The LAST publicly available comparison test of HDTV's and HD-STB was way back in 2005, but model numbers were NOT disclosed:
The biggest disparities in performance were Sensitivity (Noise Figure) and how many of the ATSC A/74 "Field Ensembles" the different boxes could receive.
The 50 "Field Ensembles" are a set of Lab & OTA signal recordings captured during various field tests representing WORST CASE Multipath. Some of the captures are known to be damaged and hence unrecoverable. And a few may be so "bad", they'll probably never be successfully decoded:
In early 2007, CECB's (Coupon Eligible Converter Boxes) are the first (and ONLY) Tuners REQUIRED to actually MEET ATSC A/74 Receiver "Guidelines":
Here are results of comparison tests, but again, model numbers are NOT disclosed:
Section 3 includes the Field Ensemble test results, showing a big improvement wrt to HDTV & HD-STB's for 2-years earlier.
NAB also sponsored a comparison test, but again model numbers were NOT disclosed in test results:
Note that ALL CECB's had the SAME Sensitivity, give or take 1 dB, so no clear winners......or losers for that particular spec number.
And that's the last we've heard on this subject...and BTW, all of the above tests were on Single-Conversion Super-Het Tuners, except for the first-ever Prototype 1st Gen ATSC Tuner that was Double-Conversion. So we still don't know if Double-Conversion Tuners, like Microtune and a few Samsung Tuners have better or worse (my suspicion) resistance to strong out-of-band signals.
Zenith (who sold the ATSC Patent Rights to Korea) created their own "Generation" designation:
Gen Ghost Post-Ghost Pre-Ghost Dynamic
Amplitude Length Length Speed
First 70% 10-20 us <3 us 5 Hz
Second 80% 40-44 us 3-5 us 8 Hz
Third 90% 40-44 us 5-10 us 10-15 Hz
Fourth 95% 40-44 us 10-20 us 20-50 Hz
From: "Advances in 8-VSB and Receiver Performance: The Digital Evolution",
Wayne C. Luplow (Zenith), at CITEL PCC II, Fortaleza, July 2002.
In 2004, LGDT3303 chip was labeled "5th Gen", pushing the Pre and Post-Ghost Lengths to +/- 50 us, an IMPORTANT improvement:
From my experience with Adaptive Equalizers for HF Communications, 20-years EARLIER, we learned that it is IMPORTANT to have an Equalizer that is TWICE the length of the maximum delay expected. When the Equalizer is initialized the first time or when recovering from a severe fade, it doesn't KNOW whether the received signal preamble signal is a Direct Path signal, or is a Delayed Path signal. The early Equalizers assumed that they were ALWAYS locking onto a Direct path signal and failed whenever the Direct path signal had momentarily faded away. A Double-Length Equalizer always puts the Preamble signal in the MIDDLE of the Equalizer, so it is ready to accept a late arriving Direct Path signal.
In early 2007, CECB "6th Gen" chips were REQUIRED to meet ATSC A/74 "Guidelines", and subsequently were head and shoulders better than earlier chips. We should hope that any HDTV or HD-STB designed since about 2008 will have equivalent "6th Gen" (and maybe slightly better) performance. FYI: I enumerated the various 6th Gen chips in the "CECB FEATURES", found in my signature line below.
Unless manufacturers disclose how much "better" they are at meeting various requirements, any claim beyond 6th Gen is pure marketing hype....or an admission that their earlier 6th Gen chips weren't quite as good as the others and had room for improvement...
BTW: I doubt that there has been much advancement in tuner technology since the 6th Gen chips.....they were THAT good....and apparently no one has any interest in reporting what's in their HDTV or HD-STB.