Atoll DAC200 Signature - Review & Measurements (DAC)

[amanieux]

The DAC100 has been tested and measured by the German magazine Audio, issue 08/2010 (back issues do surface from time to time on Ebay). From the published FFT, I have calculated a THD over H2 to H5 of about 0,057 %. Thus, performance is on par with the DAC200 despite the fact that the DAC chip in the DAC100 is totally different - a Burr Brown PCM1796. This observation about the DAC100 is one more evidence that the harmonic distortion spectrum of the DAC200 is produced by the analogue output stage rather than something else.

is dac100 relevant as of today ? i can get one used for 100 euros but is it worth it ?

 

[VintageFlanker]

is dac100 relevant as of today ? i can get one used for 100 euros but is it worth it ?

You can get some SMSL SU-1 for 79€. Just sayin'.

 

[amanieux]

You can get some SMSL SU-1 for 79€. Just sayin'.

thats what i guessed, thank you

 

[Tonio2000]

Interesting discussion, isn't it?

A theory to explain why end users look happy of the DAC200 despite the technical performances are disastrous ? How cannot these weaknesses be heard on high end systems ?

Then, is it common to observe that performances are that far away from specification requirements ?

 

[Johnkeizer]

This would be a perfect partner for the PSAudio BHK Signature preamp.
Although I seem to recall some test results which suggest PS Audio has it’s own matching problem dac.

 

[RandomEar]

Interesting discussion, isn't it?

A theory to explain why end users look happy of the DAC200 despite the technical performances are disastrous ? How cannot these weaknesses be heard on high end systems ?

This isn't difficult to explain. Most people - including me, probably you and definitely 99.9% of all audiophiles - vastly overestimate their hearing capabilities and lack a basic understanding of the tech behind audio as well as the science behind signals processing and psychoacoustics. The measurements of the DAC200 are desastrous but if you look at the performance in detail, it's pretty clear that most defects will be difficult to hear as is and definitely very difficult to hear without a direct A/B comparison to a competent DAC with instant switching. Let's look at some specs:

• THD + N -68.2dB: Really bad, but luckily distortion-limited. Most people fail to identify distortion in actual music below -40 dB in controlled testing, with some reaching a bit below -50 dB with pure tones. That means anything below -50 dB is pretty safe to go unnoticed outside of a controlled testing environment and values below -60 dB are very likely to go unnoticed even in controlled testing. So -68 dB is just at the edge of "good enough".
• IMD -55.7dB: Worse than THD and could potentially be detected by some people in controlled listening tests using pure tones and quick A/B switching. But that's not how people use DACs: They use them to listen to music or podcasts. So these IMD levels are still extremely unlikely to get noticed.
• Noise is -115.2dBA, DR is 112.6dB. Both are not state of the art, but still respectable values and should not lead to any clearly audible defect. Both are better than Red Book material (audio CD), so not something to worry about.
• FR is down 1.3 dB at 20 kHz, but only 0.35 dB @ 10 kHz and 0.74 dB at 15 kHz. Only about 10% of people over 30 can hear past 17 kHz at relevant SPLs. At 45, that limit drops to 15 kHz for the same percentile. The detection limit for broader FR changes is around 0.5 dB as far as I remember. Therefore, I would expect this to be audible for some people with some tracks, but it will not be blatant. In fact, outside of controlled testing and without a direct comparison to a good DAC this will likely go unnoticed or at best create the impression of very slightly "mellower" treble.
• Jitter looks bad, but is still close to CD quality material and will be drowned in distortion anyway on this device.

And that's just one side of the coin. Most people don't conduct detailed listening tests with their DACs or if they do, they don't know how to do so properly and with the necessary controls. So they might "hear" differences, but those are not the defects of the DAC discussed here, but simply placebo.

Then, is it common to observe that performances are that far away from specification requirements ?

No. This is uncommon for mainstream manufacturers, but happens occasionally with DIY and "boutique" manufacturers. I would guess that they simply lack the engineering expertise to properly design and/or measure their products.

 

[Tonio2000]

This isn't difficult to explain. Most people - including me, probably you and definitely 99.9% of all audiophiles - vastly overestimate their hearing capabilities and lack a basic understanding of the tech behind audio as well as the science behind signals processing and psychoacoustics. The measurements of the DAC200 are desastrous but if you look at the performance in detail, it's pretty clear that most defects will be difficult to hear as is and definitely very difficult to hear without a direct A/B comparison to a competent DAC with instant switching. Let's look at some specs:

• THD + N -68.2dB: Really bad, but luckily distortion-limited. Most people fail to identify distortion in actual music below -40 dB in controlled testing, with some reaching a bit below -50 dB with pure tones. That means anything below -50 dB is pretty safe to go unnoticed outside of a controlled testing environment and values below -60 dB are very likely to go unnoticed even in controlled testing. So -68 dB is just at the edge of "good enough".
• IMD -55.7dB: Worse than THD and could potentially be detected by some people in controlled listening tests using pure tones and quick A/B switching. But that's not how people use DACs: They use them to listen to music or podcasts. So these IMD levels are still extremely unlikely to get noticed.
• Noise is -115.2dBA, DR is 112.6dB. Both are not state of the art, but still respectable values and should not lead to any clearly audible defect. Both are better than Red Book material (audio CD), so not something to worry about.
• FR is down 1.3 dB at 20 kHz, but only 0.35 dB @ 10 kHz and 0.74 dB at 15 kHz. Only about 10% of people over 30 can hear past 17 kHz at relevant SPLs. At 45, that limit drops to 15 kHz for the same percentile. The detection limit for broader FR changes is around 0.5 dB as far as I remember. Therefore, I would expect this to be audible for some people with some tracks, but it will not be blatant. In fact, outside of controlled testing and without a direct comparison to a good DAC this will likely go unnoticed or at best create the impression of very slightly "mellower" treble.
• Jitter looks bad, but is still close to CD quality material and will be drowned in distortion anyway on this device.

And that's just one side of the coin. Most people don't conduct detailed listening tests with their DACs or if they do, they don't know how to do so properly and with the necessary controls. So they might "hear" differences, but those are not the defects of the DAC discussed here, but simply placebo.


No. This is uncommon for mainstream manufacturers, but happens occasionally with DIY and "boutique" manufacturers. I would guess that they simply lack the engineering expertise to properly design and/or measure their products.

Thanks a lot for your detailed reply.
I am 49 years old and I own a DAC100 Sig associated to an IN100
That is the problem of expensive audio system: when you can eventually buy them, your ears are already altered.

 

[AudioSceptic]

This isn't difficult to explain. Most people - including me, probably you and definitely 99.9% of all audiophiles - vastly overestimate their hearing capabilities and lack a basic understanding of the tech behind audio as well as the science behind signals processing and psychoacoustics. The measurements of the DAC200 are desastrous but if you look at the performance in detail, it's pretty clear that most defects will be difficult to hear as is and definitely very difficult to hear without a direct A/B comparison to a competent DAC with instant switching. Let's look at some specs:

• THD + N -68.2dB: Really bad, but luckily distortion-limited. Most people fail to identify distortion in actual music below -40 dB in controlled testing, with some reaching a bit below -50 dB with pure tones. That means anything below -50 dB is pretty safe to go unnoticed outside of a controlled testing environment and values below -60 dB are very likely to go unnoticed even in controlled testing. So -68 dB is just at the edge of "good enough".
• IMD -55.7dB: Worse than THD and could potentially be detected by some people in controlled listening tests using pure tones and quick A/B switching. But that's not how people use DACs: They use them to listen to music or podcasts. So these IMD levels are still extremely unlikely to get noticed.
• Noise is -115.2dBA, DR is 112.6dB. Both are not state of the art, but still respectable values and should not lead to any clearly audible defect. Both are better than Red Book material (audio CD), so not something to worry about.
• FR is down 1.3 dB at 20 kHz, but only 0.35 dB @ 10 kHz and 0.74 dB at 15 kHz. Only about 10% of people over 30 can hear past 17 kHz at relevant SPLs. At 45, that limit drops to 15 kHz for the same percentile. The detection limit for broader FR changes is around 0.5 dB as far as I remember. Therefore, I would expect this to be audible for some people with some tracks, but it will not be blatant. In fact, outside of controlled testing and without a direct comparison to a good DAC this will likely go unnoticed or at best create the impression of very slightly "mellower" treble.
• Jitter looks bad, but is still close to CD quality material and will be drowned in distortion anyway on this device.

And that's just one side of the coin. Most people don't conduct detailed listening tests with their DACs or if they do, they don't know how to do so properly and with the necessary controls. So they might "hear" differences, but those are not the defects of the DAC discussed here, but simply placebo.


No. This is uncommon for mainstream manufacturers, but happens occasionally with DIY and "boutique" manufacturers. I would guess that they simply lack the engineering expertise to properly design and/or measure their products.

Just to add that with an item like this, which is expensive and looks very nicely made, we tend to be misled by our eyes and assume that at least basic/good performance *must* be present and correct.

 

[VintageFlanker]

Just to add that with an item like this, which is expensive and looks very nicely made, we tend to be misled by our eyes and assume that at least basic/good performance *must* be present and correct.

That and also reputation. Many times, the brand itself only is one of the main factor to build expectation bias.

 

[VintageFlanker]

Then, is it common to observe that performances are that far away from specification requirements ?

No. Here, we're talking about some disastrous case. This should not happen whatsoever.