Tandberg MPS vs MCG

[gbroo]

Has anyone seen/used a Tandberg MPS?...I was wondering how it stacks up to the MGC?....

[Entropy3XD]

I have just spent the past two weeks testing out an MPS with IP and serial connectivity. Unfortunately, I am not an expert on the MGC, so I cannot give you an accurate comparision. If you have any specific questions on the MPS I might be able to help you out.

[BPVTC]

I use the MGC product line daily. Yes, specific questions would be helpful. Both have the capabilities to do IP calls. Depending on the IP card in the MCU and what speed calls you do, we can help you find additional information.

[ChuckF]

Up until February of this year, the Tandberg MPS platform was really not on the same level as the Polycom MGC. In February of this year Tandberg released the support for ISDN and Serial connectivity within the MPS.

Definetly would like to hear your thoughts on the MPS platform - Entropy - if you could be unbias on your eval results in this forum. We perform similar work (secure), so I can relate to your interest in the MPS's ability to do serial. I had a couple of days with the MPS and have written my own eval (for our internal sales engineers). At the time, I didn't get to see the serial interface config. But here is my two cents on the IP and ISDN parts.....

The MPS is basically a chassis that supports the originally designed Tandberg 16x16 MCU. However, since it was designed with the newer chipset (same as in MXP's), it has more capability than their original MCU. Tandberg basically designed a 16x16 MXP MCU in a blade form, and built the chassis to support upto 8 of these blades. Therefor it has some of the same "limitations" as the 16x16 MCU. One being that each blade only supports 3 simultaneous conferences. That may or may not be a limitation, but to me it is. The MAX capabilities are such - 16 ports per blade, 8 blades per chassis, therefore 128 ports (connections) total, and 24 simultaneous conferences. All-in-all a pretty nice size and form factor. Bundle sizes are, of course, 32, 48, 64, etc upto 128.

Pro - :thumbup: extremely easy to setup and configure. This alone was nice enough for us installers (luv the small lcd display).
Pro - :thumbup: Easy to quote and configure as well.
Pro - Standards, standards, standards - full compiance across the board (as far as I can tell - H.264, H.239, H235, etc.)
Pro - :thumbup: I like the fact that you can "re-use" the same Serial cables as you would use on a Tandberg endpoint. That cuts down on some engineering time needed to figure out what interfaces to use. The MPI panel (not sure what Tandberg calls it) is somewhat smaller than the MGC version and the cable reuse allows for a much "cleaner" solution.

CON - :thumbdown and this is a major one for me - Each chassis only supports ONE ISDN switch type. It supports many switch types, but you can only select one for the chassis. In the Federal space, you sometimes are in need of connecting to multiple carriers - Commercial, FTS, DVSG, etc and most of the time, they come in from different carriers. This could be an issue for some customers - but maybe not a show stopper. It also doesn't support NFAS.

CON #2 - :thumbdown and this is just my opinion, I don't like the fact that you have to have an ethernet connection for each blade. That means for every install, I'm going to have to request not only two ports off of the customers switch (for management and first blade), but 1x times the number of blades I have. Tandberg should almost throw in a Ethernet switch with every MPS, especially a fully loaded one. Granted, these ports can be DHCP, but it still looks a little awkward.

One thing that the MGC has over MPS is all of the IVR/DTMF Audio type conferencing features. I have many customers that use the MGC strictly for that alone, and some that now use the Audio only conferencing features more than video. Its a nice dual role switch.

I have more in my eval write-up, plus a side by side comparison with the MGC (and RAD), but this is some of the more important issues. Both sides will argue major TX and CP differences, but I'll let them both throw out that FUD.

[Entropy3XD]

Well it looks like Chuck covered most of it. I agree with all of his pros to a tee and I found it to be one of the easiest MCUs to setup. Con #1 was a big WOW. I haven’t worked much with the PRI blade, so I didn’t know you could have only one switch type per chassis. This could definitely be an issue for many. CON #2 is not a big deal to me personally, as I am used to the Radvision and Tandberg lines where all of the equipment uses Ethernet to interconnect.

On the serial side I don’t see how it can be any easier. You simply go to the Serial Board Configuration page and set the port to “dual� for 530, 449, V.35 and “single� for X.21 (page 49 of MPS manual). That’s it as far as the web interface goes.

I used Adtran 512s, Criticom Dial Isolators, and Kiv-7 HSBs for testing. I tested 366 dialing, dialing from the imux, and inbound calls. The Adtran configuration I used is basically the same one Marty Marlow posted here:

http://www.vtctalk.com/showthread.php?t=14159 (http://showthread.php?t=14159)

DTR works as follows……DTR is low when the port is not in use. For 366 dialing simply add the phone number you wish to call to the phone book entry or manual dial, select H.320 for the network, then add the participant to the conference. If the endpoint disconnects from the far end, the port stays in the conference and DTR remains high. DTR will only go low if you disconnect the port from the conference via the MPS. For those who dial from the imux or have all dial-in participants, it is basically the same. The only difference is when you add the port to the conference you do not use a phone number. This will add the port to the conference and DTR will go high. In talking with some MGC gurus I am told the MGC uses DTR the same way as the MPS. Please correct me if I am wrong as I don’t have an MGC at my disposal.

Clocking………If you have an MPS with a Serial and PRI blade, it is best to use the PRI for clocking. For those who only have the Serial blade, you must use one port for clocking and this port must always be connected. There is an Xcommand that can be used via telnet to set a specific port as the clock.

The high and the low……….This one was a bummer at first, but there is a work around. The MPS only supports two rates per conference, a high and a low. This means that if you have a conference with 384k and 256k callers and a 128k site joins, the 256k users down speed to 128k. For many this may not be an issue, but there are some out there who communicate with VTC via a 64k satellite links using H.264 (which is surprisingly good). If I have users in my conference that get down-speeded to 64k and they are not using H.264, chances are their video is gone. There is a workaround, but at a cost. What you have to do is cascade two conferences. Start you first conference and add your two highest bandwidth sites, say 384k and 256k. These are the high and low for the first conference. Start a second conference and add the low bandwidth sites, but it can only be one bandwidth, say 128k only. Now add the second conference to the first conference at 384k. This makes your high/low for the second conference to be 384/128. You now have a cascaded conference with three bandwidths, with the first conference as master and the second as slave. If you are using continuous presence you will still have it on the master conference. The slave conference however will be voice switched, which I personally don’t have an issue with for the low bandwidth sites. This work around does waste a conference, but if you need more than two rates in your call it does work.

There are a few bugs here and there, but I have already seen beta code which addresses most of the things I came across. Overall I found the MPS to be extremely easy to setup, and the quality was excellent. It may lack some of the bells, lights, and whistles that the MGC has, but I am interested to see what Tandberg has in mind for the future. One thing I do like about the MPS is that a port is a port, meaning that you do not have to break out the calculator to figure out the capacity of the MCU at certain bandwidths. This means if you buy a 32 port MPS, you get 32 ports all the time, no matter what the bandwidth or if encryption (IP) is used. The unit I tested was put in a deployable Hardig case with a Tandberg gatekeeper, Cisco switch, and server running TMS. The cables Tandberg uses to connect the blade to the connection plate are slim, which made for a very clean install. The MPS also shipped with all of the tools needed to connect the cables to the connection plate. They included a nice hex driver for the barrel connectors used on the serial connections and I was sad to have to give up.

Between you and me Chuck it looks like we have a good review of the MPS. How about you MGC folks out there? Gbroo needs to hear things from the other side of the fence. Gbroo……we still need some details as to what you are looking to do.

[gbroo]

Thanks for the scoop, great info, by virtue that the 1 ISDN switch per chassis limitation exists, that is a show stopper for my requirements. We (government) need the ability to mix and match our switch types, DSN(national) and commercial(euro), so the idea that I would need a chassis per switch type is a no go.

[Sean Lessman]

Thanks for the scoop, great info, by virtue that the 1 ISDN switch per chassis limitation exists, that is a show stopper for my requirements. We (government) need the ability to mix and match our switch types, DSN(national) and commercial(euro), so the idea that I would need a chassis per switch type is a no go.Today we support a switch type PER PRI card, not per chassis. Improvements through software are always happening, so don't let that give you heartache. Make sure your TANDBERG rep knows what you need and we will make sure those features make it into the box. Its only version 2 :-) .

I am sure your rep would give you a good price on a second PRI card if that was your only showstopper today.

Sean

[Sean Lessman]

The MPS is basically a chassis that supports the originally designed Tandberg 16x16 MCUNot true. Our products share a common code base. While the software is similar today, the hardware is completely different, the MPS is capable of much more and you will see this as subsequent software is developed and released. Our competitors love to use this FUD, so I will go on record as saying this is completely false.

Tandberg basically designed a 16x16 MXP MCU in a blade form, and built the chassis to support upto 8 of these blades. Therefor it has some of the same "limitations" as the 16x16 MCU.Again, not a true statement. See above.

One being that each blade only supports 3 simultaneous conferences. That may or may not be a limitation[3 simultaneous fully featured conferences. Start loading up the MGC with full features and the number of conferences will be comparable. TANDBERG believes in a port is a port and all features should be available. The Accord is designed with a strategy that the conferences are basic and you add features and its conference and port capacity are rated at 'basic' ports. Its realy apples and oranges.

Should we remove features so the port count/conference count goes up? There will continue to be improvements in these areas.

Bundle sizes are, of course, 32, 48, 64, etc upto 128.You forgot 16 :)

CON - :thumbdown and this is a major one for me - Each chassis only supports ONE ISDN switch type.Today we support one ISDN type per ISDN blade, not per chassis. This can certainly be improved through software.

It also doesn't support NFAS.We have recieved this feedback and its on the wishlist.

One thing that the MGC has over MPS is all of the IVR/DTMF Audio type conferencing features. I have many customers that use the MGC strictly for that alone, and some that now use the Audio only conferencing features more than video. Its a nice dual role switch.We are by design not an audio bridge today. Anything is possible through software in the future :happy:

Sean

[Sean Lessman]

The high and the low……….This one was a bummer at first, but there is a work around. The MPS only supports two rates per conference, a high and a low. This means that if you have a conference with 384k and 256k callers and a 128k site joins, the 256k users down speed to 128k. Its important to remember that this functionality is FREE. You have to pay $4500 per site on the Accord to get any transcoding (video+) cards. Two streams is good enough for most (not all), but of course there are possibilities in the future for improvements in software.

If and when we improve this to more streams, should we also charge $4500 per site on top of the Free feature to get more streams, would the customer be willing to pay? What are your thoughts?

Sean

[gbroo]

Sean,
Interesting nuggets of knowledge, we are trying to convince folks to look at the MPS vice assuming the MCG is the only game in town.

Greg

[ChuckF]

Sean - I understand the high/low MPS "rate or speed matching" TX method Entropy described, but how does the MPS handle Video and Audio Algorithm TX? Is it the same? A high and a low per conference?

Say you have a conference with a H.264, H.263, and H.261 callers. What happens in this case? Not sure if this would ever happen, but there are certain legacy systems still out there that only support H.261.

And when you setup the conference do you have to specify an audio or video algorithm or is there an "Auto negotiate" type feature?

[Sean Lessman]

Sean - I understand the high/low MPS "rate or speed matching" TX method Entropy described, but how does the MPS handle Video and Audio Algorithm TX? Is it the same? A high and a low per conference?

Say you have a conference with a H.264, H.263, and H.261 callers. What happens in this case? Not sure if this would ever happen, but there are certain legacy systems still out there that only support H.261.

And when you setup the conference do you have to specify an audio or video algorithm or is there an "Auto negotiate" type feature?
The MPS and MCU (and MXP endpoints for that matter) will all take in any number of video and audio algorithms. It will encode (today) two outgoing streams. The properties of those streams are determined by the combination of endpoints that are connected today.

Sean

[Entropy3XD]

Sean, thanks for stepping in and making some clarifications, especially on the switch types. It is only version 2 and I know from dealing with Tandberg on many occasions that the wish list never ends and development is constant, so I look forward to the evolution of the MPS over time.


I will admit that only two streams is fine for most people, however, for some this is an issue and that is simply the way it is. For now we have been cascading the conferences in order to support those who need more than two rates and at the moment this is working out fine. Would a customer be willing to pay extra for this feature? In some instances I would say yes, as long as the price is still competitive with the competition. Are there enough people who require this function to justify major changes to the MPS? I doubt it. However, I cannot stress how important this feature is for those who really need it and at least having the choice for this option, even at a cost, would be enough to please most people.

The multiple rates is really the only thing I am not crazy about, but I do not consider it to be a no-go for the MPS. Everything else about the MPS simply rocks. I find it to be an outstanding unit for "mission critical" environments where stability and zero downtime is a must. The configuration of the MPS also comes closer to following the KISS theory than any MCU I have ever touched, without sacrificing most of the features I expect in a bridge. All of this is of course my own personal opinion, but anyone who has ever had to tack up an unscheduled conference for a high ranking military or government official in an emergency situation can appreciate the brilliance of an MCU that is simple to use and can take a heavy load.

On a purely cosmetic note, I have to say the black and silver chassis with the LCD is....well......sexy. Call me a techno perv but I know I am not the only person in here who appreciates the beauty of a well wired rack of VTC gear.



***For those unfamiliar with the KISS theory

Keep It Simple Stupid

[trapehzoid]

hey the new MGC is silver too!

:)

[Sean Lessman]

...However, I cannot stress how important this feature is for those who really need it and at least having the choice for this option, even at a cost, would be enough to please most people.
Noted! Thanks for the feedback.

Sean

[ChuckF]

One last question. On the MPS, what happens in a H.264 and H.239 conference as far as port count and simultaneous conferences? I heard 'rumor' that the MPS only supports one simultaneous H.239/H.264 conference per blade. I believe that there isn't a port drop when invoking both, but only changes to simultaneous conferences per blade. Any information would be appreciated.

[trapehzoid]

no changes due to h239 or h264

[trapehzoid]

w00 h000 I can post again :) board was broke for a bit

[Sean Lessman]

One last question. On the MPS, what happens in a H.264 and H.239 conference as far as port count and simultaneous conferences? I heard 'rumor' that the MPS only supports one simultaneous H.239/H.264 conference per blade. I believe that there isn't a port drop when invoking both, but only changes to simultaneous conferences per blade. Any information would be appreciated.I have also seen the docs from our friends. Its completely false info. None of the features of the MPS reduce the port count.

Sean

[djackson]

I will admit that only two streams is fine for most people, however, for some this is an issue and that is simply the way it is. For now we have been cascading the conferences in order to support those who need more than two rates and at the moment this is working out fine. Would a customer be willing to pay extra for this feature? In some instances I would say yes, as long as the price is still competitive with the competition. Are there enough people who require this function to justify major changes to the MPS? I doubt it. However, I cannot stress how important this feature is for those who really need it and at least having the choice for this option, even at a cost, would be enough to please most people.


Have you looked at the cost of the box already? Either way, customers getting there own delayed video back because of the limited streams reduces the experience dramatically. Tandberg is hurting themselves without knowing it. I make all my bridge operators sit in the rooms where videoconferences occur as part of their training so they understand what is going on in a conference. They have all noted how customers say "there is a five second delay" because they see there own video back. I know the delay is closer to 300 ms, but they shouldn't see their own video back ever in a delayed state.

I have supported videoserver/ezenias, accord/polycom, tandberg and codian. Tandberg is the easiest and one of the most stable MCU products, but this lack of power should come at about half the price based on today's market.

[trapehzoid]

the 'I hate seeing yourself' arguement is stupid in my opinion. I like it because you know exactly what everyone is seeing.

Fewer and fewer systems are dual monitor anymore.. people don't like the PIP and I'm tired of hearing people say 'do you see it now? can you see me?' and that stuff. Having the video in the quad is like built in troubleshooting.

The people who get hung up on it need a slap IMO.

[Yukikaze]

Tandberg is the easiest and one of the most stable MCU products, but this lack of power should come at about half the price based on today's market.

A rather pointless and confused statement considering that Tandberg's MCU business is growing, don't you think? For whatever reason, their growing customer base think it is acceptably priced given the features.

They are even selecting the product over others that don't loop video, which suggests that this feature is pretty low on many buyers priority list and that these customers feel they are in fact paying for other aspects of the MCU.

[Entropy3XD]

I can't really argue with djackson's statement. I personally don't have any issues with getting my own video back and would not get hung up on it, however, I did have a client mention it to me this week when they were testing the Tandberg MCU. I don't think it will be a reason for them to not buy a Tandberg bridge, but it did come up as a small concern.

[djackson]

Those who don't think seeing your own video back delayed are people who use videoconferencing on a daily/weekly basis - it something you just expect. It doesn't bother me at all during a conference. My customers that have used unstable boxes in the past are just happy that it stays connected.

However, a complete newbie user will notice their own delayed video, point-it out and be distracted - and for some reason they always say "5 second delay". Then I have to explain the delay is closer to 300ms or less. The newbie user is the majority of users - but not the person who buys the equipment. If you disagree with me it is only because you do not sit in on enough real conferences with complete newbie users. Chances are you are a dealer or manufaturer that deals with technical staff, or your customers have done so much videoconferencing in the past they are just happy that it works. Newbie users expectations are much higher and these are the people that cause the industry to grow.

"The people who get hung up on it need a slap IMO." - opinion is the reason people are scared to use videoconferencing after their first experience.

Tandberg makes great products, but as for "their customer base is growing and people most love to see their own delayed video" arguement - recently Codian's MCU customer base is growing faster and for an unkown company - and Codian does not send your own video back. It seems everytime I talk to someone setup a conference, they told me they just bought a Codian. Heck, Codian doesn't even support H.239 yet, DuoVideo or People+Content. Tandberg has had an MCU on the market for 3 years. Codian has had one on the market for about a year. I think if Tandberg was a bit cheaper not as cheap as Codian and as powerful as the Codian MCU counterpart, then there would be no Codian company.

This is discussion forum topic is who is better - Tandberg or Polycom for an MCU. The Polycom is more powerful. Is this in the hardware and Tandberg will catch up with the next software release or with an additional board, I don't know. However, I would go with Tandberg because of stability and the easy user interface, I wish they had a more powerful box - please don't slap me. :banana: :banana: :banana: :banana: :banana: :banana: :banana: :banana: :banana: :banana: :banana: :banana: :banana:

[jland]

At least in my setting, our customers don't want to see themselves "on TV" at all - it's distracting and points out to the end user that there's "technology" in the way of their message. They just want to have their meeting, give their lecture, etc., without feeling that they're "under the microscope" of a camera.

My organization spent many many months researching the various MCUs available (Polycom, Tandberg, Codian, etc.) to replace our legacy Ezenia/VideoServer MCU. We were quite surprised to discover that any MCU would send the speaker's video back to themselves. Soon, we found that one of our top three questions when evaluating any MCU device was "Does your MCU send video of the speaker back to themselves?" In the end, we chose the Polycom solution because the speaker didn't see themselves and, in my opinion, the MGC Manager interface (for actively "directing" each videoconference since we don't typically use voice activated switching, nor do we expect our customers to select screen layouts using their endpoint remote control.) was the easiest to use.

Although the majority of the endpoints deployed on our network are Polycom, we would have chosen another MCU if the Polycom solution sent the speaker's video back to themselves - delayed or otherwise.

[Sean Lessman]

Have you looked at the cost of the box already?
I happen to think the price of the MPS is very competitive for what it offers. Pricing discrepancies are usually due to differences in feature sets and abilities of the two boxes.

Remember, the MPS looks like a scaled up TMCU today because it is running essentially the same software block as the MCU. TANDBERG has always been very good at scaling software and hardware throughout the portfolio. Others have followed suit and you will find the only manufacturers that are successful today are doing the same thing. However, the hardware architecture is completely different. As software is written to take advantage of the new hardware, you will see improvements over the features offered today. Stay tuned, some nice stuff coming.

Sean

[Kevin]

the 'I hate seeing yourself' arguement is stupid in my opinion. I like it because you know exactly what everyone is seeing.


As discussed elsewhere on this thread this really does depend on the previous experience of the users.

When meeting users I have come across a definite trend:

1. Those who have done multipoint before and are used to seeing themselves delayed in one of the small panes may see little benefit in not seeing themselves. In fact many like yourself prefer to see themselves as they are used to it.

2. Those who are new to video conferencing find it incredibly off-putting. They instantly see the delay in the conference (which is obviously unavoidable) which they are totally unaware of if they do not see themselves. They complain about this to the guys that have sold them the equipment.

Hopefully with the help of us on this board there will be lots of new users to video conferencing so lets try and keep these new users happy!

For this reason by default users do not see themselves on the Codian bridges (though of course you have the option for people to see themselves).

There are other important technical reasons why it is best to send seperate streams to each user rather than just one or two stream to everyone (even if people are seeing themselves). Here are a few in no particular order:

a) With just the one or two streams a large percentage of participants in the conference may have to view video that is not the best video they can see (either lower bandwith, lower resolution or lower complexity codec that they can cope with).
b) Everyone has to view the same layout so they can't choose what or who they see.
c) If one participant is having packet loss and requesting key frames from the MCU everyone gets sent the key frame and the corresponding loss of video quality. If the MCU has to respond to flow control messages to send lower bitrate to this participant because of packet loss everyone has to view this lower bitrate.

All IMHO of course :)

Codian doesn't even support H.239
Codian will support H.239 in the very near future. If anyone would like to see a demo of this please let me know and I can arrange it now.

Kevin

[trapehzoid]

I can understand 'newbie' users not being used to seeing themselves.. but I don't see them getting hung up on it.

The industry has gone for years with this type of 'penalty' and yet only recently is it even mentioned.. that says to me its because people are 'selling it'. Users are being conditioned to 'not accept it' rather then actually having a meaningful position on it.

Yes, I've heard my users mention it.. but I've never had one not go 'oh, ok' and be fine with it after being explained to. Honestly the only people I've ever encountered in my travels that don't like it are big ego-types.. I honestly think its because they don't want to see themselves 'in a bad light'.. in that they think they look lesser in that view.

I do know (as someone mentioned) some people don't like being they are reminded they are on camera at all (shy) but unless you using operated driven conferences I don't see how you can use a system and not have any confidence-monitor at all. How many times have you watched people ramble on when they aren't even on camera?? Or people ramble on while they are muted? You gotta have feedback.

To the people that are camera shy.. I say get over it. Obviously you don't exaggerate it for them (tight zooms on big screens) but being comfortable with the stuff is a requirement if you are going to increase usage. If you are going to walk away and let your user just be shy and not try to engage with them to help them get comfortable.. you are loosing a user and eventually have your investment sitting in the corner and you looking for a new job.

Back to the continuous presence conversation.. if people want to give users that choice.. that's fine. But I'm certainly not letting it steer my next MCU refresh.

I find my users want to see everyone more then they want to exclude seeing someone.

[jland]

At the risk of hijacking this thread...


The industry has gone for years with this type of 'penalty' and yet only recently is it even mentioned.. .

That's not entirely true. Until our organization began considering a new MCU purchase, to replace our legacy MCU from the early '90s, we had never heard of - nor imaged - why anyone would want to see themselves, let alone as a delayed image, while using valuable real estate in a CP conference where they could be viewing their colleagues instead.


I do know (as someone mentioned) some people don't like being they are reminded they are on camera at all (shy) but unless you using operated driven conferences I don't see how you can use a system and not have any confidence-monitor at all. How many times have you watched people ramble on when they aren't even on camera?? Or people ramble on while they are muted? You gotta have feedback..

I don't think that there's an endpoint on the market today that doesn't already have a confidence monitor... it's PIP. Usually, it can be enabled/disabled by choice. It's not using valuable screen real estate in a CP conference.

I think one of the primary issues we are seeing with the adoption of the VC technology is that all too often we (VTC professionals) throw new users into the environment without the hand-holding/pre-planning that needs to occur. Speaking for my organization, if it's a large conference (greater than 4 individuals at 2 or more sites) then we are absolutely there to provide the "operated driven" environment that you speak of. When it's just talking head, 1-1, pt-to-pt, again, we're absolutely there to assist and remind our customers that camera composition, mic placement and lighting is crucial. That's the feedback our customers need. If an individual has no idea what they're transmitting, if they are framed properly, or whether their mic is muted, then that's the sole fault of the VTC professional at their location for not being there to plan and guide that event. After all, if anyone can do it, then the VTC professionals aren't really needed.


To the people that are camera shy.. I say get over it.

I don't think "camera shy"ness has anything to do with it, and I think your comment is both knee jerk and closed-minded. I spent 7 years working in broadcast news prior to joining to VTC field, even then, the "professionals" didn't want to see themselves on a monitor during production. It isn't natural to look at yourself while addressing a group of people. In the broadcast field you relied on the technical director and director to ensure you looked the way you're supposed to look. You relied on the audio engineer to ensure your microphones were on at the appropriate time. The speaker's job is to communicate an idea or a series of ideas and not be encumbered by the technology. If they didn't need to see themselves in a monitor in a traditional meeting room, then why should they be told that they do now? Our job is to make sure the technology is there, that it works properly, and that it's being used effectively - in the most natural way possible. We're already fighting an uphill battle with the acceptance of the technology... Why make our jobs even more difficult by making the experience anything but "traditional" for our customers and telling them that they have to adjust to "our way" or, as you so eloquently put it - "get over it"?

[trapehzoid]

I don't even need to go line by line in your quote.. you summed it up with replacing the feedback with operators. If every conference needs an operator to go smoothly then your technology isn't doing your job.

Do you have someone sitting at your desk helping you type every email? What about dialing your phone for you? VTC will never be mass deployable if I need a tech sitting over my shoulder to make it go smoothly.

Your comment about never having to deal with this before.. and the 'self' quad is taking up valuable space. What do you think the PIP does when its there? Its worse.. not only does it take up the space.. its indiscrimate and can overlap into several squares..

So you say 'well the PIP only comes up when its needed'.. which means its not there unless you a) ask it to be or b) move the camera. Neither of which give you unsolicited feedback that something is amiss.

What happens when someone moves in their chair and get moved off camera.. unless the PIP is up (taking up space and potentially overlapping several squares).. you have no idea. So, sorry, I don't follow your logic how the PIP is a suitable alternative to UNSOLICITED feedback that seeing yourself can give.

PIP is good for moving/changing your sources.. the PIP isn't good for feedback, and is in fact in this case of talking about a CP call worse then a 'self-view' image on the CP image IMO.

And as to your broadcast comparison.. again.. the people had a whole team of people taking care of the details for them. Remove that team and then see how well they do on their own and how their needs change.

How do you plan on scaling VTC deployment in your organization if you need one person per meeting to sit and hold their hands.. do you have unlimited headcount?

And my 'get over it' comment was that you either need to adapt or get left behind. If your user said 'oh I will not use that computer...' what are you going to do? Sit there and do all their work for them?

No, you work to getting either the technology more accessible and training the user to make them comfortable with it.

If your solution is just do it for them.. good luck to you

[Entropy3XD]

I have always been a proponent of using a confidence monitor, but that's only because I like to remind myself how good I look :silly:. Seriously though, jland has made some excellent points on why others would want to do without it. It can be useful for some, but just as easily a distraction for others. I never really gave it much thought until this thread, so I appreciate everyone opening my eyes to this subject.

[jland]

I think we shall agree to disagree. Perhaps there's a time and place for having the speaker see delayed foldback of themselves.

However, if it's camera composition and source selection that the end user is concerned with, is relatively easy to engineer a simple solution to show the user what they're transmitting via a standard monitor that's positioned next to their VTC monitor. It doesn't, necessarily, need to be a function of the MCU.

Back to the original poster's question...

During our comparison of both the Tandberg and Polycom MCU solution, the Polycom was the device that didn't require foldback of the speaker's video to the speaker. I have no idea if this has anything to do with the horsepower of the Tandberg solution, or if this is simply by design.

Anyhow, that worked best with the way we, and our customers, had been using the technology for years and that's one of the primary reasons we chose the Polycom MGC.

[trapehzoid]

However, if it's camera composition and source selection that the end user is concerned with, is relatively easy to engineer a simple solution to show the user what they're transmitting via a standard monitor that's positioned next to their VTC monitor. It doesn't, necessarily, need to be a function of the MCU.

I know.. and if you read up higher in the thread.. I said you are seeing less and less dual monitor systems.. especially as the price points continue to come down.

In a Point to Point call its not as big an issue simply because you typically are both engaged and someone says something without being too much interuption.

Put that in a MCU call.. and the problems simply multiply by the # of people in the call. Just think of how many times you've seen people do stupid things because they think no one sees them.. or starts sending some source they aren't supposed to.. or basically just sits there and blatently shows they aren't paying attention.

The less re-enforcement they get reminding them they are being stupid.. the better chance they'll do it and continue to do it and not know it :) Until the presenter or the people talking interupt to remind 'Hey Joe, you might want to turn off the TV behind you and watch the president while he's talking' kind of stuff.

Dual monitor (or now emmulation) are the best methods.. but are becoming less common.

Another good example is when sending content (like powerpoint) when you get the image back from the MCU, you know exactly what everyone else is seeing.. and how readable it is.

If people don't want it.. their perrogative.. I find it very useful for my calls.

[djackson]

I happen to think the price of the MPS is very competitive for what it offers.

.... Stay tuned, some nice stuff coming.

Sean

Of course you think it is fairly priced, people are buying the box, and you are making money off of it. :) I buy the boxes, so I always think they should be cheaper. If the rest of the markert is sustaining your product, then that is justification to keep selling it.

I'm looking forward to the "nice stuff coming"

[ChuckF]

Well that was a long dissertation on some market opinions about "delayed video", PIP, and DualMonitor. I could barely get through reading it all. However, it brings up another good question - what are the CP comparisons of the MGC and MPS? I don't know all of the MPS selections, but can recite the MGC offerings in regards to Continuous Presence.

Besides the Audioconferencing "bells & whistles" of the MGC line, I believe that the next biggest PRO for the MGC is that it has the most extensive "bells & whistles" when it comes to Continuous Presence. Someone correct me if I'm wrong or explain to me what else is out there in the market.

The MGC has the most extensive CP layout selection on the market today. But hey, whats the practicallity or who wants to see 10+ squares on thier 32" monitor anyway.

It provides the conference creator the ability to select any site to be in any square (or auto populate).

To combat Videoservers original ability to scroll through sites, Polycom added the "Lecture" mode selection which allows the Lecturing site to see all the sites via a timed interval. Now they added Lecture Show and Presentation mode.

As mentioned previously, by default, Polycom does not allow the users to see themselves (in a delayed broadcast stream). But there are some users that hate seeing a grey screen in one of the squares, so Polycom added the "Same Layout" radial button. This allows users that want to see themselves (delayed), just like all of the other mcu's on the market. (someone mentioned earlier in this thread that Polycom does not do this, but the MGC does support this if that button is checked).

The MGC also allows the creator the ability to change the background colors, the layout border colors for the squares, or just the speaker colors. I admit that some of this stuff is crazy, but there are some customers out there that want and use this stuff.

Just an example of the bells & whistles. Not sure what the ROI is for the video+ cards that are needed to support this, but its worth something to someone or else these options wouldn't be available.

I believe that the MPS when purchased with the AVO - advanced video option, increases the CP layouts to include a 5+1 and 7+1 option. These are probably the most widely used CP layouts chosen within the market. But I'm not sure if the MPS allows the user to select which sites go in what squares.

Any comments on CP comparisons?

[Entropy3XD]

I believe that the MPS when purchased with the AVO - advanced video option, increases the CP layouts to include a 5+1 and 7+1 option. These are probably the most widely used CP layouts chosen within the market. But I'm not sure if the MPS allows the user to select which sites go in what squares.

Any comments on CP comparisons?[/QUOTE]
The MPS does not allow the user to select which sites go into what squares. I know this has been brought up with Tandberg, so I guess we will see if it is included with a future release.

[trapehzoid]

Do people have nothing better to do?

Honestly.. I don't understand how people expect to be able to do more video if everyone expects to micro-manage every aspect of the meeting..

[Entropy3XD]

In some ways I would agree with you Trap, but there are many situations where micro management of VTC is required. Take a military environment for instance. A military General typically does not and will not have any interest in how a VTC call is managed. They simply want to come in and discuss more important matters without the distractions of having to learn about the equipment or the technology. This is why many locations have dedicated VTC personnel. Talk to anyone here who supports those rooms and they will tell you what happens when the General does not have a smooth conference.

This is also why having the ability to choose who goes in what square when using CP is an important feature. If I am using a 5+1 or 7+1 scenario, I would like to have the ability to put the General on the large block, as they are typically the most important person in the meeting.

[djackson]

Do people have nothing better to do?

Honestly.

Honestly, Umm... no.



I don't understand how people expect to be able to do more video if everyone expects to micro-manage every aspect of the meeting..

If you need to setup a view a certain way, then it is very handy. One of the drawbacks of the Tandberg 7+1 and 5+1 view is that it puts the first small image in the upper right hand corner. This is the same exact place that the very large microphone muted icon goes on the Tandberg MXP product line. Not very smart. I noticed in a conference the other day Codian leaves the upper right image blank. Of course, if you can move around the picture like on the Polycom then it wouldn't be an issue.

[trapehzoid]

If you need to setup a view a certain way, then it is very handy. One of the drawbacks of the Tandberg 7+1 and 5+1 view is that it puts the first small image in the upper right hand corner. This is the same exact place that the very large microphone muted icon goes on the Tandberg MXP product line. Not very smart. I noticed in a conference the other day Codian leaves the upper right image blank. Of course, if you can move around the picture like on the Polycom then it wouldn't be an issue.

So why not complain to the vendor that the allocation is less then superior? Why think that specifying quadrants is the solution? Why not just ask them that the quadrants be allocated better based on assumptions on what is displayed on the screen.

I mean.. if my car gets a flat tire.. you fix the tire.. not buy a new car.

I think your point is valid.. I don't think it justifies the overhead of specifying quadrants tho.

I don't actively monitor any of our conferences. We rely on our management suite to tell us of failures or user calls to the 'help' number.

[djackson]

So why not complain to the vendor that the allocation is less then superior? Why think that specifying quadrants is the solution? Why not just ask them that the quadrants be allocated better based on assumptions on what is displayed on the screen.

I have mentioned this to the vendor. I think specifying quadrants could solve the problem, but that involves more administration, which is great if you have the time. I think having a smarter MCU is also more useful as you pointed out.

My question to you, is why is more powerful less useful?
More power would only be less useful if the box is (1) overly complicated and/or (2) unreliable. If you read my earlier statements I suggest the Tandberg box over Polycom for this reason. This should not be an excuse for Tandberg to become complacent.


I mean.. if my car gets a flat tire.. you fix the tire.. not buy a new car.


Flat tire implies not working. This is more a v4 (tandberg) versus a v6 (polycom) discussion. If your car has a v4 and you are happy, then that is great. If you need a box with more power go for a v6. I prefer a reliable v4 over an unreliable v6. What's more important reliability or experience? Experience of course, but experience is dependent on reliability. Technicians think reliability equals experience, which is not true.


I don't actively monitor any of our conferences. We rely on our management suite to tell us of failures or user calls to the 'help' number.

My point exactly! Basically you only know if the sites are connected or disconnected, but have no idea of the quality of the conferences until someone gets so annoyed they call a help number - which is typically when the system is completely unusable.

I suggest you not make statements on the quality of features until you monitor some of the conferences. We monitor our conferences at off peak times when man power is available (this is man power intensive for all conferences), and have surveys for users to fill out after a conference. Until you do that you are only speaking from ignorance on user’s experience.

[trapehzoid]

My question to you, is why is more powerful less useful?
More power would only be less useful if the box is (1) overly complicated and/or (2) unreliable.

You are confusing 'power' with 'complexity'. Not everything 'simple' is less powerful.. in you goal should be the simplier solution to just be more elegant. Giving me 1000 switches doesn't make it more powerful.. it just means someone took the time to give you that flexibility.


Flat tire implies not working. This is more a v4 (tandberg) versus a v6 (polycom) discussion. If your car has a v4 and you are happy, then that is great. If you need a box with more power go for a v6. I prefer a reliable v4 over an unreliable v6.

This is an over generalization.. leading to my point above. I'd say the Accord is less powerful IMO.. the tandberg unit does more with less. Now, if you throw enough Accord at it.. it can do more.. but thats only after you put everything under the sun in there.

My point exactly! Basically you only know if the sites are connected or disconnected, but have no idea of the quality of the conferences until someone gets so annoyed they call a help number - which is typically when the system is completely unusable.

No.. I don't have to sit on top of them to know the quality is good. The quality is always good. Failures are the EXCEPTION, not the norm. That's why we don't have to hand-hold every conference.


I suggest you not make statements on the quality of features until you monitor some of the conferences. We monitor our conferences at off peak times when man power is available (this is man power intensive for all conferences), and have surveys for users to fill out after a conference. Until you do that you are only speaking from ignorance on user’s experience.

Uhh.. whatever. I do know what the users are experiencing.. and its our evolution and focus on task-built solutions (rather then just throw it in there) that allows us to expect and deliver the quality we do and not expect every call to need help. I spend the time UP FRONT rather then chasing my tail for the next 3 years.

[djackson]

Uhh.. whatever. I do know what the users are experiencing.. and its our evolution and focus on task-built solutions (rather then just throw it in there) that allows us to expect and deliver the quality we do and not expect every call to need help. I spend the time UP FRONT rather then chasing my tail for the next 3 years.

trapehzoid - thanks for sharing, we seem to usually disagree, but I appreciate your opinion :nervous: .

If your customers are happy, please continue doing what you are doing. And if you don't mind, one day I will steal the phrase "our evolution and focus on task-built solution"... very nice wording :nervous:

About the "chasing tail" comment, am I the only person in a videoconference support position that gets customer feedback? Maybe so, and if getting customer feedback is "chasing my tail" I guess I will keep going in circles.


Now, if you throw enough Accord at it.. it can do more.. but thats only after you put everything under the sun in there. .

Good point, not disagreeing.

Anyways, the Tandberg MCU/MPS is underpowered in functionality as long as it sending out only 1 (or 2 at best) video streams to all partipants. This is a trick manufacturers do in endpoint units with built in MCU capability or in software only MCUs. I think Tandberg is the only major hardware vendor to do this in an expensive hardware solution.

As Kevin from Codian posted earlier there are a few problems with doing this:



"a) With just the one or two streams a large percentage of participants in the conference may have to view video that is not the best video they can see (either lower bandwith, lower resolution or lower complexity codec that they can cope with). ....
c) If one participant is having packet loss and requesting key frames from the MCU everyone gets sent the key frame and the corresponding loss of video quality. If the MCU has to respond to flow control messages to send lower bitrate to this participant because of packet loss everyone has to view this lower bitrate. "



I have had problems with both Points A & C today on our Tandberg unit - please don't tell me that it is not under powered. (Of course, if we didn't monitor the conferences, we would have never known that the MCU was receiving an XGA image, but was only sending out CIF to all participants).

[trapehzoid]

About the "chasing tail" comment, am I the only person in a videoconference support position that gets customer feedback? Maybe so, and if getting customer feedback is "chasing my tail" I guess I will keep going in circles.

No, the difference is progress vs in circles. We have a bit of an advantage that I work within a closed network, but my point was we minimize the 'free willy' nature of some users who want every gizmo, and we build our solutions to be solid and reliable first.. fancy later. Honestly.. I can't tell you the difference between a $100 handset on a desk and a $500 handset on a desk in terms of actually making calls, etc. They all work the same.. the rest is bells and whistles that for most people and most situations are fluff. All they do is complicate the issue. I put the task before the 'wants'.

Anyways, the Tandberg MCU/MPS is underpowered in functionality as long as it sending out only 1 (or 2 at best) video streams to all partipants. This is a trick manufacturers do in endpoint units with built in MCU capability or in software only MCUs.

I don't think that's fair at all. One, no one has transcoding built into the embedded MCU, nor do any send out seperate images.

Plus, if you think about it, the MPS is sending more then 2 images.. because its doing things like H.239 as well. I think people are getting hung up on 'mine is bigger then yours' vs practicality of getting things done. Personally, I think having to use Video+ boards in the Accord to achieve the same basically 'any' connectivity is a crime in its own. Anyone who doesn't think so needs to pull their head out of .. you know. The Codian followed the approach right following the 'any' apporach and took it further by adding more flexibility. I'm sure others will follow as well.

Personally, I wouldn't jump through hoops to have the any layout stuff because I don't want to have to do it for users, and I don't think most users would use it. VideoCommander has been out forever.. how many of you actually use it?

The point about 'isolating' streams is valid.. but we avoid that in my setup by fixing the problem endpoint. The current conference may be affected, but we resolve it before future conferences. Basically we don't allow crap sites in to begin with if we can help it :) So in our situation, it would be a nice to have, but does not demand a refresh from us at this point.

[djackson]

The Codian followed the approach right following the 'any' apporach and took it further by adding more flexibility. I'm sure others will follow as well.

I totally agree.


The point about 'isolating' streams is valid.. but we avoid that in my setup by fixing the problem endpoint. The current conference may be affected, but we resolve it before future conferences. Basically we don't allow crap sites in to begin with if we can help it :) So in our situation, it would be a nice to have, but does not demand a refresh from us at this point.

Lucky you, I'm jealous. We connect to all kinds of junk... We do more of our fair share of international conferences and third world countries over IP or ISDN. I seem to find every imperfection in a product, so the the more powerful MCU works better for our conferences.

[Sean Lessman]

...I seem to find every imperfection in a product, so the the more powerful MCU works better for our conferences.
More powerful or more bug free?

The MPS certainly doesn't lack power. It can be and will be improved with features some of which you and others view as deficiencies. But to say a product lacks power because it doesn't have the features you want isn't really a valid statement. What about the features it has that others do not?

Sean

[djackson]

More powerful or more bug free?

The MPS certainly doesn't lack power. It can be and will be improved with features some of which you and others view as deficiencies. But to say a product lacks power because it doesn't have the features you want isn't really a valid statement. What about the features it has that others do not?

Sean

It is good to hear that the MPS will be upgraded. Will that be a software upgrade or hardware?

[William]

The MPS is NOT the only game in town to make the MGC platform look out of date. Have you seen the Codian MCU. It can do up to 40ports of video, at speeds of upto 4mbs. Any combination of ports can he H.264 (up to 2mbs). Each port is transcoded so that you can freely mix and match slow connections with high speed conections, old endpoints and new ones. Users can choose individual layouts, and there is NEVER a drop in port count. It is by far the most flexible and powerful MCU on the market roday. http://www.codian.com

[Entropy3XD]

I agree with you William that the Codian is an excellent MCU. Unfortunately, many of us here are in a field of work where a serial interface is a major requirement. This makes Tandberg, Polycom, and Radvision the only viable options.

If someone truly requires a Codian, yet needs serial connectivity, a Radvision serial gateway could be used as an interface. This might not be a cost effective solution however.

[Sean Lessman]

It is by far the most flexible and powerful MCU on the market roday. http://www.codian.com

Very strong comments Will. But I would expect nothing less from the company's founder! :)

Sean

[Sean Lessman]

...and there is NEVER a drop in port count....

When you want encryption, port count drops to zero :)

Sean

[Team EITC]

We currently have both on site and have been using the MGC as our primary and the MPS as a secondary which can possible replace the MGC.

One major benifit of the MGC is Meeting Rooms. With less and less staff available I've setup allot of my conference's for full automation. Its not possible to setup meeting rooms on the MPS. But I heard from a Tandberg SE that something like Meeting rooms will be added in there new release.

Another benifit for me is Conf on Demand when using Polycom Pathnav. Its only good for IP sites but allow the users to build Multi point conferences without any help from me. Well maybe some training and an Address book. Which you can push out from an ILS server or TMS for most codecs.

[Vtech]

Sean,

Who actually uses AES encryption?

[svanwhy]

Unless i'm wrong you can't move participants between conferences with th MPS like you can with the MGC. I use that feature alot since we allways have multiple conferences going on at the same time and often use that feature.

[trapehzoid]

Sean,

Who actually uses AES encryption?

I do :)

if you are trying to infer its useless... well I'll refer the questions from my director to you when he asks me about who can see him or not. "Ahh.. don't worry about it, no one will bother" doesn't fly :)

If you are trying to infer its not secure.. then I'll just ignore you.

If you are trying to infer no one uses it.. then I'll just say that's because you are using crap endpoints or bits.

If you are trying to infer its not secure enough.. no for military use its not, but there are other solutions for that. And I don't think anyone here will argue that what the military requires to get a task done is always the best way to get it done :)

Just because its not KG or KIV doesn't mean its not good enough for the rest of the world.

[tom9933]

I agree that AES can be very important, although I’ll also say we are currently not using it. The reason for this is that for our users they are not worried about today and we tend to see problems when calling legacy sites that don’t support it. So today I’m buying equipment that supports it, but then turning the feature off so I can still call the legacy sites. The biggest problem with call completion seems to be with legacy gatekeepers and firewalls rather than legacy endpoints. Having said that, the users don’t care they just want it to work so we disable it in the endpoints. Ah growing pains, you gotta love them :)

[Sean Lessman]

We currently have both on site and have been using the MGC as our primary and the MPS as a secondary which can possible replace the MGC.

One major benifit of the MGC is Meeting Rooms. With less and less staff available I've setup allot of my conference's for full automation. Its not possible to setup meeting rooms on the MPS. But I heard from a Tandberg SE that something like Meeting rooms will be added in there new release.

Another benifit for me is Conf on Demand when using Polycom Pathnav. Its only good for IP sites but allow the users to build Multi point conferences without any help from me. Well maybe some training and an Address book. Which you can push out from an ILS server or TMS for most codecs.

Adhoc and Personal Conferences are both available with J3 on the MPS.

Sean

[Sean Lessman]

Unless i'm wrong you can't move participants between conferences with th MPS like you can with the MGC. I use that feature alot since we allways have multiple conferences going on at the same time and often use that feature.

This is available in J3 software for the MPS.

Sean

[svanwhy]

Is the J3 software currently available? If not when is it released and how do we get it?

Thanks

[Vtech]

No I was actually just asking if anyone used it. Yes I am a DOD contractor and it does not meet NSA requirements but I was unaware of any civilian agencies that do. I beleive eventually AES will become a NSA certified type 1 encryptor with a smart card of some sort. Sean should be suggesting that as we speak. We usually turn it off because some of our users freak out when they see encryption not possible on the other side.


I do :)

if you are trying to infer its useless... well I'll refer the questions from my director to you when he asks me about who can see him or not. "Ahh.. don't worry about it, no one will bother" doesn't fly :)

If you are trying to infer its not secure.. then I'll just ignore you.

If you are trying to infer no one uses it.. then I'll just say that's because you are using crap endpoints or bits.

If you are trying to infer its not secure enough.. no for military use its not, but there are other solutions for that. And I don't think anyone here will argue that what the military requires to get a task done is always the best way to get it done :)

Just because its not KG or KIV doesn't mean its not good enough for the rest of the world.

[trapehzoid]

AES isn't as much the issue as the requirements around how the device is handled, setup, etc. AES is strong enough, its just the military requirements are just too retarded for the general population. There are certified systems that use AES encryption.. just not the video codecs we are talking about. My guess is that type of encryption will always be left to the specialists. Its too far off mainstream for a product intended for the mass market. That is unless they build special products for military customers.

[Sean Lessman]

Is the J3 software currently available? If not when is it released and how do we get it?

Thanks

Should be within 1 week from now. Documentation etc is being finalized.

Sean

[Vtech]

Yes national security is very retarded so an unconventional method of encrytpion would be crazy using the same thing as the rest of the world makes perfect sence.

AES isn't as much the issue as the requirements around how the device is handled, setup, etc. AES is strong enough, its just the military requirements are just too retarded for the general population. There are certified systems that use AES encryption.. just not the video codecs we are talking about. My guess is that type of encryption will always be left to the specialists. Its too far off mainstream for a product intended for the mass market. That is unless they build special products for military customers.

[trapehzoid]

Yes national security is very retarded so an unconventional method of encrytpion would be crazy using the same thing as the rest of the world makes perfect sence.

Oh stop being retarded.

AES is plenty secure. The reason these systems are certified are they don't go through all the hoops you must do to seperate and isolate the systems.. key control.. etc. Its got nothing to do with AES not being secure.

My comment was towards the levels of obsurd that most military procedures require around security.

The fact is most installations have tons of 18yr old kids they can use like monkeys so the fact you must manually plug 18 different cables to make one connection doesn't bother or phase them. There is always another enlisted guy they can use to hold the two cables for them.

You're just defending it because you work for a company that sells products to satisfy their stupidity.

[Vtech]

I guess I can asume you do not work for DOD?


Oh stop being retarded.

AES is plenty secure. The reason these systems are certified are they don't go through all the hoops you must do to seperate and isolate the systems.. key control.. etc. Its got nothing to do with AES not being secure.

My comment was towards the levels of obsurd that most military procedures require around security.

The fact is most installations have tons of 18yr old kids they can use like monkeys so the fact you must manually plug 18 different cables to make one connection doesn't bother or phase them. There is always another enlisted guy they can use to hold the two cables for them.

You're just defending it because you work for a company that sells products to satisfy their stupidity.

[trapehzoid]

I guess I can asume you do not work for DOD?

Doesn't matter.. you still avoided

"You're just defending it because you work for a company that sells products to satisfy their stupidity"

And there are AES approved products...

[Vtech]

No actually I'm not defending them for that reason I also was in the military and understand the importance of CCI and controlling such items. 128 bit encyrption is not an NSA approved standard. But a spearate device with encyption keys generated and controlled by JICMO is no one has access to the crypto keys accept properly cleared individuals. The chance of out cryptographic keys being compromised is very slim where we see standard commercial encyption broken all the time. The idea of saying the DOD's measures for cryptographic control are retarded is not correct perhaps to your standards they are but when it comes to the lives of men and women in harms way it's perfectly acceptable.

Doesn't matter.. you still avoided

"You're just defending it because you work for a company that sells products to satisfy their stupidity"

And there are AES approved products...

[trapehzoid]

No actually I'm not defending them for that reason I also was in the military and understand the importance of CCI and controlling such items. 128 bit encyrption is not an NSA approved standard.

AES is not limited to 128bit. In fact, most use 256bit. Just so happens the implementation by several vendors in the video world is 128.

But a spearate device with encyption keys generated and controlled by JICMO is no one has access to the crypto keys accept properly cleared individuals.

Key point.. 'individuals'. People are always your weakest link. So, thats why I believe its retarded to go as far as say requiring your RS-366 leads be disconnected physically when its a ZERO risk and other obsurdities when 99% of your problem is the guy in front of the system.

The chance of out cryptographic keys being compromised is very slim where we see standard commercial encyption broken all the time.

Apples and oranges. Commercial encryption being broken is typically due to poor keys or key handling, not necessarily 'stronger encryption'. The military's policies on keys simply remove that problem (which is a good thing). My point is.. don't throw AES out because its not applied the same way a military encryption is. You could use the same methods with Bob and Joe inside Amway.. but it would never fly because it would be too cumbersome.

Look at PKI. Every message we type today could easily be encoded and signed with PKI technology today and nearly no one could break it (except ft meade types).. but its not done. Why? Because the key handling is too cumbersome for the mass market. Every system to make key handling more easy.. reduces the security's effectivness.. because you are only as safe as your key is.

My gripe is not with the crypto technology.. its as I stated in the first place.. the policies the securities officers come up with are absurd.

The idea of saying the DOD's measures for cryptographic control are retarded is not correct perhaps to your standards they are but when it comes to the lives of men and women in harms way it's perfectly acceptable.

Koolaid is over on ailse 5...

[Entropy3XD]

You're just defending it because you work for a company that sells products to satisfy their stupidity

The correct statement would be that some of us work for companies that sell products to satisfy their REQUIREMENTS. Even the military branches have to conform to a basic set of requirements passed down to them. Now I will agree that the policies do not always make much sense as most of them were put into place during the cold war, but they were put into place for good reason. We might be able to change the minds of a few security officers for some policies, but most of us don't wield enough power to debate the NSA. If you have this kind of power Trap, please let me know because I have a few suggestions to pass on as well.

I will definitely agree with Trap that the biggest risk is the individual in the room. Most of the risks I have seen were due to human error and it usually came down to lack of training and education. Because of high turn over rates and lack of time or money for training, most locations have a basic set of guidelines for room build-out and a SOP. As long as the guidelines and procedures are followed (regardless of how stupid they may seem) you at least make the uneducated less dangerous. Unfortunately, sometimes it does not matter how “idiot proof” you make things, because the world will always build a better idiot.

I will also agree with Trap that there are many other means of encryption which are just as secure, but unless they are approved to go into DOD environments you can forget about them for those particular customers.

As far as the KoolAid……..Yea, it sucks to have to drink it, but it goes down smoother if you spike it with a little alcohol.

[trapehzoid]

you at least make the uneducated less dangerous.

yeah.. that's a good point, but that's all policy and procedures.. nothing to do with the technology or 'capabilities'.

I just can't stand it when someone trashes something for the wrong reason.

[Vtech]

Explain how the below quote is bashing anything, I feel it will eventually be used once a smart card (possible use for Tandberg's USB port) of some sort can be used to make the algorythm unique. I never said AES was bad nor did I intend that simply stated is doens't meet current NSA requirements.


No I was actually just asking if anyone used it. Yes I am a DOD contractor and it does not meet NSA requirements but I was unaware of any civilian agencies that do. I beleive eventually AES will become a NSA certified type 1 encryptor with a smart card of some sort. Sean should be suggesting that as we speak. We usually turn it off because some of our users freak out when they see encryption not possible on the other side.

I just can't stand it when someone trashes something for the wrong reason.