Et innlegg fra sjefen i SlimDevice, Sean Adams.
Et aldri så lite spark til tweakerne av SB? Mer interessant er det at han prøver å sammenligne målinger av Transporter med Benchmark DAC1. Hvis det viser seg at Transporter er i samme liga eller bedre en Benchmark så er kanskje ikke prisen ille? Transporter kan jo brukes som vanlig DAC, da den har digitale input.
"Lots of people have speculated as to whether Transporter could sound better than [insert expensive DAC here]... Well, spending big bucks or using some popular part doesn't automatically make it sound better, which is why I always encourage people to rule out the placebo effect when doing listening tests. There is a lot of real science under the hood in transporter - I didn't throw some expensive parts in there and call it a day. I also didn't design it overnight, and I didn't do it on my own. I had help from Andrew Weekes on the power supply and op amp stages, and Richard Kulavik from AKM advised along the way with ideas to maximize the performance of the AK4396. And that wasn't even the first DAC I tried... I tested all the best parts from Analog Devices and TI/Burr-Brown, and accumulated a mountain of scrapped prototype boards along the way.
Getting to the point: while Transporter excels at the usual basic measurements like THD+N, SNR, etc, if we look closer at some of the characteristics that we think will determine subjective sound quality, there are some pretty substantial differences between transporter and any other DACs I've tested. And they are no accident. Jung regulators, low-jitter clocks, quality passives, etc - this stuff takes many months of design and testing to get right. You aren't going to get this by slapping a bybee filter or a black gate cap in there.
Now, of course measurements dont't tell the whole story, but until you hear it for yourself, just have a look:
http://www.slimdevices.com/temp/tran..._benchmark.gif
This is comparing a 10KHz sine wave played on three different DACs. In green is the analyzer's built-in signal generator. In red is the benchmark DAC-1. The blue shows Transporter running from its internal clock (the normal case), and yellow shows transporter running as a simple DAC, taking its clock from the s/pdif input signal (no clock sync).
First note the scale of the graph. ALL of these dacs are very good. We are looking way down below -120dBu, relative to an input at +8.4dBu, and we are zoomed to +/- 500Hz in order to reveal the effects of clock jitter.
There is a lot of information here - let's start with the noise floor. With an idle channel noise level of 4.4 microvolts RMS, Transporter is by far the quietest DAC I've ever measured. There is no single factor responsible for it - every component, the layout, EMI - all contribute to the noise floor. This is not just about the "blackness" of the background, but is also the limit to the level of detail you can hear. Here's a plot showing the noise floor across the whole spectrum.
http://www.slimdevices.com/temp/noise_floor.gif
Going back to the first graph, the other thing to notice here is the effects of jitter - this is determined by the width of the 10KHz spike. As you can see, both the benchmark (red) and the analyzer's internal signal generator (green) are very narrow, which is good. This is because they are each running from their internal oscialltors - the DAC1 uses sample rate conversion to isolate it from s/pdif jitter. For transporter, I ran it two different ways so you can see the effects of jitter introduced by the s/pdif interface. In the yellow, it is running as a slave - the clock is being generated by the dScope analyzer and transmitted along with the data over s/pdif. In blue, Transporter is the master, and I am using the word clock _output_ feature to synchronize the analyzer's digital output to it. This is equivalent to what you'd get when Transporter is playing audio streams from the PC - its internal clock is used.
As I get the time to document more measurements I will put together a more complete set, but for now I thought these would be interesting. By the way, while these were done using a special analyzer, you can get useful comparative measurements using just a PC sound card. If you want to see how your DAC or CD player stacks up to Transporter, try the Rightmark analyzer software:
http://audio.rightmark.org/index_new.shtml."
Hele tråden kan leses her:
http://forums.slimdevices.com/showthread.php?t=28126
