[reginalb]

   
 
 
What you're both saying makes perfect sense of course and I'm certainly not ignoring the facts. What I'm trying to understand is how many other significant and audible variables might be factors when comparing a low end and high end amp.
 
Hypothetically let's say I take some measurements from both a $100 DAP and a $3000 DAP and I discover that both have a very flat Frequency response and both react exactly the same with various multi  BA loads. In this scenario, what would be the next comparison I might make, if accurate sound reproduction is my only real interest? 

 
 

  [Castle's standards]

 
Castle has even more than I do in there. For me it's just a few things:
 

• Measures flat
• low distortion
• low EMI (alternatively, good shielding to interference. I have had a couple phones and laptops - I go through a lot, and the Hifiman HM-101 that were very prone to interference, nothing else ever has been)
• Sufficiently low output impedance*

 
*This number will depend on your headphones, I need really low because of my Merlins, which I love. But the original Hifiman line, and the first A&K AK100 had stupidly high, inexcusably high, output impedance, as do a few phones. But if you have a more standard impedance headphone, most portable sources today will have sufficiently low impedance for your needs. 
 
 
And that's pretty much it for me, until we get in to features. And the overriding feature for me is that it supports Google Play Music All Access, because I subscribe to that, and use Google casting for multi-room audio throughout my house. So for me, it's basically: Flat, low distortion, emi, and impedance, and runs Android. Unfortunately, my phone misses on the impedance thing. 

 

[castleofargh]

oh it's my wish list to Santa(as apparently Xmas starts the day after halloween now) for the stuff I'd like to know about, nothing more.  in the end the DAPs I use the most are the ones with long battery life and UI/size that I enjoy ^_^. not exactly the objectivist in action.
 
 
and +15554435435421 for shielding against EMI/RFI. but I have no idea how to set up a reliable test.

 

[reginalb]

  oh it's my wish list to Santa(as apparently Xmas starts the day after halloween now) for the stuff I'd like to know about, nothing more.  in the end the DAPs I use the most are the ones with long battery life and UI/size that I enjoy ^_^. not exactly the objectivist in action.
 
 
and +15554435435421 for shielding against EMI/RFI. but I have no idea how to set up a reliable test.

 
Yeah, unfortunately, for EMI it's more, if I don't like it, sell it and replace. I pay a lot more attention for complaints of interference in reviews, though. the HM-101 it was a pretty common issue that I just ignored. 

 

[CraftyClown]

Thanks fellas, really useful feedback.
 
Last question... for now 

 
If the manufacturer of a particular DAP isn't forthcoming with the output impedance, is there a simple way to either calculate or measure it?

 

[reginalb]

  Thanks fellas, really useful feedback.
 
Last question... for now 

 
If the manufacturer of a particular DAP isn't forthcoming with the output impedance, is there a simple way to either calculate or measure it?

 
If you happen to have a dummy load and a multimeter, then you can. You can also do it with more expensive, fancier equipment, but you can also do it with those inexpensive simple tools. Or do an RMAA measurement with your favorite iems as a load (be careful not to damage them). 
 
If I was going to use an offline MP3 player every day at this point, I wouldn't buy anything but a low end Fiio. X1 II probably. Low impedance, and a solid player (UI is probably lacking, though). 

 

[Ruben123]

   
If you happen to have a dummy load and a multimeter, then you can. You can also do it with more expensive, fancier equipment, but you can also do it with those inexpensive simple tools. Or do an RMAA measurement with your favorite iems as a load (be careful not to damage them). 
 
If I was going to use an offline MP3 player every day at this point, I wouldn't buy anything but a low end Fiio. X1 II probably. Low impedance, and a solid player (UI is probably lacking, though). 

Have you ever tried a Sandisk Sansa Clip player? Even cheaper and smaller, although Rockboxing isnt possible anymore unfortunately.

 

[ClieOS]

   
If you happen to have a dummy load and a multimeter, then you can. You can also do it with more expensive, fancier equipment, but you can also do it with those inexpensive simple tools. Or do an RMAA measurement with your favorite iems as a load (be careful not to damage them). 
 
If I was going to use an offline MP3 player every day at this point, I wouldn't buy anything but a low end Fiio. X1 II probably. Low impedance, and a solid player (UI is probably lacking, though). 

 
RMAA won't actually tell you the  output impedance.
 
To be more specific, you will need a test tone (1kHz is the usual frequency of choice, can be generated via Audacity), a load (a pair of 16ohm resistors is the usual choice, but you can pick any sensible resistance), and a True RMS multi-meter that can measure Vac in the test tone frequency (cheap multi-meter might not able to measure the test tone frequency very accurately, but they are passable if you don't need a very precise number. If you are using a cheap multi-meter that you are not sure if it can measure 1khz, then use 500Hz signal).
 
First you need to measure the DAP (playing back test tone of course) unloaded's output, then the loaded outout. Calculate the output impedance with the following formula: [ (Unloaded Vac -  Loaded Vac) / Loaded Vac ] x resistance = output impedance.

 

[CraftyClown]

   
If you happen to have a dummy load and a multimeter, then you can. You can also do it with more expensive, fancier equipment, but you can also do it with those inexpensive simple tools. Or do an RMAA measurement with your favorite iems as a load (be careful not to damage them). 
 
If I was going to use an offline MP3 player every day at this point, I wouldn't buy anything but a low end Fiio. X1 II probably. Low impedance, and a solid player (UI is probably lacking, though). 

 
Well I am planning to use Chris's guide to take FR measurements when I get a free moment, so that would be a perfect time to check impedance as well I guess?

 

[CraftyClown]

   
RMAA won't actually tell you the  output impedance.
 
To be more specific, you will need a test tone (1kHz is the usual frequency of choice, can be generated via Audacity), a load (a pair of 16ohm resistors is the usual choice, but you can pick any sensible resistance), and a True RMS multi-meter that can measure Vac in the test tone frequency (cheap multi-meter might not able to measure the test tone frequency very accurately, but they are passable if you don't need a very precise number. If you are using a cheap multi-meter that you are not sure if it can measure 1khz, then use 500Hz signal).
 
First you need to measure the DAP (playing back test tone of course) unloaded's output, then the loaded outout. Calculate the output impedance with the following formula: [ (Unloaded Vac -  Loaded Vac) / Loaded Vac ] x resistance = output impedance.

 
Just read this after I posted. Thanks ClieOS

 

[reginalb]

 
Have you ever tried a Sandisk Sansa Clip player? Even cheaper and smaller, although Rockboxing isnt possible anymore unfortunately.

 
 
I measured the old Clip+ back in the day to be 1.06Ω, haven't measured the newer version. 
 

   
RMAA won't actually tell you the  output impedance.
 
To be more specific, you will need a test tone (1kHz is the usual frequency of choice, can be generated via Audacity), a load (a pair of 16ohm resistors is the usual choice, but you can pick any sensible resistance), and a True RMS multi-meter that can measure Vac in the test tone frequency (cheap multi-meter might not able to measure the test tone frequency very accurately, but they are passable if you don't need a very precise number. If you are using a cheap multi-meter that you are not sure if it can measure 1khz, then use 500Hz signal).
 
First you need to measure the DAP (playing back test tone of course) unloaded's output, then the loaded outout. Calculate the output impedance with the following formula: [ (Unloaded Vac -  Loaded Vac) / Loaded Vac ] x resistance = output impedance.

 
You're correct, I should have elaborated. Won't tell you the impedance, but it will give you what you need in terms of HF interaction more directly. 

 

[Samueru Sama]

  Thanks fellas, really useful feedback.
 
Last question... for now 

 
If the manufacturer of a particular DAP isn't forthcoming with the output impedance, is there a simple way to either calculate or measure it?

You can still use the ADC of your PC to measure output impedance. By comparing the amplitude drop at the line in when you add a load to the device being recorded. I've found that method to be quite precise, you need custom cabling though. And make sure to prevent clipping either from the ADC or whatever device being tested.
 
For example, hook "X" phone to your PC, play a test signal and adjust the levels until you get 0 dBFS at the input, Then add a known load (32 Ohm for example) in parallel to the phone while it is still connected to your PC. The amplitude at input will drop, lets say in this case it dropped down to -6.7 dBFS. And this is all we need to figure out the output impedance.
 
0 dBFS could be any voltage. really. It is the amplitude drop that tells you the output impedance, not the measured voltage. So we can assume whatever voltage to do this. It will give you the same result.
 
Lets say that 0 dBFS (our unloaded voltage) is 1 Vrms.
 
So to know the voltage at -6.7 dBFS (with a load). Use this: 1 Vrms * 10^(amplitude change/20)
1 Vrms * 10^(-6.7 dB/20) = 0.462 Vrms.
 
And the rest is basically ClieOS' formula: (Unloaded Vac -  Loaded Vac) / Loaded Vac ] x resistance
 
(1 Vrms - 0.462 Vrms) / 0.462 Vrms * 32 Ohm = 37.26 Ohm.
 
Also these might be of your interest:
 
https://www.reddit.com/r/headphones/wiki/tech/output_impedance_database
http://monoadc.blog64.fc2.com/blog-entry-99.html

 

[Raketen]

possibly beside the point, and I reckon not relevant to measurements presumably volume matched- but, for user experience volume adjustment granularity might make two otherwise acoustically identical devices effectivly sound different using the same headphones, depending on the headphones

the other thing I don't think has been mentioned are DAC correction filter settiings as many dacs have different modes... most DAP DACs seem to be fixed but some devices allow you to switch these which gives some consistent base for comparison- For instance archimago's measurements of the GOV2 were an interesting read though a little beyond my grasp of what actually becomes audible (and further beyond my memory since I read it so long ago ), though i believe it does affect the actual waveform in some subtle way? http://archimago.blogspot.com/2015/09/measurements-light-harmonic-geek-out-v2.html

 

[zareliman]

If your DAP is capable of providing at least 2V RMS line output and it's not obviously broken or terribly designed (better than 0.9% THD) then it's probably good enough if  your headphones are within today's portable standards (that is ~100dB/mW and 32 ohm). The Xduoo X2 price point is probably when you start getting extremely diminishing returns IN TERMS OF SOUND QUALITY. Of course higher end DAPs still offer more features but the raw untouched sound reproduction differences will most likely be inaudible in a proper A/B test.

 

[TheoS53]

I'm liking this thread as it asks some very important, but most importantly, objective questions. 

As some have stated, the output impedance can have a considerable effect on the sound between 2 DAPs. However, there are some things that are assumed which I feel should be discussed. 

Firstly, the OP assumes that the FR is the only thing that makes a difference in how we perceive sound. But, we also need to consider the various orders of harmonic distortions. These too can have an effect on how "warm" or "bright" a DAP can sound. then, we also need to consider how the DAC decodes the audio, and more importantly the incredibly small timing delays that it introduces between sounds. To me, that might very well be the tipping point of how natural a DAP sounds. 

By this I mean I think too often we take for granted how very very small changes can make a difference to how we hear things. Think of it like this, Perspex and glass look almost identical, but if you tap on them, they give off a different sound, and those small differences allow you to differentiate between the 2 different materials. 

Also consider how our brains work. The tiny little differences in delays allow our brains to figure out where a sound is coming from, how near i is to us, whether or not the source is moving, and if there are 2 different sources, what are their positions and vectors relative to one another. If those delays and amplitudes are changed, it changes how our brains perceive them. I think this plays a large part in what *can* separate the cheaper from the more expensive DAPs. However, just because it is more expensive doesn't mean it's going to sound better. For example, I prefer the sound of the DX80 over the AK70. the AK70 costs roughly twice as much as the DX80, but the DX80 just has a wider, more natural sound stage to me. On the flip side, the Oppo HA-2SE is a completely different device compared to the DX80, yet they sound identical to me. Sound stage, signature, instrument separation....all sounds identical to that of the DX80.

 

[gregorio]

  we also need to consider how the DAC decodes the audio, and more importantly the incredibly small timing delays that it introduces between sounds.

By this I mean I think too often we take for granted how very very small changes can make a difference to how we hear things. Think of it like this, Perspex and glass look almost identical, but if you tap on them, they give off a different sound, and those small differences allow you to differentiate between the 2 different materials. 

Also consider how our brains work. The tiny little differences in delays allow our brains to figure out where a sound is coming from, how near i is to us, whether or not the source is moving, and if there are 2 different sources, what are their positions and vectors relative to one another. If those delays and amplitudes are changed, it changes how our brains perceive them.

 
You seem to be confusing "incredibly small", "small", "very very small" and "tiny little". A 1mm (thousandth of a meter) speck of dirt on your piece of glass is "tiny" but easily visible under the right conditions. What about a small group of unwanted molecules on your piece of glass of just a few hundred pico meters (trillionths of a meter), how visible do you think that would be, how would it affect your perception of the glass? Of course it's going to make no difference whatsoever. In fact our 1mm piece of dirt in this context is not tiny at all, relative to our group of molecules it's massive, orders of magnitude more massive! You are correct in saying our brains use "tiny little" timing differences to determine position, differences of just a handful or few tens of milli seconds. However, the timing differences that even a cheap DAC should introduce are in the order of a few hundred pico seconds. In comparison to this, the "tiny little" timing differences used by the brain for positing are actually massive! If we don't just limit ourselves to positioning but to any form of audible detection, a modern cheap DAC would still introduce timing delays/errors about 1,000 times below that level and relative to purely positioning, many tens or hundreds of times lower again!
 
Using your analogy, how about a piece of glass and another piece of glass with say a few thousand perspex molecules on it. Do you think you'd hear a difference if you tapped them or be able to tell them apart with any of your senses? There has to be some context to "tiny" or "very very small" otherwise we can end up with completely erroneous analogies and completely ridiculous conclusions!
 
The audiophile community often appears particularly bad in this regard. The logic often seems to go something like this: 1. Compared to a galaxy, our sun is tiny. 2. Compared to the moon, a football is tiny. A logical conclusion is therefore: As both the sun and a football can be described as "tiny", they are roughly of the same size. Obviously (hopefully!) this conclusion is ridiculous, but shockingly, no more ridiculous than some audiophile claims of what audibly affects sound quality!
 
G