[DarginMahkum]

How many new technical features DACs would need? If DAC chips would just do their simple job, we would simply have few cost-effective solutions, and that is it. In some sense, Apple USB DAC is along these lines.

Also, the fact that different DAC chips can be implemented in a similar way with the major standard parameters measured similarly does not need to be disputed.

Then I wish all DAC/DAP implementation would sound the same, alas, it is largely not the case in portable devices.

Is AKM "velvet sound" a gimmick?

My "conspiracy theory" is that harmonics and perhaps phase differences between the channels are subtly manipulated to sound more pleasing similar to lamp sound, as the most successful DAP manufacturer likely does for their special sound.
Before you menrion THD - if only the series of harmonics are processed and manipulated as a group (enhancing/depressing some n-th harmonics, etc.), it won"t be detected with a single-frequency probing.

Colouring the sound in a pleasing way is likely to be appreciated by consumers to enjoy the music.

There are some contradicting expressions often used by audiophiles:

- I want to hear the music exactly as it was produced, that is why I don't use DSP and EQ.

Then they go and buy some device with altered sound and say it sounds better. They have no clue about why it sounds the way it sounds (assuming it sounds really different), but when you ask them "then why don't you use EQ or a high quality DSP", they tell you they are against EQ because 1) EQ is bad 2) they want to listen to music as it is supposed to sound like.

Very contradicting behavior, especially when you consider EQ is a much more deterministic than trying to match "you have no clue what goes on in here" devices.

To me, this type of decision making is already questinable. That person probably 1) likes to socialize more than being an audio enthusiast. 2) is more interested in buying the next expensive gadget to confirm the marketing and fan club hypes floating around it. 3) is lazy and does not want to learn "new tricks" that can be solved by just spending a few hours of reading and experimentation. That is also in sync with being so much offended by measurement data, as one has to learn before talking about it. 4) likes spending money as a theraphy.

Of course, there's nothing wrong with all those. Everything you mentioned and things like stage can be easily tweaked with simple DSP in a much more deterministic and reproducible way. There are so many excellent VSTs for that, but no. It is simply impossible to convince someone who likes to have 5 devices with tweaked sounds to even try this method. Have seen it so many times.

So I question if the idea is really having a pleasing sound etc. or buying the next most hyped gadget. Otherwise if one buys the a technically perfect amplifier and DAC (assuming he is not looking for additional features like Roon endpoint etc.), and adds a good EQ / DSP on top, the possible options of sound, including that so called "pleasing sound" would be vastly larger than the number of DACs and amplifiers in the market for much less money.

PS: Portable devices are a different story, as in many cases the output goes through the headphone amplifier.

 

[gregorio]

The most probable cause for audible differences is the use of different reconstruction filters which can shape magnitude spectrum of the highest half-octave so much that a trained ear can notice it. While I have never done proper listening tests, I am very certain I can hear the differences between reconstruction filters at 44.1 kHz, but not at higher frequencies.

Maybe you should do some proper listening tests, I’m quite confident you can’t hear the differences between reconstruction filters, except in a few pathological exceptions.

Most DACs have a 44.1kHz reconstruction filter with the transition band starting around 19-20kHz, some though have a transition band starting around 18kHz or even 16kHz. However, a just noticeable difference at 16kHz is around 3dB but that’s with a pure tone (so no masking) at a high level (95dB if I remember correctly) and with young and perfect ears. Even taking these optimal conditions, you would need a slow roll-off filter with a transition band starting probably somewhere around 10kHz, so it’s at -3dB by 16kHz. There are such reconstruction filters but they are rare (sometimes as an option on DACs with switchable filters) and even then, with a musical signal, not so young ears and a more reasonable listening level you would struggle at the very least. I can’t say it’s absolutely impossible but it is very unlikely. The only exception I can think of would be a filterless NOS design at 44.1kHz, which typically start to loose HF response far lower than 10kHz.

I don’t think you would be so “very certain” after a DBT/ABX but it would be interesting to know for sure.

G

 

[71 dB]

Maybe you should do some proper listening tests, I’m quite confident you can’t hear the differences between reconstruction filters, except in a few pathological exceptions.

Most DACs have a 44.1kHz reconstruction filter with the transition band starting around 19-20kHz, some though have a transition band starting around 18kHz or even 16kHz. However, a just noticeable difference at 16kHz is around 3dB but that’s with a pure tone (so no masking) at a high level (95dB if I remember correctly) and with young and perfect ears. Even taking these optimal conditions, you would need a slow roll-off filter with a transition band starting probably somewhere around 10kHz, so it’s at -3dB by 16kHz. There are such reconstruction filters but they are rare (sometimes as an option on DACs with switchable filters) and even then, with a musical signal, not so young ears and a more reasonable listening level you would struggle at the very least. I can’t say it’s absolutely impossible but it is very unlikely. The only exception I can think of would be a filterless NOS design at 44.1kHz, which typically start to loose HF response far lower than 10kHz.

I don’t think you would be so “very certain” after a DBT/ABX but it would be interesting to know for sure.

G

As I said, if I really hear differences it is pretty insignificant in regards of music enjoyment, so it doesn't matter.

I also have to add, that the most differences I think I hear is with headphones.

 

[PhonoPhi]

I’m not sure they do have different beliefs about what “transparent” means, they all seem to implicitly agree that it means “fidelity”. Although I agree that some seem to prioritise certain aspects of fidelity over others and therefore have different implementations. However, this does NOT imply different “colourations” because these are all inaudible aspects of fidelity. Chord’s approach to filters would improve fidelity but only above the frequency threshold of human hearing. Topping’s approach improves fidelity but only below the volume/level threshold of human hearing. Benchmark’s approach seems the most rational, sacrificing some inaudible SINAD to make sure there are no ISP issues, which can be audible under certain conditions, although even cheap, half decently designed DACs have been able to deal with ISPs for many years, so there’s no audible fidelity benefits here either.

In all cases, we’re talking about approaches to fidelity/transparency that are beyond the thresholds of audibility, with the occasional pathological exceptions mentioned, “tube” DACs or some NOS DACs for example.

Isn’t it obvious that DAC chips/DACs need an infinite number of new technical features? Once all the other DAC chip manufacturers have the same feature set as your DAC chip, you only have two options: 1. Compete by lowering your price or 2. Come up with a new technical feature your competitors don’t yet have.

#1 is a poor long term option because it squeezes profit margins and leads to an unsustainable race to the bottom (a price war), which is bad for share price and can ultimately lead to insolvency.
#2 is an obviously better long term option because it maintains or even improves profit and can continue ad infinitum as long as marketing can keep convincing consumers there’s some actual benefit. For example, 16/44 is more than adequate for human hearing. Then we had 96/24, way beyond human hearing and not even reproducible by the best transducers, let alone audible. Then we went to 192/24, in some cases we’re up to 384/32, next will be 768/32 and so on ad infinitum, until even the most gullible audiophiles can no longer be convinced there’s any audible benefit but there’s no sign of that yet!

G

Very much true.
MQA-supporting DAC chips highlights and exacerbate this issue...

There are some contradicting expressions often used by audiophiles:

- I want to hear the music exactly as it was produced, that is why I don't use DSP and EQ.

Then they go and buy some device with altered sound and say it sounds better. They have no clue about why it sounds the way it sounds (assuming it sounds really different), but when you ask them "then why don't you use EQ or a high quality DSP", they tell you they are against EQ because 1) EQ is bad 2) they want to listen to music as it is supposed to sound like.

Very contradicting behavior, especially when you consider EQ is a much more deterministic than trying to match "you have no clue what goes on in here" devices.

To me, this type of decision making is already questinable. That person probably 1) likes to socialize more than being an audio enthusiast. 2) is more interested in buying the next expensive gadget to confirm the marketing and fan club hypes floating around it. 3) is lazy and does not want to learn "new tricks" that can be solved by just spending a few hours of reading and experimentation. That is also in sync with being so much offended by measurement data, as one has to learn before talking about it. 4) likes spending money as a theraphy.

Of course, there's nothing wrong with all those. Everything you mentioned and things like stage can be easily tweaked with simple DSP in a much more deterministic and reproducible way. There are so many excellent VSTs for that, but no. It is simply impossible to convince someone who likes to have 5 devices with tweaked sounds to even try this method. Have seen it so many times.

So I question if the idea is really having a pleasing sound etc. or buying the next most hyped gadget. Otherwise if one buys the a technically perfect amplifier and DAC (assuming he is not looking for additional features like Roon endpoint etc.), and adds a good EQ / DSP on top, the possible options of sound, including that so called "pleasing sound" would be vastly larger than the number of DACs and amplifiers in the market for much less money.

PS: Portable devices are a different story, as in many cases the output goes through the headphone amplifier.

True. I would not be sure about conventional band equalizers, but DSP should able to compensate or achieve desired coloriration with DACs and can do quite a bit to help adjust transducers.
Yes, the mantra of "no equalizing" is running strong with many, with some (like me), it is more lazyness/simplicity. I do have two USB dacs: one warmer, one more analytical, using the same DAC chips. Could I use just one with DSP - most likely.
At some point, I got about 20 USB DACs trying to match the colouration of my older DAP. From many perspectives, quite a bit of waste of money, but I learned few things and also became more open for different sound signatures.

 

[GearMe]

My conspiracy theory is that science is largely correct and the equipment manufacturers utilize misleading marketing, leading to many imaginary reasons owners feel their purchase is "special".

Of course, there is the real possibility that "audiophile" brands intentionally color their output. Not sure why anyone would pay more for less transparency rather than applying EQ...

Personal preferences -- it sounds better to them...improving their individual listening experiences. Which is cool.

After all, it's their $$$ they vote with and their personal listening preferences they're catering to; preferences that are both real (i.e 'Pathological' Tubes, Transparent, basshead cans, treblehead cans, neutral cans, bass boost, crossfeed, 'mis-applied' EQ, etc.) and imagined (hi-res, cables, etc.).

Some of us enjoy both 'Transparent' as well as 'Pathological' systems; attaching to them a variety of cans with different sound signatures...switching configurations up for any number of 'illogical' reasons.

BTW...you don't have to shell out mega-bucks for solidly built 'Pathological' amps, etc.

Now...'Imagined' differences? That's another story! One should expect to shell out significant coin for those as there's often a direct correlation between the amount paid and the perceived benefit.

 

[bfreedma]

Personal preferences -- it sounds better to them...improving their individual listening experiences. Which is cool.

After all, it's their $$$ they vote with and their personal listening preferences they're catering to; preferences that are both real (i.e 'Pathological' Tubes, Transparent, basshead cans, treblehead cans, neutral cans, bass boost, crossfeed, 'mis-applied' EQ, etc.) and imagined (hi-res, cables, etc.).

Some of us enjoy both 'Transparent' as well as 'Pathological' systems; attaching to them a variety of cans with different sound signatures...switching configurations up for any number of 'illogical' reasons.

BTW...you don't have to shell out mega-bucks for solidly built 'Pathological' amps, etc.

Now...'Imagined' differences? That's another story! One should expect to shell out significant coin for those as there's often a direct correlation between the amount paid and the perceived benefit.

Agreed on all points. If statements were couched as "personal preference" rather than "reference", these discussions would be much easier.

I certainly have buying criteria that goes beyond baseline performance, but those are feature, aesthetic and personal purchasing preferences. The key is realizing that all of my purchasing criteria other than baseline audio reproduction are indeed, my personal preferences and not universal reasons to select gear.

For example, I like my GSX-MkII but certainly wouldn't tell someone that they needed to spend that much on an amp to get the same transparency. If someone asked for a recommendation for a well built, good looking (IMO), high resale value product that was also a competent technical component, I'd certainly include it as an option.

 

[bigshot]

The important thing is that transparency is a calibrated baseline. It guarantees that if you have a preferred EQ or DSP setting, you can swap in different amps and DACs and your system will still sound the way you want it to. If every component was a different color, it would require a complete rebalancing every time you bought something new.

I have no quibble with people who prefer a specific coloration. We all have different ears. Just because a system is transparent, it doesn't mean that you have to listen to it that way. There are tone controls, equalizers and DSPs that are designed to take the calibrated sound and sculpt it any way you want. My personal EQ curve is a couple of dB different than flat. Nothing wrong with that. The components upstream from my DSPs are all transparent though.

I studied design in college. They taught me that form followed function, and efficiency in solving the problem is beautiful. When I look at systems that are covered with bling and conspicuous consumption, they don't impress me. They evoke the same reaction as velvet sofas covered in plastic furniture covers, gold plated toilets and handbags that cost a fortune because of the logo printed on them. But I can be delighted by a system that is well thought out and inexpensive.

 

[castleofargh]

I’m like that too. Using EQ and more on just about any setup, and given how much work it can be to find just the settings I like most(now that I’m pretty clear in what it is), I naturally put gear stability high on my priority list and wish for the next one to sound the same.
A reasonable amount of fidelity seems like a good practical reference to that effect.

And of course I’m all in with @bfreedma wishing for people to discuss how they feel and what they like instead of constantly trying to redefine those feelings and preferences as objective facts about sound.
People believe there is one objective reality and use objective means to know more about it.
Or they believe in subjectivism and don’t try to define the objective reality because they don’t believe it exists.
Thinking that something about the sound is real for all because I think I heard it(sighted), that’s one weird third option and I’m not sure how it should be called?

 

[DarginMahkum]

@GoldenOne, just saw your video "Why you can't trust audio measurements", with a cover title "Don't trust me". Such a misleading click-bait title that rather than contributing to the debate just gives more confusion to people, especially those 90% that does not watch the content or try to understand but just stick to the title. It is just silly title, really.



PS: Though I think the content is OK.
PS2: I guess I am late to the game. It is an old video. Sorry. But that doesn't change the fact that it has a misleading click-bait title.

 

[redrol]

I will DSP/EQ when i have no other choice (for IEMs) but mostly I take them apart and modify them to hit my favorite frequency curve. In the end, if you can do it, it makes using the IEM much easier.

 

[bigshot]

I will DSP/EQ when i have no other choice (for IEMs) but mostly I take them apart and modify them to hit my favorite frequency curve.

I like to be a little more precise than that. I know exactly what my preference is as a deviation from Harman with the frequency band and dB. I can precisely dial that in on all of my cans. Stuffing Kleenex in it or shaving off foam rubber would be too much of a clumsy kludge for what I am trying to do.

 

[redrol]

I dont shove crap in my iems man. I have very precise curves I intend to hit: for instance here is one: (L/R)

 

[bigshot]

This is a measurement from one of your modded IEMs? Or is it just a curve from something you like? If it's the latter, how are you sure that the correction you make matches that curve?

You see, I can take a published curve and dial in an exact correction with my equalizer. I'm not guessing.

 

[redrol]

It's a modified IEM. Thats my curve, as per listening. I modify until its good for me and then take the formula and repeat. I acutally have 12 of these 14.8mm planar IEMs which I am going to be selling.

 

[redrol]

Here is a front view. I did have to take the shell apart and change the back damper on the driver.. and poke out the front vent precisely or if I don't hit the mark, have to apply a damper to the vent. The rubber insert there is designed to move the mid peak up and also squash treble. And combined with a wide bore tip as shown, creates a horn type shape which increases the midrange to where I want. This has taken me months to achieve and now I do the mod for people.