How to compare amps more fairly - my experience

I don’t believe that’s a valid conclusion. SINAD is not a comprehensive statement of quality of a device–it’s a measurement at a single reference point (typically 1kHz at certain fixed loads). It’s a good indicator, but we’re not listening to 1kHz sine waves, and headphones don’t behave like simple fixed loads. Actual performance can deviate when you move away from the measured reference conditions.

All of the amps I own are > 100db SINAD, but under the right conditions I can hear repeatable differences between them. The 600 ohm Beyerdynamic DT880 is a notoriously difficult to drive headphone. It also features a non-removable single-ended cable, meaning that without modifications you’re stuck with an amp’s SE performance. Despite the extraordinarily high SINAD of my THX 887 amp, its power rating at 600 ohms is significantly lower than the power rating of my V280. I hear a clear difference in performance between those two amps with those headphones.

Interestingly, I do not hear such a dramatic difference between those amps with my Beyerdynamic T1.2s, which are also 600 ohm headphones. What gives? Well, take a look at the sensitivity ratings for these two headphones:

DT880 Edition, 600 Ohm

  • Nominal Impedance: 600 ohm
  • Nominal Sound Pressure Level (Sensitivity): 96 dB

Source:

T1, 2nd Gen

  • Nominal Impedance: 600 ohm
  • Nominal Sound Pressure Level (Sensitivity): 102 dB (1mW / 500 Hz)

Source:

That might not look like a big difference, but 6dB difference in sensitivity means that 880s need 4x the power to match the same volume as T1.2! The similar impedance means that loudness should scale similarly. But 880s need a big head start. The math is described here:

I wasn’t aware of the difference in sensitivity between these two models until I looked into specs to see if I could explain my subjective listening experience. Sure enough, those numbers tell me a good story:
The 880s seem to need power just outside the ideal operating capabilities of a lot of mid-fi amps, whereas the increased sensitivity of T1.2 means they fall just inside the bounds of what those amps do well.

I don’t know that ASR measured the V280, but I’ll assume that it’s probably similar to the V281 that they did measure, given that both share the same amplifier section (the 281 has XLR outputs and preamp capabilities, I believe). In any case, the V281 has a SINAD* of 111 and the 887 is shown on the same graph with a SINAD of 119, but 880s clearly sound better on my V280.

* Note the chart actually says “Best Case SINAD.” The 600 ohm 880 does not allow the 887 to operate under best case conditions.

Incidentally, I have spoken to a number of people who have balance-modded their 880s and all have reported dramatic improvements in quality while listening balanced on amps like THX models and A90. It’s not the “balanced” part that’s doing that, but the fact that balanced outputs tend to have more power because of the differential signals.

I certainly wouldn’t say so. Although I have found a lot of cases where cheaper amps sound just as good as more expensive ones. There are good explanations for why they might sound the same:

  • bottlenecks in the audio chain (try YouTube music on a good amp, lol)
  • headphones are a good match for operating capabilities of the cheaper amp (see above example)
  • lack of critical listening experience or inattention to the right details
  • expectation biases, etc.

Another anecdote from my recent experience:

I mentioned my Holo May DAC above. Since I wrote that initial impression–the one where volume matching collapsed perceived differences on my loudspeakers–I have enabled the following audio chain: Both DACs → Mapletree Audio passive XLR switch → Goldpoint level control → McIntosh 8207 amp → Susvara. I have both DACs chained together in Roon, so whenever I play anything through the two DACs, I can use the Mapletree switch to immediately toggle between the two DACs. It’s not perfect, because as I discovered before, the May has a higher output voltage, and is thus about 3dB louder than my Ayre DAC (this chain doesn’t have the benefit of level trimming like my pre-amp chain does). So I have to fiddle a little bit to volume match.

Nevertheless, I have noticed that the May produces a larger, more spread out, more holographic soundstage than the Ayre, which sounds smaller and more intimate. For some tracks–notably stuff busy with spatial detail like “Down to the River to Pray” by Alison Krauss from the O Brother Where Art Thou? OST–this bigger stage with better imaging is awesome. But for tracks like “Krigsgaldr” by Heilung I actually prefer the more intimate staging, which plays well with the tension of that track as well as conveying a greater sense of being a participant in the music, rather than an observer.

I have no good explanation for why May sounds more holographic, though that does agree with some other impressions I have collected from others who have heard it. And I still didn’t notice the difference out of my loudspeakers, which might simply mean that it’s not important or that my loudspeaker setup is not configured optimally to convey differences. Anyway, I know what I hear, but I cannot explain why.

It’s also worth underscoring that neither one of these presentations is objectively better. I do have a great deal of respect for the heroic engineering that went into May to make such a clean R2R DAC, but that still doesn’t change the fact that I prefer the Ayre for some music. For that matter, the ADI-2 → A90 chain on my desk has a similar more intimate presentation, and is considerably cheaper than either of the DACs in my loudspeaker setup.

I’m personally thrilled that I do hear differences between my two DACs, because it lets me use the dial on my XLR switch like a soundstage selector.

8 Likes