I'm not really a can man, I'm just saying for example if the sound lacks clarity, proper timbre, dynamics, ect, ect over headphones then speakers will not improve the situation except by the possibility of adding distortion via the room, crossover, or mechanical properties.
But it's more difficult to tell when there are the added variables of speaker system in place.
IMO it's best to ensure the absolute capabilities of the amp/dac system by using a close to ideal listening situation such as that of high end headphones before moving on to the speaker system optimization.
Of coarse there are some things headphones just cannot do like speakers can do, but for everything else headphones are a good tool.
But it's more difficult to tell when there are the added variables of speaker system in place.
IMO it's best to ensure the absolute capabilities of the amp/dac system by using a close to ideal listening situation such as that of high end headphones before moving on to the speaker system optimization.
Of coarse there are some things headphones just cannot do like speakers can do, but for everything else headphones are a good tool.
Last edited:
So you have your lecturing/playback system, then (cables...!) the processing of the signal and the transformation from electrical to mechanical ( air impedance, any...? ). The sound travels in the air ( the...room ) which is a gas/fluid and you think that between the speakers and your ears something happens?
It's called propagation, after the emission. Not distortion.
Why is this important to know? Your 5 khz distortion plot shows distortion at 10, 15, 20 and so on. Anyone who's hearing is limited to10 kHz, wil hear non of that distortion.
What I would expect is high amounts of distortion above 22 kHz as you haven't used an analogue filter after your DAC.
What I would expect is high amounts of distortion above 22 kHz as you haven't used an analogue filter after your DAC.
As I said in that post, the distortion products shown are from the analyzer itself and not the amp or dac. I need a notch filter to get rid of those harmonics, they are there even if I short the inputs and outputs of the QA401 to itself.
But also it appears to me that you don't understand that harmonic distortion is meaningless for audio?
HD has almost no correlations to sound quality at all. This is easily discoverable with basic tests. The same harmonic profile can sound wildly different in listening tests and very high levels of harmonic distortion can become inaudible in some cases and very audible in others. HD profiles are a symptom, not a diagnosis.
Amps and dacs can have a huge variation in SQ even though they have low measured HD. I can make my system sound horrid while also producing extremely good measured results. HD is a marketing lie. An easily measurable benchmark that is used to sell products and also the only real standard of measurement that we have.
The truth is we don't know what or how to measure what really needs to be measured. HD measurements itself is only of a single sinewave, sometimes a few sinewaves. But music is infinitely dynamic. However there is a long list of other types of distortion that an HD analyzer will not show you directly or at all.
Good point about the 22khz though, I'll check when I measure next time.
But also it appears to me that you don't understand that harmonic distortion is meaningless for audio?
HD has almost no correlations to sound quality at all. This is easily discoverable with basic tests. The same harmonic profile can sound wildly different in listening tests and very high levels of harmonic distortion can become inaudible in some cases and very audible in others. HD profiles are a symptom, not a diagnosis.
Amps and dacs can have a huge variation in SQ even though they have low measured HD. I can make my system sound horrid while also producing extremely good measured results. HD is a marketing lie. An easily measurable benchmark that is used to sell products and also the only real standard of measurement that we have.
The truth is we don't know what or how to measure what really needs to be measured. HD measurements itself is only of a single sinewave, sometimes a few sinewaves. But music is infinitely dynamic. However there is a long list of other types of distortion that an HD analyzer will not show you directly or at all.
Good point about the 22khz though, I'll check when I measure next time.
Last edited:
You can't hear the difference in filters? No surprises there then if they are competently designed, ie satisfy Nyquist.
Do use an analogue low pass though, as others have said, because the ultrasonic hash coming out of the DAC could potentially cause problems for the stuff it's connected to (they could explode) and cause EMI issues otherwise.
Do use an analogue low pass though, as others have said, because the ultrasonic hash coming out of the DAC could potentially cause problems for the stuff it's connected to (they could explode) and cause EMI issues otherwise.
Here's the issue I'm having though. If I disable the OSF on the dac and then use HQplayer on my PC with filters disabled with 44.1k music files...it still sounds the same. 44.1k with NO filters should have audible effects I would think. I haven't done overly extensive testing on no filters vs filters with 44.1k but I will say that it is not easy to hear the difference if there is one.
Yeah I attached a small first order filter to the analog stage today since there was no harm in it.
HD of <1 kHz is clearly audible. YouTube. Hope this link works.
Does it matter where the distortion originates from? Could be anywhere in the chain, it doesn't change its audibility. Distortion of higher frequencies is going to impossible to detect by ear.
The filters are removing frequencies that are above the hearing limit.
Does it matter where the distortion originates from? Could be anywhere in the chain, it doesn't change its audibility. Distortion of higher frequencies is going to impossible to detect by ear.
The filters are removing frequencies that are above the hearing limit.
Last edited:
I'm not entire sure of the point you are attempting to make mark. HD is not really a factor in anything I've said in this thread, as I've expressed on the last page, it does not matter in any decently designed amplifier because it is not audible. It's the other types of distortion that matter. The end result of the sound is the only concern. Does it sound real or does it not? Now my point which I think you are attempting to address was that amps and dacs have a wide variability in sound quality and sonic traits, even if the HD measurements do not show it. Speakers mess with the sound too much to easily get an truly objective gauge on the performance characteristics of an amp or dac and it's too easy to get caught up in adjusting speaker parameters when the issues could lie at the source, aka the amp or dac. It's always best to minimize variables when optimizing. That's why I suggested that headphones are a better option when attempting to decide whether your amp/dac is up to par because headphones are all-round superior to speakers in almost every way they can be when it comes to critical listening.
Of course speakers and headphones are apples and oranges and some things can only be achieved on speakers and some things can only be achieved on headphones but as I've said, their advantages should be exploited when searching for absolute optimizations.
Of course speakers and headphones are apples and oranges and some things can only be achieved on speakers and some things can only be achieved on headphones but as I've said, their advantages should be exploited when searching for absolute optimizations.
Last edited:
If you have very high bandwidth amplifiers all down the line , the only ones that can be negatively affected are the speakers if their filters aren't properly designed.
Truth is that you hit the right question: what if all the analog chain can properly reproduce anything you're given to ?Well...your ears will only hear what they can hear and if there's no significant IMD or THD present in the system, you'll only hear anything you can in the audio band and that is all!
On the other side, the commercial audio equipment is made to be highly versatile, to be compatible with any other equipment, that is why it has lots of filters.Today we have very high bandwidth circuits at hand and very high bandwidth amplifiers too for diy.
The fact that you are listening to headphones might be also a reason why you weren't bothered by the amp bandwidth, because it's much easier to have very high slew rate amplifiers at low power.
I'm experiencing right now a sony cd player with tda1541 and no low pass filter at all with only an lm6172 as i/v and a opa1641-tpa6120 composite amplifier with 2x slew rate for headphones amplifier(250 ohm Bayerdynamic dt880pro)...so nothing out of the audio band will be heard until my headphones will be burned in case of any strange high freq content that might heat the headphones.I already listen to this system for 2 days continuously and i can't hear anything weird...
Truth is that you hit the right question: what if all the analog chain can properly reproduce anything you're given to ?Well...your ears will only hear what they can hear and if there's no significant IMD or THD present in the system, you'll only hear anything you can in the audio band and that is all!
On the other side, the commercial audio equipment is made to be highly versatile, to be compatible with any other equipment, that is why it has lots of filters.Today we have very high bandwidth circuits at hand and very high bandwidth amplifiers too for diy.
The fact that you are listening to headphones might be also a reason why you weren't bothered by the amp bandwidth, because it's much easier to have very high slew rate amplifiers at low power.
I'm experiencing right now a sony cd player with tda1541 and no low pass filter at all with only an lm6172 as i/v and a opa1641-tpa6120 composite amplifier with 2x slew rate for headphones amplifier(250 ohm Bayerdynamic dt880pro)...so nothing out of the audio band will be heard until my headphones will be burned in case of any strange high freq content that might heat the headphones.I already listen to this system for 2 days continuously and i can't hear anything weird...
I had a similar argument about filters as i can't hear any damn thing over 13.8khz and i said that i'd be using a 16khz filter if possible .There was a gentleman that put a good light on the subject as he argued that although he can't hear anything over 15khz , a 21khz brickwall filter is easily recognized in a chain, because it takes the energy of the lower spectrum too with it, which is true. There's no high pass or low pass filter that won't affect the lower or higher bandwidth.
I only used a 6db first order filter RC-2k4 ohm+1nf in the chain after the i/v stage and i might just make the capacitor 2n2F or 3.9nf if i don't feel loosing too much in my own system but actually i prefer the system this way. I have the 2k4 resistor before the rca output and the 1nf capacitor at the other end of the rca cable(with its own capacitance) , in the headphones amplifier. It is reproducing everything that is thrown at it and i don't need any other cosmetics for this.My own brain is smart enough to reduce any noise or ignore any distortion due to dac or poor mastering if it can hear any.I can even change the CD if i don't like the music
I only used a 6db first order filter RC-2k4 ohm+1nf in the chain after the i/v stage and i might just make the capacitor 2n2F or 3.9nf if i don't feel loosing too much in my own system but actually i prefer the system this way. I have the 2k4 resistor before the rca output and the 1nf capacitor at the other end of the rca cable(with its own capacitance) , in the headphones amplifier. It is reproducing everything that is thrown at it and i don't need any other cosmetics for this.My own brain is smart enough to reduce any noise or ignore any distortion due to dac or poor mastering if it can hear any.I can even change the CD if i don't like the music
Last edited:
For what it's worth, I don't hear any difference between the various digital filter options of my valve DAC either, not even when using the apodizing filter on 44.1 kHz sample rate material, which makes the treble roll off well before 20 kHz. Lampie519 does hear that difference, at least in an informal non-blind test. In all cases there is a fixed analog filter at 82 kHz.
There are those that run without any of those output filters on all the time with NOS DACs. Technically you're going to be introducing aliasing products, things that shouldn't be there, but if they are low enough in level and otherwise masked by other stuff then you aren't going to hear them. Won't look pretty in measurements either.
Good.
That's why I recommended headphones as an option. They are equivalent to an anechoic chamber and playing speakers in an anechoic chamber actually sounds like listening to headphones. Except headphones have the additional benefit of superior performance characteristics due to their low power, low excursion requirements, small size, and linear phase response. Anyway I used my scope to check out 44.1k no filters and overlayed the dac output signal to the amp output signal and the signal is a blatant aliasing stair-step signal. No visual difference between the input and output of amp so I guess it doesn't care about the HF content too much. When using high bitrates the aliasing goes away but still can't hear a difference. I guess I'll just keep the bitrate at 384k.
Last edited:
Hehehe, no. He's on the right track with the DAC and headphone thing. The amplifier and speaker are unnecessary variables that will ruin his experiment. If only he knew why he's right, he might be able to design the right amplifier and the right speaker and he will never worry about treble quality, bass quality or the likes like Sumotan and everybody else
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.