jfk

'Have to'? If I had to choose a machine with Win 8 installed, I would buy a copy of Win 7 too. To properly prepare it for WO, you should reformat and re-install Windows anyway (then tune). Might as well install a tried and true variant while you are at it. I personally would pay the Windows retail copy price NOT to be a pioneer. ymmv You can always pick out the pioneers, their the ones with the arrows in their back.

That is a legacy message aimed at Windows XP, not Windows 8. Windows 8 provides all functions available in Windows 7. Message should probably be updated to ... "Windows 7 or higher required for full functionality" That said, Windows 8 provides no additional benefit over Windows 7. More likely the opposite, Windows 8 contains more potential new things that are not needed but will still muck up reliability. So you will pioneer a bit as the experience pool on Windows 8 is small. Windows 7 on the other hand is mature and limits WATCHOUT in no way. Dataton publishes tuning guidelines for Windows 7, Win 7 is still a current product, and it is a tried and trued platform.

Uhh, MiniDisplayPort and DisplayPort are both the same as it relates to active adaptor / passive adaptor use. Very possibly the nVidia card has the needed extra support for the MDP/DP to work without an active adaptor and three or more ports lit up, but that is not related to the DP connection type. I know that ATi cards with DisplayPort still need the active adaptor when more that two ports are used. It stops working when two DVI outputs are used by the non-MDP/DP outputs. On ATi, the same MDP/DP port will work with a passive adaptor when two or less ports are used, so testing with pasive adaptors needs at least three ports lit up to be a valid test on ATi. Did you test the passive DP->DVI-D adaptor on nVidia with all ports lit up? nVidia's info on this, How To Setup NVIDIA Surround on GeForce 600 Series GPUs makes no mention of active/passive adaptors.

Yes, search Echo's support, they indicate the Echo driver is sensitive to the Firewire interface type. (item 6 at that link) We have encountered issues with Asus motherboard Firewire interface and have used PCIe Firewire cards to overcome it. May not be your issue, but worth looking into.

I am still having difficulty following what is happening, but no matter, it sounds like your WATCHOUT PCs are seeing EDID interruptions on their outputs. Any change in EDID during operation will force a graphics card reset. Anytime a graphics card reset occurs, the outputs can and will be reassigned to new output numbers. The only way to avoid this is to provide stable / constant EDID. An EDID manager on each output can isolate any bad behaviour in the display chain and prevent random output resets. Have seen this occur when the transceiver of a display extender was crashing and restarting. Have seen this occur if a display is powered down while watchpoint is running, etc.

Sounds like a touchscreen issue, but I think you need to clarify some. You seem to indicate 1 WATCHOUT Display PC for each screen - a total of 11 display computers? Is each output playing back from an independent aux timeline, or are the outputs played back in groups? Where does the touchscreen software reside, you seem to indicate it is on the WATCHOUT production PC ? How does the touchscreen system trigger the WATCHOUT playback? What WATCHOUT version?

And that seems to be the key, assuming quicktime decoding is not involved, the movie decoders are multi-threaded. How multi-threading is implemented could vary by the codec (decoder). But I do not know how dynamic that multi-threading is, i.e. Does one big clip get more threads to do the work based on its pixel count? Or does multi-threaded really just mean two per movie, or some other ceiling lower than 4 or 6 times for larger pixel counts? With six outputs and a six-core hyper-threaded processor you have 12 threads to play with. If the thread utilization is 'per clip limited', then four or six HD/30p movies could utilize four to six time the processor resources than one aggregate movie - same number of total pixels either way, but total pixel count appears to be secondary to processor utilization in this case.

First off, I respect Omar has done his homework and appreciate his good communication. Could you expand on that? Are they cloned or simply individually tuned to the same list? If you did clone, did you virus check your master before cloning? I doubt it is a virus ... Do you have the ability to revert to your original clone image?At some point you stop chasing this stuff and revert to 'known good state'. If 'known good state' still exhibits the issue, it is back to the test bench. Starting watchpoint computer(s) from shortcut in startup? Any command line switches used in the watchpoint shortcut? i.e. startup script, noLogo, etc. That is a tell, as NS correctly pointed out. Even before any attempt to connect with watchmaker is made?with no IP on a virgin watchpoint logo screen . with no connection attempts or autostart scripts, got to stop right there, eliminate production from the equation, and determine why that is happening. That is strange, and you are correct, the appearance of the display name as a new (third) text line on the watchpoint logo screen is derived from the show data and is an indication that watchmaker connected to watchpoint and attempted to load a show file. But I do not think the IP address should disappear in this case. As this is indicative of a network issue, FireWall is always the first suspect. I believe the initial show load to watchpoint is done via documented IP ports, but the next steps involve dynamically assigned IP ports. Never seen a problem with that unless some form of port blocking is occurring. Firewall is the usual suspect, and in Win 7, the FireWall has a nasty habit of turning itself on when related settings are tweaked. Still, I do not think FireWall is what Omar is seeing. ----- For consideration by those who may be seeing a firewall issue ... It is possible to run WATCHOUT with Firewall On by always granting permission when asked. Issue is, with dynamically assigned ports, you will be periodically asked again. watchpoint prevents that message from taking the foreground. You might try quitting watchpoint when this occurs to see if any Windows error messages are lurking behind. Then check for for the presence of watchpoint and/or watchdog log files in the log folder at watchpoint.exe's location. All but the waiting to go Online part, it is not getting that far, it has not completed all the checks it runs through before going online. When you see that, it would be interesting to go offline, open WATCHOUT Remote, connect to watchpoint, and open the show. Bet it will work as there is only communication on the documented port, watchpoint does not need watchmaker to open / load a show that has been previously successfully transferred from watchmaker. It means no connection has been made. A network issue would not display the IP address, it either suppresses it, or you may see the local host address. In general, yes to all of the above, but if port blocking comes in to play - not necessarily, as that is a very specific network issue. Connecting with VNC could result in different dynamically assigned ports on the next online attempt. Wake On LAN is very different than anything else in the system it is not even IP, as it is done with the NIC's MAC address which can not route outside the physical network.

Are you using this iPhone app -> Dataton WATCHOUT Remote By Dataton AB or one of the other WATCHOUT iPhone apps?

How does that make sense? Pretty sure an HDCP compliant EDID manage simply passes through the HDCP handshake. Otherwise, HDCP is not worth much if it so easily defeated for a lossless digital output and the major studios would be after anyone offering such a device in a hurry. Only way I know to work around HDCP is to use an output method that does not support HDCP, which means analog output instead of digital output. That is how people work around cable TV sources, they use the alternate HD analog component outputs of the cable box as a source. With a computer, VGA would provide the same workaround.

Hej mosu, Where are you located? Does that eliminate the HDCP issues with a Macintosh computer as a source? Can you say 'think different' ? DVI supports HDCP just like HDMI does, and no capture card is going to accept an HDCP restricted signal.

What does that mean - " The middle screen still crashed. " From a single computer, how does one screen of three crash?

I do not see any mention of Windows tuning, did you follow the instructions in the Windows_7_Tweaking_list_2.1.pdf after you reformatted the drives and performed a fresh Win 7 install? And what do you hope to gain with RAID 1? Everything is already backed up by the production machine, I sure would not add the unnecessary complexity of RAID 1 on a display computer. Over the years, RAID has been a real sore spot, especially mb RAID controllers, and especially when RAID is used as a boot drive. We have encountered so many issues with RAID 0, we will not build a machine with RAID without a separate boot drive, and up until now, we have always used a PCIe hardware RAID controller, ignoring the mb RAID support altogether. ymmv

don't want to leave you hanging ... There are a lot of variables there from the simple to the sublime. Simple - put the data on a conditional layer as it is in synch with the 'points of interest' and reveal it with a simple condition change. A bit tedious to code, but doable with a sensor trigger as simple as a switch closure -> MIDI Note interface -> Task trigger -> aux timline that talks to watchout itself and sets the appropriate condition. As you are venturing into the world of control (vs content assembly) such an endeavor assumes you have mastered Appendix D of the User Guide.

Yes. Using a MIDI foot pedal interface generating a MIDI note command to connect the GPI switch closure to WATCHOUT. Just place the MIDI note input in the task window trigger field for the timeline you wish to trigger. Of course this generates a blind "go" which is extremely easy to get out of synch, so you must monitor operation carefully. With WATCHOUT 5.5 (coming soon) you could get clever and use multiple GPI triggers / MIDI note inputs in combinations to trigger unique cue points, a bit tedious, but a lot more reliable. i.e. 8 GPI triggers could provide 8! unique cue triggers. We have not, but customers have reported success with AJA in the past. You will need to rely on AJA support for any assistance if you run into issues. On the other hand, there is a wealth of information in this forum about using BlackMagic Design and Datapath capture cards. If you do go the AJA route, please post back with your experiences

Correct Simple answer is yes, but to be clear, after the changes are made and saved with channelshifter.air, you must go online, or update if you are already online, to transfer the new files to the display computers. WATCHOUT will detect the file has been changed when going online / update and automatically transfer the new version. Yes.

No, there is no provision to receive timecode via TCP/IP. By definition, SMPTE/EBU timecode is an audio signal. The specs allow for anachronistic analog anomalies like wow and flutter, off speed, etc. When we tested timecode chase, we were surprised how fast the timeline locked to timecode. Of course, with a random jump there may be some "paint up" time. but the stills we tested with were pretty snappy. I would recommend you set the timecode specific format instead of 'auto detect', this seems to help in repeatability and lock up speed. --- Not that this may apply to RayS, but often when I am asked about TC lock up time, the user intends to trigger cues (start / stop) with timecode. Note, while TC does lock very quickly, it deliberately releases slowly. So timecode is not well suited to cueing / stopping during performance. If you must, allow for some coasting (and snap back) at the pause point.

That is correct. Because VNC conflicts with the software installed by WATCHOUT.

You want your cake and eat it to? No matter, DO NOT INSTALL VNC on ANY WATCHOUT COMPUTER. And you do not want to run production over a wireless network either, production is the timing master when present, and WiFi will muck with the timing. Either move your production computer and hard wire it, or punt.

Do not install VNC on ANY WATCHOUT computer. You may be overcomplicating things. If you have a laptop plugged into the WATCHOUT network, just shut down production and run the display computer directly with WATCHOUT Remote.

No, as stated in the User Guide, the DMX output device(s) can only use one ArtNet universe. Chapter 11 Inputs and Outputs page 205 You could drive more than one universe using the DMX record / playback functions added in v5.2, but not with the DMX device function.

Back on topic, Roman, from a Reliability point of view, experience in the field suggests the DataPath wins that comparison hands down. It is far less finicky about the host system (motherboard compatibility, etc.). DataPath drivers do their job cleanly without mucking with things that effect WATCHOUT. BMD changes a few less used decoders like MJPEG causing those codecs to fail in WATCHOUT. DataPath does no such nonsense. Furthermore, DataPath Live Media object settings are far more forgiving than BMD. Of course, the BMD Decklink series is far lower cost (and lower bandwidth) and with proper care can be made to work. But as such, they would have to rate as less reliable. The DataPath accepts up to 1080p, DeckLink series tops out at 1080i/720p. BMD has an extreme series that exceeds 1080p and is priced in the same neighborhood as the DataPath, but BMD extreme drivers are comparable to the DeckLink series (and in the comparison that is not a good thing).

Off Topic FYI - I may have done something a little confusing here, so let me clarify. The post by SkyIsFalling (Roman) above in this thread (datapath sdi2 vs deckline DUO) was not originally posted here. It was submitted as a new thread. Rather than tell him to search, I went ahead and did it for him, then merged his submission into this thread. (My hope is to keep the knowledge base here as compact as possible.)

That has nothing to do with WATCHOUT and everything to do with the BMD drivers. If the BMD WDM drivers will support three Intesity cards in one machine, WATCHOUT will do it too. I am having difficulty confirming the BMD Intensity can do up to four when using WDM drivers. I did find a post in the BMD support forum that indicates up to four Intensity cards can be used, but it is not clear which driver it references.

There is nothing in WATCHOUT that will prevent it. This does challenge the graphics card, but it should be possible if your GPU is up to the task. Not so much a graphics sub-system memory size issue, more about GPU speed and memory bandwidth. If all must appear at the same time, you may need to manually "pre-roll" them to avoid a bottleneck. Obviously you would need to test on your hardware to confirm.