Dear our backers,
We wanna share with you the good progress we’ve made on both chassis production and voice quality improvement.
We’ve finally been able to nail down the chassis issues with the right vendor, which solves the quality discrepancy through both machine fine-tuning and manual inspections. It’s a labor intensive process with lower production rate. But, since the initial production volume is not too high, we are able to manage it with our vendors and get the process expedited, in order to deliver the first shipment to our lovely backers who have been waiting patiently for so long.
The first batch of chassis mass production is almost done now, pending the final inspection, and will be delivered to our contract manufacturer for final assembly. Due to the holiday season, our contract manufacturer will be closed for a few days during Christmas, and we are unable to pin down the assembly schedule before Christmas. We’ve continued to work with our contract manufacturer for the schedule now, and hopefully the assembly can be done right after the Christmas/New Year holidays.
Also, we’ve made many changes in the software in order to improve voice quality. Originally, we were using Codec2 from open source for voice compression, and have spent quite an effort to optimize its performance on our DSP chip. But, due to the higher error rate of our analog modem under GSM channels, we were unable to deliver intelligible quality voice with Codec2. To address the problems, we’ve invested our time in improving our modems with error corrections, as well as voice compressions with better error resistance and lower bit rate. So far, for voice compression in the range of 300-600 bps, the only solutions we can find are commercial offerings, but not open source solutions. We’ve been testing with some of the commercial stacks, and adjusting our system software to integrate with new solutions here.
The preliminary test result is promising with the new commercial offerings, but the catch is that it’s from US vendors with export regulation. If we were to use such commercial offerings, then we’ll have to abide by the US “Export Administration Regulation”, which makes JackPair subject to national security controls as identified on the “Commerce Control List”. This means JackPair cannot be exported the following countries, as of year 2015:
- Group E1: Cuba, Iran, North Korea, Sudan, Syria
- Group D1: Armenia, Azerbaijan, Belarus, Burma, Cambodia, China, Georgia, Iraq, Kazakhstan, North Korea, Kyrgyzstan, Laos, Libya, Macau, Moldova, Mongolia, Russia, Tajikistan, Turkmenistan, Ukraine, Uzbekistan, Vietnam
As the goal of JackPair is to make it universally available and affordable to everyone, we really don’t like such a regulation imposed upon JackPair. But so far we couldn’t find any other affordable solutions to solve the voice quality issues. In the long run, to reach the goal of JackPair as an universal solution, we would have to help improve open source solutions like Codec2 to deliver good voice quality in the low bit range of 300-600bps. Meanwhile, to avoid any further delay on the initial delivery of JackPair, we’ll probably have to bite the bullet and limit JackPair’s availability outside those countries listed above.
Overall, the initial batch of JackPair will be assembled by the first week of January, and we are wrapping up software integration efforts for voice quality improvement. Hopefully we can have JackPair at our backer’s hands by end of January.
The JackPair Team