This is a very common type of comment here and also a fallacy. It is intentionally ignoring DACs that are “broken” or just very old, or amplifiers which are extremely underpowered and rich in distortion.
Not all devices are like the highest SINAD ones reviewed here. And there was a reason the first DACs were considered to have a harsh sound.
So I would not discard certain comments with laughs. But ask first what they listened to. Yes, sighted comparison is biased. There is confirmation bias. We all know that. And: Two decently built recent DACs or amplifiers since a few decades, when used in similar conditions, do not contribute to audibly different final outcomes, this is established. There are, however, other conditions. Or are we to assume that a 2A3 single ended amp will provide the same sound out of a pair of LS 3/5a than a Purifi module at reasonable listening levels? Because in this case science tells us it cannot be the case, because of the high distortion when you start needing 5-6W of power.
You are misinterpreting my statement.
I was referring to the comment made by the user a few posts back—#637—where he mentioned specific modern DACs with which he heard differences. If you hear differences with devices like that, either something is defective, or the differences are extremely marginal. The latter usually refers to minimal differences in resolution rather than tonal characteristics.
Over the last 26 years, we have compared well over 1,000 devices in level-matched blind tests. I am very familiar with the claims guests make about the differences they hear in the equipment they bring along. Yet, in a level-matched blind test, those differences suddenly vanished—or, in a few cases, remained only marginally perceptible.
Differences between identical units—or those caused by component variance—are also extremely rare, if not unlikely.
We have frequently compared multiple units of the same model from different production batches for this very reason—models like the E30, E50, SU-1, SU-10, DO400, etc., and even six different D1 units. It was a very arduous process but yielded no results; in other words, they were indistinguishable. Of course, one must ensure that the settings on the DACs are exactly the same.
Exceptions clearly include defective units—though the fault is usually very audible—as well as extremely poorly designed devices.
Another factor involves devices that have been intentionally (or unintentionally) "voiced" to produce a specific sound.
And that brings us to a crucial point.
Here, we are concerned with reproducing the music exactly as it was recorded, not with interpretation.
Please take a close look at the source data. You will find that it contains absolutely precise information, whether analog or digital.
If a device is incapable of reproducing that information accurately, it can only be considered defective; it has failed to perform its primary function.
This applies equally to older DACs and digital devices that produce a harsh sound. Most of that is absolute nonsense; if such devices actually sound "harsh," that would be evident in the measurements as well. Ironically, this "harsh" sound almost always originates in the analog stage.
I still use a Cambridge Audio DACMagic from 1994 (designed by John Westlake), which clearly demonstrates—even in direct comparisons—that the challenge of DAC design was already solved 32 years ago.
Devices with a deliberately (or unintentionally) "voiced" sound signature are acceptable, provided the manufacturer discloses this and can reliably explain how it affects the music. However, such units can no longer be compared to standard equipment.
If the manufacturer does not explicitly disclose this, the device must be considered defective—and the situation regarded as a fraud against the customer.
