The DG-28 is a digital graphical equalizer, which corrects in the frequency domain only with 64 linked filters. This is like the Z-Systems RDP-1 but with more filters. The Sigtech and Tact units correct in the time domain, that is the arrival times of the sounds that are "distorted" due to room reflections and speaker phase anomalies. The approach and results of these two signal processing technologies is worlds apart. Frequency domain correction only will not give satisfactory correction.
Frequency domain correction only will not give satisfactory correction for these room interactions. Adjusting the frequency response without correcting time anomalies results in a processed sound that degrades well produced recordings.
Forgive me for reposting some previous material, but it addresses questions raised in this thread. The reason DSP doesn't degrade the signal is that the Sigtech algorithms have a bit precision that is greater and calculation rates that are faster than the Redbook CD standard. Sigtech’s internal bit precision is 56 bits compared to the CD’s 16 bits and can calculate a full bandwidth 2544 tap filter at the Redbook 44.1 kHz sampling rate. The particular filter implementation is software determined based on the actual time domain measurements of your room and speaker.
The sound at your ear should be both and simultaneously phase accurate and with a flat frequency response relative to that encoded in the original bit stream. The speaker and the room interaction “degrades” that sound through standing waves, first refection bounce, etc. are the primary targets for the Sigtech algorithms. Once these are addressed the unit can shape the overall frequency response if desired but if your speakers are flat to begin with most of the frequency correction happens as a result of the time correction.
With good room treatment (Physical or DSP) the goal should be to smooth the frequency response while damping the first and most of the second reflections while leaving the direct sound and later reflections to provide air. Sound travels at about 1 foot per millisecond. Reflections under 5-7 ms are perceived as colorations. Over 30 ms the perception is of ambiance, reverberation or echo. The best DSP units use algorithms that are designed with the psychoacoustics in mind as well as with variable amounts correction with respect to frequency so as to not over correct a speakers inherent response. If your speakers are already impulse coherent to begin with, like the Dunlavys, all of the DSP power can be applied to room correction. The Sigtech can correct up to the first 50 milliseconds.
If your primary source is Analog, I highly recommend the ASC products. By absorbing and diffusing reflections they do reduce time domain distortions. I use them along with the Sigtech, as I listen to a mix of analog and digital.
Frequency domain correction only will not give satisfactory correction for these room interactions. Adjusting the frequency response without correcting time anomalies results in a processed sound that degrades well produced recordings.
Forgive me for reposting some previous material, but it addresses questions raised in this thread. The reason DSP doesn't degrade the signal is that the Sigtech algorithms have a bit precision that is greater and calculation rates that are faster than the Redbook CD standard. Sigtech’s internal bit precision is 56 bits compared to the CD’s 16 bits and can calculate a full bandwidth 2544 tap filter at the Redbook 44.1 kHz sampling rate. The particular filter implementation is software determined based on the actual time domain measurements of your room and speaker.
The sound at your ear should be both and simultaneously phase accurate and with a flat frequency response relative to that encoded in the original bit stream. The speaker and the room interaction “degrades” that sound through standing waves, first refection bounce, etc. are the primary targets for the Sigtech algorithms. Once these are addressed the unit can shape the overall frequency response if desired but if your speakers are flat to begin with most of the frequency correction happens as a result of the time correction.
With good room treatment (Physical or DSP) the goal should be to smooth the frequency response while damping the first and most of the second reflections while leaving the direct sound and later reflections to provide air. Sound travels at about 1 foot per millisecond. Reflections under 5-7 ms are perceived as colorations. Over 30 ms the perception is of ambiance, reverberation or echo. The best DSP units use algorithms that are designed with the psychoacoustics in mind as well as with variable amounts correction with respect to frequency so as to not over correct a speakers inherent response. If your speakers are already impulse coherent to begin with, like the Dunlavys, all of the DSP power can be applied to room correction. The Sigtech can correct up to the first 50 milliseconds.
If your primary source is Analog, I highly recommend the ASC products. By absorbing and diffusing reflections they do reduce time domain distortions. I use them along with the Sigtech, as I listen to a mix of analog and digital.
