4pin XLR connector.

The middle graph is my early 'B' set of 800's that have minimal mods that include a cryo'd mondo gauge set of balanced cables (that will be hardwired during the next round of mods) that also have a rhodium 4 pin xlr connector, and the protective screens on the outside of the driver have been removed.

The bottom graph is a set of 800's with the Anax mod and that have been hardwired as well.

One way to view these graphs is to follow the line (the trailing edge) that descends off to the right, and extend it up and to the left until it reaches the leading edge of the square wave signal.
Notice how that line changes it's relative position with respect to what a square wave should look like, when you compare these 3 different responses.

Another way to perceive these graphs is to see how much 'overshoot' and 'undershoot' exists on the leading edge of the response of the diaphragm to the square wave driving it, with respect to that line that descends off to the right (the trailing edge of the response) if it were continued all the way to the leading edge the same as the description above.

I'm close to adding the next gen of mods to both my primary ('A' set) and my 'B' set of 800's, and then taking more measurements to see what results this next attempt will yield.

I'm hoping it'll be Gen-5.

JJ