Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: 24-bit audio proposed for iTunes (Read 44443 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

24-bit audio proposed for iTunes

Reply #75
Though I'm becoming increasingly annoyed with my Clip+ and may try to find an old Clip on eBay: you can "lock" the Clip with the power off. You can't "lock" the Clip+ unless it's switched on. Which means... it gets switched on all by itself in my jacket pocket, and when I want to use it the battery is flat. Again. and again. and again. Oh, it crashes sometimes too.

I haven't encountered either problem, but I usually keep mine either clipped to me while using it, or in small pocket on my backpack when not (or laying around somewhere).

Are you running the Sansa or Rockbox firmware? If you're not running the latest Rockbox you should give it a try and it might not crash? I doubt Rockbox can change the power switch behavior, however. And if it's still crashing, report it to Rockbox. They do new builds nightly I think. But it might be your hardware? I didn't read all 1500 reviews on Amazon but I don't remember crashing being among the common complaints.

You could also always try the latest Sansa firmware as well if you haven't already. It's not under the support/firmware part of their website, but in their forum.  But I've never had either one crash. My old e260 would crash fairly regularly--sometimes in really odd ways... like dropping one channel while the other kept playing with all the controls locked up.

Getting somewhat back on topic, I'm also likely to test the next generation of the Shuffle and/or Nano as well and report on those. But, for now, I'm really happy with the Clip+ (and its bigger brother the Fuze). And I'm looking forward to the Touch 5G mainly because I hope they fix the output impedance problem but also to see if the new 24 bit capability is pure marketing fluff.

24-bit audio proposed for iTunes

Reply #76
Sansa Clip+ vs iPod Touch 3G Detailed Measurements
Very nicely done!  Thanks for making the effort, its a breath of fresh air to see someone deal with data, and nice conversations around output impedance, etc.
Yes, that is great.

Though I'm becoming increasingly annoyed with my Clip+ and may try to find an old Clip on eBay: you can "lock" the Clip with the power off. You can't "lock" the Clip+ unless it's switched on. Which means... it gets switched on all by itself in my jacket pocket, and when I want to use it the battery is flat. Again. and again. and again. Oh, it crashes sometimes too..


Never happens to me, but I don't use my clip a lot, mostly only camping.

What else do you carry in those pockets? 

I've had serous problems with me destroying flash drives that I carried in my pocket until my wife bought me a flash drive cozy a couple of years back.  I can't find a listening for the plain black one that she gaie me, but this is about as close as I hae found on the web:

Link to USB Flash Drie keychain cozy

Problem went away. Maybe if you carry your Clip+ in a small bag like the one that comes with some earphones...

24-bit audio proposed for iTunes

Reply #77
Apple is famous for using either proprietary (or relatively uncommon) standards so they could invent their own flavor of 24 bit audio. And they like DRM (copy protection) and have been reluctant to fully get rid of it. So you could be stuck playing the new 24 bit format on only Apple hardware. It could be only the new 24 bit hardware (to help drive more hardware sales), or it could be anything you can upgrade to support the new format (Macs and iOS devices and possibly older iPods). Or, if they wanted to be nice for a change, they could use an industry open standard like 24 bit FLAC and then you could play your purchased music on lots of things.



That's very baseless speculation. DRM was mandated by music labels. Apple pushed forward on February 6, 2007 to get rid of it. First there was some resistance by the big labels, as of April 2009, all music on the iTunes store is available DRM free. If you call AAC "uncommon", it is the official and advanced successor of MP3. Tech companies should not be criticized for not being reactionary. ALAC already supports 24 bit, so speculation about a new "24 bit format" is without basis. While ALAC wasn't documented there are free implementations that work fine.

Sansa Clip+ vs iPod Touch 3G Detailed Measurements


Most portables are used with IEMs. Since those are most of the time very sensitive, the RMAA measurements don't reflect actual performance. You'd need a sound card with matching input sensitivity to get meaningful results.

24-bit audio proposed for iTunes

Reply #78
Clip+ crashes on all Sansa firmware, including latest. Happens only when going into podcast list (I have lots in there). It's occasional (maybe once every five hours?). It reboots itself just fine.

Thanks for the suggestions for an external mechanical solution to prevent the Clip+ power control getting pushed. I suppose that would do. But the Clip had a perfect mechanical control built-in, and they removed it for the Clip+

Cheers,
David.

24-bit audio proposed for iTunes

Reply #79
Sansa Clip+ vs iPod Touch 3G Detailed Measurements

Most portables are used with IEMs. Since those are most of the time very sensitive, the RMAA measurements don't reflect actual performance. You'd need a sound card with matching input sensitivity to get meaningful results.


I generally like the tests, but find the weight given to measurements made near FS to be excessive,  for the reason given above.  The AES allows measurements to be made at FS -3Db which I think is more representative of how equipment is used. I also think that measurements made at FS -10 dB are very relevant to actual use. 

To me the most important thing about operation near FS is that the clipping is reasonably clean and lacks major nasties such as bursts of oscillation around 400 Hz, which I have seen at times.  The possiblity that a digital player or DAC is going to see signals above -10 dB  @ > 15 KHz is IME about zero.  Reality is more like -40 dB.  Oh, I'm sure that there is one second of some track some place that violates this, but...

24-bit audio proposed for iTunes

Reply #80
Apple is famous for using either proprietary (or relatively uncommon) standards so they could invent their own flavor of 24 bit audio. And they like DRM (copy protection) and have been reluctant to fully get rid of it. So you could be stuck playing the new 24 bit format on only Apple hardware. It could be only the new 24 bit hardware (to help drive more hardware sales), or it could be anything you can upgrade to support the new format (Macs and iOS devices and possibly older iPods). Or, if they wanted to be nice for a change, they could use an industry open standard like 24 bit FLAC and then you could play your purchased music on lots of things.


That's very baseless speculation. DRM was mandated by music labels. Apple pushed forward on February 6, 2007 to get rid of it. First there was some resistance by the big labels, as of April 2009, all music on the iTunes store is available DRM free. If you call AAC "uncommon", it is the official and advanced successor of MP3. Tech companies should not be criticized for not being reactionary. ALAC already supports 24 bit, so speculation about a new "24 bit format" is without basis. While ALAC wasn't documented there are free implementations that work fine.

Sansa Clip+ vs iPod Touch 3G Detailed Measurements


Most portables are used with IEMs. Since those are most of the time very sensitive, the RMAA measurements don't reflect actual performance. You'd need a sound card with matching input sensitivity to get meaningful results.

Well the DRM topic is open to debate, but other companies somehow were able to drop DRM before Apple did with some of the same music lablels. And it's generally well accepted that Apple tends to do things in a way that keep you buying other Apple products. A good example is the video chat with the iPhone 4. There were widely accepted open standards they *could* have used, but they chose a proprietary closed solution instead that only works to other iPhone users. The same is also true of some of their Bluetooth implementations--they only support Apple authorized devices. I could keep going, but you get the idea.

And yes, I was referring to inventing ALAC when there were perfectly good open lossless standards already in place. But yes, a lot of what I posted is speculation. I admitted as much.

As for your IEM comment relative to my testing, I agree 100%. I wrote an entire blog article on why RMAA testing often doesn't work well because there are no absolute levels, among other problems. That's why I do most of my testing with a $10,000 Prism Sound dScope using a variety of loads including IEMs. The dScope is vastly superior to *ANY* soundcard that I'm aware of and I explain why in the links below:

RMAA Testing

Testing Methods

24-bit audio proposed for iTunes

Reply #81
A good example is the video chat with the iPhone 4. There were widely accepted open standards they *could* have used, but they chose a proprietary closed solution instead that only works to other iPhone users.


Which would be?

24-bit audio proposed for iTunes

Reply #82
A good example is the video chat with the iPhone 4. There were widely accepted open standards they *could* have used, but they chose a proprietary closed solution instead that only works to other iPhone users.


Which would be?

I was thinking of DVB-H and Nokia's video calling (which predates Apple's by several years) but, in looking into it further, they're not (yet at least) fully interoperable either. Aylus does have an interoperable standard, but it's newer than the iPhone 4.

So I rephrase my comment to "so far Apple's FaceTime is a Apple-only solution". But we're getting further and further OT here.

24-bit audio proposed for iTunes

Reply #83
I don't think it's too far fetched that DRM will come back with lossless music though. But also I don't doubt outrage from many consumers.

24-bit audio proposed for iTunes

Reply #84
Sansa Clip+ vs iPod Touch 3G Detailed Measurements


Most portables are used with IEMs. Since those are most of the time very sensitive, the RMAA measurements don't reflect actual performance. You'd need a sound card with matching input sensitivity to get meaningful results.


I don't understand what you're getting at here.  "a sound card with matching input sensitivity" would mean something like a preamp right?  Why is one of those needed to get meaningful results?

24-bit audio proposed for iTunes

Reply #85
Sansa Clip+ vs iPod Touch 3G Detailed Measurements


Most portables are used with IEMs. Since those are most of the time very sensitive, the RMAA measurements don't reflect actual performance. You'd need a sound card with matching input sensitivity to get meaningful results.


I don't understand what you're getting at here.  "a sound card with matching input sensitivity" would mean something like a preamp right?  Why is one of those needed to get meaningful results?


I've been testing all sorts of things with RMAA, and did similar things with Spectra Lab before that. FWIW many of the screens and menus shown on the NwAvGuy's web site (apparently screen shots from the Prism test set he loves so dearly) look very much to me like Spectra Lab.

I see many of his complaints with RMAA as being procedural things. For example he says:

"RMAA has no concept of absolute levels. So it can't measure voltages, power outputs, etc. So you have no way of knowing, for example, how loud that portable player or headphone amp can play without obvious distortion"

The real problem is not RMAA, its the fact that audio interfaces aren't formally calibrated. To me this is just a matter of common sense - if you want to know the voltage, get a voltmeter!  Ihave a Fluke 85 that has very flat frequency response to over 100 KHz.

When I use RMAA (or Spectra) to test equipment whose voltage is outside of the reasonable realm of use of the audio interface, I add an external amplifier or attenuator. 

It turns out that a high quality mic preamp makes an flexible and effective  gain block for measuring low voltage signals. When used at reasonable gains in the 0-20 dB range they are  typically very clean.

I use RMAA to test power amps by means of a high-powered dummy load (8 pieces 8 ohm 300 watt non-inductive resistors with some switches, connectors and a fan) that has a built-in attenuator. I also have a loudspeaker simulator with similar features and capabilities.

If I want low-residual measurements, I use a really good audio interface. For example the M-Audio Delta AP 24192 has residuals that are about 110 dB down and costs less than $200.  This gives you a goodly amount of wiggle room if you want to use it to measure signals that aren't near FS at its input.

High frequency IM tests can be done with RMAA by simply setting the IM test frequencies to 19 and 20 KHz.  The test frequencies can be reset and jitter can be analyzed from the spectrum analysis plots.

Etc., etc.


24-bit audio proposed for iTunes

Reply #86
I don't understand what you're getting at here.  "a sound card with matching input sensitivity" would mean something like a preamp right?  Why is one of those needed to get meaningful results?


A typical IEM already outputs far over 100 dB of sound pressure (106 dB in my case) for 1 millivolt input voltage. At these low levels output performance may differ considerably from the kind of measurement one usually gets from a high impedance line-level input. No, loading with a parallel resistor doesn't help.

Since this is OT, we could continue the discussion [a href='index.php?act=findpost&pid=86306']here[/a].

24-bit audio proposed for iTunes

Reply #87
A typical IEM already outputs far over 100 dB of sound pressure (106 dB in my case) for 1 millivolt input voltage. At these low levels output performance may differ considerably from the kind of measurement one usually gets from a high impedance line-level input.


Ok but you said:

the RMAA measurements don't reflect actual performance. You'd need a sound card with matching input sensitivity to get meaningful results.


It sounded like you had results or a concrete theory here, not just speculation.  Is that the case?

24-bit audio proposed for iTunes

Reply #88
It sounded like you had results or a concrete theory here, not just speculation.  Is that the case?


I have posted my findings with the Sansa Clip+ in the linked thread. They were independently confirmed. RMAA results are very good, but the described noise is clearly audible, I'd say at max -50 dB down. When I try to record it into the high-impedance line-in of my sound card there is at least -90 dB of silence.



24-bit audio proposed for iTunes

Reply #91
I've been testing all sorts of things with RMAA, and did similar things with Spectra Lab before that. FWIW many of the screens and menus shown on the NwAvGuy's web site (apparently screen shots from the Prism test set he loves so dearly) look very much to me like Spectra Lab.

I see many of his complaints with RMAA as being procedural things. For example he says:

"RMAA has no concept of absolute levels. So it can't measure voltages, power outputs, etc. So you have no way of knowing, for example, how loud that portable player or headphone amp can play without obvious distortion"

The real problem is not RMAA, its the fact that audio interfaces aren't formally calibrated. To me this is just a matter of common sense - if you want to know the voltage, get a voltmeter!  Ihave a Fluke 85 that has very flat frequency response to over 100 KHz.

When I use RMAA (or Spectra) to test equipment whose voltage is outside of the reasonable realm of use of the audio interface, I add an external amplifier or attenuator. 

It turns out that a high quality mic preamp makes an flexible and effective  gain block for measuring low voltage signals. When used at reasonable gains in the 0-20 dB range they are  typically very clean.

I use RMAA to test power amps by means of a high-powered dummy load (8 pieces 8 ohm 300 watt non-inductive resistors with some switches, connectors and a fan) that has a built-in attenuator. I also have a loudspeaker simulator with similar features and capabilities.

If I want low-residual measurements, I use a really good audio interface. For example the M-Audio Delta AP 24192 has residuals that are about 110 dB down and costs less than $200.  This gives you a goodly amount of wiggle room if you want to use it to measure signals that aren't near FS at its input.

High frequency IM tests can be done with RMAA by simply setting the IM test frequencies to 19 and 20 KHz.  The test frequencies can be reset and jitter can be analyzed from the spectrum analysis plots.


I say in the first paragraph of my article RMAA is certainly useful when it's used correctly and people understand the limitations. But a lot of people using RMAA don't have the tools, skills or knowledge to use it correctly. Yet they still post their results with no disclaimer indicating they were taken at unknown levels, often with an unknown load, unknown sound hardware, etc. So for the Average Joe using Google to find test results on some piece of equipment, the odds actually favor finding poorly done RMAA results. People like you are the exception among RMAA users, not the rule.

You're correct some issues are procedural, and at the end of the article I talk about, for example, using a good DMM like your Fluke, to set levels. And, with enough effort, knowledge, and a good variety of external hardware, voltage dividers, other software programs, etc. you can extract a decent set of results from a good soundcard. But most don't do all that correctly, or even at all, with RMAA.

And, even in the best of hands, RMAA still falls flat in several other areas. It's rather fixed in what it does, and AFAIK, it's not well documented how it's doing some things internally. This makes comparing RMAA results, to results obtained other ways, more difficult (or some would argue impossible). With an instrument-grade analyzer like an Audio Precision or dScope you have total control over most everything right down to the amount of jitter and sample rate deviation you want in the digital output signal for testing a DAC's ability to handle less than ideal digital inputs. You can configure the inputs and outputs just about any way imaginable with any combination of digital/analog/soundcard generated signals and similar on the analysis side.

The dScope measures the level at its inputs from microvolts to 159 volts with 0.05 dB accuracy without needing any dividers, meters, etc. It has seamless autoranging that keeps the A/D running within 2 dB of it's 0 dBFS level. Try THAT with a soundcard! External dividers are typically far more "coarse" and can be a source of considerable error depending on resistor tolerances, sound card input impedance, device source impedance, etc.

It's incredibly useful to have digital and analog outputs, and digital and analog inputs, that are always referenced to known levels and each other. There's no need to constantly use a meter to set levels on both sides. And, of course, you can't "meter" a digital signal at all. But dScope always knows what dBu, dBv, etc. are. They're not just relative abstract numbers like they are with a soundcard. And it can relate those numbers on both sides to dBFS in the digital domain which is really useful.

And sound cards with analog level controls (like say the popular E-Mu 0202, 0404, Pre, M-Audio, etc.) make it difficult to change the levels without also messing up the channel balance. So you need to get your meter out every time you even bump those controls. And the controls on the 0202 and 0404 cover such a wide range (65 dB I think?) they sometimes change themselves they're so touchy. They're the opposite of being accurate.

And sound cards that use digital "mixer based" controls typically have fixed steps that might be too coarse to say do something like correct for a 0.7 dB channel imbalance. You're stuck with 1 dB or 2 dB steps for each channel. And many are not calibrated in dB at all. They're "mystery steps".

The dScope has precise selectable input impedance settings which sound cards lack. And perhaps most important, it has floating differential input and outputs to avoid ground loops, or worse, that can be a serious problem trying to use a typical soundcard. Some power amps will self destruct if you ground their output "grounds" to real ground like most sound cards do. And ground loops can easily contaminate sound card measurements.

And the dScope (unlike the Audio Control AP5xx) has what they call a Continuous Time Analyzer. This lets you do true real time analysis without needing to first run FFT passes like RMAA and Spectra Lab do. This can be incredibly useful during testing in many different ways--especially when combined with the scripting built into the dScope software. The dScope can, for example, regulate the input signal driving an amplifier to maintain a constant THD level at the output of the amplifer while doing a live frequency sweep so you end up with a very useful power at xxx % THD vs frequency plot.

The dScope also has, built in, a wide variety of industry standard measurement filters and test signals. Like CCIR, A Weighting, C Weighting, SMPTE, etc. And it can implement those filters in real time rather than applying them after the fact to a completed FFT. You can also listen to the live distortion portion, or even just the Nth harmonic, via the monitor output live if you want. I could keep going, but hopefully you get the idea.

To put it another way, the people designing the soundcards, or any other piece of respectable audio gear, you admire are very likely not using a soundcard to design and measure them. Odds are they're using an Audio Precision or dScope analyzer.

24-bit audio proposed for iTunes

Reply #92
It sounded like you had results or a concrete theory here, not just speculation.  Is that the case?


I have posted my findings with the Sansa Clip+ in the linked thread. They were independently confirmed. RMAA results are very good, but the described noise is clearly audible, I'd say at max -50 dB down. When I try to record it into the high-impedance line-in of my sound card there is at least -90 dB of silence.


If I understand you correctly, you comparing an unloaded RMAA test to a loaded RMAA test?  I think its expected that the loaded test will be much worse, which is why people generally only compare loaded RMAA results.  If I understand what you are suggesting, I think you would have to use a good preamp, a parallel load, and then test at different volume settings.

24-bit audio proposed for iTunes

Reply #93
And, even in the best of hands, RMAA still falls flat in several other areas. It's rather fixed in what it does, and AFAIK, it's not well documented how it's doing some things internally. This makes comparing RMAA results, to results obtained other ways, more difficult (or some would argue impossible). With an instrument-grade analyzer like an Audio Precision or dScope you have total control over most everything right down to the amount of jitter and sample rate deviation you want in the digital output signal for testing a DAC's ability to handle less than ideal digital inputs. You can configure the inputs and outputs just about any way imaginable with any combination of digital/analog/soundcard generated signals and similar on the analysis side.


While I always like to see better testing performed, reading your results I'm not sure I see anything being revealed that we didn't already know from RMAA?  You've certainly measured parameters more accurately then previous tests, but the more accurate results don't really seem to change anything we thought we knew from RMAA.

Which makes it a little odd that you're so concerned about people using RMAA to compare things.  If anything, your own results seem to have confirmed that its a good enough tool for what people want to do with it.  Or am I missing something?

24-bit audio proposed for iTunes

Reply #94
[To put it another way, the people designing the soundcards, or any other piece of respectable audio gear, you admire are very likely not using a soundcard to design and measure them. Odds are they're using an Audio Precision or dScope analyzer.


Interesting factoid: I happen to own an Audio Precision System 1 test set that I hardly touch. I kinda sorta inherited it from Julian Hirsch.

Relevent fact: Last time I looked the DScope was about $10 grand, while the most expensive piece of equipment I've described so far is under $200.

Here's how I approach the issue of test equipment - one of my goals in life is to encourage people to equip themselves to better understand how audio works. I think that hands on experience with hardware is very helpful.  By sticking with budget approaches I keep myself familiar with the slings and arrows of going down that road. And on occasion

I've published auiod equipment test reports that it is said proved performance in excess of that possible with the best equipment then available.

24-bit audio proposed for iTunes

Reply #95
While I always like to see better testing performed, reading your results I'm not sure I see anything being revealed that we didn't already know from RMAA?  You've certainly measured parameters more accurately then previous tests, but the more accurate results don't really seem to change anything we thought we knew from RMAA.

Which makes it a little odd that you're so concerned about people using RMAA to compare things.  If anything, your own results seem to have confirmed that its a good enough tool for what people want to do with it.  Or am I missing something?

I partly agree with you. But there are some important things that RMAA can't directly measure, such as the maximum output level and output impedance, which are critical measurements when comparing portable players. It also can't measure the famous pitch error of the Sansa. Yeah, you may have known about all 3 of those numbers, but they don't show up in any RMAA table, graph or report. It also doesn't measure the DAC linearity (unless you count the -60 dB spectrum plot but virtually any DAC has near zero error there) or square wave response. And remember I was using a really high-end $1800 ADC with RMAA to make those measurements. And I was doing it at known levels, with known loads, etc. Others may not get nearly as similar results with more typical RMAA usage.

Because I picked one of the more popular (non Apple) players on the planet, there's no shortage of test data for it. It also helps that it's a Rockbox player as that means a higher percentage of serious geeks are involved. But if you look at a niche product, like the Cowon i9, it's a lot harder to scare up valid numbers. And the sort of "peer review" process the Clip+ results have already been subjected to is often totally missing. So you may well just get a single set of RMAA numbers and graphs with little clue if you can trust them or not. And, for the reasons I've outlined, I'd bet on "not" most of the time.

24-bit audio proposed for iTunes

Reply #96
Interesting factoid: I happen to own an Audio Precision System 1 test set that I hardly touch. I kinda sorta inherited it from Julian Hirsch.

That's kinda fun! We're getting way OT here, but the AP System 1 is the beast that started it all, and so in some ways, was Julian Hirsch. Both, in their own way, are legends.

If I had a System 1 I wouldn't use it either except perhaps as ballast. It's big, heavy, and uses a proprietary interface that originally only worked with a compatible ancient 8 bit ISA slot (aka as found in the original IBM XT). And the software is DOS based. I'm not sure what AP offers today in terms of interfaces, but it's probably not worth the expense to buy a modern interface for it if one is even available.

24-bit audio proposed for iTunes

Reply #97
It also doesn't measure the DAC linearity (unless you count the -60 dB spectrum plot but virtually any DAC has near zero error there) or square wave response.


Sigma Delta DACs (the industry standard for about the past decade or more) can't possibly have linearity problems. That's one reason why virtually every DAC has low error at FS-60 dB .

Of course if every DAC actually had near zero error at -60 dB then they would all give the same results in a standard dynamic range test. But they don't.

As far as square wave response goes, this sort of thing should really be done with an ADC that is sampling at a fir higher rate then the UUT or an actual oscilloscope.

Square wave testing has gone out of favor because any related sonic problems would show up in a frequency response test.



 

24-bit audio proposed for iTunes

Reply #98
I partly agree with you. But there are some important things that RMAA can't directly measure, such as the maximum output level and output impedance, which are critical measurements when comparing portable players.


But are those really so useful to know?  Pretty much every player in existence has an output level with 3-4 dB of every other player, since they're all running off the same supply voltage.  Yeah its interesting to know if your player is marginally louder then another player, but theres so little difference between players I'd hardly consider it critical.  Likewise, if a player a sufficiently flat frequency response, is it really that important to know the precise output impedance?

It also can't measure the famous pitch error of the Sansa.


Thats a fair point, although this is a pretty rare problem.

Yeah, you may have known about all 3 of those numbers, but they don't show up in any RMAA table, graph or report. It also doesn't measure the DAC linearity (unless you count the -60 dB spectrum plot but virtually any DAC has near zero error there) or square wave response.


Wouldn't DAC nonlinearity show up in the distortion tests though?  Yeah you won't get the exact number, but do you really need that?

And remember I was using a really high-end $1800 ADC with RMAA to make those measurements. And I was doing it at known levels, with known loads, etc. Others may not get nearly as similar results with more typical RMAA usage.


Yeah, but heres the thing:  the stuff people actually want to measure doesn't require a good ADC like that.  A good onboard sound card will tell you if the frequency response is flat from 100Hz to 16kHz, if the noise floor is unreasonably high, and if theres a lot of distortion.  If I think about audio quality at all when buying a portable player, those are the only three numbers I'd really care about. 


24-bit audio proposed for iTunes

Reply #99
But are those really so useful to know?  Pretty much every player in existence has an output level with 3-4 dB of every other player, since they're all running off the same supply voltage.  Yeah its interesting to know if your player is marginally louder then another player, but theres so little difference between players I'd hardly consider it critical.  Likewise, if a player a sufficiently flat frequency response, is it really that important to know the precise output impedance?

Yes they are important! The load interacts with the output impedance in a large ways that have little to do with the power supply voltage. And some amps simply don't have the ability to drive low impedance loads to typical levels before the distortion goes off the scale. For example, most portable headphones these days are 16 ohms. And into 15 ohm loads, I've measured anywhere from 179 mV RMS at 1 % THD up to 800+ mV. That's a range of 2.1 mW up to 42.7 mW--a 13 dB difference not a 3-4 dB difference as you suggest.

For example, here's what happens with a 6 ohm output impedance device into typical balanced armature (Ultimate Ears SuperFi in this case) headphones:



Does 4+ dB of variation matter? Some devices have 20 - 200 ohm output impedance--far greater than the 6 ohms show above. And, obviously, cause far greater deviations from flat frequency response.

Wouldn't DAC nonlinearity show up in the distortion tests though?  Yeah you won't get the exact number, but do you really need that?

That's a fair question, and based on what I know and have seen, the answer is: "not usually". It's really tricky to take accurate distortion measurements at very low levels as many of the harmonics are lost in the noise. And if it's THD+N the entire measurement is dominated by random noise, not the non-linearity error. And when you test at higher levels, other sources of distortion usually dominate the measurement. So DAC nonlinearity can be lost either way.

Yeah, but heres the thing:  the stuff people actually want to measure doesn't require a good ADC like that.  A good onboard sound card will tell you if the frequency response is flat from 100Hz to 16kHz, if the noise floor is unreasonably high, and if theres a lot of distortion.  If I think about audio quality at all when buying a portable player, those are the only three numbers I'd really care about.

So far I've been concentrating on portable equipment, and for that, I mostly agree with you. But I don't have a dScope just to test $29 portable players. That said, even the iPod Touch 3G will give many sound cards a good run for their money in terms of noise and distortion. So it's useful to get real numbers and not be flirting with the limitations of the PC's sound hardware.

It's also worth pointing out there are all sorts of other measurements I could be showing results for that I haven't (yet). They include  linearity plots, digital deviation and jitter (for digital interfaces), the output power vs THD, various phase measurements, impulse responses, etc. My reviews are already long enough for now.

So the bigger issue is all the things that RMAA and soundcards don't measure directly at all. And part of it is convenience. If it takes 3 days to run a full battery of tests using a kludged together collection of "tools" you're a lot less likely to ever do all the tests in the first place. If it's a matter of connecting the device, setting a calibration level in the dScope software, and clicking the "GO" button to run the script, that's very different.

RMAA is great in that it's easy to use. But, as I've explained, it's very much "garbage in/garbage out" and you're still missing lots of important things (including what we just discussed above). To get the rest of the numbers with a soundcard, meter, etc. is considerably more work. And even with all that there are still typical specs you can't measure (or measure very well). Like I said, RMAA and soundcards are certainly useful. But I don't think it's accurate to imply an audio analyzer has little advantage making real world measurements.