RBeddig

Hi Hugo, I've just sent you an email about this. Best regards, Rainer

WATCHPAX 4 has framelocked outputs and can drive 4 LED controllers. Having said this, I would not suggest using it for this scenario for two reasons: 1) WP4 can play 4x 4K with 30fps encoded as HAP but for more data throughput the internal raid is too slow. It will just deliver some 1GB/s max. or less. Running those files with internal looping or fading into 4 new videos somewhere might hit the border. 2) WP4 can not use cross-server sync since it does not have a sync board inside which would deliver a signal to the next graphic card.

Most codecs use the cpu for rendering (in WATCHOUT systems), except for h264 when using WATCHPAX servers with enabled hardware acceleration and for HAP which also uses the gpu. For large files and when you use many outputs with high resolutions from one server the memory size of the gpu can be an issue. I doubt though that any card with 6 outputs can really give you 6 output screens with 4K and 50/60fps. The gpu might be fast enough but there are plenty of bottlenecks in the processing chain of a server. Be aware that AMD graphic cards do not support more than three 4K outputs under Win7 due to a driver issue!

You will definitely have to pre-split the content to make it playable. How much you have to split depends on the codec, the frame rate of the video material and the performance of your WO servers. Splitting the content helps a lot to level out the load on each cpu core. For HAP the same can be achieved by using chunks, but you'll still have to reduce the number of pixels per stream. I would maybe go for two or three servers with 4 outputs each and use the outputs in portrait mode, i.e. each 4k output would play a file with 2160x3465px. Make sure to use a sync board to framelock/genlock the outputs. DO NOT render any transparent (empty) parts, e.g. do not render slices with 2160x3840 (because this is the UHD resolution) but render it to 2160x3465px as this is representing the real content size if using vertical outputs.

It's always wise to not change things in an uncontrolled way, but MPEG2 is rendered using a WATCHOUT direct draw filter. The same goes for other supported codecs. That's why you do not need to have QT installed on display computers anymore. Regarding the license key, if you own a WATCHOUT 6 license key, you will be able to use this with all versions 6.x.x. The latest version is 6.3.1 right now and there have been quite a few bug fixes in the early versions of V6 as well as some new features on the way. https://www.dataton.com/watchout-release-notes

Hmmm.... It's not really possible to change this through the timeline but what could help is to use different TIERS for the different screens. There is a setting in WATCHOUT to frame the displays and this will always frame all displays of the active tier and scale this up to the size of the stage window.

It looks like an EDID problem to me. It is not advisable to copy the EDID from the display to the EDID manager directly since this will copy all resolutions and frequencies at the same time. The graphic card discusses the optimal resolution with the display and if all frequencies are available, they will agree on what they decide is best. Our approach is to use an EDID manager which can be set to a fixed resolution/frequency through dip switches, a dial selector or software. Then the graphic will only see the resolution/frequency we want to use and will switch to that.

Are you sure that the DP adaptors are active??? Is it this model? https://www.mwave.com.au/product/ugreen-20401-displayport-to-hdmi-mf-converter-aluminum-ac07162 Does not state anything about "ACTIVE" and the price would rather indicate a passive adaptor.

Hi, When looking at the specs of your computer, I have some questions about your SSD? Is it the EVO or the PRO version of the M.2 960? The PRO would be the one to choose here. Is Windows residing on the same SSD or on a separate SSD/HDD? I do not know which mainboard you're using, where does the SSD sit? Flat on the mainboard? Is there a good airflow around the SSD? Those M.2 SSDs tend to heat up a lot when they deliver video data. This even more if you use large and less compressed codecs. ProRes files are usually rather huge and the same goes for HAP-Q. Out of those I would rather go for HAP-Q and not for ProRes. ProRes is a codec designed for post-production and not really meant for playback. You could also try H264/MP4 if encoded following the guidelines found in various threads in this forum. When the SSD heats up it will start to throttle, means it's performance will go down. We have some servers using the 960 PRO M.2 SSD but we are now using PCIe slot carriers with a big passive heat sink and we have an extra cooler (vent) blowing air onto it. Since then we did not see overheating anymore. Benoit's hint of cutting the large video into smaller portions might help. In fact it is not really the same when it comes to performance whether you use a 4K video or 4x fullHD files. Using 4 files will help since then it would make use of more cores instead of squeezing the whole video through one core. Of course, using chunks does the same in a way, but this only works with HAP codecs. Hope it helps Rainer

Hi Misterk, We have built a couple of servers using 1607 Enterprise LTSB and the update KB4057142. We did some additional changes in the group policies and scheduling but basically if you follow the tweaking guide you should be fine. I doubt that anyone in Sweden can tell you right now whether the updates from 11/13/2018 are working. Looking at the date, they are 24 hours old now. To make sure that an update works well under all circumstances you will need to build a master image first and then have a couple of people testing the system over a lengthy period of time. All recent WATCHOUT servers are using the master build based on version 1607 and it makes no sense for Dataton to have various operating software versions in the same sort of products.

NvTmMon, NvTmRep and NvTmRepOnLogon are not needed. They only spy on your computer. The other Nvidea items should also not be neccessary. Also I would not install all those Nvidea drivers, except if you use the 3D capability or want to do gaming. On my computer I only have the graphic and audio drivers installed. Apple software update has caused issues on some computers built by my clients in the past.

Search will not work if you turn it off in SERVICES as described in chapter 11. Drag and drop should work.

Create your show with your production software and go online. Test that it works from your production computer. To make things easier use a cluster name and a computer name instead of an IP address. Click onto the header of your stage window, then into one of your screen rectangles and then use the "Remote Access" coomand thorugh your menu or by sending the keys Ctrl+Shift+A Your display computer should now show a smaller screen with a menu bar. Go to File>Edit Startup Script and edit the auto start script. All lines starting with ";" are comments only. Search for the block ; authenticate 1 ; setLogoString "The show will begin shortly" ; delay 5000 ; load "MyShow" ; wait ; run and delete all ";". Then swap "MyShow" with the name of your project without the ending ".watch". Watchout is case sensitive! Save the file and restart the WATCHPAX without the production computer being online. It should now start WATCHOUT, show the cluster name, computer name and the text "The show will begin shortly" and should then load your show and run the main timeline. If you use auxiliary timelines, you will need to use run "name of your timeline" instead of just run. All described on pages 159 and 173 in the manual.

Have you checked your startup items (run shell:startup) to see whether there are any other things which run in the background stealing the focus? While tweaking a win 10 computer I also noticed that the Nvidea card installed a couple of scheduled tasks to run daily. I deleted those of course but I'm not sure whether your consumer card also has tracking or auto-update components which might be triggered once a day.

The output resolution has nothing to do with the capture card resolution. Selecting a capture card depends a bit on what sort of signals/players you plan to use, sending their output into the WATCHOUT system. If you use computers as sources and maybe semi-professional camera systems, a capture card with HDMI 2.0 should be fine. DVI does not support 4K! There are also capture cards supporting DisplayPort which also goes up to 4K. I have not seen many professional camera systems with 4K on SDI in the event and staging sector. This is due to the very high costs of professional 4K cameras, mixers, class-A monitors etc. So people tend to stay with what they have invested in for longer times. Also 4K@50/60fps is not (yet) a standard for TV due to bandwidth and infrastructure.

The Windows Antimalware is part of Defender. Should be the executable MsMpEng.exe. The servicegateway.intel.com is part of what you posted (see above). So it seems that your computer tries to connect to the outside and does things in the background which you do not really want.

I would take the following steps... a) connect the KISSBOX to the network and find it with your KISSBOX editor software. Check the IP address and port of the unit and take a note. b) use a telnet programm (e.g. Putty) and try to send the commands from there to the IP address/port of the KISSBOX. Use UDP to send the data! Probably you do not need to add the CR/LF (x0D) when using Putty since pressing the Enter key just adds this by itself. If it works from there it should also work from WATCHOUT.

While WATCHOUT can now output different resolutions from one server, it can not output different frequencies in one project. Could it be that your third output asks for a slightly different frequency? Have you tried using EDID minders behind all three active adaptors with identical EDID sets?

Hmmmm... Are there any services running in the background? Port 443 is not used by WATCHOUT, it's the https port. It tries to open a connection to servicegateway.intel.com This looks to me as a problem with insufficient tweaking. Things like malware detection and Microsoft Security Client should also not be active. I would suggest to go through the tweaking list one-by-one and disable everything which is not needed for very basic computing. Also delete any bloatware from the PC manufacturer if there is any installed. Rainer

What do the log files say? Check the log files in the logs folder inside your WATCHOUT folder on the display computer and on your production computer. Also check the Windows events logs. Culprits could be hardware issues, encoding issues or tweaking issues. I had the latter in a client's installation where some players would show such behavior exactly once a week. Here it was the Windows scheduler for QT updates which tried to search for updates once a week. (tweaking issue, though the client insisted that he fully tweaked his computers) It could as well be a problem with your RAM. Have seen such things before although the RAM was selected from the QVL of the mainboard manufacturer. Run a test with a good tool, e.g. MemTest, not only using the internal Windows memory test app.

Hi Songyost, The RS2TR basically works with simple UDP command strings ( like the IOCC units ) The datagrams are : - 4 bytes identifying the type of message / the command / the physical port ( “connector" 1 or 2 ) / the amount of following string of data bytes - X bytes containing the actual serial data string You set it up with the Editor software ( IP address / Port ) and set set it to Automatic UDP for watchout The 2 ports are identified by the 4 byte message header So to control "Fog Level 1" through RS232 port 1 you should probably send UDP $A8$00$00$05$3A$46$31$4E$0D The red byte defines the remaining bytes - could be $04, just try it. An extract of the manual is attached. It works for the 2 port version as for the legacy 4 port version. Hope it helps Rainer RS2TR.pdf

Why are you using two outputs to display 128 x 128 pixels? All LED processors should be able to grab a part of a larger resolution, e.g. , 1024x768, 1920x1080,... If you use only one output and distribute the signal to two processors to do what I described above, then it is the processor itself which is getting you off-sync.

WATCHOUT has a setting for frame blending in each media cue. Double click onto the name of the media in the Media window. Due to it's nature, frame blending will always look a bit blurry. If your projectors can work with 29.97 Hz, you should turn the frame blending setting off. Unfortunately there is no global switch for this in the preferences, has to be turned off in each media. BTW, the second setting there (hardware acceleration) only supports servers (WATCHPAX) built by Dataton.

I haven't tested to encode HAP Q Alpha yet but I can confirm that HAP Q Alpha is not supported under WATCHOUT 6.

I'm a bit confused with the displays and projectors. How many outputs are you using on your display computer? Which type of graphic card are you using? Are you using active adaptors from DP to HDMI/DVI? Some graphic cards, e.g. AMD Radeon Pro, can only use two outputs with passive adaptors. If you use more, you will need to use active adaptors! Sending a ping to the display computers will only check whether the computer is reachable over the network. WATCHOUT uses various ports, some of those being dynamically numbered broadcast ports. If your network blocks UDP/broadcast you will be able to ping the computer but not to use WATCHOUT. The "Test connection" button in Jim's mail (next to the greenOK field) is a better way to check the network. This button will check that WATCHOPOINT is on and that the ports are open.