Home Propeller Head Plaza

Technical and scientific discussion of amps, cables and other topics.

Analysis of "Freezer Effect" sample files - if you're interested.

Okay,

If we're going to beat people over the head with scientific analysis, then we had better do some analysis when we're balking at someone's claims.

Posy Rorer said that "open minded" individuals had the courage to try the "Freezer Effect" and post their "results" on the internet for all to see. So I downloaded them.

http://www.mympxplayer.org/4-vt4234.html?postdays=0&postorder=asc&&start=45

The first character, moniker "lattesurf" described how he was going to do his test, as follows:

OK, after more than a month's of procrastination, i've finally dumped my player into the freezer earlier this afternoon.

To test on the freezing effect, i've used a line-out audio cable to record 3 different mp3s stored in my player, onto my laptop. This was before i popped it into the freezer.

When i'm done with the process in a few days, i'll record the 3 same mp3s onto my laptop again. Then compare the before and after quality. And if i find anything interesting, or if it really does work, i'll post the before and after mp3s to this thread."

Okay, so lets get this straight. This guy is recording the ANALOG OUTPUT of his MP3 player before and then after the "freezer effect" is applied. This means he's going to add the effects of a D/A convertion AND an A/D conversion of a track compressed with LOSSY compression.

Why did he not record TWO samples before and TWO samples after? At least if we could establish that his recording process could create to identical (if not similar) files in the first place, then we could perhaps attribute differences between the files to the "freezer effect". But no, we're left with this:

He is making the ASSUMPTION that any differences (and trust me - the differences are VAST) were caused by the 'freezer effect', but fails to realize that making two recordings of the same file at two different times are going to be UNIQUE anyways, due to the D/A and A/D conversion processes made by digital components with two completely independent clock sources. These clock sources are almost certain to have a certain degree of clock skew (drift) which is going to result in a unique sample set, even if both recordings had all converters synchronized at time=zero. But this is mute because such synchronization is simply not possible with Mr. Lattesurf's recording and playback equipment.

And what's up with the "fade in" added to the "Before" track? Did he just realize his levels were pooched and jack up the level mid test? Or did he have some sort of fading option selected? Anyways, I doubt the fade-in happened in anyones freezer...

What's better is Mr. Lattesurf's appraisal of his "results".

Read them for yourselves. I like the part where he can see "improved step response" by looking at a MP3 file. This guy must be able to do FFT's and convolution in his head - because I've never "seen" step response by looking at a file in a wave file editor. Then again, I would like to ask Mr. Lattesurf - "The step response of what?". I also want to know what his definition of 'instantaneous response' is when he says: "the After clip had an instantaneous response".

The other guy, Michiganjfrog, is a real clever trickster. The little devil put two IDENTICAL FILES on the net and asked people to hear the difference. :o) Man after my own heart...

We salute you Michiganjfrog! You belong here in PHP you skeptical naysayer you!

But no listening tests were necessary. Subtracting the "black" and "white" MP3 files results in zeros nice and neat across the board.

Well folks, I figure the D/A and A/D conversions Mr. Lattesurf is putting his test files through pretty much renders them USELESS to come up with any substantiation of the "freezer effect". I guess you could listen to these files and listen to the differences, but that's not going to provide ANY insight whatsoever as to what caused these differences. I think they're mostly caused by crappy recording techniques and use of MP3's, but hey - I've been wrong before.

Now this does not mean I am offering this as "proof" the 'freezer effect' does not work. I'm only suggesting we have not seen any tangible proof (or substantiable evidence anyways) via those sample files that were linked to. There are simply too many variables at work with Mr. Lattesurfs recording methods for them to be of value.

What this does mean is now we can discuss the following issue in a TECHNICAL AND SCIENTIFIC manner:

"Was Mr. Lattesurfs recording method useful for the purpose of demonstrating the 'Freezer Effect' or not? And why?"

I say NO for the reasons I have already outlined. I find it amusing that Posy thinks of Lattesurf as a person qualified to test audio concepts, but I am completely unqualified. I'll let my peers be the judge of that, and YOU Posy, are not presently among them.

So you guys (A)gree or (D)isagree?

Cheers,
Presto


This post is made possible by the generous support of people like you and our sponsors:
  Michael Percy Audio  


Topic - Analysis of "Freezer Effect" sample files - if you're interested. - Presto 00:15:46 04/12/07 (18)

FAQ

Post a Message!

Forgot Password?
Moniker (Username):
Password (Optional):
  Remember my Moniker & Password  (What's this?)    Eat Me
E-Mail (Optional):
Subject:
Message:   (Posts are subject to Content Rules)
Optional Link URL:
Optional Link Title:
Optional Image URL:
Upload Image:
E-mail Replies:  Automagically notify you when someone responds.