View Single Post
Old 4th Mar 2021, 3:57 pm   #2179
knobtwiddler
Octode
 
Join Date: Mar 2016
Location: London, UK.
Posts: 1,046
Default Re: The Audiophoolery Thread.

Modern speakers, even fairly modest ones, can have surprisingly low distortion if you're talking about steady tone 1KHz measurements. I don't work in speaker R+D, but I have casually slung test mics in front of decent studio monitors and observed that the highest harmonic when hit with 1K to be around -50 (inc. test mic's THD) relative to the fundamental. Where speakers really distort is at the limit of cone excursion, whereupon what was quite a neat looking FFT plot turns into a many toothed comb of all harmonics. Watching this behaviour on an FFT is quite instructive. It confirms what your ears will have already told you, i..e most speakers typically sound sweetest at modest levels. There is very little warning: the plot looks nice, and you increase the level by only a handful of dB and all hell breaks loose - it comes on brutally. There is also power compression to think about.

As for phono carts, if you know of one that renders the 2nd harmonic (highest one in the vinyl world) at -50, please let me know the model number. Testing carts has an enormous amount of variables. Because of inherent distortion in test records, alignment, generator used to make said test record, tonearm compatibility with DUT - and about a thousand other variables, it's pretty much impossible to give a general THD figure. However, if you have a trusted record and a bunch of different carts, again, the FFT can be highly instructive. The best cart I've measured at 1Khz gave 0.4% THD (an expensive MC cart). With the same record and tonearm, common-or-garden MM carts have given anywhere between 1.2% and 2% THD... I've seen a posh Japanese MC cart start out on the same test track with around 0.5%, and as the weeks go by and it's used more, it winds up looking just as ugly as a DJ cartridge (when new, the harmonics are nearly all 2nd and 3rd, and as it ages, the upper order harmonics grow, as do the lower ones). Again, this doesn't take tracking performance into consideration, so even if 2 carts give identical steady-state THD readings on the same groove, the one with the more agile cantilever and better quality tip will sound far cleaner in actual use.

And if you think vinyl is high in distortion, don't even think about running any form of analogue tape machine through your FFT analyser...particularly cassette...

I write the above with the disclaimer that any measurement of electro-mechanical devices is a 'generalisation' - and others with similar kit will get different values, depending on test media, sample variability etc. I'm also talking about steady-state tone testing at a given level, which is a vastly simplified model compared to real music. However, I think it's fair to say that - with regards to modern speakers - in terms of the ones I've played around with (studio monitors, as I like to hear imperfections such as ground hum and clumsy mute solos - call me weird), they are surprisingly good. As I wrote earlier in the thread, although the conditions of the value might be a little non-specific, I can believe ATC's 0.3% quoted THD figure.

Having said the above, if you define frequency response anomalies as 'distortion', then the speaker is the chief culprit my a mile (in combination with the room). Even high-end recording studios can have peaks and dips that are almost into double figures dB-wise... But the brain compensates very well for this, or the NS10 wouldn't be so ubiquitous, would it?
knobtwiddler is offline   Reply With Quote