Guess Who Holds the Keys to 4K Television and Broadcast?

[Nick Morrison]

BTW, I think one of the first ?'s that needs to be answered for 4K broadcast is "What is 4K,?" in the context of broadcast. I'm assuming it means quad-HD. So looking at it from an Avid-centric perspective, I'm assuming that would mean having quad-HD versions of both project format and DNxHD.

I still see 4K as a finishing solution. And 1080/2k as a mastering solution. What I mean is you finish and grade in 4K. And then deliver 2k/1080 for broadcast. I can see that becoming more common place.

But for a network to have a whole channel broadcast 4k would require that they have hundreds of titles and episodes delivered in 4k. And I don't see that happening right away. For History to have a 4K channel, that means they'd need at least 15-20 shows with 10-20 episodes each at 4K. So that's upto 400 episodes. That's a lot. I just don't see it. Most production companies doing reality are using 1080p camcorders and other run and gun systems. I dont see Swamp People being shot on 8 epics. Or Ice Road Truckers being filmed on 50 epics. Believe me, I've been up there, IRT is shot by dozens of dudes, and they have cameras all over. They would all have to be 4K to have the show master in 4K.

Again, they still haven't moved to the next logical stage after 720p. Which is 1080i. And that's not even 1080p.

Now do I see online material like Youtube, Vimeo, Hulu, making 4K available? Sure. That's more alacarte. But a huge network like AETN? I think they'd rather convert their back library to 3D first. That's attainable. But an entire 4K network? I can see that happening in 10 years. All though I'd like to be surprised and have it come in 5. And that would be great if it did. But I delivered the first IRT trailer in HD in 2007. That was sent to theatres in 720p. All the promos went up SD. Four and a half years later, we are now sending up promos 720p full time. But that's as far as we've gotten. That seems like pretty slow appreciation.

Am I making any sense here? One thing is what is possible. Another thing is a gigantic corporation changing all their specs across dozens of platforms and departmens. There aren't enough REDS out there yet to make that possible. Yet.

Hate to be the bearer of bad tidings. I just don't see this.

 

[M Most]

Just to emphasize a point I made earlier about the real reasons for the digital television conversion (and thus the creation of HD broadcast), this appeared today:

http://gcn.com/articles/2012/02/17/fcc-spectrum-auction-public-safety-network.aspx

In case anyone here still thinks that HD was brought about because of some sort of consumer demand for better television pictures, or that higher resolutions in broadcast will happen anytime soon for that same rather suspect reason, I suggest you read the above article, which corroborates what I was saying about what happened then, and what's happening now and in the future.

 

[Peter Moretti]

Nick, you are making perfect sense!

But Avid acknowledged in July that 4K (all the way up to 28K) is in their plans. Here's a link to an article summarizing the Burbank "Committed to the Professional" event:

http://www.postmagazine.com/Post-Bl...o-Avid-Media-Composer-64-bit-4K-and-high.aspx

So the question becomes HOW should Avid do this in a manner that makes the most sense? Are we really going to have a new project format every time another frame size is released? Should they stop at quad-HD? Will there be a larger than HD DNxHD? How will effects and plug-ins work with larger than HD? Do shared storage and air servers work with these larger files? It gets very messy and complicated, which is probably why Avid is lagging the pack on 4K. Well they had to become 64-bit first :idea:.

So given that 4K+ IS in Avid's roadmap, how should they go about implementing it that makes the most sense?

 

[Peter Moretti]

Mike, when I read that in the news a few days ago, I thought of you!

 

[Marc Wielage]

But don't kid yourself in regards to HDCAM tapes. Sure you can toss them around and whatnot, but they're still just as vulnerable as anything when it comes to electromagnetic fields or heat. They still suffer the same form of magnetic dispersion that we see on HDD platters or any other digital tape format. Over time, they will degrade and eventually be unreadable. None of the HDCAM tapes out there are old enough to start testing theoretical age limits, which start to kick in at about the 25-35 year mark.

An even bigger problem are the machines. I know of quite a few tape formats that are increasingly difficult to play in 2012. Try to find a working Panasonic MII machine, let alone an IVC 9000. Both of those were very popular and being used in the early-to-mid 1980s. Heck, playing back Betacams is starting to get very iffy.

Once the heads and parts are gone, it's going to be very hard to play back archival masters in 25-30 years. I had a situation about ten years ago where we tried to play the 1/2" 48-track digital audio master for a major, Oscar-winning film from 1990. Couldn't be done. Luckily, the old analog Dolby SR mags were fine -- and that's what was used for the final Blu-ray, now in stores. This was a case where the analog backup to a digital mix saved the day. Sad.

 

[Dave Blackham]

Its interesting that several comments have been made about tapeless delivery. The point Mike makes about there being a financial reason to make change is very relevent here. The principle reason in my view is tapelsss delivery is slow on take up isnt that its MUCH less costly to operate for broadcasters. Its measured in 10's of percent and its a very big saving so its a obvious area to implement to save cost. The reason is that Post houses are resitant as they are able to make margin out of Video Tape based delivery and have existing costs to recover from capex int he past. Also the process isnt that well understood by the majority so is a real issue getting up to speed with these systems. So there is resistance to it, it will prevail eventually but with the resistance from old school its slower uptake than it should be.

And for my money, LTO has proven to be fine for the most part, but YMMV.

 

[Stuart English]

WRT to delivering 4K to broadcasters in the future, LTO certainly sounds like a good solution - its a file based system, its cheap and it can be delivered by sneaker net if you want to. But also hopefully by that time there could be a dedicated fibre optic line between Hollywood and NYC so the issue of who pays for the file transfer costs to a network becomes a moot point.

 

[Dave Blackham]

WRT to delivering 4K to broadcasters in the future, LTO certainly sounds like a good solution - its a file based system, its cheap and it can be delivered by sneaker net if you want to. But also hopefully by that time there could be a dedicated fibre optic line between Hollywood and NYC so the issue of who pays for the file transfer costs to a network becomes a moot point.

I think that sounds a good plan, Id keep it file based though. Any guesses to codec, wrapper and bit depth ?

 

[Stuart English]

I think that sounds a good plan, Id keep it file based though. Any guesses to codec, wrapper and bit depth ?

Depends on whether you want to edit it again, or just transmit it...

 

[M Most]

Its interesting that several comments have been made about tapeless delivery. The point Mike makes about there being a financial reason to make change is very relevent here. The principle reason in my view is tapelsss delivery is slow on take up isnt that its MUCH less costly to operate for broadcasters. Its measured in 10's of percent and its a very big saving so its a obvious area to implement to save cost. The reason is that Post houses are resitant as they are able to make margin out of Video Tape based delivery and have existing costs to recover from capex int he past. Also the process isnt that well understood by the majority so is a real issue getting up to speed with these systems. So there is resistance to it, it will prevail eventually but with the resistance from old school its slower uptake than it should be.

The practical needs of post houses have absolutely nothing to do with broadcast delivery format decisions. What post houses have or want was a non-issue when ABC and Fox decided to go 720p (at the time supported only by Panasonic's D5 format) and everyone else went 1080i (supported primarily by Sony). It has nothing to do with TNT, for instance, deciding to use ProRes (the worst possible choice, IMHO - although not nearly as bad as it once was, now that there's an open source version of it) as a delivery format, and it has nothing to do with at least one broadcast network deciding to offer suppliers a choice of delivery codecs. Post houses have no power or influence in this area, they are, for lack of a better description, struggling to keep up with whatever the broadcasters decide, and that changes weekly. The one constant, however, is SR tape, because that's a stable format that offers both broadcast formats, has a very good history in terms of reliability and robustness, and a proven codec. It's survived because it represents the physical asset that can be put on a shelf, used directly, or used for quick retrieval of entire programs or parts of programs. That isn't possible with LTO tape - even if LTFS is used - because a program is usually formatted as one file and thus only has random access once it's been restored, which takes time. Longer than real time in the case of a one hour television program. It is a good archival element, but not necessarily a sensible broadcast delivery element.

It's not easy to get a consensus on things like broadcast delivery formats. File based deliveries have been talked about for a number of years now and to this point there still isn't a consensus. As a side note, the HPA Tech Retreat was held this week. I wasn't there this year, but at least 3 years ago, I sat at a roundtable discussion about the then-proposed Digital Video Package, which was basically being modeled on the DCP. The table was moderated by Annie Chang from Disney, and the general feeling was that the format was in its final stages of planning and would be implemented within a year. I've heard nothing about it since. The lack of consensus, and the myriad of choices, is in part what has led to just about everyone staying with SR tape until someone - anyone - actually sorts out the issue and comes up with something everyone can agree on. And that's where an entity such as SMPTE comes into play. Sooner or later there will be such a format. But I can almost guarantee it won't be ProRes, or for that matter, any proprietary codec (and yes, that includes RedRay unless Red decides to open it up). It will be a codec that is open and thus not owned by any single entity. The industry was willing to put up with SR being solely a Sony product because it was the only thing around that really suited the need. In the world of files and codecs that's not the case, just as it wasn't the case when the DCP was developed. JPEG2000 wasn't used for the DCP because it was the best (although it largely was), it was used at least in part because it was open and flexible, which were the absolute requirements of the DCI. That's why they also used XYZ as the color space, MXF as the container (also open) and FIPS as the basis for the security (also open). The television industry has many more players than the motion picture industry, though. Ironically, many of those players are now the same since essentially every studio is now co-owned by a broadcast network....

 

[M Most]

WRT to delivering 4K to broadcasters in the future, LTO certainly sounds like a good solution - its a file based system, its cheap and it can be delivered by sneaker net if you want to. But also hopefully by that time there could be a dedicated fibre optic line between Hollywood and NYC so the issue of who pays for the file transfer costs to a network becomes a moot point.

Getting files to networks and broadcast entities isn't a problem and hasn't been a problem for a long, long time. There are many high speed lines already in place, both private and open, and essentially every player, from post companies to networks to broadcasters, have access to reliable high speed data deliveries. There are also satellite based private networks already in place for many years. And for those not directly connected, colocation facilities exist in every major and even many minor cities to provide extremely high bandwidth lines. This is, in fact, already being used for location based productions who need to upload their original camera files to Los Angeles for editorial and further post work. Uploading the files to "home base" has, in fact, replaced location based dailies systems on a number of productions because it has ultimately proven to be quicker and less costly, as well as minimized the need for location based personnel to do that kind of work.

And, just as an FYI, ABC is in the process of moving most of its network broadcast operations to Los Angeles. That means that for both Fox and Disney/ABC owned networks, the point of delivery will be in L.A. and not New York going forward (Fox has always had L.A. based deliveries as they have always been headquartered here).

 

[Brian F Kobylarz]

It's not Sony, Pansonic, Apple, Fox News, SMPTE or any of the players you'd normally think of.

It's a moderately sized company ($400 million market capitalization).

It's company who back in the day Red ignored, but that's ancient history.

It's a company that's been downsizing and struggling to turn a profit, but it just did announce a GAAP profit.

It's a company that is passionately loved but also reviled.

It's not in Asia or Silicon Valley.

It's Avid.

The fact of the matter is that until Media Composer and Symphony can work in full 4K resolution, there will be no 4K TV.

How did I miss this thread!
Thankfully, it has moved from the quote above onto more provocative and insightful thoughts about the future. Stuart, I believe you nailed it with LTO. Even when dedicated fibre is in place, the need for backup and alternate "delivery media" to and from producers is essential. LTO has another advantage - built in encryption that uses NIST/FIPS approved algorithms. Whether distribution of camera original, work in process or final delivery, protecting our assets is more important now than ever.

Need to disagree with the thought that started this thread though. Using the logic proposed, Kodak would have needed to lead the charge for movie theatres to switch to digital. Obviously, that is not the way the history books are being written. An industry leader today can become an afterthought blowing in the wind so quickly it's mind-boggling. Whenever AVID joins in the adoption of 4K, and I'm sure they will, they will be playing catch-up. They do have the talent but need to be more nimble in the marketplace to stay relative.

The industry has moved on. Guess who we get to blame for that?
The game has been changed forever due to this upstart company that wasn't afraid to go up against giants in the industry.
That was good for all of us.

 

[Stuart English]

Getting files to networks and broadcast entities isn't a problem and hasn't been a problem for a long, long time. There are many high speed lines already in place, both private and open, and essentially every player, from post companies to networks to broadcasters, have access to reliable high speed data deliveries. There are also satellite based private networks already in place for many years. And for those not directly connected, colocation facilities exist in every major and even many minor cities to provide extremely high bandwidth lines. This is, in fact, already being used for location based productions who need to upload their original camera files to Los Angeles for editorial and further post work. Uploading the files to "home base" has, in fact, replaced location based dailies systems on a number of productions because it has ultimately proven to be quicker and less costly, as well as minimized the need for location based personnel to do that kind of work.

And, just as an FYI, ABC is in the process of moving most of its network broadcast operations to Los Angeles. That means that for both Fox and Disney/ABC owned networks, the point of delivery will be in L.A. and not New York going forward (Fox has always had L.A. based deliveries as they have always been headquartered here).

All I can say about that is at HPA this week file based delivery of programs to the networks was still a point of contention. i.e. who would pay for the cost of the file transfer being and what the file format(s) required was ... so if the cost of using existing high speed lines is too high, it's essentially the same issue from a practical point of view... physical tape is cheaper. Which if actually true is daft.

 

[M Most]

All I can say about that is at HPA this week file based delivery of programs to the networks was still a point of contention. i.e. who would pay for the cost of the file transfer being and what the file format(s) required was ... so if the cost of using existing high speed lines is too high, it's essentially the same issue from a practical point of view... physical tape is cheaper. Which if actually true is daft.

Why do you think it's "daft?" What makes you assume that manufacture and distribution of a single cartridge of videotape, which has quite a bit of storage capacity, should be more expensive than construction and maintenance of high speed data networks, which require physical cable, rights of way purchase, high speed switching centers, vast amounts of physical storage, and enormous security infrastructure, not to mention constant upgrading and implementation of new technology and protocols? Especially when the purveyors of such networks also need to make a profit? Are you implying that high speed data transfer should be free?

It seems to me that whatever was being discussed at HPA is more about the fact that nobody wants to pay for just about anything related to computer data these days, because movement of a compressed video file representing a one hour HD television program is pretty minimal, since it's only around 150GB tops, and can be dropped into a secure FTP site, or sent securely by software such as Aspera. Of course, it has to be created (usually about running time to do that), copied to a local server, uploaded, and then copied again at the receiving end. That's time at the supplier, and time at the receiver. Time is money. Or at least it should be.

 

[Dave Blackham]

All I can say about that is at HPA this week file based delivery of programs to the networks was still a point of contention. i.e. who would pay for the cost of the file transfer being and what the file format(s) required was ... so if the cost of using existing high speed lines is too high, it's essentially the same issue from a practical point of view... physical tape is cheaper. Which if actually true is daft.

Well thats what Id expect but being a cynic, there are unrecovered capital costs in SR decks and so on, the argument surely is why do we have to spend money to send out same or smaller invoices. I completely agree this isnt free delivery and there is a realistic cost to bear. I also think the telco's don't understand the costs or needs of the required service well. One example of Telco's mis quoting is I once need a Gbit link for some hundreds of yards down the street, annual cost as quoted at 50k UKP rental PA. I contacted the cable company's external contractor had wayleaves issued, dug up the street laid our own multi Fibre for 4k UKP. We had more potential bandwidth than we could ever use. I realise this was a local issue and may not be doable for a big MAN. Mike is right, Tapeless deliver has been possible for some years, industry needs to get behind it.

 

[Nick Morrison]

Nick, you are making perfect sense!

But Avid acknowledged in July that 4K (all the way up to 28K) is in their plans. Here's a link to an article summarizing the Burbank "Committed to the Professional" event:

http://www.postmagazine.com/Post-Bl...o-Avid-Media-Composer-64-bit-4K-and-high.aspx

So the question becomes HOW should Avid do this in a manner that makes the most sense? Are we really going to have a new project format every time another frame size is released? Should they stop at quad-HD? Will there be a larger than HD DNxHD? How will effects and plug-ins work with larger than HD? Do shared storage and air servers work with these larger files? It gets very messy and complicated, which is probably why Avid is lagging the pack on 4K. Well they had to become 64-bit first :idea:.

So given that 4K+ IS in Avid's roadmap, how should they go about implementing it that makes the most sense?

Ok that's different, I agree. Avid needs to embrace 4K fast. We have an Avid Symphony, and I'd love to cut 4K native, especially when Jim's projector comes out.

I just don't see a 4K broadcast network coming out any time soon...

 

[Nick Morrison]

Its interesting that several comments have been made about tapeless delivery. The point Mike makes about there being a financial reason to make change is very relevent here. The principle reason in my view is tapelsss delivery is slow on take up isnt that its MUCH less costly to operate for broadcasters. Its measured in 10's of percent and its a very big saving so its a obvious area to implement to save cost. The reason is that Post houses are resitant as they are able to make margin out of Video Tape based delivery and have existing costs to recover from capex int he past. Also the process isnt that well understood by the majority so is a real issue getting up to speed with these systems. So there is resistance to it, it will prevail eventually but with the resistance from old school its slower uptake than it should be.

And for my money, LTO has proven to be fine for the most part, but YMMV.

I agree that tapeless delivery is on everyone's mind. Some networks in NY already send up their promos tapeless. I do see tapeless as the future. But until there is a cloud big enough to store everything, and I mean everything, networks will still want to store everything on tape. And LTO isn't good enough. It needs to be mastered to tapes that can be cloned and sent around to different vendors. As far as I understand, LTO is a mass storage device that's expensive to access. AETN needs HDCAM masters they can send to four vendors tomorrow.

 

[Brian F Kobylarz]

And LTO isn't good enough. It needs to be mastered to tapes that can be cloned and sent around to different vendors. As far as I understand, LTO is a mass storage device that's expensive to access. AETN needs HDCAM masters they can send to four vendors tomorrow.

HP LTO-5 with SAS connectivity can be had for less than $3000.
HDCAM deck - multiply that many times.

LTO-5 cartridge = around $65
Holds 1.5 or 3 TB

HDCAM tape costs far more for the content - a 94 minute cassette is around $80...

There is a good reason that financial institutions, insurance companies and any others that consider data archiving a priority have chosen LTO as their format.
Many others in the film / video / broadcast are moving their archives to LTO as well. The "Huge" expense is when it is a massive on-call library system - that runs over the six figure cost - but then again, that is a specialized system only needed by a few.

 

[Stuart English]

Why do you think it's "daft?" What makes you assume that manufacture and distribution of a single cartridge of videotape, which has quite a bit of storage capacity, should be more expensive than construction and maintenance of high speed data networks, which require physical cable, rights of way purchase, high speed switching centers, vast amounts of physical storage, and enormous security infrastructure, not to mention constant upgrading and implementation of new technology and protocols? Especially when the purveyors of such networks also need to make a profit? Are you implying that high speed data transfer should be free?

It seems to me that whatever was being discussed at HPA is more about the fact that nobody wants to pay for just about anything related to computer data these days, because movement of a compressed video file representing a one hour HD television program is pretty minimal, since it's only around 150GB tops, and can be dropped into a secure FTP site, or sent securely by software such as Aspera. Of course, it has to be created (usually about running time to do that), copied to a local server, uploaded, and then copied again at the receiving end. That's time at the supplier, and time at the receiver. Time is money. Or at least it should be.

I think I am agreeing with you that FTP transfer work fine, my point is if its costing more than $90 to transfer a 1 hour program ( a guesstimate for the average cost of a 64 minute HDCAM SR tape ) over that IP network then that's daft.

 

[M Most]

I think I am agreeing with you that FTP transfer work fine, my point is if its costing more than $90 to transfer a 1 hour program ( a guesstimate for the average cost of a 64 minute HDCAM SR tape ) over that IP network then that's daft.

Really? So creating the file, checking it for errors, housing it on a local server, using security software like Aspera over a very high speed connection - all of those things should be what - free? I think you and possibly others here need to stop thinking the professional world of network television consists of individuals who use nothing other than desktop software and start understanding that there is an industry out there that requires significant infrastructure to service a very demanding clientele (namely, networks and studios). And who need to be paid when creating products and providing services to that demanding clientele. We're not talking about your home Internet connection on a cable modem here.

And BTW, the cost of raw stock is not the only cost in creating an acceptable network delivery element. There's an SR tape machine involved, as well as infrastructure to get a signal to that machine, people to operate it, people to QC the result, 12 channels of sound to lay back - most of which also needs to be done even if there isn't a tape to deliver. Not to mention that network deliveries consist of more than just a single SR tape master. There is an entire package of deliverables that gets created and sent out on every episode of every show. They include physical as well as file deliverables, it's not just one or the other. Oversimplifying all of this down to "it shouldn't cost more than $90" is a clear indication that you're probably not very familiar with what network program deliveries actually consist of and require, and have for a long time.