Ferrum Wandla GoldenSound Gen 2 DAC

 WANDLA GoldenSound Gen 2 It’s all about finding your sweet spot While keeping all great features of the original, WANDLA GoldenSound Edition Gen 2 manages to improve on three great features, called Impact , Tube Mode and Spatial Enhancement. In Gen 2 our engineers completely overhauled them and each got their own Ferrum Sweet Spot Tuning! Something great just got even better.The EISA awarded WANDLA continues to be one of the best DA converters in its league. WANDLA GoldenSound Edition managed to add three great features. One to enhance the sheer impact in the lower frequency, one to simulate the warmth of a tube stage and one to offer an enhancement in the third dimension. With WANDLA GoldenSound Edition Gen 2 these three features, called Impact , Tube Mode and Spatial Enhancement, were completely overhauled and each got their own Ferrum Sweet Spot Tuning! Something great just got even better.Three great features just got even better. Impact , Tube Mode and Spatial Enhancement were completely overhauled and each got their own Ferrum Sweet Spot Tuning! Gen 2 Impact You can now control your low-end sound signature by choosing one of seven predesigned settings of bass and dynamics and adjusting its level from 10% to 130%. Gen 2 Tube Mode You can now choose your tube sound signature by choosing one of five popular tube designs (EL34, KT88, 300B, 2A3 and 7062) and adjusting the level of its effect from 10% to 200% in search of your personal perfect setting. Gen 2 Spatial Enhancement You can now independently toggle the spatial effect, select your output mode (Headphones/Speakers), and engage the new Transient Compensation to fine-tune the imaging in search of your preferred setting. Gen 2 Elevated digital headroom Higher immunity to intersample overs even in bypass mode at full output level, allowing you to play back any music free from additional distortion, and enjoy a high-quality sound. Gen 2 New Voltage Adjustment The WANDLA GoldenSound Edition benefits from a hardware level output voltage adjustment of standard WANDLA, allowing the user to swap between just below 10 Vrms output to a just below 4 Vrms output. This is facilitated via a dedicated hardware voltage divider, allowing the user to select lower output voltage and ensure compatibility with all amplifiers, without a need to use any digital volume control, or to have analog volume control components in the signal path. Gen 2 Dynamic Digital Filtering WANDLA has three ESS digital filters to choose as standard and, as a world first, two digital filters created for Ferrum by renowned filter-maker Signalyst, know for their HQPlayer app. More filters will follow after interacting with WANDLA users. Gen 2 Proprietary amplification Fully balanced modified IC pre-amplifier, fine tuned for WANDLA. Gen 2 Optimized Digital inputs Specially programmed USB, and tuned coaxial and optical S/PDIF, AES, ARC (TV) and I2S ports, optimized for audio. Gen 2 Made for HYPSOS WANDLA performs beautifully right out of the box, but you can take its performance above and beyond when pairing it with HYPSOS. Using the proprietary Ferrum Power Link (FPL) connection with 4TSD voltage sensing, both components will perform to their absolute maximum, unleashing unheard-of musicality. Gen 2 SERCE Digital audio system module - Our optimized ARM chip offers the shortest signal path possible, making 5 chips redundant. Impact in the original GoldenSound Edition featured a base shelf and an additional peak to provide a beefier, more realistic sound signature. WANDLA GoldenSound Edition Gen 2 evolves this by offering seven distinct shape settings (Reference, Ref , Smooth, Kick , DD-Comp, DD-Comp , and Sub-Bass). Instead of a single fixed enhancement, you can now select from these profiles – including specialized compensation for dynamic drivers – and regulate the total impact level from 10% to 130%. This allows you to perfectly tailor the output to your specific sound system, headphones, or soundtrack. Enter Gen 2’s Impact with Ferrum Sweet Spot Tuning. Tube Mode  - In the original GoldenSound Edition added the sweetness of the second harmonic. With GoldenSound Edition Gen 2 apart from adding artificial 2nd harmonic, we introduce the sound signature of five tubes (EL34, KT88, 300B, 2A3 and 7062) with reproduction up to the 7th harmonic. Experience the tube sound, without the associated negatives like big and bulky components, excessive heat production, not to mention the costs of tubes. To take things further, you can also regulate the levels from 10% to 200% in search of your personal perfect setting. Enter Gen 2’s Tube Mode with Ferrum Sweet Spot Tuning. Spatial Enhancement - You can now independently toggle the spatial effect, select your output mode (Headphones/Speakers), and engage the new Transient Compensation to fine-tune the imaging. Spatial Enhancement originally had two modes, which were tailored for speakers and headphones separately. GoldenSound Edition Gen 2 introduces Transient Compensation, or T-Comp in short. This setting fine-tunes the spatial enhancement algorithm and reduces how it affects high frequencies, resulting in a vastly more realistic three-dimensional staging. Enter Gen 2’s Spatial Enhancement with Transient Tuning. WANDLA GoldenSound Edition Gen 2 WANDLA GoldenSound Gen2 Edition, fully balanced DAC/PREAMPNew features include elevated digital headroom, spatial enhancement, tube mode, impact Proprietary fully balanced modified IC pre-amplifier, fine tuned for WANDLA GoldenSound Gen2 EditionSpecially programmed USB, and tuned coaxial and optical S/PDIF, AES, ARC and I2S ports, optimized for audioProprietary optimized ARM chip offers the shortest signal path possible, making 5 chips redundantFour proprietary Signalyst filtersProprietary 4-pin Ferrum connector for HYPSOSBalanced XLR and unbalanced RCA line outputsDigital and Analogue volume controlEES Sabre ES9038PRO with selected filterRemote control included Ferrum WANDLA GoldenSound Gen 2 Edition DAC/PREAMP specifications:DAC chip: ESS Sabre ES9038PRODAC resolution: 768 kHz / 32 bit, DSD 512Digital inputs: • USB Type-C (up to PCM 768 kHz / 32 bit, DSD 512, DoP 256) • I2S (up to PCM 768 kHz / 32 bit, DSD 512, DoP 256), PS Audio® compatible • ARC (up to PCM 192 kHz / 24 bit), TV input with CEC • AES (up to PCM 192 kHz / 24 bit, DoP 64) • Coaxial S/PDIF (up to PCM 192 kHz / 24 bit, DoP 64) • Optical S/PDIF (up to PCM 96 kHz / 24 bit*) *may work up to PCM 192 kHz and DoP 64 – depending of customer optical cable and transmitterAnalog inputs: RCAAnalog input Vmax: 9.5 VRMS (2 – 3.5 VRMS recommended)Analog input impedance: 47 kΩLine outputs: balanced XLR, unbalanced RCAVolume control: analogue with bypass option / digital for DAC operation onlyOutput level (@ 0 dBFS / 1 kHz sine): • PRO: 8 VRMS balanced, 4 VRMS unbalanced • Red Book: 3.5 VRMS balanced, 1.75 VRMS unbalancedFrequency response: 10 Hz – 200 kHz /- 0.1 dB analog inputsDAC THD: -121 dB (0.00009 %), THD N: -118 dB unweightedAnalog input THD: -123 dB @ 2 VRMSDynamic range (A-weighted): 127 dB analog, 119 dB digitalCrosstalk: -120 dB for 1 kHz, better than -100 dB for 20 Hz – 20 kHzOutput impedance: 22 Ω unbalanced, 44 Ω balancedPower consumption: 10 W idle, 15 W maxPower inputs (22-30 VDC): • 5.5/2.5 mm DC connector, center positive • proprietary Ferrum Power Link (FPL) 4-pin DC connectorPower adapter: 100-240 VAC to 24 VDCRemote control: includedDimensions: 21.7 cm x 20.6 cm x 5 cm / 8.6″ x 8.1″ x 2.0″Weight: 1.8 kg / 3.97 lbs WANDLA DDF DYNAMIC DIGITAL FILTERING explained What is DDF DDF stands for Dynamic Digital Filtering. WANDLA includes several built-in digital filters to tweak the sound — but why dynamic? Because they will change: YOU will determine which filters are the best. We are inviting WANDLA users to vote — tell us what your favorite filter is, and why. We will gather this data to declare the most popular filters, so, as a reward for your experimental efforts, we can progressively replace the least popular filters with new ones. This is a great chance to collaborate and to deeply engage with how filters affect sound, so that you can fine-tune your WANDLA to be perfectly tailored to your preferences. In WANDLA’s menu, you can choose from a set of digital filters. One is a part of the ESS DAC, and four were designed especially for WANDLA by filter specialist Jussi Laako of Signalyst, best known for the HQPlayer app. Out of the box, WANDLA defaults to the team’s favorite filter — in a sense, we already gave this setting our stamp of approval. But to deserve the name ‘The Converter’, WANDLA needs to go further. Just adding a static set of filters was not going to do it for us — and more importantly, not for you. We want YOUR stamp of approval. We will add future filters via firmware upgrade — once you’ve voted on your favorite in our survey, you’ll be notified when additional filters are released. In the end, by exchanging the least popular filters for the best ones, WANDLA users will arrive at the ultimate set of highly tailored digital filters. This is unique to Ferrum, to DACs in general, and to the audio business as a whole. Welcome to the world of digital filters! WANDLA DDF DIGITAL FILTERS explained ESS digital FIR filters WANDLA makes use of the ESS Sabre 9038PRO DAC chip. Via WANDLA’s menu, the user can choose from three ESS filters in addition to our HQ filters. We encourage you to experiment to find out which is your personal favorite. Linear Phase Fast Roll-off (ESS Lin-Ph) A linear phase filter with steep slope after passband (fast roll-off). It has very good aliasing rejection and it has constant delay of all frequencies in the audio band.HQ digital filters Specialist filter-builder Jussi Laako from Signalyst has adapted digital filters from the HQPlayer app to be implemented in our SERCE module, that lies inside of WANDLA. The appropriate digital filter settings can be made via the menu where the user can make his choice from the two HQ digital filters which are included at launch. We encourage you to experiment with these HQ digital filters to find out which is your personal favorite. Short filter (HQ Short) Short time domain response, slow roll-off, partially apodizing, single stage filter optimized for transient response. Especially suitable for jazz, blues and similar recordings. Apodizing Minimum Phase (HQ Apod. MP) Minimum-phase version of HQ Apodizing filter. For studio multi-track recordings with focus on transient attack. Gauss filter (HQ Gauss) Apodizing filter with focus on optimal time-frequency response. Provides very good transient response and space information. Good general purpose filter fit for most source content and genres. Apodizing filter (HQ Apod.) Apodizing filter especially for modern recordings. For recordings made in real acoustics or otherwise having notable space information. More to come in future firmware updates DIGITAL FILTERS in general Digital FiltersThe use of digital filters in the audio industry is broad nowadays. They are used from equalizing the frequency response of speakers or headphones to interpolating filters in oversampling DACs. In this text we will focus on the latter one. The basic task of every filter is to suppress unwanted information from the signal or shaping it to desired waveform. It may be DC component removal, specific frequency or band of frequencies attenuation and amplification, averaging or even calculating a median of signal. Filters (analog and digital as well) are described in two domains: time and frequency. Time domain is the real one. It describes how signal takes specific values in time. On the other hand, frequency domain is a mathematical construct in which only sinusoidal signals exist and those sinusoids are components by which signal in time can be described. It is not a real domain, because signal is physically propagating in time domain, however it allows simple analysis of what frequencies are apparent in measured signal. It simplifies analysis of frequency properties of filter too. In filter design both time and frequency domains are used because there are effects which are easier to observe in time domain than in frequency domain and vice versa. However, it does not mean time domain contains different information than frequency domain. They are mathematically interchangable (signal in time domain explicitly corresponds to its frequency spectrum), but information is presented in other way, therefore different aspects are emphasized.Interpolation filtersSometimes also called reconstruction filters. They are used in upsampling of the signal from lower to higher sample rate to attenuate the aliasing artefacts. They are lowpass filters, so they are used to reject unwanted high frequency components of signal. Upsampling and aliasing are more extendedly described in next paragraphs, but most important thing here is, in WANDLA those filters which can be picked are interpolation filters. Therefore, we will describe some of their most important characteristics, which dictate their performance.Linear Phase vs Minimum PhaseLinear phase filters have as name suggests linear phase, which results in constant delay of all frequencies in passband. Therefore, there is no difference in time of arrival of different frequencies of the signal at the output. However linear phase filters have drawback in its impulse response, which is time domain characteristic. They have pre- and post-ringing in its impulse response. If the signal going through this filter have high enough frequency components, they will excite oscillations before and after the impulse itself, which are not present in signal before filter.Especially pre-ringing is detrimental, because it is artefact added before the impulse itself, which in practice may deteriorate transient reproduction. It is completely artificial to have oscillation before the transient.Minimum phase filters do not have linear phase and in case of low pass filter they have higher delay of higher frequencies in passband. Post-ringing is usually higher than in linear phase filters, but there is no pre-ringing. Therefore, minimum phase filters are better for transient reproduction, because there is no abnormal oscillation before transients.It must be said that pre- and post-ringing is usually a problem at lower sample rates (44.1 and 48kHz). At higher sample rates frequencies in audio band (and even beyond) may not be able to excite those pre- and post-ringing oscillations, therefore differences between linear and minimum phase filters are getting lower or can even become unnoticeable at higher sampling rates. However, it is the case only in ‘native’ high sample rate files. If the ‘native’ file is 44.1 or 48 kHz and it has been only upsampled to higher sample rate, it means digital filter used in native file preparation may have added pre- and post-ringing. Upsampling those files to higher frequencies does not solve the problem fully, ‘the damage has been done’. It can even add additional pre- and post-ringing, but there are methods, such as apodization in upsampling filter, which can reduce errors to some extent.ApodizationApodization is technique of lowering down pre- and post-ringing in impulse response of the filter. It is done by so called windowing functions, which frequency spectrum is multiplied by frequency spectrum of the base filter. This modified frequency spectrum corresponds to impulse response with lower pre- and post-ringing. However, this operation is making the slope of the filter less steep and attenuation in stopband is a little lower, therefore unwanted higher frequencies are less rejected.This is causing filter to reject aliasing in worse manner. How much the pre- and post-ringing is attenuated and how much slope of the filter is deteriorated depends on the used windowing function, therefore there are many ways of achieving apodized filters with different properties. Both linear phase and minimum phase filters can be apodized.Apodizing filters are generally used to correct/reduce errors in the source data, introduced by the ADC digital decimation filters, or at later stage conversion tools used to produce the final deliverables. Such errors can for example add artificial sheen on the highs, such as cymbals. It is like looking through a scratchy and dirty car windscreen in bright front sunlight.Fast Roll-off vs Slow Roll-offRoll-off describe the slope of the filter after passband, so it means how fast filter starts to attenuate when going from passband to attenuation band. Fast means steep slope and slow means mild slope. Fast roll-off results in better attenuation of unwanted frequencies, but those filters are computing the signal longer and the pre- and post-ringing is usually higher than in slow roll-off. On the other hand slow roll-off filters are worse in attenuating higher unwanted frequencies, which is detrimental in terms of aliasing rejection, but they are computing output signal much faster.Oversampling vs non oversampling DACThe most general factor differentiating DACs (especially in audio) is oversampling. It defines the frequency at which DAC finally performs signal conversion from digital to analog (DA) form. Non oversampling DAC adjusts to input stream sampling rate and makes DA conversion at this specific frequency. Although this technique allows real bit-perfect conversion, because theoretically there is no need for digital filtering, it has some drawbacks in analog realm. First, if the sampling rate is changing, the analog antialiasing filter on the DAC output must change its cut-off frequency to accommodate to different sampling rates. Therefore, the antialiasing filter must be tuned, which deteriorates its performance, or there must be independent filter for each sampling rate, which strongly elevates the price and complexity of the design. Additionally, non-oversampling DACs with enough resolution in hi-fi audio (24 bits) and high enough sampling rate (192 kHz) are almost non-existent. It is because, that it is technologically hard to achieve non oversampling DAC with those parameters. This is why oversampling DACs are used.Oversampling is a technique, which uses much higher sampling rate, than it is needed from the standpoint of usable bandwidth. For example, if we are interested in 24 kHz bandwidth 48 kHz sampling frequency is theoretically completely sufficient. However, if we are sampling with 64×48 kHz = 3.072 MHz frequency and we are interested only in frequencies up to 24 kHz it means oversampling by factor of 64. Oversampling paired with specific modulation extends the resolution of DAC. This is why 6-bit DAC with Delta-Sigma modulation is able to achieve 32 bits of effective resolution, as it is with ESS SABRE DACs. Additional benefit of Delta-Sigma modulation is noise shaping. Quantization of analog signal, which makes it possible to digitally represent signal, induces quantization noise. It is because we are encoding only specific levels of signal, and we cannot represent what is between those levels. It is quite intuitive, that more bits result in less quantization noise, because difference between each quantization level is smaller. Therefore, if the DAC itself has 6 bits it should have quite high quantization noise, but because of Delta- Sigma modulation this quantization noise is pushed into higher frequencies, and it is almost non-existent in usable (audio) bandwidth. This effect is called noise shaping (Fig. 1).Fig. 1 Impact of noise shippingNoise shaping is not present in every modulation. PWM (pulse width modulation) does not have this characteristic, although at first sight signals of PWM and Delta-Sigma modulation might look similar. Differences between PWM and Delta-Sigma exceed range of this text.Main advantage of oversampling DACs is possibility of using lower resolution and higher sampling frequencies because those devices are simpler and more repeatable in production. When combined with oversampling and Delta-Sigma modulation they can surpass non oversampling DACs in effective bit resolution. What is more, there is only need for one analog antialiasing filter (Here this filter is additionally filtering the noise on higher frequencies from Delta-Sigma modulation), which doesn’t have to be tuned and is much simpler. Even when the input sampling rate is changing, only the oversampling factor must change and sampling frequency of DAC itself stays the same. Additionally, because usable bandwidth is many times lower than effective sampling frequency, the slope of the analog antialiasing filter (sometimes also called reconstruction filter) can be much lower, which makes filter much simpler. What is more, this filter can have cutoff frequency much higher, therefore this filter has less impact on signal in audio bandwidth.Digital filter in oversampling DACAlthough oversampling DACs are solving many problems in analog realm, they require more attention at digital side than non-oversampling DACs. Let’s assume, that we want to stream 48 kHz sampling rate file. In the case of oversampling DAC, before signal will be converted to analog form, it must be upsampled from 48 kHz to frequency at which DAC is operating. This operation is called interpolation and that is why digital filters used for this purpose in oversampling DACs are called interpolation filters. However, digital interpolation itself in its simplest form has inconvenient characteristic, which is mirroring signal relatively to old sampling frequency. For example, we have 1 kHz signal sampled with 48 kHz clock. When we will upsample this signal to 96 kHz sampling rate, there will be signals at 48 – 1 = 47 kHz and 48 1 = 49 kHz (Fig. 2). This mirroring characteristic is called aliasing and it requires usage of additional digital filter, which will eliminate those aliasing ‘artefacts’, before signal will be applied at DAC input.Fig. 2 Interpolation ‘artefacts’ and filteringHowever, there is price for using this filter because it has significant impact on the signal in the audio frequencies. There are many factors, which can influence the final sound signature, like slope of the filter, its cut-off frequency, pre- or post-ringing of its impulse response. It is a complicated topic and that is why we are offering many interpolation filters, which have different characteristics for sound signature tuning. Now we have broadened our set of filters by cooperation with expert in that topic. WANDLA does what no other DAC does. Our new Ferrum Streaming Control Technology lets you control your streamer with your WANDLA remote control and also display title and artist info on its touch screen. Being the most important part of the WANDLA 2.0.0 firmware update, let’s focus on our Ferrum Streaming Control Technology for a minute. Just imagine. You made the decision to upgrade your computer’s sound by connecting our WANDLA to its USB port. Your streaming content has never sounded this good and you will probably play more of your favorite music than ever before. Wouldn’t it be more convenient for you if you would be able to also put the basic controls and album info in the place where your great sounding music is flowing from? With our latest firmware update WANDLA can do exactly that (and more), and puts your music center stage conveniently at your fingertips! Right from WANDLA’s touch screen With our Ferrum Streaming Control Technology (FSCT) you’ll be able to control your streaming music right from WANDLA’s touch screen and its remote control. A quick look on WANDLA’s screen will tell you title and track information and give you the ability to play, pause and skip your beloved songs. While our finely tuned UX makes it a breeze to operate, FSCT puts the control over your music right where it belongs. On WANDLA and its remote control.The complete list of new features our WANDLA 2.0.0 firmware will present looks as follows:FSCT: convenient control of your streaming content at your fingertips*.Volume Synchronisation: control DAC built-in preamp volume from your computer and vice versa.Improved user experience (UX): focused on presenting relevant information the best way.I 2 S compatibility: support for more devices connectable via I 2 S input.Favorites input: choose frequently used inputs and select between them easier. *At launch Windows, MacOS, and Volumio OS will receive full support, provided that their applications, plugins, and services include all necessary features. Partial support may be available on other platforms, streamers and Operating Systems depending on their implementations. Please consult our website for the latest information.