Step by step surgery

[fas42]

Just reading this thread,

 

 

, I was reminded of the very common bottleneck of getting volume control working well, when you drive the power amp directly with the DAC. Audio friend up the road has been 'fighting' this for years - and each time I visited him it was a "near but no cigar" situation, for one of his rigs most of the time.

 

Analogue volume adjustment and digital source are not good partners - the flaws of the physical nature of most level control mechanisms do a nasty job of highlighting weaknesses in the integrity of such a chain ... so, it may come down to expending a lot of effort getting this right - the difference between getting it working well, or not, is quite huge in subjective terms.

[fas42]

Just had to put a pic up of this, for those who like the idea of modding stuff - think your soldering skills are up there with the best? This is an opamp, four terminations, one on each corner, and one more on the bottom ...

 

[fas42]

Just came across this post,

 

This is as good an example as any of where the belief that conventional measurements "tell you all that you need to know" just doesn't cut it ... the Topping brand is touted as a value for money solution for getting optimum numbers - but it doesn't deliver, for many people.

 

Why I put this in this thread, is because this is probably something that can be sorted by some "step by step" working through with a sample unit - working out where in the engineering of the overall some key lacking occurred, which may, or may not, be able to be rectified by one or two low cost mods ...

 

[fas42]

I will respond to this post,

 

here, because I don't want to come across as being negative on that thread.

 

Frankly, this is well short of being good enough. Using MDF as a base is better than nothing - but one could do far, far better. What one needs is mass, lots of it - that's as inert as possible. Personally, I would use the heaviest piece of stone/concrete that I could muster - if I wanted it to look fancy, I would use thick as possible polished marble, say - the closer I could get to it being equivalent to resting the speaker on top of a virtual rock ledge that the house had its foundations set into, the better.

[fas42]

This post,

 

is a good example of what I Iook for, when making changes. Incrementally, the negatives in what I hear are attenuated - the result for this member is a perfect description of what a Subtracting Badness step does - in most cases, the number of remaining Badnesses are not infinite 😀 - so, it's just perseverance, and patience, pinpointing the last gremlins, and doing something about them that gets the desired outcome.

 

The ear/brain doesn't need perfection - but it demands a certain level of attenuation of irritating anomalies to trigger the illusion presentation. Less "perfectly" recorded tracks provoke the playback chain to add more annoying, obvious artifacts to what you are trying to connect to, emotionally - and your mind says, "This sounds like crap!!" 🙂. The amazing thing is that the brain is able to discard the annoyances at a certain, critical point of replay SQ - for any particular recording - and the "magic" switches  on 🙂. This would be very hard for many people to accept - but if one has experienced this over and over and over again, then it just becomes part of the landscape. ... Especially note, just because you have heard a certain recording come over well, say, 20 times in a row; if you then hear it on a sub-par setup, all the previous attuning to it will be for nought, and it will sound "Bloody awful!!" - this is how it becomes easy to assess the capabilities of an unknown system ...

[fas42]

This post,

 

is as good as any as an example as any of how many audiophiles approach this hobby, activity, the completely wrong way, IME. He wants to "tame the highs" - which is code for there being far too much distortion in the replay, which is particularly obvious in the treble content of the recording. And that there are 'magic' devices out there, which "solve everything!" ... by adding some specialness to part of the chain, which makes all the bad stuff go away. Well, it may work, help, but it will be a fluke if it does this decisively ... it's the lack of greater insight by most in the audio game which holds back the possibility of achieving really excellent SQ, in many cases - the thinking that one brings to the process of optimising the sound is so, so important; and one that doesn't directly tackle the underlying issues will mean that it could take a long time to get satisfying results.

[Archimago]

3 hours ago, fas42 said:

is as good as any as an example as any of how many audiophiles approach this hobby, activity, the completely wrong way, IME. He wants to "tame the highs" - which is code for there being far too much distortion in the replay, which is particularly obvious in the treble content of the recording. And that there are 'magic' devices out there, which "solve everything!" ... by adding some specialness to part of the chain, which makes all the bad stuff go away. Well, it may work, help, but it will be a fluke if it does this decisively ... it's the lack of greater insight by most in the audio game which holds back the possibility of achieving really excellent SQ, in many cases - the thinking that one brings to the process of optimising the sound is so, so important; and one that doesn't directly tackle the underlying issues will mean that it could take a long time to get satisfying results.

 

Sometimes, "tame the highs" is just "tame the highs" depending on the system. Maybe a little EQ would be more effective and easy to refine to taste.

 

I agree that there's no need for any magic devices like this. But also no need to interpret the anomaly as "distortion in the replay" is there? Unless one can show actual distortion present...

[fas42]

9 minutes ago, Archimago said:

 

Sometimes, "tame the highs" is just "tame the highs" depending on the system. Maybe a little EQ would be more effective and easy to refine to taste.

 

I agree that there's no need for any magic devices like this. But also no need to interpret the anomaly as "distortion in the replay" is there? Unless one can show actual distortion present...

 

The key words there are as bolded ... if the system status is such that the high frequencies are disturbing to listen to, then the playback has a problem. A trivial example is a very ordinary car radio, which has a 'tone control' - something comes on with lots of treble, and it sounds awful! So the passengers yell at the driver to "turn it down!" - and he winds back the tone control; which is merely a treble cut circuit - ahh, nice and cuddly sound!

 

The problem was too much distortion - simple fix, dump the HF ... well, it turns out even the most high end replay rig can suffer the same subjective issue, though it may be far more subtle in its impact.

 

So, DSP it away - but have you really solved things? IME, no ... a recording which is unbearable on some rig, with razor sharp edges tearing at your ear drums, can be played on a competent setup at any volume, with complete ease in the listening - your brain is balancing, 'EQ'ing' what you hear; it even adjusts for the non-linear distortion within the recording itself, if not too severe ... this is where the 'magic' happens; exploiting how human hearing can compensate for deficiencies in the source, if you don't overtask its abilities to do this.

 

 

[Archimago]

On 1/30/2021 at 8:53 PM, fas42 said:

 

The key words there are as bolded ... if the system status is such that the high frequencies are disturbing to listen to, then the playback has a problem. A trivial example is a very ordinary car radio, which has a 'tone control' - something comes on with lots of treble, and it sounds awful! So the passengers yell at the driver to "turn it down!" - and he winds back the tone control; which is merely a treble cut circuit - ahh, nice and cuddly sound!

 

The problem was too much distortion - simple fix, dump the HF ... well, it turns out even the most high end replay rig can suffer the same subjective issue, though it may be far more subtle in its impact.

 

So, DSP it away - but have you really solved things? IME, no ... a recording which is unbearable on some rig, with razor sharp edges tearing at your ear drums, can be played on a competent setup at any volume, with complete ease in the listening - your brain is balancing, 'EQ'ing' what you hear; it even adjusts for the non-linear distortion within the recording itself, if not too severe ... this is where the 'magic' happens; exploiting how human hearing can compensate for deficiencies in the source, if you don't overtask its abilities to do this.

 

 

 

But that's touching on my point (bolded). How do you know there was too much distortion? Obviously not all recordings are of good tonal quality. Some have "razor sharp edges" baked into the signal itself and an accurate reproduction of the sound would let listeners know that the recording was simply of poor quality.

 

Has there been consistent demonstration that the situations you're alluding to actually are primarily distortion-related? Not saying it's not a possibility, rather I'm not sure I've come across this much...

 

[fas42]

3 hours ago, Archimago said:

 

But that's touching on my point (bolded). How do you know there was too much distortion? Obviously not all recordings are of good tonal quality. Some have "razor sharp edges" baked into the signal itself and an accurate reproduction of the sound would let listeners know that the recording was simply of poor quality.

 

Because you've heard that recording sounding much better ... every time you hear a particular track, and it sounds different from how it sounded on another system,then at least one was distorting the playback. An accurate reproduction means you only hear the recording, which is the music as it existed in front of the microphone, etc - plus, the recording chain distortion ... and it turns when you hear this , and this only, the brain can sort out what belongs to the music, and what doesn't ... this is the 'magic' that happens when replay is good enough.

 

If one has never heard this happen, then it makes sense to disbelieve. Unfortunately, objectivists do hear this happening, and disregard it, believing that "the rig is distorting the SQ to make it sound better" ... ummm, no - if one does it carefully, then it's easy to demonstrate, to oneself, that in fact the accuracy has improved ... and this is what causes the subjective change.

 

3 hours ago, Archimago said:

 

Has there been consistent demonstration that the situations you're alluding to actually are primarily distortion-related? Not saying it's not a possibility, rather I'm not sure I've come across this much...

 

 

It's largely distortion in the sense of being noise related, from a variety of sources - I have not yet come across any experiments or papers, that I recall, that have properly investigated this ... something still in the To Do basket ... 😉.

[fas42]

Good post,

 

This reinforces understanding how important it is to maintain scrupulous electrical 'cleaniness' with digital replay - the ongoing 'battle' is to ever improve the integrity of the overall setup ... because the wins are so great! 🙃

 

Digital reproduction can be staggeringly good - but one has to be fully aware how fussy one has to be ... "near enough" is miles from what's needed to get the best subjective results.

 

"The greater the potential of the hardware, the more effort is needed to extract that full potential" ... that rule is embedded in thick concrete, at the moment - it will be a while before every good value for money device will deliver close to its best straight out of the cardboard boxes, unfortunately.

 

 

Interference effects are a curse in digital, I'm afraid 🤒 - as the 'transparency' improves, it becomes easier and easier to hear the impact of every tiny electrical anomaly; every one of these has to be knocked over to get the best subjective performance - which is intense, fully convincing, immersive experiencing of the recorded musical event.

[fas42]

This post,

 

again emphasises what makes getting the best SQ hard ... first one needs components which are intrinsically capable of greater resolution; and these show what is possible - but often lack the refinement which makes them robust against interference and other environmental effects. So, the frustration often is that "everything matters!", when it shouldn't, but this is purely because it takes extended development of the product to make it have the highest level of integrity. Which the consumer then has to pay for ...

 

The workaround, currently, is to accept these 'shortcomings' - and do the step by step tweaking which largely eliminates them ... this will be cheaper, and just as effective if well done.

[fas42]

After a bit of a longer break from doing this, just has a solid listening round with the audio friend up the road, N. . And it was a good one, with a high standard overall. Mentioned on another thread that the main work done in between had been on improving coupling caps on his two Naim amps - turned out that the original ones, some awful tantalum things, had been ditched ages ago - and he had been using good quality bipolar electrolytics in their place. So, the next, latest move had been to replace with back to back Cerafines ... which seemed to doing all the right things! 😉

 

The most noticeable thing for me, was how much  some of the CDs I brought over sounded precisely like the Edifiers, in current status! For those who think different rigs must have different signatures, this round highlighted that very different configs should end up meeting in the same place - at the content of the recording. Tonality, space, separation were so close - yet, one was modern digital actives using class D amps; the other a chain of separates, quite old DAC, class AB amp, and speakers ... turn your back, you wouldn't know which was playing, type of thing. Which of course is how it should be ...

 

Any issues? Yes ... connections are still a problem, refreshing them cleaned up the SQ when the presentation started to degrade, in several areas - will be looked at in the next working bee by N. . Also that old bogey of some sort of, possibly, static buildup in the DAC was audible - resetting the unit by pulling the power plug, and letting the power supplies discharge was needed several times to refresh the quality.

 

But, highly satisfying - no matter what was thrown at the two systems, they worked. The "bad CDs" I brought were to no avail; they came over as well as I could have expected ... another strike against the SRBs, 😝.

[fas42]

Something to add ... N. uses normal consumer grade DVD players, as one type of source for the main rig, with Toslink talking to the DAC - same method as I use for the actives. I say players, because he now has two - picked up one since last time at an opportunity store for a few dollars - both connected to the D/A unit. The new, backup unit was definitely not as good as the previous one, which he demonstrated early on.

 

Something that registered later during the listening was that the one that he had had for some time did a poor job when sourcing from burnt CDs - flat, unconvincing presentation. And he remembered that a couple of sessions ago the same happened with another CDR of mine - right, aha!! moment ...

 

Well, what about the other DVD player? And, you guessed it! It did a better job with written CDs than the other one - not decisively so, but certainly on one key disk ... yet, it did poorly on commercial CDs. The obvious thought is that the variation in laser, etc, mechanisms was enough to favour one read assembly in terms of the ease with which it could pick up what was on the disk - depending on the nature of the pits.

 

To complicate understanding, my own DVD player is not so brilliant at reading - depending upon everything, it struggles some times; an irregular tick starts occurring in the sound, which may or may not go away, as the tracks are gone through. Yet, this seems to do nothing to degrade the SQ - that remains consistent as far as I can tell.

 

Which says what? Well, that coming up with simple rules on what works better is probably not going to happen - every rig will be unique, and the strengths and weaknesses of the parts of the whole need to be worked out - and then follow the path of least resistance to getting a setup which does more things right; enough for satisfactory sound.

[fas42]

In case @Abtr  looks over here, to respond to his question,

 

 

No, everything that matters happens in the playback area of your system - what's being streamed is data over the net, and the precise timing of chunks of it to the app on your streamer will vary, per service - and how it's then processed by the server app will be difference, because different protocols will be used ... the "noiseness" is related to exactly how the data is buffered, and finally assembled to be fed to the DAC - the pattern of electrical activity in your listening environment will vary, and depending upon everything that could impact the SQ.

 

This is why the analogy of the CD transport was used - what varies is the electrical noise of the reading assembly; not the noise of the device that produced the CD master.

[Abtr]

So what you are saying is that the Tidal client software (app) may be audibly more noisy than the Qobuz client software (or vice versa). I seriously doubt that.

 

The timing of the USB data packages (chunks) is standard 8 KHz and doesn’t vary. What varies is the amount of data per package, which is determined by the USB receiver of the DAC which requests more, or less data per package in case of buffer under, or overruns respectively. This is not determined by the server. Furthermore, with lossless streaming, the data is simply relayed. Both clients will not perform DSP and both Tidal and Qobuz use FLAC as a data container. Finally both clients use the same USB kernel functions for a given OS (e.g. Mac or Windows) to drive the DAC’s USB receiver.

 

Hence, I don’t see how one client could produce so much more noise or such a different noise profile that (according to @AudioDoctor) the difference is easily audible in the analog domain. Indeed, that is not my experience. In my system, Tidal and Qobuz running in exclusive lossless mode sound exactly the same when the same file is streamed. If there is a significant audible difference then my guess is that something else is going on such as a different master is played or the clients are differently configured. For example, The Tidal client has the nasty habit of resetting exclusive lossless mode to system controlled mode after installing an update. This actually does introduce an audible difference, although even then the difference is not *easily* and *immediately* audible to my ears.

[fas42]

The clients being differently configured is the type of area I'm thinking of ... the 'weirdest' subtle differences can occur even on the simplest devices - that audio friend of mine demonstrated on a couple of occasions with a DAP that the precise sequence of keystrokes, the exact method of selecting a track to be played, varied the quality ... what??!! 😵

 

The software within the device is operating in a slightly different mode, is positioned in its memory in not quite the same way for the different scenarios; on a device that is sensitive to such - is a possibility for this behaviour. Whether it's an audible factor in a particular setup will be very hard to predict - but is a plausible explanation.

 

Masters being very slightly different, or that the FLAC compression data may not be 100% identical, even though when decoded they are then bit for bit the same - are other things to consider.

[Abtr]

Well, @AudioDoctor tried various DACs and streamers that apparently all suffer from this *easily* and *immediately* audible difference between Tidal and Qobuz 'lossless' streaming. He also tried galvanic isolation to no avail. To be honest, I find this hard to believe. I think something else must be going on. These software clients are not that architecturally different and (as said) they use the same OS kernel functions to process USB streams.

 

Anyway, in my system all electrical noise problems are solved by simply installing a decent USB to Toslink DDC before the DAC. 

[AudioDoctor]

4 hours ago, Abtr said:

Well, @AudioDoctor tried various DACs and streamers that apparently all suffer from this *easily* and *immediately* audible difference between Tidal and Qobuz 'lossless' streaming. He also tried galvanic isolation to no avail. To be honest, I find this hard to believe. I think something else must be going on. These software clients are not that architecturally different and (as said) they use the same OS kernel functions to process USB streams.

 

Anyway, in my system all electrical noise problems are solved by simply installing a decent USB to Toslink DDC before the DAC. 

 

All anyone needs is a Topping D10 to convert USB to Toslink... Gosh where were you years ago, you could have saved me so much money...

[Abtr]

45 minutes ago, AudioDoctor said:

All anyone needs is a Topping D10 to convert USB to Toslink... Gosh where were you years ago, you could have saved me so much money...

Yeh, I know what you mean. I own a lot of audio equipment that didn't live up to expectations. But I can recommend the Topping.

[fas42]

N. rang just before - we could have had a get together this afternoon; but I'm tied with doing other stuff ... 'til next time!

 

Anyway, he decide to pursue the path of doing something about the mains power that the main living area setup is using - from me saying I was thinking of using the inverter bought just to do back up for the house, if the power dropped out. A low powered unit was purchased by he, and via a route I'm not 100% sure I have a handle on, feeds mains to the Quad CD player - which now has been completely disabled, apart from the DAC; it's effectively, a stand alone converter, only. And, feels, he's getting best digital sound ever!

 

We've been here before 🤒 - so will wait for a listen to see how it's doing ... definitely has changed the sound, as most would suspect. But it's doing experiments like this that advance one's understanding - even if it turns out to be a bit of a backward move, one's always learning ... 🙂.

[fas42]

Bob posted this comment, about mains filtering, here

 

The key bit,

 

Quote

I have also done some basic testing with their power cords.  In my last test I used three cords to the Farad3 feeding my DAC and the Puritan cable was better.  I am sure that I could make this even better, but when does it stop?  Is it madness?  Puritan has some new power cords.  I am thinking about working with the distributor to test a couple.

 

No, it's not madness. And it stops when you can't hear anything change when you add something new ... it's all pretty simple, really ... 😉.

 

OK, you have a system already with some mains conditioning, filtering; that's built up over the years - pull all of it off in one go; put all of it back on again in one go; if you hear absolutely nothing change from doing this, and listening to the two states of the system, then why are you using it? Next step, switch some really nasty electrical noise makers in the house on, preferably plugged in adjacent to the socket used for the system - repeat the mains cleaning circuitry out, and back in, exercise - if you hear absolutely nothing change, then I repeat, why are you using it? 😁

 

If you do hear changes - as typical - then it's doing some good - and, there is definitely a length to this string - you keep improving the filtering that you use until it becomes impossible to hear any benefit from anything more being tried, or added. It has now become "sufficient" - at least for the current level of 'transparency', 😜.

[fas42]

Just came across this post,

 

which is an excellent example of how the behaviour of the system gives one clues about problem areas.

 

How so? Okay, he says,

 

Quote

... the differences are more audible through my full system than through my headphones

 

This is saying his "other DACs" are more sensitive to the components they are connected to, or to power supply noise generated by the operation of those components; almost certainly, the power amp. This is fairly typical misbehaviour - the much higher current draws of the amp cause stronger interfering electrical behaviour; when this is taken out of the equation by dropping back to headphones. the "other DACs" behave better; and the converter units converge more to a common quality.

 

Which means what? Well, either get a DAC that is more 'robust' - like, apparently, the Chord - or improve the mains filtering and other isolation strategies, so that the "other DACs", being less well sorted, can perform at a higher standard.

[fas42]

This is particularly pertinent ... because it is the story of digital ...

 

 

The paragraph to note,

 

Quote

I bought a Chord DAVE DAC only so I can do some power supply experiments and tests. Didn't like the stock DAVE at all. But after replacing the stock power supply, it got a lot better. Then I realised that feeding the DAVE with 16fs files made it sound better (as it bypassed the upsampling internally). And then I tried PGGB, which made it even better. The combination of all these 3 things turned the DAVE from an average performing DAC to something incredibly musical. It completely transformed the stock DAVE, which to be honest I have no idea why people even like. I'd say PGGB is a must for DAVE owners. And I can absolutely see how PGGB would boost up the sales of Chord DAVE DACs!

 

This is how it works, when aiming to get the best from something like CDs - one key thing wrong, and you are still way away from the finishing line! Everything has to be in place - or else you will keep wondering why you bother listening to the mongrel !!

 

It can be a thin slip of paper between "average performing", and "something incredibly musical" - it's knowing this that motivates one to keep trying things, to find the critical bottlenecks which prevent the good stuff coming through ...

 

[fas42]

Was just reminded by reading some posts, that an absolutely critical rule of thumb if one is aiming for the highest standard of subjective SQ, is that finessing a low cost, "what you already have", setup is vastly superior on a ROI basis, then outlaying on some new, "miracle cure" device or approach ... unless, of course, your thing is the fun of playing with expensive bling ... 😜.

 

Why? Because, the "new thing" is usually adding complexity, giving a greater chance of introducing completely new flaws into the chain - and overriding many of the gains that what you had previously made, by optimising what you had earlier ... it's classic, one step forward, three steps backwards - at times ...

 

Get what you have right now working as well as you possibly can!! Then, and only then, jump down another rabbit hole if you really, really feel there is an 'impossible' bottleneck, that can only be fixed by making major changes.