Henno
ex 00120
did score one of the 40

>They've fixed this with the mark2's:) (digital noise)

Just checked. Bloody hell, they have. Completely. Player on pause, amp on the end stop, and silence. Given how little time I spent listening to my Mk I play silence at full gain, it's not going to make a lot of difference, but it's nice that it's gone.
(Anyway, all you sound obsessives want to get yourselves a proper car to put it in.. finally found a picture of one of the cars I run the Mk I in, see if this works...)

Problems with attachments unfortunately

____________________________
Murray

I have to say that I am quite happy with the sound stage I have in my car. I first put in an Alpine CD before my Empeg arrived , and still have it with the Empeg now. It worked a treat and I did a listening test a while ago using the same track on MP3 and CD.
I could hear a difference in tone, subtle, but still different. I do prefer tracks off the Empeg though. for one they are normalized better to my system. (please don't start that debate here, I just like to have them normalized to 95%-98%).
All my tracks have been Ripped with Audio Catalyst, with occational tag edits by MP3 studio. Thanks tfabris, It's a great find.

I enjoy good quality audio but haven't found any reson to go with oxygen free cables, gold plated connecters and solder every joint. But then neither have most studios which create the music we listen to.

I concur that the only real way of finding the better system is to play WAVs from the Empeg.



____________________________
Murray

>Problems with attachments unfortunately

Oh well. It's also at http://www.race-cars.com/carsold/other/ffsx01/ffsx01pa.htm , but with older graphics...

steve

The isolation pad may not be improving the digital data from the CD itself, but all of the other components (power supply, etc) between the laser and the audio output.

Which I'd grant, if the components were something like a tube amplifier. I could see how transient vibrations could affect a vacuum tube. But everything inside a CD player is solid-state.

I could also see how mechanical isolation could prevent ground loops between pieces of equipment. Or how additional physical distance might prevent electromagnetic interference. But in both of those cases, there would have to be something wrong with components in order for an isolation system to be of any help at all.

I also found a url of a review of some isolation pads

I read those, and in neither case did the reviewer do a blind comparison.

Here is a fairly straightforward article on the need for blind testing in subjective tests. Without a blind test, you're opening yourself up to the possibility of subjective validation, wishful/selective thinking, etc. No matter how hard you try to be objective during such a test, if you know the expected outcome then you're going to have a tendency to fool yourself into thinking the evidence is there.

The soundstage of an audio recording is a very subtle thing. Almost as subtle as the dowsing example given in that article. The difference is that there are known factors which affect the soundstage of an audio recording and they can be reproduced in controlled tests.

It would be fairly easy to design a protocol that would blind-test the isolation hypothesis:

1) The protocol requires two people: an audiophile willing to take the test, and a tester willing to administer the test. (This is a simplified protocol which assumes both the tester and the testee want to be honest about it. If things were more serious, you'd need to verify that the tester and the testee aren't in cahoots, you'd need independent judges, you'd need to make it double-blind, etc.)

2) Agree on a reference recording- a specific passage of a specific recording where the soundstage is assumed to improve with isolation. Make sure the audiophile agrees that the passage is a correct one with which to determine the soundstage. It should be a fairly short passage, since you're going to have to run through it a bunch of times. I'd say no more than 30 seconds or so, unless the audiophile says he really needs a longer reference passage. Ideally, it would help if there were an easy way to cue up the exact beginning and end of the passage on the test equipment.

3) Visually isolate the equipment rack from the audiophile. Such as placing a partition between the audiophile and the rack. Make sure no mirrors, glass, chrome reflections, or shadows can give away what the tester is doing behind the partition.

4) Test the speakers and the listening position to make sure the audiophile agrees that they accurately represent the soundstage.

5) Agree on the number of tests to run. I know little about statistics, but for argument's sake, let's say that thirty runs would be enough. You would also agree on a number of "hits" that would constitute a positive outcome. If the difference between padding and not padding is as striking and unmistakable as you say it is, there should be an incredibly high number of hits: I'd say at least 25 out of 30. I don't know the formula for determining what would be expected due to chance, though, you'd have to look that up in a statistic textbook. I'd guess that anything under 22 would be chance, and 23-24 would be inconclusive.

6) The tester and the audiophile both have a piece of paper and a pen. The tester stands behind the partition where the audiophile cannot see. The tester calls out the number of which test run it's going to be, for instance, "This will be test number five". Both the tester and the audiophile write down the number. From this point on, the tester does not say a word or make any other noises.

7) The tester flips a coin. Heads=padded, Tails=unpadded. It is important that this coin flip happen AFTER the tester calls out the test number so that no voice inflection cues are given. Before doing anything else, the tester writes down next to the number whether it's supposed to be padded or unpadded.

8) Regardless of the outcome, the tester fidgets with the equipment so that the audiophile can't tell whether the tester is adding or removing the pad. (Here is a place where the tester could give unconscious clues if they weren't careful or if there were some kind of collusion between tester and testee- I don't know how to solve this one without significantly complicating the protocol.) The tester either places the pad under the CD player or removes it, as dictated by the coin flip. Absolutely no other changes are made to the equipment except the removal or addition of the isolation pad. Not even volume changes.

9) Without saying another word, the tester plays the test passage. After the test passage is done playing, the audiophile writes down his guess next to the number on his piece of paper. He calls out to the tester "Ready" once he's done recording his answer.

10) The tester calls out the next test number, flips the coin, etc. etc. for thirty runs.

11) At the end of thirty runs, the two pieces of paper are compared. Total the number of "hits" with "misses".


Now, to get this thread back on topic: A nearly identical protocol could be used for testing the Empeg's sound stage in comparison with a CD player. You would have to be extremely careful to compensate for differences in equalization between the two units. For instance, when I hooked up my Empeg to my home stereo, it's EQ curve when "flat" was significantly different than my CD player's. You'd probably need some frequency-test equipment to sort that out. But once it was done, then the test itself would be even easier since you could just use an A/B switchbox, or two different input channels on the amplifier.


___________
Tony Fabris

I left out one step in that protocol. Before the test starts, you have to play the test passage a couple of times both with and without the pad, but with the audiophile's knowledge. Once the audiophile agrees that he can hear a difference and that the test is working, then you can move on to the blind part of the test. All parties have to agree and be sure that the test is valid before you actually perform the test. If you don't, then the testee could come up with an after-the-fact excuse as to why he failed the test.

___________
Tony Fabris

> That's a nice looking car, is it street legal?

But of course :) It was my only (working...) car last year - drove it 400 miles a week, rain or shine - wet arse, big grins.
There's _nothing_ in the world like blasting open-topped through the English countryside at 5am, mid summer, dawn breaking, mist rising, empeg playing. Happiness overload.
Not tried the empeg on the track (yet).

Steve

The isolation pad may not be improving the digital data from the CD itself, but all of the other components (power supply, etc) between the laser and the audio output.

Yep, I agree. No matter how good the bits are going to the DAC, the Digital to Analog Converter produces Analog. Every DAC i've ever seen has seperate power and ground for the analog and digital sides. Good DACs (like in CD players) provide massively excellent isolation between these two sides. However that does not prevent the circuit designer from allowing some noise to seep into the analog portion from the outside. Now, any good analog chip (DAC, OpAmp, etc.) also provide amazing noise rejection from the power supplies. None the less, I believe he said the sound staging improves, not that he could hear lots of noise in the audio. I submit that it could be that a TINY amount of noise is getting into the analog side of things and thereby clouding the subtle high frequencies that the ear uses to locate a sound. I bet that the noise is comming from whatever the CD player sits on (without the pad) and enters through the feet of the CD player. The pad is probably just providing electrical isolation.

It's just a guess, but could be easily tested by placing a piece of foil under the pad and wrapping it around to the top side and then setting the player on top. I doubt the the foil would greatly affect the mechanical properties of the pad, while providing a nice electrical path similar to the CD player without the pad. You could just use a single wire, but that's not quite the same.

Steve

The soundstage of an audio recording is a very subtle thing.

Absolutely. Just think about it: in proper soundstage 'images' can be located a close as one foot apart! Our ears (and brain) interpret this from the tiny differences in delays between sounds reaching our left and right ears. Even at the speed of sound (as opposed to the speed of light , these time differences must be extremely small (I don't have a calculator ready: maybe someone with more Physics 101 than I have, could calculate the difference).

If one wants proof if mechanical isolation from a mat as used by Paul does make a difference, it is essential that the play-back chain is capable to maintain these subtle timing differences: it's time accuracy needs to be higher than the differences induced by properly isolating the player (and/or D/A converter; pre-amp; amp; speakers; etc). Therefore, one needs a pretty decent system to detect the effects that Paul's ears have heard. Thus, in order to do a blind test, one also needs to define the playback chain, and set-up.

Also, some equipment (especially higher-end stuff) has been built to minimise mechanical feed-back.Thus ideally one needs a good player that is poorly isolated. The magazine Stereo (Germany - April 2000) did something like this when they published a series modifications to a Sony CDP-XA7 player. I can't remember all the details, but besides improving the power supply they changed the player's feet, placed it on a heavy slate of stone, and added a load of bitumen padding to the inside of the box. This caused it to play like devices many times its costs.

Of course, all of this is pretty meaningless to the empeg discussion: empeg is not a CD-player; MP3 is not WAV; and in-car amps, speakers and wiring, together with the difficult environment (noise, electrical noise, lack of space, odd shape, variety of reflection/absorption) will induce so many other artifacts, that it will hardly be worthwhile to place Emma (or Mark) on Paul's mat .



NB: there is also evidence that compression changes the timing of recordings: immedeately prior to a transient there is a kind of pre-echo that in effect announces the transient's imminent arrival.

HiFi News / Record Review (from the UK) did articles on this in May and June. They used the clacky sound of a castagnet to compare the timing eight codecs in including Dolby Digital,MP2, MP3, TAC and MS-Audio. MP3 (at 128 bps) started to announce the clack some 10 milliseconds early. MP2 at 128 kbp is more accurate (pre-echo at 4 milliseconds), just as Dolby Digital at 128 kbps; MS-Audio started as early as -20 milliseconds; TAC was the earlies at -50.

This may indicate that some of the loss in soundstage is not due to the empeg, but caused by MP3 technology.

Henno
ex 00120
did score one of the 40

Just have to share this story of CD isolation with you.

I was at my old girlfriend's house and she was complaining about having to replace her CD player. I said, "What's wrong with it?"

"It keeps skipping, but only on some CDs." Naturally, testing those CDs on other players proves the CDs are OK.

So we test it out. She puts on Nine Inch Nails and we sit back. After only a few seconds of "Closer to God", the CD player is skipping and having a lot of trouble.

I stand up, and lift the CD player off the speaker on which it had been resting. Suddenly the skipping stopped. No big surprise to find out that the pieces on which it had been skipping were the loud ones...

But would an 'isolation pad' have been useful?

Save the whales. Feed the hungry. Free the mallocs.

Okay, I'll take the bait.

I'll set something up with my g/f with me looking the opposite way to the system.

Watch this space.....

Paul.

Paul Haigh, 6GB, Blue
Reg: 4120 - Serial 00254

I stand up, and lift the CD player off the speaker on which it had been resting. Suddenly the skipping stopped. No big surprise to find out that the pieces on which it had been skipping were the loud ones...

Cute story. Has nothing to do with an isolation pad improving the soundstage though. Like I said, transient vibrations which reach the CD player will manifest themselves as skips, not as soundstaging problems.

___________
Tony Fabris

MP3 (at 128 bps) started to announce the clack some 10 milliseconds early.

10 milliseconds is an eternity in audio terms. That's insane.

This may indicate that some of the loss in soundstage is not due to the empeg, but caused by MP3 technology.

"May indicate"? If those numbers are accurate, you don't even need to test the soundstage. There's no question it's been completely obliterated at that point.

My only question is:

Did "HiFi News / Record Review" test other bitrates besides 128kbps? I wouldn't have expected the soundstage to be anywhere close to preserved at 128kbps anyway, but I'm curious about other bit rates.

___________
Tony Fabris

I'll set something up with my g/f with me looking the opposite way to the system.

Cool, but wouldn't facing the opposite direction ruin the perception of the soundstage?

___________
Tony Fabris

Well, it should just reverse it. I'm mainly looking for if I can hear the difference - i.e. does the isolator pad make any difference.

Besides with the setup I've got (and no barriers bar sheets) this is the best way to ensure that I cannot tell wether the isolator works or not.

We've set it up, so I should have some results by tomorrow.

Watch this space!

Paul.

Paul Haigh, 6GB, Blue
Reg: 4120 - Serial 00254

Okay the results of the jury are in:

Raw statistics (of thirty runs)

(as determined by a coin throw):
17 with isolator
13 without isolator

Actual noticed results:
21 correct answers
9 incorrect answers (4 of which were thinking there was an isolated when there wasn't - 5 thinking that there wasn't an isolator when there was).

Song used: first 30 seconds of Porcelain by Moby (an awesome track, but I don't think that I'll ever be able to listen to it again!).

Enough said, I'm knackered!

Comments please....

cheers,

Paul.


Paul Haigh, 6GB, Blue
Reg: 4120 - Serial 00254

Cool! I'm impressed!

Like I said, I don't know the mathematical basis for determining what one would expect by chance. In my previous post, I pulled some numbers out of the air, but I don't know if they were the right ones to go by.

Anyone know anything about statistics and random sampling? Out of 30 random runs, is a 21-9 ratio (slightly better than 1/3) significant or not?

More importantly, though: How do you feel about the isolator now? After listening carefully to the same track 30 times in a row, and after 9 incorrect guesses (split evenly between the two possible guesses), do you still think there's a difference?

Oh, and out of curiosity, how long did it take you to go through it all?

___________
Tony Fabris

Hmmm

Just thought.

Is it me that has poor ears, basic electronics knowledge -> (ie. understands that equipment can be interfered), dont't think that this interference is 'enough' to 'destroy' the sound. Or is it that someone has so much better ears, and have listened much more criticaly to sound all their lifes, and they have a much more indepth knowledge of electronics/sound than me (and others) have???

(Not trying to be a smarta** here, just a very 'deep' thought) :)

I guess this subject must be very subjective, because when some people I know
claim to hear the difference on signalcables, I can't.


Cool test anyway.

TommyE

...dont't think that this interference is 'enough' to 'destroy' the sound. Or is it that someone has so much better ears, and have listened much more criticaly to sound all their lifes, and they have a much more indepth knowledge of electronics/sound than me (and others) have?

Right, there's a word for those people: "Audiophiles".

Practically any interference at all in the audio signal is enough to 'destroy' the sound for an audiophile. My contention was that mechanically isolating a CD player's casing didn't do anything to prevent interference because there isn't any interference to prevent in that particular instance.

___________
Tony Fabris

Cool! I'm impressed!

So am I!

Like I said, I don't know the mathematical basis for determining what one would expect by chance. In my previous post, I pulled some numbers out of the air, but I don't know if they were the right ones to go by.

Anyone know anything about statistics and random sampling? Out of 30 random runs, is a 21-9 ratio (slightly better than 1/3) significant or not?


I really hope so!


More importantly, though: How do you feel about the isolator now? After listening carefully to the same track 30 times in a row, and after 9 incorrect guesses (split evenly between the two possible guesses), do you still think there's a difference?

I can safely say that I'm sick to death of that song now - every time I hear it I get goosebumps! I still believe in the isolator pads to still use them, given that I could tell the difference in the majority of cases (70%). The amount of times I get to listen to music though, (i.e. in a quiet environment) is limited although I'm still quite lucky in that respect (no kids!).

I'd definitely recommend the pads (especially since mine were free), since the difference is probably siginificant enough for most people to bother.

Oh, and out of curiosity, how long did it take you to go through it all?

All in all about 3 hours. And without any doubt - it was very hard work (and all I did was listen!).

Cheers,

Paul.

PS Congratulations on the promotion to green

Paul Haigh, 6GB, Blue
Reg: 4120 - Serial 00254

PS Congratulations on the promotion to green

I'm pretty sure that just means I'm one of the moderators. Not for this thread, but for the FAQ thread.

___________
Tony Fabris

Now after all the posts - what do you think about this? Will the critic (positive and negative) affect something in some software releases ( except the decoder which will be changed as someone (i think Hugo) said)?
i.e. changing volume level will affect changing of the base-level?



TeeMcBee

Rite, first off i`m not a expert like some of u seem to be, nor do i own an Empeg, although i know what sounds good, and I am very very fussy.
However, after reading through this thread (very intersting btw :) I can see one huge problem.
You all seem to be comparing mp3`s made in Audiocatalyst, which without a DOUBT, is the worst sounding encoder u can get.
Its pitiful, it crops more high frequencies that the others, I`m not surprised people are finding treble lacking.

Usin an encoder such as Lame (http://www.sulaco.org/mp3/), I really can barely tell the difference between mp3 and wav (usin HQ VBR mode 1, roughly a 180kbps file)
Check out this page for proper graphs and such, highlighting Xings (ACAT`s) below par performace.
http://www.r3mix.net/

Says it all really, even if u dont belive graphs, or others results, you must be deaf if u cant hear the difference between ACAT and Lame or any Fraunhofer encoder.
I`d recommend Fraunhofer (Radium release, or Opticom Mp3 Producer Pro)@ 192kbs, or Lame set to VBR mode 1.
Give it a go, you have nothing to lose and i think you`ll be pleasantly surprised.

There's not a lot to say - the new decoder is the one from ARM, it's very efficient and small, & has the same accuracy. It'll sound identical.

The mk2 has a cleaner sound output, tantalums in the output path to ensure better stability, etc, which may well improve the audio quality though.

Hugo

Have you tried Audio Catalyst recently? It got MUCH better about a year ago.

Rob

I don't think there have been any real conclusions about the sound stage issue, which is what interested me in the first place :) I guess the only real conclusion is that low bitrate MP3 files definitely destroy the sound stage, that higher bitrate files may or may not destroy it, and that what we really need is WAV playback to prove for sure how good the empeg hardware is.

Rob