“High Resolution Audio”, what do we mean with that?

Like many others in the music and High End audio industry, we at The Spirit of Turtle claim to deliver “High Resolution Audio" recordings. This term might seem straightforward and easy to understand, but in practice it's used for far too many formats of which several do not always qualify. Because of that, for some people, it has become a rather confusing term, causing many misconceptions amongst non-experts. Therefore, we'd like to explain what we do at The Spirit of Turtle, and why we call our recordings and audio files “High Resolution Audio”.

It might not come as a surprise, but a good recording starts with good musicians, a suitable recording space or studio, impeccable artistic producing, inventive microphone techniques and an optimal implementation of “state of the art” tools like microphones, cables, amplifiers, digital converters and playback monitoring. Also the control room where the recorded material is auditioned and evaluated is incredibly important, as this is the place where the music is frozen into the bits on the tape or Hard Disc. Many decisions made at the time of recording cannot be altered or eliminated afterwards. We at The Spirit of Turtle mostly work with custom tailored tools which perform according to our wishes and specifications developed over more than 25 years of trial and error and this has lead to a highly proprietary facility. Recording tools and microphone techniques that guarantee an optimum in transparency, dynamics and temporal resolution. The source formats we have used and currently use, and their inherent “high resolution”, have always been according to the highest standards at that moment in time, and several ground breaking formats even have been introduced as a worlds first by ourselves in collaboration with hardware manufacturers like Data Conversion Technologies - dCS (UK) and Merging Technologies (CH), companies which have been leading in the field over the past decades.

" All High-Resolution Audio starts with a good recording ... "

Please, never believe anybody who will tell you that the delivery format on a physical disc or audio file is responsible for a “good sound”. The format only communicates and/or translates the source material - which cannot be high resolution, if microphones and the other tools further up the line are not dealing properly with the enormous wealth of colors and temporal transients which music consists of.

What we try to explain here is that the format in which a recording is released comes at a rather late stage of the process of bringing the music from the artist towards the audience, and is not and can not be responsible for a good result other than simply allowing for the quality of the performance and recording to become apparent. It is not DSD, DXD or PCM or whatever number attached to it which is responsible for a good recording! A bad recording is still bad in DSD or DXD, it is often even more apparent that it is bad because one hears it better due to the higher “resolution” of the delivery format…

Having said this, the delivery format can certainly take away something from the original quality if the original recording resolution supersedes the delivery resolution. For instance, a recording done in DSD or DXD will not translate 100% to a traditional CD as its resolution is roughly 8 times less in terms of accuracy in the frequency- and time domain. These numbers do count, however a good recording in DXD or DSD is still a good recording on CD…

All recordings we share with you as “High Resolution” audio have been recorded with a sampling frequency of at least 96kS/s (kHz) with a bit depth of 24 bits, but for the majority of the projects with 352,8 kS/s (kHz) at 24 bits, the so-called DXD format. All these recordings are offered in their original format, but ultimately also as DSD. Some have been recorded as so called “Native DSD”, however this format is proven rather unsuitable for several post production processes, which in some cases causes the format to loose its initial advantages as rigid filtering is needed to keep the high frequency noise spectrum within the SACD specification." ... the vast majority of our projects [are recorded] with 352.8 kHz at 24 bits... "

Note: It must be stated that many SACD’s on the market are not what they seem, as the productions consist of stacked DSD channels which consist of too much high frequency noise and therefore are “limited” in bandwidth.

Our reason to output all productions in DSD is that playing back through DSD DACs of a DSD stream generated from a PCM source inherently translates the music ever so often better – in our experience - than even the original does, due to more elegant filter characteristics in theseDACs and a rather different “jitter” behavior of DSD vs. PCM. Furthermore, a dramatically different response from monitoring equipment to the DSD signal supplied makes that it is perceived as a very musically attractive format. In essence, we believe that in most cases the best playback comes from DSD - for reasons described above.

" ... we believe that in most cases the best playback comes from DSD256 ... "

So we decided to offer the 96/24 repertoire also as DSD as the benefits outweigh the fact that the source bandwidth and impulse response are theoretically less optimal than the playback format DSD. The data rate is approximately the same however and our implementation of 96/24 has always been one with gentle filtering which allows for an impulse response and temporal accuracy which suits DSD playback perfectly. This is one exception to our rule that that source material should outweigh the delivery format or at least equal it.

The other exception we've experimented with is delivering our DXD source material in the DSD128 and DSD256 format. The reasoning there is that the proprietary filtering we use in many of our DXD recordings has an optimal impulse response with a Gaussian like characteristic. This delivers a very wide pass band characteristic which benefits greatly when played back in a format from the DSD family with far less, by noise shaping induced, correlated noise. The clean frequency response up to 176kHZ of DXD and its elegant filtering deliver a fantastic DSD128/256 stream which in turn benefits from DAC implementations capable of these formats to output. The best of both worlds in our experience!

What is important to consider however is that the term DSD does often stand for many different permutations of the format. How the so called “bitstream” is conceived in the first place differs greatly from one ADC to another and even within the same analogue to digital converter there often are more than 10 different “modulators” to choose from. So again, never be deceived by a “term” or “number” of the format, but listen for yourself and decide which format communicates the musical message best under the circumstances available…

- Bert van der Wolf

Photo: Brendon Heinst