Canadian TV, Computing and Home Theatre Forums banner

HDTV must not be degraded: CRTC

19K views 75 replies 29 participants last post by  studlygoorite 
#1 · (Edited)
This thread is to discuss the current and growing practise by most TV broadcast distributors (cablevision, satellite, telco, etc) to re-compress HDTV channels to use less bandwidth.

In many cases this additional compression results in degradation of the image by causing soft edges (mosquito noise), macroblocking and adding other errors and artifacts.

This additional compression is in violation of CRTC directives that state:
program signals should be of the same quality and in the same format as those received by the BDU, without any degradation.
In fact, not only HD but also SD digital TV stations are covered. For reference I will post the relevant sections of the CRTC notices.

Do you want the distributors (Shaw, Rogers, etc) left to decide how much compression is too much? Or do you want your HDTV channels in the original quality that the broadcaster intended and provided?

You can submit a complaint to the CRTC stating that you object to the image degradation caused by additional compression the distributors are applying to the HD and SD digital TV signals you receive from them, in violation of CRTC notices 2006-74 and 2003-61.
 
#2 · (Edited)
relevant CRTC notices

For HDTV pay and specialty services see CRTC 2006-74 sections 118 thru 130.

In particular, these sections spell out the issue very clearly:
121. The Commission added that HD picture formats do not, by themselves, guarantee the highest picture quality. For example, the digital compression of a signal may not change the format (i.e., the number of lines), but if taken too far, may introduce picture degradation that could result in disappointment to viewers who have purchased expensive HD receivers and displays. In this regard, the Commission considered that the program signals of pay and specialty services distributed by a BDU should be of the same quality and in the same format as those received by it, without any degradation.
...
128. Bell Canada and MTS Allstream et al. suggested that technical standards are unnecessary, as the marketplace will be the ultimate judge of what quality standard is adopted. However, as discussed earlier, the Canadian marketplace is characterized by a small number of key distributors, with the result that pay and specialty services do not have the option of withholding their programming services from distributors who do not maintain adequate signal quality. Likewise, there are viewers who are reliant on a particular distribution platform to receive their video and audio programming services, for example, those in remote areas served only by DTH.

129. For these reasons, the Commission will impose the same quality standards for the distribution of the pay and specialty services as it did for the distribution of the over-the-air services, as proposed in Public Notice 2004-58. Specifically:
...
* the program signals of pay and specialty services distributed by a BDU must be of the same quality and in the same format as those received by it, without any degradation.
For other digital TV signals see CRTC 2003-61 sections 100 thru 107.
107. Accordingly, the Commission adopts the following principle:

# An over-the-air digital television signal distributed by a BDU to its subscribers should be of the same quality and in the same format as that received by the BDU, without any degradation.
There is additional reading of interest (CRTC Call for Comments notices) in CRTC 2004-58 sections 95 thru 100, and CRTC 2002-32 sections 58 thru 61.

My thanks to voyager6868 for his post in the Rogers thread that motivated me to go digging through the CRTC notices.
 
#3 ·
Certainly the move by Rogers has been vetted by its corporate lawyers who would argue that the CRTC principle only applies to over the air redistribution not to pay or specialty channels. (where degradation may appear at source) Just when you think the HD world has settled in.......a new disruption appears on the horizon. This will become a very interesting "technical" arguement that unfortunately may appear before the commission many months after the changeover.
 
#5 ·
3 channels in 2

Hugh found an article with real-world examples of this over-compression on Comcast cable in the US. There is also a picture that clearly shows the effect it has on the image quality.

Comcast Compressing HDTV Signals to Fit Three Shows into Two Shows' Bandwidth

This may not be regulated in the US, but it looks like we have the appropriate regulations in Canada to put a stop to it here.
 
#6 ·
technut;

Alberta doesn't have Class Action Law but BC does and, once started, all customers could join in no matter where they live. Do you know any lawyers that might be interested in taking this on? It appears we would have a cut and dry case.
 
#8 ·
program signals should be of the same quality and in the same format as those received by the BDU, without any degradation.
I don't see numbers anywhere in this quote. That is going to be the problem if you take BEV and others to court. Unless there is some way to quantify quality it is a subjective matter open to interpretation and practically unenforceable. For each person who notices degradations there's another who doesn't. Hardly a cut and dried case.

If you think I'm defending the BDU's you're wrong. There's no doubt that the regulations were deliberately written in this language at the request or direction of the BDUs with the CRTC's concurrence in order to protect themselves from lawsuits like these.
 
#11 ·
I don't think a lawsuit is the way to go with this.

I suggest we start with filing complaints with the CRTC and see how they respond first. The more people that file complaints, the higher priority this should get with them.

If you are on a smaller BDU make sure to get a complaint filed... it may require at least one complaint for each BDU to ensure they are all called to task and none overlooked. In my complaint I tried to encompass all BDUs but I did mention Shaw specifically, since that's my BDU.

At the least, I expect they'll ask the BDU(s) to respond to the complaint(s). The BDUs will probably get a month to respond, and then we'll see how the ball rolls from there.

I think this is well worth pursuing, but don't expect changes soon... any fixes to the delivery of HD probably wouldn't happen until several months after a CRTC decision. We could be talking about this still a year from now.
 
#12 ·
RE: Post #5:

NicoHockey9 was able to do a quantitative representation comparing Comcast to Fios:
I was able to pull a comparison: first bit rate is Fios second is comcast. then the % diference

AETV HD 18.66 Mbps 14.48 Mbps -28.9%
Discovery HD 14.16 Mbps 10.43 Mbps -35.8%
Discovery HD Theater 17.45 Mbps 12.60 Mbps -38.5%
Food Network HD 14.32 Mbps 13.73 Mbps -4.3%
National Geographic HD 11.39 Mbps 10.32 Mbps -10.3%
Universal HD 12.72 Mbps 11.01 Mbps -15.5%
I'm sure this could be done for Rogers pre-compression and post-compression.
 
#13 ·
Here's the text of my CRTC complaint form, in case anyone wants to just copy/paste... but substitute your own BDU name in place of Shaw:


I wish to object to the image degradation caused by additional compression that the broadcast distributors (including Shaw Cable) are applying to both the HD and SD digital TV signals I receive from them, in violation of CRTC notices 2006-74 and 2003-61 (Quality Standards).

I believe the additional compression has been applied to allow the BDU to carry more broadcasts within a channel (eg. 3 HDTV signals in a channel instead of 2). They may claim that statistical multiplexing enables this without any degradation. However it has had a visible effect on the image, creating additional mosquito noise and macroblocking, especially during scenes with fast action or continuous motion.

I would like the CRTC to intervene and enforce your quality standards regulations on behalf of myself and other customers on Shaw and the other BDU systems, most of which are already using or are in the process of implementing similar additional compression.

Thank you.

 
#14 ·
#15 ·
technut, although Rogers will be using statistical muxing (and we haven't yet seen those results), I don't believe Shaw uses it. I believe they simply provide just under 13 Mbps to each channel (or some fixed number).

The other question that arises is whether Rogers will be using the higher bitrate (I believe around 50 Mbps stream) available for some of these channels, vs a max of 19.4 available for OTA.
 
#16 · (Edited)
Thanks 57, I've always wondered exactly what Shaw are doing. They said they did some "upgrades" late last year but I don't know what they were. They aren't very transparent about how they carry their signals.

If I get into a dialog with someone at the CRTC about this, one of the things I will suggest is that they direct the BDU's to submit a list of their incoming signal characteristics, their current channel/frequency assignments on their plant, and what methods they use to divide the bandwidth (TDM, stat muxing, etc).

I imagine the BDUs might claim they don't want to reveal that information for competitive reasons, but I think we as consumers have a right to know what we are getting for our money.

If they don't reveal it then we might be able to get the source signal info ourselves from the sources (with a lot of work), and we can get the channel/freq info from the STB (with a bit of work), but that still doesn't give us the whole story. If they are forced to reveal what they are actually doing, then their compression methods and ratios will be much more obvious. And I think we'll confirm that not only is the HD overly compressed, but also a lot of the SD as well. There are many SD digital channels on Shaw that look far worse than analog ones.

And before someone says "But you'll lose a bunch of channels if they reduce their compression." (eg. 14 HD channels instead of 21), let me say that I'd rather have fewer good quality signals than a lot of low-grade ones. Besides, the BDU's will still find ways to add more channels... they'll juggle their priorities and hurry up their plant upgrades. Maybe even make the leap to MPEG-4 sooner rather than later.

We've moved beyond 20" TVs... these big screens need cleaner images. Leave the over-compressed stuff to the Internet.
 
#17 ·
I've referred this thread to member Kaphyr and asked him to post instructions to members who have a DVR on how to calculate bitrate for recorded programs and how to keep a spreadsheet record.

The above will provide specific data to those who intend to file a complaint to the CRTC.

Having worked before on this with Kaphyr, I have a spreadsheet with the appropriate formulas and it works very well. But Kaphyr is the expert.
 
#20 ·
Calculating bitrate with a PVR like the 8300HD

There is a easy way to calculate the bitrate with a PVR like the 8300HD. This method has been tested and the results are quite precise. If somebody can get some before and after values, this will be wonderful. If you can read french (if not, it's not hard at all to figure out how to use it!), you can also use the bitrate calculator on this page : http://illico2.tripod.com/CalcBitrate.htm

In short, you just need to get the free cluster value, before and after you delete something in the recording list. You can get this value on the page "DVR HDD INFORMATION" (page 18 of the info pages), in the "FILES SYSTEMS" section, "AVFS" column, "Free Space:" line.

So, you recorded something, 30 min long or more. When the recording is finish and there is no other recording in progress (quite important!), you tune to an unrecordable channel (anything with text) or go directly to channel 998, which is better. Get the free cluster value. To see the info pages, you need to hold down the "select" button on the front of the terminal until the "message" blink and press the "info" button. A few "Page Up" on the remote will get you to page 18. Write down the free space value, in the AVFS column. Hit "exit" to get back to the 998 and delete your recording. Go back in the info pages to get the other free space value. The difference between the two will give you the amount of cluster used by this recording.

The formula to get the bitrate is :

(TotalCluster*4700*512*8)/(Duration*60*1000000)

The answer will be in Mbps and yes, the conversion in Mbps is really 1000000 and not 1048576. Since a lot of HD channel use variable bitrate, don't expect the same result all the time but going from 2HD/QAM to 3HD/QAM will clearly show a big difference.

So, if your eyes see a drop in quality, you will be able to get some real value to explain the difference. Don't hesitate to complain to the CRTC if this is the case ( http://www.crtc.gc.ca/rapidsccm/register.asp?lang=e ). You can also write to the affected channels directly.

Kaphyr
http://illico2.tripod.com/
 
#21 ·
Another less precise way is to do a fairly long recording and look at the % available on a 160 GB HDD (less precise if you have more space). Each hour of HD at "full bitrate" (say about 18 Mbps) takes about 5% of the 160 GB HDD, so a 2 hour recording would result in you getting back about 10% of your HDD when you delete that programme. If the bitrate is 2/3 of that, then you'd get back 6-7% of the HDD.

This is how I first spotted the PBS Buffalo lower bitrate - I removed a 1 hour recording and only got back 3% space. I now have a 500 GB external drive, so I would use the more precise diagnostics method.
 
#22 ·
... this is how we used to spot the TMN HD problem, where deleting 1-hr typically only freed up 4%.

But normally I get 6% from 1-hour on CTV, CITY, etc. not 5%. Perhaps I should record a 10-hour chunk while my PVR is pretty empty. I'd guess it's about 5.7% or so per hour based on how many times it's been 6% instead of 5%.
 
#25 ·
... this is how we used to spot the TMN HD problem, where deleting 1-hr typically only freed up 4%.
That only confirms that the movie channels deliver less than 19.4 mbps to the cable company (in fact, something like 13 mbps). This may be true for other providers as well.

If stat muxing were used to combine three 13 mbps signals into one 38 mbps pipe, is that really a problem?

My point is just that what we want is the cleanest, crispest, most error- free HD pictures and sound that can be delivered with the current technology and with no visible/aural degradation introduced by the cable or satellite company. How they get there is less important.

In spite of the posting a few days about Rogers plans, have any channels actually been combined? That's an easy thing to check, isn't it?

Whereas we do know Bell ExpressVu most certainly has driven three channels into about the space of 30 mbps across much of its HD line-up since mid-Nov. In that case, the "no visible/aural degradation" test has not turned out well for viewers.
 
#23 ·
Disk space is not a good indication of program compression. Various PVR recording methods create different sized files. Also, space measurement does not truly reflect quality of the data in the files. For example, EV HD programming includes a fairly high "NULL" content in the output data stream. A couple of years ago it was about 33%. Now it is closer to 60% but the raw files are the same size. Stripping the nulls creates a smaller file that is still playable but may be of different quality. Further MPEG2 compression also produces smaller files but with different methods (and are sometimes unplayable.) If the PVR software does any processing of the raw data stream, results could be invalid. It also wouldn't indicate the quality of the compression being used. The raw file must be analyzed for quality but I don't know which software would do that. I do know that measuring disk space will likely not stand up as evidence.
 
#24 ·
Various PVR recording methods create different sized files.
On the PVRs like the SA8300HD, the BEV92XX and the *C units, there is no processing, it's a bit for bit recording. The disk space used is therefore a perfectly accurate (average) indication of the incoming digital channel stream. From the file size, you can accurately calculate the bitrate (on average)
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top