DAW @48k...alternative DAC comparision

  • On my DAW (Cubase 8 Pro) I record at 48k 24 bit always because of the sonic performance advantage and therefore cannot use the SPDIF output of the KPA being 44.1k only. I've tried and compared the SPDIF vs. Analog out from the KPA and found the SPDIF output superior using clean profiles. But I can't use it... or can I?


    While reading a recent thread in this forum on this subject I recalled I have a DAC (4-6922 tubes in a Class A output stage) and I could therefore compare the KPA DAC to this other external DAC.


    Setup:
    1. KPA Analog out. To RME UFX Analog In. RME UFX USB Output to Cubase.
    2. KPA SPDIF Out to Audio Research Corporation (ARC) DAC3 SPDIF In. Audio Research DAC3 Output to RME UFX Analog In. RME UFX USB Output to Cubase.


    Stratocaster-Clean MBritt and TAF profiles used-Record both 1 & 2 simultaneously in Cubase 48k 24b wave format.


    Listen to matched level playback and blind switch between the two and note sonic differences over Neumann 310's and Sennheiser HD600s.


    The purpose of this was not to subjectively judge which was "better" or "worse". It was to note the differences and be able to have alternatives.
    While listening back and switching between the two it was easy to tell which was which because the tonal balance and the transient response was quite different. Looking at the magnified waveforms showed differences in the structure of the peaks.


    I recorded both the "Stack" (not effects at all) output and later I recorded "Main Stereo" output with some reverb.


    Tonal Balance:
    KPA has more energy in the low bass region, less lower midrange (warmth), more forward upper midrange and a brighter top end.
    ARC in comparison (the opposite) ie. Less low bass, more lower midrange, less upper midrange forwardness, and a less bright top end.
    Transient Response is obviously related to the tonal balance (and the "look" of the magnified waveforms).
    KPA has an edginess to the initial transient that the ARC does not. Last year when I compared the KPA Analog Out to the KPA SPDIF Out and recorded both at 44.1 this was a noted difference.


    Again, no judgment as to better or worse. Every DAC, microphone, pickup, guitar, speaker, amplifier, vacuum tube etc has it's own euphoric coloration and transient behavior.


    Some general information of the ARC DAC3 (not the mkII version) I used. Keep in mind the ARC when new cost more than two KPAs (discontinued in 2000):
    The DAC3MKII uses a high-performance monolithic input receiver and a
    delta-sigma converter supplied to Audio Research by Crystal Semiconductor. The
    differential converter is capable of 107 dB dynamic range in the audio band, and
    includes 8x interpolation and a 64x oversampled delta-sigma modulator which
    outputs to an all-tube (four 6922 twin triodes), class A operation analog output
    stage. Constant-current source technology and highly regulated power supplies
    similar to Audio Research's best preamplifiers contribute to the natural
    musicality of the output stage. The suppression of sonically destructive jitter
    in the DAC3MKII is accomplished via three primary methods: by two
    voltage-control oscillators used to reclock data in the digital interface and in
    a phase-lock-loop circuit just prior to the analog stage; and by extensive
    internal damping of critical components and chassis areas to minimize mechanical
    resonances.A detachable AC line cord has built-in filtering.
    Among other conveniences, the DAC3MKII offers selectable phase inversion via
    a front-panel toggle switch. The DAC3MKII offers a full array of digital inputs
    for compatibility with any high-quality CD/LD transport or player, and provides
    both-single ended and balanced analog output connectors for similar
    compatibility with high-quality line-level preamplifiers."

    Will

    Edited once, last by WillB ().

  • Some addition thoughts:
    While recording and listening to MBritt Voice 15 1 I added some hall reverb. The differences in texture were significant. It is difficult to describe and the reverb itself is artificial but it is like different grain structures to an image. Very different coherence and placement (distance from listener or depth).


    Some items measure well and some don't. Take for example a 1950's Neumann U47 with a VF14 (original tube) recording a human voice. There is something's that captures sound and converts it to electrical energy in it unique and is subjectively pleasing to the ear I assume due to the continued use of this microphone over 60 years later..

    Will

    Edited 3 times, last by WillB ().