What should a speaker sound like?
I have used quite a few technical terms here. Unfortunately sound reproduction is very technical. Because there are so many terms I have not provided links, but you should look up any where the meaning is not clear.
A more interesting question might be: "what should a sound system or hi-fi sound like?'
High fidelity reproduction essentially means high faithfulness (to the original). Therefore if something is hi-fi it should be as similar to the original as possible, given the constraints. Some constraints might be cost, size of the speakers, room acoustics of the listening room, the neighbours, because they might not like the 1812 Overture at full volume.
So to reproduce something and then say it sounds 'light and airy' or some other description, grates on me, for two reasons, because these things (the terms light and airy) are not defined, and because a good system should not sound of anything except the music or soundtrack that was played during the recording process. I say 'played' not 'recorded' because I think the audio and mastering engineers should be as transparent as possible, too.
I am still not sure why a valve amp with 0.5% total distortion and noise can sound 'better' than a transistor amp with .0005% total distortion and noise. What I do know is that if a loudspeaker exists with 0.1% distortion over the frequency band, then we will all have something exciting to listen to. After all the speakers are still the weakest link in the audio chain, well some might argue after the drummer, or the producer.
So I think music or a soundtrack should sound like it did in the recording studio, not even the control room, as the music has already been through several amplifiers, an analogue to digital converter or two, some affects 'boxes', and many, many op amps. As well as the studio's amplifiers and monitors.
So who knows what the music sounded like prior to you buying or downloading it (hopefully as CD or LP, not MP3) and listening.
So electronics that sound "better" do not do it for me. Instead I want to hear about electronic components (and speakers) that are not audible. I want a violin to sound like the violin that was recorded, a guitar to sound like a guitar, a singer to sound like they did in the studio or stage. I want all of the electronics that the signal went through to add and subtract as little as possible. And I want it in my living room, or wherever I am listening.
My point of comparison is the way the music sounded when recorded. For many listening tests we use classical music which has generally been through fewer processes in the studio than many other types of music. It is not that we don't like Bowie or Nirvana or Knopfler, but rather we understand what a piano should sound like.
One way to do comparisons
Recently a friend was telling me that he set up his very high end system, and that a small group of friends did some comparative listening. One test they did was to compare a CD and with the vinyl of the same recording. Having conducted some subjective testing in the dim distant past, I had to ask rather cheekily if they did some double blind tests. Well I think that it was a bit too late in the evening on that occasion to consider double blind testing. The issue with this type of comparison is that the CD was most likely remastered, and may have sounded different, but certainly the album would have sounded different.
I am sure they enjoyed themselves, but a late night listening session after a few drinks is not the way to do serious listening tests.
How we do our comparisons
About a year ago I went into a hi-fi shop with the intention of listening to some high quality speakers essentially as a point of comparison, but then I started to consider that a reference pair would be a good thing. Well I ended up building a pair, (no they won't appear here for sale, but they were given a model number 003 if you are interested). They were made up from some slightly older drivers that I had in the workshop that were very good in their day, with some described as some of the best made. They consist of an 8 inch woofer, 5 inch mid-woofer and 1 inch tweeter. Crossover duty is taken care of with a 'box of DSP', at 300 and 3kHZ, at 24dB/octave LR crossovers.
Our next step will be to move a piano into the listening room. When conducting comparisons, part of the procedure will be to compare the piano with a recording done at the same time.
Currently we compare to the in house reference pair, some surround sound speakers, some studio monitors (that you will find in many of the ABC studios) and anything else I can get my hands on. For amplifiers we use a gain clone amp using an LM3876 chip, a good quality class AB amp, a THX certified surround sound receiver (that weighs 19kg), and anything else I can find. Please don't ask what brand they are as we don't believe we should endorse one brand over another.
Ultimately we will be moving to comparison to musical instruments where we can compare the sound from the 'source' to the reproduced in the same environment (room).
I hope that this adds something to the debate, rather than adding noise to the discussion.