Industry Voice

2. What are the most important features of your product that directly affect sound quality, and can you talk about why they offer an improvement?

Jonathan Reichbach, President, Sonic Studio (Amarra)

We feel there are many sound engineering principals that can positively or negatively effect the way music sounds. For over 25 years now, we've studied these concepts and applied our findings in the development of our Sonic Studio Engine (SSE), which is recognized as the most accurate and most 'analog' sounding audio playback system available for professional and consumer audio applications.

Two areas where we focus our efforts are 1) advanced digital signal processing for audio applications and 2) in optimizing the interaction between the SSE, the computer hardware and its operating system with the goal of attaining the very best music experience possible.

Some of the advanced digital signal processing features of SSE include our world class Sonic Mastering EQ which is a very high quality minimal phase frequency equalizer, as well as digital volume and dithering algorithms that are at least as good or superior to the best analog based solutions. The minimal phase topography preserves imaging while improving the music.

The second area addresses the optimization of audio playback in the very noisy environment of a modern day computer like the Apple Macintosh. Amarra and SSE are designed in a manner which optimizes to a degree all the competing resources for memory, CPU, and USB devices. These techniques result in reducing system interaction and overhead which can lead to an overall lower electrical noise and emissions and better sound from music reproduction systems.

Damien Plisson, Founder, Audirvana, (Audirvana Plus)

No alteration of the recorded audio signal (unless voluntary chosen by the user).

"Bit-perfect" playback with 64-bit internal processing and Automatic sample rate switching.

Exclusive access to the DAC to prevent other applications (e.g. system beeps) from interfering.

Shortest audio signal path to minimize computer induced jitter & interferencesp

Full memory play: Loading, decoding, and converting to the DAC native format the track to play entirely in memory before playback. So that no disk or any intensive CPU activity is interfering during playback.

Shortest audio signal with Integer Mode to optimize the sound path inside the OS X kernel. Now available (in beta version currently) also for Lion & Mountain Lion.

SysOptimizer to automatically deactivate unessential services detrimental to audio quality

Worldclass audio filters from iZotope for upsampling and dithered volume control.

And the "High quality" tunes are lossy compressed files…

Tim Murison, Co-Founder & CTO, BitPerfect Sound Inc., (BitPerfect)

High quality playback is all about efficiency - getting the music from the disk to the DAC with the fewest steps in between, which really means no features! If extra steps are required, for instance for volume control or to match a supported sample rate, then high-quality algorithms are important.

Stephen F. Booth, Founder and Developer, sbooth.org, (Decibel)

The most important features of Decibel affecting sound quality are automatic sample rate switching and minimal audio processing. Listening to music is a personal experience; what sounds amazing to one person may not sound equally amazing to another. With this in mind, my goal with Decibel is to take the audio bits from the disk to the DAC in an efficient manner without altering the source audio in any way, to remain faithful to the original. This is why the DAC sample rate is adjusted to match the file—to ensure the audio remains untouched—and that no gain or other processing is applied to the audio unless specifically set by the user. In cases where the output volume is adjusted digitally I use 64-bit computations for greater fidelity. Finally, to minimize the possibility for audio glitches Decibel provides the option to load a file in memory completely before it is played.

Jussi Laako, Owner, Signalyst, (HQ Player)

Most features are fairly technical...

The original first goal of HQPlayer was to provide software-based upsampling for RedBook material with a number of different algorithms, especially minimizing length of ringing while maximizing stop-band attenuation, since this combination helps the DAC to reconstruct the original waveform represented by samples as accurately as possible. Since this process can produce sample values that have lot of decimals (including irrational numbers), it was equally important to represent these in the maximum bit depth supported by the DAC as accurately as possible. This requires optimal dithering and/or noise-shaping algorithms. Which also allows optimal digital volume control. There is also support for downsampling for the purpose of playing back hires files at lower resolution, when DAC doesn't support the original resolution.

The second most important feature is software-based delta-sigma modulator for upsampling to "native language" of modern DAC chips. This is usable with DACs supporting DSD playback (DSD is one-bit delta-sigma). Since doing this process in an optimal way is very resource intensive, there's a benefit of doing it with processing resources of a modern computer. There is also support for playing DSD content from DSDIFF and DSF files, either to a DSD-capable DAC or through PCM conversion with a number of different conversion algorithms. Delta-sigma modulation is an elegant way of reaching very good low level linearity without excessively heavy and expensive hardware tolerance requirements.

Other important features are inclusion of convolution engine for applying digital room correction filters in real time, and support for multi-channel music playback with speaker delay and level adjustments.

HQPlayer moves a lot of functionality traditionally implemented in hardware to the playback software. This provides more processing resources, configuration flexibility and easy software upgrade-ability. Hardware implementations tend to be constrained in these areas.

Josef Piri & Marcin Ostapowicz, JPlay (JPlay)

This is how we describe this on our website: JPLAY was built with only one goal in mind—optimal music reproduction. We introduced a number of world's-first features for the ultimate computer audio music playback:

- Full memory-based playback: most other memory-based players dynamically load tracks into memory during playback. In contrast, JPLAY pre-loads complete playlist into RAM guaranteeing no disk activity during playback (zero disk I/O). In our opinion any disk utilization during playback compromises sound quality.

- DirectLink: For the first time CD material is guaranteed to work with lowest latency of 1 sample on most PCI or USB audio interfaces.

- Large Page Memory: superior memory management provides minimal CPU latencies.

- JPLAY as a Windows Service: JPLAY is the first and only audio player that operates as a Windows Service. By leveraging ‘Session 0 Isolation’ JPLAY can get to non-fragmented memory sooner than with current approach of manually tweaking loading order of startup programs. No user intervention is required as Windows guarantees JPLAY will be started at earliest possible moment for best results.

- Maximum System Timer: reduce operating system latency by making Windows switch tasks faster. (0.5ms instead of default 15.6ms)

- Maximal Priority Scheduling: ensure uninterrupted flow of music data by running music playback at highest possible priority.

- Throttle Mode: when Throttle mode is activated, JPLAY will reduce to absolute minimum both CPU and I/O priorities of all running processes and individual threads in the system it can get to, without risking the stability of Windows.

- Hibernate Mode: cancel OS ‘noise’ by eliminating dozens of jitter-inducing processes and hundreds of threads.

JPLAY is the first and presently the only audio player that can totally take control over Windows in a way that nothing else is allowed to run during playback - no processes, threads or services limiting sound quality. In Hibernate mode JPLAY uses full power of the PC; all the CPU cores to provide not only bit-perfect stream, but more importantly - 'time-perfect'.

In music, timing is everything. And in digital audio doubly so: while producing bit-perfect output is easy, producing it at exact time required by digital formats (e.g. 32 bits every 22 microseconds for CD) is not. Why? Because while your PC may be really fast, it’s also doing hundreds if not thousands other things at the same time it plays music. With so many things going on, do you trust it will always ‘hit the beat’ at just the right time? Programming optimizations in JPLAY are designed to minimize both software and hardware interruptions in order to make it ‘easier’ for PC to ‘keep the rhythm’.

JPLAY also offers 3 unique bit-perfect playback engines: River, Beach and Xtream. Even though they do not modify music data, the sound they produce vary.

Jim Hillegass, Founder and CEO, JRiver (JRiver Media Center)

JRiver's internal audio path is 64bits. This allows optional manipulation of audio (DSP, Volume) without loss of precision. JRiver supports all direct audio paths: ASIO, WASAPI, Kernel Streaming, bypassing manipulation by the Windows Mixer. JRiver supports bitstreaming audio to a receiver or other device. JRiver supports multi-channel audio, high bitrate and high bit depth audio.

Dr. Rob Robinson, Director of Engineering, Channel D, (Pure Music)

I think an often overlooked aspect of good sound quality is having a product that functions reliably. When it isn’t functioning, discussions about sound quality are moot. As a seasoned Apple Macintosh programmer, I’ve learned the necessity of writing code that follows the Apple APIs “by the book” and not take risks using hidden or undocumented features that might add zing or serve as a short cut to overcome a design problem or Apple OS limitation. These will break in the long run, or lead to unreliable or unpredictable behavior. That’s not what we need for Computer Audio.

But rather than speak about specific features, how about their origin? They have come about because the person who created the product is a committed audiophile that happens to know something about computer programming, instead of the other way around. For over 16 years we have invested in the company to insure that our products will deliver exceptional performance and sound quality. For example, we have 5 turntables (two of which have two or three tonearms available, and a number of different cartridges at many price points) for supporting our Pure Vinyl product development. We have a testing inventory of over two dozen different DACs and audio interfaces. We have a dedicated listening room (see photo) for developing and evaluating our products, with a selection of different loudspeakers available. I would like to think that in the arena of companies providing digital media playback (and recording) software, we are unique in our level of commitment to “being audiophiles.”

We also “tour” extensively at audio expositions. 2012 will mark our 7th straight year mounting an exhibit at Rocky Mountain Audio Fest, for example, and the 5th trade show this year we have exhibited at (including overseas, at Munich High-End). Nothing tests a product like the stressful conditions of a live audience demo, and this is a good way to find possible unknown operational issues. This benefits the users of our products, because the more testing of the products, the better. We also can cater to user suggestions and enjoy face to face discussions with our users. Many other exhibitors prefer to use Pure Music as well, with the net result being several top sound of show and other awards.

Besides computer programming and digital audio, we also understand analog, having designed our unique 5 MHz bandwidth Seta phono stage, top rated as a Class A Recommended Component by Stereophile (see photo of our shop where circuit board assemblies undergo final QC testing, and some custom phono cable adapters are assembled). I design products to please myself and like minded people. So the features in the products are considered from an audiophile perspective, things that audiophiles can appreciate, designed by someone that has a thorough knowledge of computer programming, digital, and analog, as well as a music lover (my music collection can be measured in units of tons).