This feature looks unusual and insteresting. Meridian MQA claims something similar in the encoding phase. Can you explain it with more detail? Is this a feature of the general HQ Player or proprietary of the SGM?
The look ahead processing is a feature of HQ Player written by Jussi Laako in Finland. HQ Player does the detection and elimination realtime during play back.
MQA does something similar based on the known parameters of the AD converter used for the original recording and this is carried out during the MQA data file creation process. HQ Player does this to every track it plays, realtime, and royalty free. I won't repeat Jussi Laako's comments about the MQA business model.
HQ Player can be run on any powerful enough PC. However not all PC's are equal when it comes to RF noise generated, and the consistency of the timing.
On the SGM team, we became big fans of HQ Player software 30 months ago. We used HQ Player with a Kodi media player front end at Munich 2015 and garnered this review from Michael Lavorgna.
http://www.audiostream.com/content/...-garde-digital-done-right#FQws6TsvxuvSRXkb.97
That was the catalyst to develop and manufacture a no holds barred computer platform to run HQ Player. Built by driven audiophiles, to deliver listen pleasure and goosebumps to fellow audiophiles.
Send me a PM and I can email you a slide deck with more details
The sound quality achieved is dependant on the CPU core speed. Jussi recommends a minimum of 3.6 GHz. The SGM CPU runs at a clock speed of 4.2 GHz. However as you go to faster core speeds, RF noise goes up. If we take a stock PC with a i7 6700 K processor, the sound you get at 4.2 GHz clock speed is noticeably worse than sound at a clock speed of 3.6 GHz. The extreme measure which are taken to get the RF noise under control in the SGM allow the benefit of higher clock speeds to shine through.
The OCXO on connected to the mother board plays a role here as well. The standard crystal has a frequency stability of about 50 parts per million. We have experimented with TCXO's and OCXO's and the best sounding oscillator which is still reasonably priced is a 5 parts per billion OCXO. We could hear the sound quality improvement as we went to more and more stable clocks. We were not sure why, but the effect on the sonics was repeatable, and consistent. Well now we can actually measure a reduction in CPU power consumption as we use more and more stable clocks. The OCXO which we use gives about a 6 percent CPU power reduction.
What I have been told is that CPU's are not precise in their operation, and error checking will often trigger a process to be repeated. A more stable clock for the CPU reduces the error rate.
There are many things which we have done on the hardware power supply side of the SGM which are pretty common practice in high end audio. To apply these same practices to a purely digital device would seem to be inconsequential, but the surprise is that the beneficial effects are of a similar order of magnitude as they are for analog electronics.
Every SGM owner who has replaced the supplied fuse with a Synergistics Research Black has noticed a nice SQ uplift ! That gives you a clue to what we have done on the power supply wiring front.
The price of the SGM is double the initial target price. We ended where we are because we kept on finding improvements that we could not in leave on the workshop table. Doubling the size of the dual chokes in the power supply greatly increased the cost and weight of the SGM, but the depth of the soundstage increased dramatically. The choice was no choice, we had to build the SGM with bigger chokes
Adam,
Apologies for hogging your thread but hope you find this post relevant to your computer audio journey