This follow up to my review of Witherscape’s The Inheritance is a bit different from the norm. In all previous vinyl follow up articles I digitized the vinyl version by recording it to my computer, and used the digital “vinyl rip” (also known as a needle drop) to measure dynamic range. In this case that wasn’t necessary, because Dan “The Man” Swanö helpfully included the vinyl mix as 320Kbit MP3 files on the CD. He also included an extra MP3 version which I’ll get to in a minute. As far as I know this has never been done before, and frankly it’s an amazing gift to the metal community that cares about sound (you guys). More interestingly, Dan also put up a challenge to see if listeners could tell the difference between the standard, DR6 CD mix and fully dynamic, DR11 vinyl mix – a challenge I was happy to accept. As I said in my review of the album, Dan did a great job of working with the space he had, and for DR6, the standard CD version sounds pretty good. As good as the fully dynamic version though? Well let’s take a look.
Starting from the top, we have the “extra” MP3 version which I mentioned earlier. You can think of this as the “iPod mix.” It’s crushed even further down to DR5, and as you can see, Sound Forge found clipping in several areas. To be honest, I’m not entirely sure what the point of including this version was. The files are 192Kbit MP3s, and they sound noticeably worse than the CD version. Back when flash based MP3 players only had 1-2GB of space, compressing to 192Kbit was necessary to squeeze as many albums on the player as possible. Now though with 16GB largely being the standard, 192Kbit just isn’t necessary. Further, if somebody wanted to create 192Kbit files from the regular CD, iTunes or Windows Media Player can easily do that in about a minute.
Related Pages:
One thing that is very illustrative though, is the visual difference between the top DR5 version and the middle DR6 version. All of those little peaks in the middle version are just gone when the track is squeezed down to DR5 – just 1dB of additional compression does that. That’s the very lifeblood of the music being stamped out, and the sad fact is that most of the music we measure looks like the top version, with as much or often significantly more clipping. It’s very dispiriting.
Now look at the vinyl mix on the bottom. Glorious, glorious dynamics! This is what music should look like, and this is what music on CD did look like until about 1993 or so. Going back and forth between the three different versions, it was easy to tell that the vinyl mix sounded best, with more weight to the guitars and much more presence from the bass. The drums were also a huge improvement, with real sparkle and shimmer from the cymbals and a nice meaty punch from the kick drum.
There is an important factor to keep in mind here. Because I’ve seen the waveform for the vinyl mix and I know that it has nearly double the dynamic range of the CD version, that creates an “expectation bias” – I can trick myself into thinking it sounds way better because I think it should just by looking at it. In order to remove that bias and properly take the Swanö Challenge, the files have to be level matched, and the testing has to be blind.
This is what it looks like when the CD version (top) and vinyl version (bottom) are level matched to within 0.3dB. This is also what actually happens when I listen to a stupidly loud CD. I control the listening volume, not the record company, and when I listen to a CD and vinyl at the same comfortable listening volume, the only effect that the Loudness War has had is to suck all the life out of the music, making it far less enjoyable and fulfilling.
To make the blind listening test a little more challenging, and to dispel the myth that you need “audiophile equipment” to be able to hear any difference between crushed and dynamic recordings, I used a Logitech Z-5500 speaker set for the test. This was a very good computer speaker set in its day (still is, actually) but they are just little plastic computer speakers with small 3″ single drivers, being fed from the on-board sound of my computer. Hardly “audiophile grade” stuff here. Making the blind part very easy is Foobar2000’s extremely helpful ABX comparator tool.
Using the tool, track “A” and track “B” are always the same, while track “X” and “Y” are randomized each time the test is taken. This is called an “ABX” test. Just like a proper ABX test, Foobar’s comparator supports instantaneous switching with continuing playback. Any amount of silence between A, B, and X or having to start the track over each time can effectively ruin the test, but that’s an article (or maybe a Headbanger’s Brawl) for another day.
Well, with all credit to Dan’s very good CD mix, I was still able to win the challenge easily, even without a $10,000 sound system. I could’ve kept going all day, but the difference was clear each time. I also asked my girlfriend who is in no way an audiophile to take the test. She had never heard “Astrid Falls” before, and more importantly she had no prior knowledge that there was any difference between “A” and “B.” She was also able to easily win the challenge, and pointed out that one version had much better sounding drums than the other.
Here’s where I get on my soapbox. The only thing that hyper compression does is make a song sound worse. Period. It doesn’t sound better on earbuds, it doesn’t sound better on laptop speakers or car stereos, and it certainly doesn’t sound better on my Logitech computer speakers. It also does not sell any more albums. Alex and I are going to keeping beating our heads into the wall until this idiocy stops.
Everybody reading this needs to show Dan some support. Contact him and thank him for including the great sounding vinyl mix on the CD, and then go buy it, or buy the vinyl which includes the CD.