[quote="Bob Ennis"]As to the question of "will GVG put in software to allow channel stealing" - I have not heard this suggestion discussed, but I will pass it on as an enhancement request the next time that I'm in there. If enough users can benefit from it, the idea certainly merits further examination from the GV software folks.[/quote] I hate the idea that I am being protected from myself. I understand the point of view from GVG, but if that is the case then give me a user overide. I like the xtend's version of whoever takes it last. That goes along with everything else related to the switcher. If I am doing any show either built by me or someone else, I should know the show well enough to understand what will happen when I hit an emem. This is no different then hitting an emem that will change ME 3, but yikes I am still on ME3, no protection there. A TE not 'moving' to the next effect and I don't notice and hit run could be even uglier, since who has time to look at the LCD screens to see if I have a TE on that keyer. Nothing else we do protects us from ourselves in a live environment I want the equipment to do what I want now. If the kids at home won't do it then maybe the switchers I work on will :) Bill

There are a couple of reasons that the transform engines are treated as floating resources - the 1st being that there are only 6 of them shared among 16 keyers (as was mentioned). But there are other reasons... Some of the postings in this thread mention the desire to manually assign the channels - like SONY does. This is really nice (along with the ability to set a flag for stealing channels), but remember that in doing this for DVE moves, the SONY requires you to build timelines for the individual DVE channels and to recall them as part of the overall timeline effect (yes you can recall them as DME wipes, but they still have their own timelines). That means that a move built in channel 1 won't work on channel 2 (unless you recall everything as DME wipes). It also means that if you built a 2-box using channels 1 & 2 on M/E 1 and later on build a split screen on M/E 2 also using channels 1 & 2, you will have a conflict if you try to recall them simultaneously. This isn't wrong - it's just a different way of having to build effects, as it requires a little more pre-planning. The Kalypso tries to free up this extra resource management requirement by the DPM channels not having any timelines of their own - they are just driven by the M/E keyer's timeline. In that way, any channel can be acquired by any keyer at any time without the risk of conflicts (except when you run out of channels). If you look at the ExtenDD line of switchers, they allocate resources on the assumption that "the last request wins". This assumes that the operator is intimately familiar with their show needs, and if they want to do something that shoots themself in the foot, then go right ahead. This logic makes sense if you built the show that you're playing back - something that isn't always the case in B, C, & D sports coverage or in most newscasts. GVG went with the philosophy that "we must protect the operator from doing something stupid". They went with the understanding that just because something doesn't have a tally doesn't mean that it's not on-air. Having a split screen in a plasma that is being shot by a camera may not give the DVE's a tally, but allowing them to be stolen as you pre-set another effect could be really career-limiting. Doing split feeds where DVE's are involved on both Primary & Secondary feeds and stealing DVE's from the Secondary client because you recall an effect on the Primary side could also really initiate some calls from the client. So the GVG system allows you to clear and re-assign channels whenever you want, but you have to conciously release them 1st through one of the means already mentioned in this thread. It's just a user protection - it's not wrong, but it's a specific design philosophy, and this philosophy was examined for still store operations too, although it was not implemented. SONY apparently took a look at both the ExtenDD & Kalypso philosophies and put in a "steal a channel if you want" flag. Because of the way they do their DVE assignments (with both the upsides and the downsides of how their DVE channels are assigned) this approach works best for them. I work on the SONY all the time, and I have never set this flag for channel stealing - I have just learned to work around resource assignments (and it works fine for me) - just like we have to work around releasing Kalypso channels before re-acquiring them. As to the question of "will GVG put in software to allow channel stealing" - I have not heard this suggestion discussed, but I will pass it on as an enhancement request the next time that I'm in there. If enough users can benefit from it, the idea certainly merits further examination from the GV software folks.

I don't know if the engineers could do this with a software upgrade or not...but what if the TEs were numbered (like the still stores, to stick with Bill's example). Then when I build an emem I select a specific TE. Let's say I build my shot clock emem with TE1 positioning the clock. I also build my double box effect using TE1 and TE2. Now the switcher knows that when I call up the boxes I want TE1 and TE2...so it gives them to me. Just like it changes SS1 from being my shot clock graphic to being my double box frames. Thoughts? Criticisms? Sean Kelly

Bill It's the same if you don't have all the Chromakeys optioned up. We have 4 chromakeyers licensed. If my 4 CKs are in use, no amount of recalling E-Mems on another M/E will enable an unlicensed Chromakeyer. The chromakeyers are floating resources, just as the TEs are. Under your scenario, if my second E-Mem only required one (used) Transform, how should the Kalypso decide which one to steal? No matter what formula you decided on, someone would complain. Incidentally, the Effects Send menu has a Lock button, which suggests that in future what you ask for might be possible. But I guarantee that the first time the Kalypso steals a Transform from your on-air effect, you would begin programming all your effects with the Lock enabled. And another incidentally, the Kalypso Technical Reference Ordering Guide from 2002 suggests that the Transforms are capable of being split from 6 "video plus key" TEs into 12 "Video Only" TEs. If this was enabled, you would have 12 "boxes" available to you. This hardware capability may have been built into the old Kalypso "Classic" TEs, but I don't know if the newer "Duo" TEs are also capable. If not, then there's probably no way they'll enable this feature just for old customers. Perhaps someone can enlighten me. brad

[quote="brad fisher"]Bill It's not so much a "protection" as the allocation of a resource. There are only three (or six) internal Transform Engines available. There are currently 16 places these can be assigned: the 16 keyers. Once they are in use, they are no longer available to be used somewhere else. If you recall an EMem that requires them and they aren't there, you simply can't have them. If they become un-assigned, they can then be grabbed and used. It would have been nicer if the Kalypso was like the ZodiaK in having a Transform Engine dedicated and available to every keyer. Probably Kalypso internal hardware limits this, even if they developed a card with 16 or more TEs on board for the Kalypso. Maybe in the next model. brad[/quote] Brad, I understand that there is only 3 or 6 TE's and 16 keyers, but I know how the show works if I call up an emem that wants the TE's elsewhere the software should give them to me, without me having macros, etc to do it. If I am using boxes in one ME, then go to camera full then to a tease in another ME with TE's, I don't want to have to release them. Just give them to me, I know what I want when I hit that button. I am sure they can write the software so that it does this. I believe the Sony gives you a choice to release or hold them. My point being that every other part of TDing and this switcher does what you want when you hit an emem or button. The TE's do not. There are many great ideas for workarounds, and I appreciate the info everyone gave on these, because I didn't even think of doing anything but calling up a generic emem, to release them. I just thought the software should handle it differently then it does, maybe I am alone in this thought. No big deal just the way 'I would like it' Maybe I should put this under the new features topic in the other forum :) thanks Bill

[quote="brad fisher"]Bill It would have been nicer if the Kalypso was like the ZodiaK in having a Transform Engine dedicated and available to every keyer.[/quote] THere was a post on here a while back about why the Kalypso doesn't have as many TEs as the Zodiak. It has to do with the fact that the Kalypso uses higher quality TE hardware which is more expensive, so a TE on every single keyer would drive the price through the roof. More details in the last post of [url=http://editsuite.com/new/viewtopic.php?id=140]this thread[/url].

Bill It's not so much a "protection" as the allocation of a resource. There are only three (or six) internal Transform Engines available. There are currently 16 places these can be assigned: the 16 keyers. Once they are in use, they are no longer available to be used somewhere else. If you recall an EMem that requires them and they aren't there, you simply can't have them. If they become un-assigned, they can then be grabbed and used. It would have been nicer if the Kalypso was like the ZodiaK in having a Transform Engine dedicated and available to every keyer. Probably Kalypso internal hardware limits this, even if they developed a card with 16 or more TEs on board for the Kalypso. Maybe in the next model. brad

[quote="brad fisher"]If the second effect (on M/E 3) also included the M/E 2 level, the E-Mem system will release the M/E 2 Transforms before trying to give them to M/E 3. So program an M/E 2 E-Mem to have the Transforms inactive (only a single keyframe is necessary), and on the Master E-Mem panel, after recalling the second effect (M/E 3), select the M/E 2 Level button and press [Learn] [Mod]. This will add M/E 2 to the Auto Recall for the second effect.[/quote] Cool idea as well, thanks Brad. I was working off someone else's stuff I think ideally just building the effect in ME 2 would make sense as well. I just find it weird that the switcher 'protects' you from this but nothing else. Which is a good thing, I don't want to be protected I want something to happen when I hit an emem, not have to look to see if the TE went with the effect. Imagine a DVEous not changing effects because it already had an effect called up. Obviously lots of workarounds thanks Bill

If the second effect (on M/E 3) also included the M/E 2 level, the E-Mem system will release the M/E 2 Transforms before trying to give them to M/E 3. So program an M/E 2 E-Mem to have the Transforms inactive (only a single keyframe is necessary), and on the Master E-Mem panel, after recalling the second effect (M/E 3), select the M/E 2 Level button and press [Learn] [Mod]. This will add M/E 2 to the Auto Recall for the second effect.

[quote="Matt Saplin"]Bill...this won't help for a non-keyframe effect, but what I do is learn the last keyframe of an effect to release any TEs that may have been used in that effect. This easily makes them available for use elsewhere on the switcher... Otherwise, I think that I'd do what Stephan suggested...just make an e-mem that is a "clear work buffer" that releases the TEs in each ME. Matt[/quote] The effects were keyframe based so releasing them on the last keyframe is a great idea.. I wouldn't mind the switcher releasing them if they aren't on the air. I know that the emem I am pressing will do something that may effect something else, but if I build an emem that needs all the TE's I will not press it or build it that way knowing that I may need other TE's at the same time for something else. If I call up my effect that needed some TE's but the switcher doesn't give them to me then that effect will tank. I feel if I ask for them give them to me. This theory is different than everything else the switcher does. If you have SS 1/2 keyed in an ME (on the air or not), if I press an emem that changes that SS channel it will change. Why get protective on the TE's, weird.. anyway thanks for an easy workaround.. Bill

Bill...this won't help for a non-keyframe effect, but what I do is learn the last keyframe of an effect to release any TEs that may have been used in that effect. This easily makes them available for use elsewhere on the switcher... Otherwise, I think that I'd do what Stephan suggested...just make an e-mem that is a "clear work buffer" that releases the TEs in each ME. Matt

The switcher will not automatically release TEs. I wouldn't want it to either, since if you have a keyer with a TE enabled, even if the key is not on the air the switcher has no way of knowing if you may be about to punch it up soon. Think of something like a basketball shot clock or football play clock. Even if it's not on air, you may not want the switcher to automatically free up that TE. I suggest learning a "blank" emem on each M/E that you can recall quickly if you need to free up some TEs. If you're in bank mode, put a blank emem in the same spot of each bank you'll be using.