Is it fair yet to say that the daisy chain system TVO went for can be labeled as a failure considering how often it reverts down to 480i?

I remember hearing about that in the video of that meeting in Toronto at that University, prior to the transition. If I recall, the driver for that system was to conserve Transport costs.
Without out it, they said they wouldn't have had the funds for Transport to all the transmitters. In essence every station could not have gone HD without it at all.
So in that sense it may be unfair to say it's a failure.
Most of the transmitters most of the time is better than None of the transmitters all the time.

Most of the transmitters most of the time is better than None of the transmitters all the time.

Really? Actually, come to think of it; do I really need 2 TVO stations overlapping coverage (Paris transmitter and London transmitter). The short answer is no.

I come from the school of "do it right the first time" unless it is a hobby or passion.

TVO's ventures regarding OTA transmissions seem like a broken template that you can repair but never really fix.

The Windsor transmitter has had problems pretty much since day one.

Last night I couldn't get tvo London at all on either of my tv's. I had full signal strength but my Sony had the error message "cannot decode signal". The toshiba had a blank screen and if I brought up the signal meter on physical channel 18 it showed full signal but no virtual channel number would come up. I'm glad I can get TVO Kitchener because the link between kitchener and london is a disaster.

Really? Actually, come to think of it; do I really need 2 TVO stations overlapping coverage (Paris transmitter and London transmitter). The short answer is no.

I come from the school of "do it right the first time" unless it is a hobby or passion.

I find these two paragraphs a bit of a contradiction. The right thing to do is have transmitters with overlapping noise bounded contours to ensure there are no holes and to have transmitters in every major market so that people with indoor antennas can still receive the broadcast. Yes people with large outdoor antennas may be able to receive the same station from more than one market that way, but that doesn't mean everyone will be able receive both (I did a quick TVFool analysis for London and CICO-DT-18 has a NM of about 1dB, certainly receivable with a good outdoor antenna, but likely impossible with an indoor one).

TVO's ventures regarding OTA transmissions seem like a broken template that you can repair but never really fix.

TVO is doing the best they can with the resources they have (don't forget they also had to add DTV transmitters in 3 non-mandatory markets where they were out of band). It is a bit of a balancing act. Money spent on OTA transmission is money they can't spend on programming. Adding an HD satellite link or fiber optic links is not a one time cost but a reoccurring fee. Eventually they will upgrade their SD satellite link to HD, but that will require upgrades to all of their transmitters province wide (or else shut them down).

For now, if there are problems with the microwave link between Paris and London, maybe they need to build a repeater half way between. This type of project takes time (and it isn't the best weather to be doing that type of work).

TVO relays the signal down the highway 401 corridor by receiving the adjacent DTV station broadcast - the same signal viewers receive at home. The original source is TVO channel 19 from the CN Tower in Toronto. Paris receives Toronto, London receives Paris, and so on down to Windsor. The cost of this relay technique is relatively low cost professional DTV receivers plus maintenance compared with monthly service costs for fibre or satelllite so it is very attractive to a company with very tight budgets dependent on government support and donations for funding.

At some point TVO will go up on satellite in HD and feed all the digital transmitters using this signal. All remaining analog transmitters will require an HD satellite receiver with down-convert capability. Internal down-converters are common even in domestic satellite receivers. Bell or Shaw Direct, or both, will carry the TVO HD signal and want to continue with an SD signal for the bulk of their customers so the HD receiver will probably be unnecessary. When the switchover to HD distribution happens, the off-air relay will be redundant except as a backup for the satellite signal when weather conditions cause loss of signal on the satellite receiver.

Makes lots of sense. TVO off the CN tower has been poor here lately. So poor to the west also.

Oh, I was under the impression it was a microwave link.

I can understand problems with the Windsor site, after all they are trying to receive channel 33 off air from Chatham, and transmitting on channel 32.
A high level of isolation is required between the two channels at the receive end, tight filtering,with steep skirts which unfortunately introduces insertion loss and group delay in the received signal.
A/D

Oh, I was under the impression it was a microwave link.
My thoughts as well.

Wouldn't quality of the signal eventually degrade? Which I suppose could explain why Windsor--presumably the "end of the line"--ends up with the worst quality; requiring the signal to be bumped down to SD frequently.

(I think I just answered my own rhetorical question)

http://www.smpte.org/sections/toronto/meeting-archive

Look for the march 8th meeting, there's a video.
Ya may wanna try and fast forward to TVO's presentation. I don't think I can test the archived video here, on linux, as it requires Silverlight.

My thoughts as well.

Wouldn't quality of the signal eventually degrade? Which I suppose could explain why Windsor--presumably the "end of the line"--ends up with the worst quality; requiring the signal to be bumped down to SD frequently.

(I think I just answered my own rhetorical question)
Well really no, as its digital so no loss of quality.

The switch from 1080i over the UHF relay to 480i from satellite is a path reliability issue. With multiple hops from Toronto to Windsor, reliability at the end of the chain is the sum of all the outages on each link.

Here is a simplified example of transmission system reliability. If the first link in the transmission system is 99.9% reliable, the first transmitter in the chain will loose signal for 8.76 hours per year. If the reliability on each path is the same, 99.9%, the second link will be out for 8.76 hours as well for a total of 17.72 hours for the second transmitter in the chain. At the end of 4 relays, the total outage will be 35.04 hours per year and reliability will be around 99.6%. Change the path reliability from 99.9% to 99.5% and the outage time for a single link jumps to 43.8 hours per year and the total system at the last transmitter suffers 175.2 hours outage per year! That sounds really bad, but at the same time it means that the transmission system works for 8,584.8 hours per year at the last transmitter.

So in the theoretical example, the viewer at the end of the line gets 8,548.8 hours of HD per year and 175.2 hours of SD.

Well really no, as its digital so no loss of quality.
I understand that. My point is that you're stretching a signal over several hops, and thus exposing it to more interference "opportunities" (for lack of a better term).

george,
Is sumthin like this outside the realm of possibility for a hypothetical relay?

Switch downstream 8VSB input from a fading 8VSB path to a more robust ATSC M/H Path, assuming it can be transcoded from H264 to MPEG2 prior to next 8VSB input
8VSB (MPEG2)------->8VSB (MPEG2)------->8VSB (MPEG2)------->8VSB (MPEG2)
ATSC M/H (H264)---->ATSC M/H (H264)---->ATSC M/H (H264)---->ATSC M/H (H264)

Great explanation GeorgeMx. The only flaw is you assumed only one link will will be down at a time, if outages were random, the difference would be insignificant, but there is a good chance the outage is weather related, and more than one link will likely suffer from the same weather conditions.

I don't think I can test the archived video here, on linux, as it requires Silverlight.

I haven't investigated it recently, but check out Moonlight (Silverlight for Linux).

Ahhh. For now Short of the presentations, which I've already seen back then, I don't think I'm missing much without it. Probably the first time I've come across it. If ever Really do need it will look for it.

I understand that. My point is that you're stretching a signal over several hops, and thus exposing it to more interference "opportunities" (for lack of a better term).
perhaps 'weakest link in the chain' will describe it better than 'degrade'