In A Blind Sound Comparison, People Can’t Tell The Difference Between A Sequential Prophet-5 Rev 3 & Rev 4

[embedded content]

One of the most hotly contested questions about Sequential’s Prophet-5 Rev4 synthesizer is how close it sounds to the classic Rev3 original, which sells for about $5,000 more than a Rev4.

To help answer this question, synthesist Alex Ball recently shared a blind test, embedded above, comparing a vintage Prophet-5 Rev3 against a new Prophet-10 rev4. The video is set up as a blind test, where you see the same clip twice, but in one you’re hearing the Prophet you see and in the other you’re hearing the Prophet you don’t see.

Now Ball has shared the results of this comparison, and it turns out that – at least in this blind test – people can’t accurately tell the difference between the sounds of the two synths.

[embedded content]

People guessed the correct synth 55% of the time, slightly better than what you’d expect from flipping a coin. And nobody had ‘golden ears’ good enough to let them consistently tell the difference between the two synths.

These results were to be expected, because the Rev4 is not intended to be a new synth, but a reintroduction of the classic Prophet-5 design.

What’s probably most interesting about the test, though, is that people did guess a little better than average, and they did better at guessing some sounds than others.

This reflects the fact that there are differences in the two synths’ designs and capabilities. And, on some sounds, these differences are more noticeable than others.

For example, the Rev4 features the ‘Vintage’ knob, which lets you vary the level of per-voice analog variation of individual synth voices, from ‘temperamental’ Rev1 to ‘stable’ Rev4. This is an awesome feature, but it’s not going to make a Rev4 sound exactly like a specific vintage Prophet-5 Rev3, just like some vintage Rev3.

The way these sort of differences show up in the test results is that there are a few sounds where users guessed a lot more accurately than others. For example, on the third sound in the test, there are more audible differences, and people guessed the correct synth about 66% of the time. This suggests that the sonic differences in some patches may make it easier to identify one synth vs the other.

But there are also patches where people did worse than what you’d expect from flipping a coin. People guessed wrong on the third sound test 57% of the time.

Doing 9 simulated coin toss tests shows how there’s a lot of variation from 50% when using small sample sizes.

Unfortunately, the small sample size of this test (148 people) means that some of variation in the accuracy of people’s guesses is indistinguishable from what you’d see if you just flipped a coin.

If you flipped a coin 148 times and guessed heads, you might expect be right 50% of the time. Try it a few times, though, and you’ll find that you may get ‘heads’ 40% of the time on one test, and 55% on the next.

It takes a lot of coin tosses to minimize this random variation. Because of this, small sample sizes make it harder to distinguish the difference between the variation of actual people making binary choices and the variation you’d see from just flipping a coin.

If we had to judge the results of this blind test, the loser would be our ears – the differences between these synths are subtle enough that people can’t accurately distinguish them from hearing their sounds. On the other hand, the winner would be our wallets. If you want a Prophet-5, the Rev4 sounds like a classic, costs a lot less and adds velocity, aftertouch, USB support and other features.

Check out the blind test and results – and then share your thoughts on the results in the comments!