One of the major challenges for a website or an online forum is to generate traffic. The all-important metric is Google Analytics, which provides all sorts of information about visitors. The higher the number of visitors and the longer they stay on your site, the more you can charge for your advertising space. That’s if you charge for banner ads and other advertising. Thus far, I haven’t tried to monetize this site. That’s not to say that I won’t at some point in the future but I started it to provide accurate information about high-resolution audio and to counter stuff that people say that’s hyperbole or just plain wrong.
In order to increase the number of visitors to your website, you need to publish articles or post comments that are controversial or otherwise compelling. These days there are a lot of pieces being written about whether “high-resolution” audio is irrelevant. There’s no better way to stir the pot than to attack a trending piece of technology. So I read a piece at AVS Forum by my friend Scott Wilkinson (I actually had a hand in that one.) and yesterday the folks at Home Theater Review posted a piece asking the same question.
Both pieces are well written and play pretty straight with the facts but there were problems with the graphics. If you watched me on Scott’s Home Theater Geek show (#204), you’ll know that I gave him some flack for getting the 16 vs. 24-bits graph wrong. It’s showed up in his AVS Forum…without being updated. So I took it upon myself to recreate it with the right information.
The original version is presented below in Figure 1:
Figure 1: The original AVS Forum Comparison Chart for word lengths and sample rate. [Click to enlarge]
As I pointed out to Scott, the chart is incorrect because the same red amplitude is being sampled at each of the three different combinations. It would seem to imply that you get great resolution with the 24-bits against the same waveform. The fact is you don’t. You get the ability to sample a wider range of amplitudes…each bit provides roughly 6 dB of additional dynamic range.
Here’s the way the charts should look:
Figure 2: The reworked AVS Forum Comparison Chart showing the additional range provided by the 24-bits and the additional accuracy provided by the additional samples. [Click to enlarge]
Then there’s the piece at Home Theater Review. I noticed their graphic on FB and immediately flinched. Here we go again, I thought. Whoever made the comparison chart of “resolution levels” doesn’t have a solid grasp on the realities of frequency response and definitely doesn’t know anything about PCM vs. DSD.
So here’s the original chart as presented on Home Theater Review:
Figure 3: This is the chart from HTR showing the “relative” fidelity merits of different specifications and formats. [Click to enlarge]
I know Sony is saying that “anything greater than CD” specifications should be considered “high-resolution”, but does anyone really think that moving from 44.1 kHz to 48 kHz is going to be audible? Or that a DSD “1-bit” recording at 2.8224 MHz (probably sourced from an analog tape?) is going to provide more dynamic range and frequency response than a 192 kHz/24-bit PCM file (or for that matter any great fidelity than a standard CD)?
So why are the DSD formats on the rightmost edge of the chart and why are the bars the longest?
Here’s a reworked version of the chart that shows things more accurately positioned. I know some DSD advocates will want to challenge my ranking but if you look at the spectra of a well done CD and a typical DSD file…they are about the same. I’m not saying the sound is the same, but the “fidelity box” that can encompass the frequency response and the dynamic range is about the same. High-resolution PCM encoding provides the potential for much greater dynamics and frequency response.
Figure 4: The HTR chart reworked to show the three relative levels of resolution…and the proper placement of the DSD format. [Click to enlarge]
These aren’t the worst charts I’ve ever seen (that distinction goes to the SuperHiRez.com placement of DSD), but they did miss the mark. We need to be able to use graphics in a way that communicates real information. I know it’s hard to avoid spin and biases but consumers need to know.
I’ll present a chart soon with the “fidelity box” approach soon.