On Jun 1, 2010, at 8:28 AM, john_ke5c wrote: > My sujective assessment from experience is that "comparable" analog and > digital systems perform "about the same". I can objectively compare analog > systems using SINAD, and I can objectively compare digital systems using BER. > There is just no standard for objectively comparing analog and digital > systems against each other
Unless... and I've been saying this for what, three years now...?... the CODEC manufacturer publishes where their drop-out point is in BER. Then you could make a judgement call about which SINAD measurement on analog is the "drop-out" point (or refer to various debated publications over the years), and consistently use that number vs. the BER drop-out point as a way to get as close to an apples-to-apples comparison you can get, without doing a full MOS voice analysis test. It's do-able if the AMBE CODEC weren't a proprietary black-box... or if DVSI would publish the drop-out point in BER where the chipset says, "I give up." The use of the one significantly good performance CODEC chipset in almost every form of digital two-way radio from major manufacturers, that also happens to be completely proprietary and closed, hampers the ability to test outside of a DVSI lab under NDA. I doubt even the Icom, Motorola, Kenwood, and other engineers have been able to do that... sadly. But if they're under an NDA, we'll never see the numbers... I hear there's a relatively new ($20-$25K) IFR test set that can record arbitrary digital waveforms off-air (clean input to the receiver port in a lab environment, assuming that the manufacturer's own user radios are clean...), and then reproduce them with the push of a button. That plus attenuation would lead to some nice starting numbers, barring published ones from DVSI. Unfortunately I don't have nor need one of those Service Monitors at that price. A local lab was trying to acquire one last year. I'll have to check and see if they ever received it. -- Nate Duehr, WY0X n...@natetech.com
