why do transports sound different? - Page 12

?... like Michael Fremer's reviews of digital players and cables that when tested were so bad as to fail the ancient and incredibly generous DIN 45 500 standard ?

Or Robert Harley who won his sinecure at SP by writing an essay ?

Or the audio rag reviewers that Sean Olive at JBL tested

The reviewers’ (poor) performance is somewhat of a surprise given that they are all paid to audition and review products for various audiophile magazines. In terms of listening performance, they are about equal to the marketing and sales people, who are well below the performance of the audio retailers and trained listeners

Yeah I trust these guys' opinions

BTW You still need to explain why Benjamin and Gannon are so far off-beam, despite their tests being very similar in findings to Manson's 1974 (BBC) thresholds at about 35ns...

But how do you know it's a high jitter source unless you measure it???? And how do you know what comes out of your Pace Car is very low jitter when you don't have the tools to measure that either??? You are making no sense.

There is a simple way you could infer the effect of a jitter-reducing device, it would not be amazingly accurate but it would give you a general idea.

You rip a CD track and verify that it is a bit-perfect rip, this is your reference point.

Load up the track on your digital device sans jitter-reducer and play back the track and record the analog output using a decent say 24 bit ADC, compare the reference with the recording, they will not be the same. Adjust the recording level until you get the recording as close to the reference as possible, it need not be perfect. Then repeat 10 times to allow for random variation.

Add the jitter-reducer to the circuit and again repeat the calibration recording and the 10 actual recordings.

Load the samples into an audio program and scope out the FR for each of the 20 samples. Export the FR data to Excel. Average the results for each combination. Then chart the final data as a line chart. Zoom to 1600% and see if you can see any differences. You can of course then do the maths and quantify any differences.

It is crude but if there are any differences that are bigger than the incipient error, and a good 24 bit ADC should give you 19 - 20 bits accuracy or about 1 - 2 parts per million, you will see them. If the differences are below the 19/20 bit accuracy I really would not lose sleep over them

You could even blind test the most typical sample from each set .

If Steve N. wants to lend me a pace-car I have several digital devices with optical outs that should be quite high in jitter such as a Western Digital HDTV and a Marantz CC4300 that bothg feed my Entech 203.2 .

This is a review and measurements on the digital player with the worst measured jitter performance in the world ! (afaik)

Stereophile: McIntosh MS750 music server

Two things are interesting about this review. First just how bad the multi $K streamer is technically, secondly how the reviewer actually *quite* liked it until John Atkinson told him just how bad it was, after which the review went south in a hurry. There is a lot of hedging still as they cannot quite bring themselves to say this is a POS avoid !

Two possible inferences exist, either SP reviewers really are tin-eared[1] or even comparatively vast amounts of jitter make little material difference.

When some pundits expound on how terrible a few 10s of ps of jitter is , it is amusing to see how digital devices that should, under this world view, be **utterly unlistenable** perform subjectively

Cheers all !


1 - see my earlier post on the value of sighted reviewer comments

+1.

Great read as always.

My favorite part of the review:

"There are still logical reasons why a consumer might buy a McIntosh MS750, especially if cost is no constraint. In that case, go ahead and get the Mac and the digital processor of your choice. But a little bit of computer savvy, a $300 Slim Devices Squeezebox, and that same processor will get you to the same sonic place for significantly less."

So are they suggesting that transports might sound the same

I think it suggests that Jitter can't/hasn't been accurately measured, at least not by Sterophile's old techniques.

I think they finally gave up on their silly Jitter measuring company. There's a new group they're promoting now (probably another thats heavily invested in by Stereophiles board of directors):

Stereophile: Breakthrough Approach to Audio Measurement

Not sure about that, but they are suggesting that the McIntosh is virtual garbage at that price and the Squeezebox is a decent device for 1/20 the price.

For the record, the Transporter is a much better source than a Squeezebox. Stereophile rated it 2nd place, with only the Chord Blu and DAC64 ahead of it out of about 20 high-end source/DAC combos. I absolutely love mine. Not that it's even necessary since the Transporter is rated at <20ps jitter, I will be upgrading the two internal clocks to 1.5ps and 5ps at 1ppm (Audio-gd JZ-1). They are a tiny fraction of the price of Audiocom clocks, only about $30 each. н¨ÍøÒ³ 1
And I know they will pair well with my RE7 DAC.

I thought that the transporter used the same circuit as the SB for the transport section? I know the DAC is better, but I swear I remember someone opened it and found the digital section to be the same?

No. The Squeezebox is very noisy in the digital section, and I'm sure the switching power supply plays a big part. The Duet is a lot better than the Classic though, and that's probably because of the vacuum fluorescent display in the Classic.

For one the Transporter has regulated power supplies and dual transformers (I think that's what those small blue things are). Two, it's digital and analog sections are completely separate from eachother - the processing is done at the front of the unit by the display, well isolated from the analog section in the rear. Third, it supports 24/96. Fourth, it has dual low-jitter clocks - one for 16/44.1 and one for 24/96. Fifth, it has AES/EBU output. Sixth, it has a word clock input and settings for asynchronous or synchronous effects loop clock mode. Seventh, you can invert the polarity. Eighth, it has 4 digital inputs............ etc.

I used the Squeezebox Duet with a CIAudio power supply and custom DC cables. Not only did each of the two DC cables result in differences with tone and soundstage, the setup didn't come close to the fidelity of the Transporter as a source.

remove room acoustics from the experiment since there is too much variation from site to site. I'd also remove DAC, amp, speakers/headphones for the same reason.

Humans are too subjective so I'd remove it from the test as well.

Measure the jitter from various devices on the input stream and the output of your device. If jitter is reduced then your device is proven to improve the chain. As far as signal, I guess the normal white noise signal spread across all frequencies would be the best way to reduce that variable. If power is a concern any of those power regeneration systems should be able to stabilize the inputs (better than the wall power of the local power company).