jfk

Could you clarify that? i.e. it plays fine in WATCHOUT on one screen without blending or scaling or in a media player? (media players use hardware accelerated movie decoding, watchout uses software decoding, so that is an apples to oranges comparison.)

Yes it is, so the last successful update under that show name will be available from the display computers. Go to the display machine, find the Shows folder and inside will be another folder with the actual show name. Inside that folder is a file called Spec, with that file, see Rebuild show file?

Flat panel calculations Convert to feet to keep all dimensions in the same unit measure. 34-8/32" = 34.25" = 2.854166667' 60-26/32" = 60.8125" = 5.067708333' since the panel is rotated 90° 1920 / 5.067708333' = 378.8694758 pixels per foot projection — 8'H consists of 1080 pixels = 135 / pixels per foot. 90° rotated 1080-p flat panel — 5.068' W consists of 1920 pixels = 379 / pixels per foot 379 / 135= 2.806440562, a correction factor inside of 300% 1920/2.8 = 684 (684.140625) 1080/2.8 = 385 (384.8291016) The projector stays native at 1920 x 1080 stage size, The flat panels 1920x1080 output become 684 x 385 stage size. To do this correctly, you would also need the measurement from flat panel image edge to the corresponding edge of the actual projected image (not your cropped image). And you would need the distance from the top of the projected image to the top of the flat panel images. You may have to use stage tiers to deal with that. The stage Display dialogs may end up looking something like this ...

You are missing the point. Please provide the exact hight and width dimensions of the projected size of the full 1920x1080 raster. In a pinch, you can get by with just height.What you crop out of it after that is not important to the calculation. Please provide the exact height and width of the visible area of the flat panel. Please provide / confirm the exact resolution of the flat panels. With that information you can calculate the number of pixels per unit measure (dpi) for each display type. Then, you can calculate the correction to make an image spanning both display types remain the same actual size. Open the Stage Display Dialog Uncheck Width and Height same as Display Resolution and Lock down the Display Resolution at the display's native pixel count. You then convert the above correction to stage size in pixels to achieve the desired result.

The projector still expects a 1920x1080 signal. So you still need to output 1920x1080, use a pair of simple black still images on the top layer (or on an aux timeline set to always on top) to mask the unused portions of the raster. There is a chance the pixel density difference between the flat panels and the projected images is outisde the magnification correction range of WATCHOUT (±300%), not enough info to make that determination.

No. But why? If there is no human to shut down production, why are you using production at all? Short of a human being making edits, there is no function available with production that is not available in display cluster operation, so why include production in an automated system?

Nor should they. Dynamic Image Server is for still images only, not movies. Granted you can update still images every second or so, but no full frame rate moving stuff.

While static IP is ok, it is best to avoid entering the IP in the Display configuration dialog. Possibly you are using names, not clear from the description. This is a change from all versions prior to 5.5 We have observed anomalous behavior when using IP addresses in watchmaker's display configuration w/ 5.5+. Best to rule that out first. Use the new Computer Name feature for the Computer: designation in watchmaker's Display configuration dialog. The names are assigned in watchmaker v5.5+ Network window (found in the Window menu). When open the window is titled "Display Computers Found on Local Network". Once assigned, the name remains with the display computer until somoene uses watchamker's network window to change them. The name is shown in watchpoint 5.5+ initial logo screen. The sample shown is configured with a fixed IP.

Simple answer is no. WATCHPAX is Zero Configuration purpose built, access to windows embedded is locked. You can not install the driver needed for any add-on not already included by Dataton. WATCHPAX includes no drivers for USB capture hardware. BTW USB video capture typically requires USB 3, WATCHPAX only provides USB 2. reference: WATCHPAX product sheet If latency is not an issue, you could use a separate computer to capture the live signal and stream that to WATCHPAX via the network. This would not be acceptable for iMag or similar live use. WATCHPAX can accept a network stream via WATCHOUT's Network Video media object. Third party software tools running on a non-WATCHOUT computer would need to perform the magic of capturing and streaming the live feed to the network in a form that is compatible with WATCHOUT. reference: Dataton WATCHOUT User's Guide - Version 5.2 — Chapter 3 — Media — Network Video — page 48—49

places to look are BIOS settings for behavior after power loss, motherboard battery, NIC driver and other hardware related issues. A good place to look for solutions is the hardware vendor's support channel or forums. Isn't that the definition of Wake On WAN? i.e. other parts of the network are at play as well. ================================================================================ ================================================================================ ================================================================================ A clean power down makes sense from an energy conservation standpoint. Restart sounds like a band-aid. Enabling workarounds masks correctable issues that may manifest in other ways. An IT professional should be able to cure that network issue for you. I am uneasy with being too helpful in taking down an entire cluster. It would also ignore or complicate the MAC address collection process.

What resolution / refresh rate signal source did you successfully use with the BlackMagic DeckLink 4K Extreme? What was the source? (camera? computer? something else?)

No. Use industry standard Wake On LAN command (which uses the MAC address) to power up the computer after a powerDown.

IP power down watchmaker's power down is a two step process, The MAC address is grabbed and held in the windows registry for the next power up (wake on lan). The IP equivalent is not done with 'one to many' cluster commands (authenticate 1) it is accomplished with 'one to one' administrative commands (authenticate 2). You must establish a direct connection with each watchpoint computer for administrative commands, these commands only affect the one connected computer. authenticate 2 getMACAddr powerDown getMACAddr Retrieves the MAC address of the WATCHPOINT network interface, if possible, as six values. A General Runtime Error is returned instead if the MAC address is unknown. Example: getMACAddr Reply 23 33 2 45 143 230 Note: the response is in decimal format, while the Wake On LAN Magic Packet will require that data in hexadecimal format. So you will need to convert that to hex for most uses. For example, the sample above would convert to 17h 21h 02h 2Dh 8Fh E6h. or $17$21$02$2D$8F$E6 powerDown with no parameter. This will immediately disconnect you, as it causes WATCHPOINT to quit, so this should be the last command you intend to send until the watchpoint computer is powered on / restarted.

You seem to indicate you are using WATCHOUT 5.5. This issue is addressed in WATCHOUT 5.5.1 .

for updates above ...

And the version 5.5.1 watchpoint implementation changed a bit, the best version of this function yet. On start up, before a show is loaded, you get a black screen, there is no logo, but the informational text portions of the logo screen remain. As always with this switch since v2, once a show is loaded, as long as you maintain full screen mode, you will never see any of the logo screen again, update or go online, with this switch enabled, the show simply freezes where it is, instead of jumping to the logo screen For reference, pre-5.5, you would still see the full logo screen on initial load until a show was loaded. BTW More information on this, including other methods of achieving similar results, are found in an existing thread, Hide Watchout logo + progress bar while updating show.

A functional equivalent exists if your control system is versatile enough. getStatus will provide the current time, have the control system add the relative jump, then execute a gotoTime just the same, in essence, I agree, I would want it for cluster protocol, relative goto time jumps might be better implemented in gotoTime

For iPad apps via WiFi, there is a third party Windows driver that accepts the network signal and provides it to Windows as standard MIDI communications. Any Windows compliant MIDI interface will do, WATCHOUT receives MIDI from standard Win OS services.

Don't know how I missed that earlier, but the answer is in the post above. Gosh, missed that one two. Also covered in the post above

We did not, wish we had. Our experience is worth mentioning. Ran for a three day show. First day ran flawlessly. Second day started out with with a repeating crash that at first was unidentifiable. What clearly appeared to be a software issue was in fact a hardware issue, specifically display signal chain. BTW I don't take credit for identifying this issue, I am just the reporter in this case Our 4k monitors only accepted HDMI 1.4 for 4k signals. From the monitor to the graphics card, our connection was 3940x2160@30p monitor 2m High Speed HDMI 1.4 cable HDMI 1.4 to MDP high speed active 4k adaptor graphics card x2 monitors on the same Radeon graphics card. And there was the rub. Three non-locking connection points per monitor and gravity can have strange affects. If a display connection was fully breaking, it would have been obvious. But in this case, a marginal connection never lost signal to the display, yet it would upset the graphics subsytem enough to tick off WATCHOUT and send it into a non-recoverable state. Seems watchdog was attemping to bring it back and failing, but don't hold me to that observation. Very misleading to say the least. Why would you suspect the cabling if the display never lost signal, never blinked, flashed or exhibited any other behavior that would point at the connections? But that is what it turned out to be. Simply reconfirming all six connections (and adding strain relief to fight gravity) eliminated the issue and the system ran fine for the day. Next/last morning, same problem, same solution, same clean run for the day. I suspect a solid EDID manager and care for the MDP connection would have masked any anomalies farther down the signal chain. Not for the content run last Fri-Sat-Sun at this time, content was only rented for the show and was encoded by the stock house. (.m4v) However, we have been testing with our current Extreme build with three different gpus - ATi FirePro W7000 (4-output), ATi FireProW9000 (6-output) and the Radeon (4-output) we used at the show. In the shop, were are testing mpeg4 videos that are encoded at different bit rates and frame rates. Our test videos ... Video 1,Bees: 4096x2304, 22.1Mbps, 24.00fps, MPEG4 (.m4v) Video 2, New Zealand: 4096x2304, 19.1Mbps, 23.97fps, MPEG4 (.m4v) Video 3, Landscape: 3840x2160, 17.2Mbps, 29.97fps, MPEG4 (.m4v) We could play all three at the same time smoothly in our testing, albeit scaling the three down into the two displays, as we don't have a third for testing at this time. Agreed, for 4k it is quite cost effective even if it is pushing the limits of PC $$.

Show Sage ran two 4k mpeg4 movies from a single computer as part of our exhibit at LDI this year (last weekend). The video was derived from a stock house 8k original and it was split and encoded into two 4k mpeg4 files by the stock house. I need to check on the specific graphics card model, I know it was a newer Radeon series card with four outputs. Computer was i7 six-core Extreme edition with SSD drive.

It can be done. The management needed for the master may best be done in the development of the iPad UI. Pretty sure you can have multiple peer control point connections authenticated at level 1. The master could lock out others by upping to authenticate level 2 if need be. Setting the master back to level 1 re-enables the other control points. No. Still images and still image renders from SWF are the only choices. Other apps like database servers, image captures, etc need to send their output to the image server via the flash scripts.

As for your workaround, an easier way to accomplish the same result is to place a live input cue the entire duration of the main timeline in a standby layer. Standby content is always active but never visible until standby is called. (credit to Mark Watzl for this clever workaround).

Could you expand on what you mean by "... remote into each display ..." ?

Yes, WATCHOUT can send user defined ACII or hex strings (and combinations of the two) to as many output devices as you wish. Wether or not the receiving device can accept and act on those strings is solely a function of the unknown software that will be controlling all the lighting cues. If they in fact support an IP or serial connection and provide a command protocol, you should be able to accomplish your goal.