A dedicated music streamer can cost as much as a car, but Jason Stoddard, co-founder of Schiit Audio, says a phone can do the same basic job for far less. His claim comes down to the data reaching the DAC. If a phone and a CD transport send the same digital signal into the same downstream system, he argues, the source itself should not change the sound.
That sounds simple until phone software, resampling, Bluetooth and DAC inputs enter the chain. The measurements help show which of those factors actually matter.
Why Schiit isn't building another CD player
The discussion grew out of a practical question about whether Schiit would ever build another CD player. Hunter, a member of the Schiit community, asked about bringing one back, and Stoddard pointed first to the difficulty of making one.
The first problem is the transport itself: the mechanism that spins the disc and the laser that reads it. Schiit would have to adapt old car mechanisms or buy drives through European firms that source them from China. Servo circuitry is the deeper hole. Almost all of it, Stoddard said, rests on old Philips chips that he is not sure anyone still makes, and he doubts Schiit can add anything to that market.
A future Schiit streamer, he added, is far more likely than a future Schiit CD player.
The $99 fix for expensive streamers
The sharper version of Stoddard's argument surfaced on Schiit's Head-Fi thread, where members had been posting photos of very costly streaming gear. The thread is nearing 250,000 posts by Stoddard's count, and he asked members to move the discussion elsewhere. His complaint was that the photographed machines are "overpriced computers." His fix, at $99, was to hook an Eitr 2 to any computer, tablet or phone, for what he called "a perfectly good streamer."
None of this was off the cuff. In its April 2026 launch coverage, audioXpress reported Stoddard pitching Eitr 2 in nearly the same terms: buyers could skip "expensive USB converters and giga-priced streamers" because the $99 box provides what Schiit calls "verifiably perfect AES and SPDIF conversion."
What the measurements actually show
In December 2023, the measurement blogger who publishes as Archimago fed four bit-perfect sources into the same Sabaj A20d digital-to-analog converter (DAC). One was a Google Pixel 8 Pro running USB Audio Player PRO, and the others were a Raspberry Pi 4, an Intel N100 mini PC and a 2014 Surface Pro 3.
Every source measured the same, phone included, with dynamic range past 120 dB, distortion under 0.0001%, and jitter sidebands more than 140 dB below the signal peak. He labeled the result "digital source invariance" and made the implication explicit.
Benchmark Media chief technical officer John Siau, whose company sells DACs rather than streamers, separates two kinds of jitter. One rides in over the digital interface, and the other sits at the converter itself. His audibility threshold is concrete: jitter-induced sidebands should sit at least 23 dB below the system's A-weighted noise to be considered completely inaudible.
The catch: when a converter changes the result
Archimago's test kept the DAC fixed and changed the digital source. But a separate question begins once another box sits between the source and DAC, as a converter such as the Eitr can send the signal through a different DAC input, and the result can then depend on how that input is implemented.
For instance, Audio Science Review (ASR) founder Amir Majidimehr tested the original Eitr that way, and the results split. In front of a Schiit Modi 2 Uber, it lowered noise and distortion. In front of a Topping D50, it was a step backward from that DAC's own USB input. Majidimehr attributed the latter result to the D50's asynchronous USB input outperforming its S/PDIF input, and judged the measured differences, around -115 dB or lower, inaudible either way.
ASR Master Contributor Veri read that split less charitably: "Step 1: Sell a DAC with schiitty jittery output / Step 2: Sell something that cleans this up in the form of a Schiit [Eitr] / Step 3: Profit? … It at least shows they can do it if they try, though. Why do it right from the start when you can charge a customer twice?"
When a phone is truly bit-perfect
Stoddard's condition sounds simple enough: the phone just has to send the original digital audio to the DAC without changing it. In practice, however, the phone's software can resample or process that audio along the way.
Android often does it in software. Most audio passes through the system mixer, which commonly runs at 48 kHz on many devices. A CD-rate recording sits at 44.1 kHz, so that path can resample it before a DAC ever sees the file.
USB Audio Player PRO, the paid app in Archimago's phone test, uses its own USB driver to route around that mixer for Tidal, Qobuz and local files. It cannot do the same for Spotify, which does not expose the same direct USB path to the app, according to GoldenSound's testing.
Android 14 opened another route. Google added preferred USB mixer attributes with an optional bit-perfect mode that can send audio to a USB device without sample-rate conversion, volume scaling, mixing or other processing. Support for that bit-perfect mode is optional, so whether it works still depends on the phone, its audio implementation and the app.
iPhones make the 48 kHz part simpler. Apple's own documentation says Apple Music Lossless ranges from CD-quality 16-bit/44.1 kHz up to 24-bit/192 kHz, with an external DAC required for songs above 48 kHz. CD's 44.1 kHz sample rate therefore falls comfortably inside the range relevant to Stoddard's comparison.
Apple's documentation does not establish that every app and wired playback path preserves the original bits, so bit-perfect playback still depends on the path being used. What it does establish is that 48 kHz is not a universal ceiling when an iPhone feeds an external DAC.
What the five-figure streamers offer
Auralic's Aries G3 runs about £9,899, sold on femto-precision clocks, galvanic isolation and low-noise power. Lumin's X1 costs $14,000 and adds an optical network input, separate power supply and its own clocking system. At £32,900 in the UK and $32,000 in the US, Aurender's N30SA splits the computing and audio sections into separate chassis. Linn's Klimax DSM reaches $48,230, with its case built around tightly regulated power, clocking and isolation.
The recipe repeats across them: cleaner power, steadier clocks and fewer paths for electrical noise around the digital source.
Moreover, the measurements earlier in this article address a narrower question. When bit-perfect sources feed the same DAC, they have failed to show a meaningful difference attributable to the source itself.
Schiit's own copy makes the divide less tidy. The company sells Eitr 2 as an escape from "giga-priced streamers," while its Unison USB implementation still advertises precision local clocks and electromagnetic and electrostatic isolation. Those are some of the same engineering categories the five-figure machines use to justify their designs.
What bit-perfect really means
Bit-perfect means the DAC receives the original digital audio without resampling, EQ, volume processing or system mixing changing it along the way. The comparison also has to keep the downstream chain consistent. A phone and CD transport feeding the same DAC can isolate the source itself. Using different DACs and analog stages adds other variables.
In practice, Stoddard's claim requires a wired connection and a playback path that preserves the original data. Android support still varies by phone and app. Bluetooth falls outside the condition entirely. Eitr 2 only becomes relevant when the DAC's own USB input is the weak link.
The takeaway for listeners
If the bits get there, the source is irrelevant. That idea has been echoed in forum comments: "Still blows my mind that people spend so much on digital source devices and streamers. 10101100 passed over a $90 CD Transport/Device, or passed over a $40k digital streamer is still 10101100. EXACTLY the same. It's the DAC onwards that have some level of impact."
Not everyone agrees. One commenter argued that a Cyrus CD-T player changes the signal before the DAC, with error correction and reclocking that improve detail. "Why do they make CD Transports and DAC's with I2s, so as too separate the timing and data so a more pure single is read by the DAC, that in itself totally changes the sound as in space around each instrument, or if you like better clarity, and then there's the Cable as well," they wrote. "Of course if you can't hear any difference then maybe your ears or your system is just not letting you."
For Stoddard, the engineering problems expensive streamers solve are real, but they do not always affect what you hear. A phone that delivers bit-perfect data to a competent DAC can match a CD transport in the measurements. The condition is strict, but it is not expensive.