Audiophiles and subjective impressions - Page 2

The objective graphs could do with a face lift of clarity, might help convince some people of their message then. I'd argue the graphs in the original post don't convey meaningful information to the casual viewer, i.e. they may not be serving their purpose in the context of this thread.

For instance, rather than plotting absolute values, might just graph the difference between what's called bad and what's called good. But even then, it needs to be established why it matters in the first place that this line runs parallel to the other.

A good graph is one that conveys the general trend in the data with as much clarity as possible. If the goal is to show that two items drawing polarized reviews are identical, a good graph might show this as two straight lines, indicating no difference between the two items. For sure, if there's no difference, you can plot the absolute values and have two squiggly lines run parallel to each other, but I'd suggest there's too much detail in a graph like that. That is, if the idea is to show the similarity, it probably doesn't matter whether there was +5 dB here and -7.5 dB there in both cases, but whether there was deviation here or there in one from the other.

The thing is though, on these forums after someone posts and explains such graphs the responses are often excuses:

Music is art, numbers don't matter.

-> Music may be art but electronics is a science. This excuse basically says that if the sound quality is screwed up it's art when in reality it's just bad engineering.

The product was initially measured but finally tuned by ear.

-> So the product was screwed up after doing proper measurements. Or maybe it wasn't measured at all..

The product was designed to sound like <enter extremely expensive/audiophile product with crappy measurements (roll-off, distortion ...) here>.

-> Even renowned companies sometimes make crap products but that's no reason to duplicate their failure. This might also be a genuine statement, because some of these badly measuring products actually get positive reviews in hifi mags (and some people take those reviews seriously!).

If you like the sound it doesn't matter (how bad it is).

-> I like my $5 ear buds.. does that make them high fidelity? No way.. If the product is of low fidelity it shouldn't be marketed, priced, discussed like a high fidelity product.

It sounds better to me than <insert non-audiophile or cheaper product with good measurements here>.

-> Bias. Volume mismatch. No blind comparison. Skewed hearing.

And so on..

Yeah and when confronted with the idea of bias in cable reviews it doesn't take long until the "then why do our impressions match?" question shows up.

Doh!

Quote:

Originally Posted by skamp 

The point is not just to show the similarities between pairs of graphs, but also to show the differences between graphs.
The OP shows that two equally bad graphs draw different opinions, but also that one (significantly) bad graph and one good graph can draw similar opinions. All the graphs convey information that shouldn't get reduced to anything.

Edit: and in case that wasn't obvious, the graphs show the frequency response of various DAPs, loaded with non-linear impedance IEMs.

You'll want to explain all of that along with the graphs, preferably explaining what's going on in each graph and what makes them good or bad, since you can bet any impressionist will not care (for solid reasons) to attempt to interpret the graphs on their own. Assuming the intent of this thread is to open some eyes rather than to bash this or that point of view.

I wouldn't call it sound quality either. Synergy is the term they like to use. It's a bit like trying to pile up pieces of rock when you could just use bricks.

Quote:

Originally Posted by skamp 

I was perhaps too presomptuous to assume that everyone would be able to read the graphs.

• First two graphs: DAPs with high output impedance, loaded with non-flat impedance IEMs, result in emphasized mids and recessed highs.
• Graphs 3 and 4: DAPs with low output impedance, resulting in a "high fidelity" signal, close to the recording, with no deviation whatsoever.
• Graph 5: DAP with high output impedance loaded with a particular set of cans, resulting in recessed bass and highs.
• Graph 6: DAP with high output impedance, loaded with various cans, resulting in widly varying frequency reponses (emphasized mids, and emphasized or recessed highs, depending on the headphones used).

Thanks. Now, if the explanations found their way into the first post, that would be even better.