iFi audio xDSD- The Official Thread
Apr 1, 2018 at 3:28 PM Post #121 of 2,505
Essentially my idea is that none of us listen to these devices without attaching headphones to them, so it seems to make most sense to test them out in as close to a real world situation as possible.
If i was to measure output power, then yes, I think your method would be best. But that won't tell me anything regarding the differences in their inherent sonic characters or the changes they bring about to headphone

Please don't take my comments as negative - that's not my intent.
I applaud your efforts here.

But, a key problem with your method here is that you are the measurement tool.
If you truly wish to compare the two sources, then measure their output and analyze that.

e.g. level match both devices on a 1 kHz (or whatever) tone. Play a song (or several!) through each and record the output.
Subtract the results from each other to observe the differences.
 
Apr 1, 2018 at 3:45 PM Post #122 of 2,505
Please don't take my comments as negative - that's not my intent.
I applaud your efforts here.

But, a key problem with your method here is that you are the measurement tool.
If you truly wish to compare the two sources, then measure their output and analyze that.

e.g. level match both devices on a 1 kHz (or whatever) tone. Play a song (or several!) through each and record the output.
Subtract the results from each other to observe the differences.

None of your comments were perceived to be negative..I enjoy these constructive convos.

I think doing things in that manner could work, but I feel that by eliminating the effect of the headphones (or the effect that the different sources would have on the headphone) we'd be eliminating a key part of the chain when considering real-world use. If we were only concerned with the absolute differences between the devices, then I would agree, your suggested method would probably be the better of the 2. But those measurements are for numbers and graphs. Whilst those types of measurements could give us far more accurate and reliable data for detailed analysis, those charts and graphs won't necessarily give us a great deal of information with regards to how we perceive the sound (not unless you're exceptionally knowledgeable and experienced in such a way that you could pin point a device's signature purely from measurements).

Generally there tends to be 2 camps on HF regarding these things......the one side says something along the lines of "we can only use accurate measurements, humans are too unreliable and different from one another", whilst the other side tends to think "measurements don't mean anything, trust your ears".
I tend to fall more towards the former camp, but I also see a need for a balance between the 2 ideas. There really isn't much point in testing any device (with regards to this topic and these types of devices) in a manner that is completely divorced from how that device would be used in a real-world scenario.

Hope that makes sense :)
 
Last edited:
Apr 1, 2018 at 5:58 PM Post #123 of 2,505
None of your comments were perceived to be negative..I enjoy these constructive convos.

I think doing things in that manner could work, but I feel that by eliminating the effect of the headphones (or the effect that the different sources would have on the headphone) we'd be eliminating a key part of the chain when considering real-world use. If we were only concerned with the absolute differences between the devices, then I would agree, your suggested method would probably be the better of the 2. But those measurements are for numbers and graphs. Whilst those types of measurements could give us far more accurate and reliable data for detailed analysis, those charts and graphs won't necessarily give us a great deal of information with regards to how we perceive the sound (not unless you're exceptionally knowledgeable and experienced in such a way that you could pin point a device's signature purely from measurements).

Generally there tends to be 2 camps on HF regarding these things......the one side says something along the lines of "we can only use accurate measurements, humans are too unreliable and different from one another", whilst the other side tends to think "measurements don't mean anything, trust your ears".
I tend to fall more towards the former camp, but I also see a need for a balance between the 2 ideas. There really isn't much point in testing any device (with regards to this topic and these types of devices) in a manner that is completely divorced from how that device would be used in a real-world scenario.

Hope that makes sense :)
The method described by Theo to evaluate different sources is the one that works best for me personally. Now, of course it is a bit subjective, but it's based on what I consider to be the essence of a proper comparison: "all other things being equal". To resume the car/engine analogy, it's like trying different engines and different chassis, on the same race track and with the same tyres. An engine might have more power but less torque, a chassis may be oversteering or understeering, in the end it's all about what combination works best for the intended driver. A purely measurement-based comparison gives a lot of useful information, which helps dividing good stuff from crap, but ultimately it's about what source makes certain music sound better through which transducers. Like with sports cars, I'm always looking for what gives me more fun for the buck. To choose wisely is not easy, but both specs/measurements together with the subjective, educated experience of trusted users can be very helpful.
 
Apr 1, 2018 at 6:30 PM Post #124 of 2,505
A purely measurement-based comparison gives a lot of useful information, which helps dividing good stuff from crap, but ultimately it's about what source makes certain music sound better through which transducers.

I would clarify that measurements are the only thing that can provide solid evidence as to why different components may "make certain music sound better".

You're not curious whether the differences you hear are manufactured internally or externally to your perceptions?

but both specs/measurements together with the subjective, educated experience of trusted users can be very helpful.

Fully agree: both approaches are helpful and can even be complementary.
But there are no published measurements in this case, only subjective comparison.
To be fair, there really is nothing objective to share if "sounds better/worse/different" is the goal.

Not necessarily a bad thing - I appreciate the efforts behind arriving at these particular opinions.
 
Apr 1, 2018 at 6:52 PM Post #125 of 2,505
I would clarify that measurements are the only thing that can provide solid evidence as to why different components may "make certain music sound better".

You're not curious whether the differences you hear are manufactured internally or externally to your perceptions?



Fully agree: both approaches are helpful and can even be complementary.
But there are no published measurements in this case, only subjective comparison.
To be fair, there really is nothing objective to share if "sounds better/worse/different" is the goal.

Not necessarily a bad thing - I appreciate the efforts behind arriving at these particular opinions.

Unfortunately the "human factor" is all too real, and there are plenty of things that can influence our perceptions....there's simply no getting around that.

However, I think, to some extent, we, as users and enthusiast, have perhaps become too reliant and obsessed with specs and the most minute of measurements that go far beyond what our senses allow us to perceive anyways. For example, if we look at amp A and amp B, we might see that A has a THD of 0.001%, and B has a THD of 0.04%. Well, technically, does B not then have 40x the distortion, and hence 40x less accurate? But would any of us ever be able to truly appreciate that? Heck it's entirely possible that some or even most of us enjoy B more than A, despite it being objectively a worse performing product. Of course, THD is just one parameter, there are other things to consider, but i think that the basis of that example could be applied in varying degrees to other parameters too .

Anyways, I get what you're saying about trying to be more objective. It's for this reason that I'm very careful with the language I choose to use when describing signatures and other arguably subjective matters. My impressions are simply intended to serve as a relative, not an absolute reference. Heck, I don't think any person's perception could/should be used as an absolute reference.
 
Apr 1, 2018 at 7:11 PM Post #126 of 2,505
You're not curious whether the differences you hear are manufactured internally or externally to your perceptions?
Actually, I'd be more curious as to the "why" of the differences. Since I'm not able to modify perceptions internal to my brain, I'm only left with the option of intervening on the external causes of perception. In a broader sense, my personal goal is to be able to hear music the closest to its original sound when it was recorded, which may have its downside when a recording sucks. Its a bit like an original, untouched hi-res photograph vs a photoshopped one: I'd always go for Richard Avedon, not David Hamilton lol
 
Apr 1, 2018 at 8:08 PM Post #128 of 2,505
My impressions are simply intended to serve as a relative, not an absolute reference. Heck, I don't think any person's perception could/should be used as an absolute reference.

Point well taken.

However, I think, to some extent, we, as users and enthusiast, have perhaps become too reliant and obsessed with specs and the most minute of measurements that go far beyond what our senses allow us to perceive anyways. For example, if we look at amp A and amp B, we might see that A has a THD of 0.001%, and B has a THD of 0.04%. Well, technically, does B not then have 40x the distortion, and hence 40x less accurate? But would any of us ever be able to truly appreciate that?

I also agree that the context of specs is very important to depth of understanding.
I, for one, would appreciate the THD of A over B in your example, from a perspective of the engineering involved in A, and lack thereof in B.
However, as a listener, knowing that my headphones have 0.1-1% THD (or more) puts the applicability of the THD of any source in context for comparison.

I am of the opinion that the specs most manufacturers use are nearly entirely marketing-driven.
Most specs reflect a single point on the audio spectrum or are averages over the entire range. Both are minimally useful to draw any conclusions from.

At any rate, thanks for the chat.
Keep up the good work.
 
Apr 3, 2018 at 8:55 AM Post #129 of 2,505
IMG_0589.jpg


So far so good with the CV5 and Cascades.
 
Apr 4, 2018 at 12:08 AM Post #131 of 2,505
I tried a rough (ie: Not level-matched) comparison with the AK380, CV5 and Utopias and the level of detail is pretty good, and I'm not picking up on anything unpleasant either.

Since the power comes at lower impedances, I'm trying it direct to the Aeon Flow Open with good results using a balanced adaptor.
 
Apr 4, 2018 at 5:32 AM Post #133 of 2,505

Users who are viewing this thread

Back
Top