Ableton Forum

Hi guys,

I am now running a poll on my website.
Basically (if you wish) it requires you to go to http://mrbillstunes.com/ listen to the 2 files and tell me which one you think sounds better.
Much appreciation for anyone who participates. It's completely anonymous also.

Nope

I tried, but got a message about needing a codec, which I think is odd, since I'm using a Windoze machine and it's a .wav file.

Hm, yeah that's strange, you should be able to DL the files man and check them out that way.

If you're using Windows Media Player it can be a bit picky about 24 bit files.

I recommend you download the free open source VLC player if you're playing it back in windows.

I have set my win preferences to open any audio files in Soundforge, so had no problems listening.
You could also open it in Live or another DAW and listen.

both sound the same here

Thanks Leeds... am trying Live here in a minute. Was at work on a locked down machine the first try.

leedsquietman wrote:I recommend you download the free open source VLC player if you're playing it back in windows.

I'm on a Mac and I still use VLC. Been a big fan for a long time.

http://www.mrbillstunes.com/post_201012 ... fb31a0c756

Here are the results of this.

ok, I'm intrigued by your conclusion though.
First you say that "These files phase out completely when played together, meaning there isn’t any difference in them at all that the DAW can pick up."

but then you go on to say that because "35.5% of people believe that realtime rendering sounds better and 24.4% of people like offline rendering better" (despite the fact that you've just proven that there isn't any difference whatsoever) it's "probably safe to say that your better off rendering your stuff in realtime".

It would seem to me that the result of who likes what better will be completely random every time you repeat the test if the files are actually mathematically the same. The only conlusions you can make from the result of this test are "it doesn't matter whether you render realtime or bounce, because the result will be virtually (and measurably!) identical" and "60% of people think they hear something that's not actually there" And maybe "you can't always trust your ears"

Yeah, well you can make that conclusion once you conduct the test over and over again .

For now, the results show that out of the 60% of people that did think they could hear a difference, that realtime rendering sounded better. You can assume whatever you want, but until you actually conduct research to substantiate your theory you can't really be sure. Even if there is no phase differences to the files, people still think they are hearing one to sound better to the other. I don't know why, maybe it's a subconscious thing? But the statistics showed that people thought they could hear a difference. In addition to this, people included comments saying what they believed to hear in differences, and they also provided what speakers they listened to these examples through. There was a fairly common statement made on the realtime rendering saying it sounded "wider" and "less constrained".

In any case, I think realtime sounds better so I'll continue to render my music this way I trust my ears.

You would need to randomise the order of playing of the files for such a test to possibly be meaninful.

60% probably isnt significant - how many responses?

90 results.

gibson_ewok wrote:Yeah, well you can make that conclusion once you conduct the test over and over again .

For now, the results show that out of the 60% of people that did think they could hear a difference, that realtime rendering sounded better. You can assume whatever you want, but until you actually conduct research to substantiate your theory you can't really be sure. Even if there is no phase differences to the files, people still think they are hearing one to sound better to the other. I don't know why, maybe it's a subconscious thing? But the statistics showed that people thought they could hear a difference. In addition to this, people included comments saying what they believed to hear in differences, and they also provided what speakers they listened to these examples through. There was a fairly common statement made on the realtime rendering saying it sounded "wider" and "less constrained".

I'm not a scientist but I'm pretty sure the reason why some people think they hear a difference is because because of the way the experiment is presented, they expect to hear a difference. By way of your explanation of the experiment ("tell me which one you think sounds better") you're already planting it in people's minds that there must be a difference. And while your single measurement of phase alignment may not be scientifically sufficient to draw any hard conclusions, it still pretty much qualifies all the statements of people who claimed one sound was "wider" or "less constrained" to be, well... bogus

gibson_ewok wrote:For now, the results show that out of the 60% of people that did think they could hear a difference, that realtime rendering sounded better. You can assume whatever you want, but until you actually conduct research to substantiate your theory you can't really be sure.

You need to google counterbalancing, significance testing, and type I errors.