This review page is supported in part by the sponsors whose ad banners are displayed below

Digital and analogue signal processing have been separately optimized to reach uncommon performance each. Audioaéro designed a top-quality circuit board with 24K gold-plated traces and 70ppm copper circuits for more precise data flow. In partnership with ex-Anagram ABC PCB firm they developed ‘ultimate’ digital processing based on 32-bit/384kHz DSP (S.T.A.R.S. Evolution) with a 2-channel asynchronous sample rate converter in balanced dual-mono mode. Its digital output drives 2 Burr-Brown 1792 DACs. The ABC PCB module features a single audio input capable of supporting PCM data up to 24/192 or stereo DSD at 2.8224 MHz. In either case DSD or PCM inputs upsample to a shared internal 8 x FS PCM format. The dedicated ultra-low jitter master clock incorporates a high-grade natural quartz crystal. This results in a further level of noise reduction. Digital artifact reduction is claimed to be less than 1 pico second of RMS jitter.

S.T.A.R.S. Evolution uses a clock synchronization scheme to resync and retime all incoming data to a high-quality local master clock. By using this stable reference, inter-component jitter is minimized. When converting from digital to analog, this jitter reduction solution is claimed to have enormous benefits for detail and clarity. This is the most distinctive feature of La Source over recently released La Fontaine.

Upsampling to 384kHz enhances Redbook material to the clarity and sweetness generally associated with SACD or DVD-A. Best-of-class dynamic range and THD+N are achieved by employing an innovative upsampling kernel with >144dB of image rejection. Adaptive Time Filtering (ATF) is an Anagram Technologies resampling process. It allows audio data to be interpolated with higher precision than via off-the-shelf SRCs.

ATF "uses an advanced polynomial curve fitting algorithm. This closer matches the original data than sample-and-hold or piece-wise linear estimation interpolators as are commonly used today". This already advanced technique underwent further refinements which Audioaéro requested to minimize jitter errors in the resampling process to irrelevance.

The diagram from ABC PCB’s standard  data sheets (thus slightly different from specific Audioaéro components) shows a comparison between normal upsampling and Anagram upsampling running at 384kHz using ATF in a theoretical case of zero clock jitter and how ATF provides superior performance in the presence of jitter and minimizes the impact of resampling jitter. One distinctive feature of ATF vs. other interpolators is the error path between output samples. This error path in a sample-and-hold interpolator won’t follow the original sample data. With ATF, the error path is determined by the smooth curve-fitting algorithm which recalculates a new path with every sample to more closely follow the original audio stream.

Having an error path more closely match original data has two benefits. Resampling errors caused by jitter and appearing on data clocks by using arbitrary non-integer sample rate conversion e.g. 44.1kHz to 48kHz will be minimized. With S.T.A.R.S. Evolution, the extremely small residual error remaining is designed to be beyond the dynamic range of 24-bit audio to far exceed sample-and-hold or piece-wise linear interpolators.

Specific data-to-system synchronization allows any incoming audio stream to be resynchronized and retimed to the local high-quality clock. Combining this synchronization process with a virtual time domain model that uses an advanced cubic interpolation algorithm to resample the incoming audio data, timing errors in the signal can be compensated for with very high accuracy. The end result is said to deliver tighter more focused bass, increased stereo imaging, better clarity and superior separation for all musical instruments and voices. The S.T.A.R.S. module output stage also features the ‘Sonic Scrambling’ technology to improve linearity of multi-DAC designs. The goal is to provide higher-quality conversion with two DACs per channel operated in differential mode. Key is that the signal sent to either +/- DAC is perfectly identical. Each becomes the other's exact counter sign with added low-level random biasing to decorrelate the signal’s less significant bits. This enhances low-level signal linearity as these signals will be random and spread out any possible signal-related distortion.

The S.T.A.R.S. module handles Direct Stream Digital (DSD) at 2.8224MHz (64 x 44.1kHz) on the input. The DSD stream then converts to PCM thanks to a specific filtering process (DSF™) to subsequently apply standard PCM audio processing techniques. The audio channel configuration supported by S.T.A.R.S. is 2-channel, not multi-channel DSD. Due to its 2.8224MHz sampling rate and one-bit nature, DSD is normally incompatible with already implemented signal processing functions targeted at standard PCM data. Anagram thus developed Direct Stream filtering algorithms that can convert DSD to PCM up to 8 x FS. The S.T.A.R.S. module integrates this feature to significantly enhance SACD support.

Given these technical considerations, La Source appears to not be compromise between Redbook and DSD standards but to have been thought through and designed to fulfil the most advanced requirements for either format. The S.T.A.R.S. module also includes phase inversion whereby the output data can be inverted. By default this function is disabled and can only be activated in software mode from the menu options. As mentioned the S.T.A.R.S. Evolution module requires a master clock which can be a sine wave input or buffered shaped clock derived from an external clock oscillator. Here Audioaéro Design selected a high-grade quartz crystal. The La Source master clock delivers <1ps RMS phase jitter @ 25MHz with outputs able to drive 12 series-terminated lines. It runs high-quality separate linear regulators for its clock generator circuitry. The master clock module features a single control input, clock enable, a digital power supply input and three clock outputs.