Thomas Leong

I have not had a major problem with WO6.1 working with the 2 capture cards I have (Magewell Xi100XE-Pro, and Decklink Studio 2 with Decklink 10.3.7 driver). Perhaps you guys should try different versions of drivers, one by one from v10.0.0 upwards. The little niggly problem I've had is when using Display PC as a standalone cluster with the Magewell card in a MSI micro motherboard with Win7 64-bit. The cluster works fine under Production PC control. As a standalone started with the autostart.txt script, the live video (HDMI capture of my laptop) comes on for 9-10 secs, then switches off to black. Putting the Magewell in my other Display PC with an ASRock Extreme 6 mobo with Win7 64-bit, all works fine in standalone mode with the same autostart.txt script. No blackout after 9-10 secs. Plays all day! After comparing the 2 Display PCs in detail, and even un-installing a couple of Microsoft C++ Redistributable 2013 in the MSI which was not present in the ASRock, comparing NIC settings, etc...I could not find a reason for the blackout in the MSI versus no blackout in the ASRock. Eventually, I stumbled on a solution: CTRL+W twice on the Display PC. The first to go into remote mode. The second to go back to fullscreen mode. Thereafter, the standalone cluster would capture live all day! Not that I'm going to use it as such very often, but could be a bug that should be solved for the museum and rental market: a blackout of live video capture in standalone cluster mode after 9-10 secs. BTW, happened also with WO6.02, but not with 5.5.2. On the other hand, did not occur with a second similarly configured Display PC using the same Magewell capture card and driver. Don't have a Datapath on hand to test but shall try it with a client's Datapath setup soon. Thomas Leong UPDATE 17 May 2016 My 'niggly problem' has been solved. My own fault!! Because the default cue duration is 10secs when added to the Timeline, with no pause cue inserted, it ran out of media! Easy to overlook when running in standalone cluster with no view of the Timeline possible! Thomas

The Dell Latitude E6500 is a laptop, albeit a reasonably equipped laptop. Since you said you are running 6 display pcs, are you saying you are using 6 of the said Dells (one output each)? Not sure if this could be the cause, but if so, check that each Dell has the NVidia Quadro NVS Graphics card selected as THE Preferred graphics card to use instead of the on-board Intel® Graphics Media Accelerator 4500MHD. You do this via the NVidia Control Panel found in Windows Control Panel.

Apparently, if need be, one can install Qucktime without the Player - https://forums.creativecow.net/readpost/2/1072246 Basically, the procedure (copied and pasted here) is - 1. In your add / remove programs, uninstall QT player completely. 2. Go to the apple QT installer download page and grab the latest package for Windows 3. During install, make sure that everything except the "critical components" is NOT installed. (Xed out). Leave the top option for install only (something like "media for third party apps" - this is the codecs only) 4. Install as normal

Maike, Please download CrystalDiskMark (Std Edition Portable w/o Ads), and run it through your machine(s) to get the Sequential Read speed of your RAID or wherever Watchout is installed on. If your RAID comprises of all 8xSSD, your Seq READ speed should be plenty fast. BTW, what RAID are you using? RAID 0 or one of the others - 1, 10, 5? Are you able to re-configure 4-8 of the SSDs to a RAID 0 and test the Seq READ speed? The other potentially contributing bottleneck is the speed of your Xeon cpu since you say you have an i7-4960X which does not stutter as much. As a comparison, my system is: - i7-4770K cpu in turbo mode at 4.2GHz - ASRock Z97 Extreme6 motherboard - 32GB 1600MHz RAM - Win7 Pro 64-bit OS on a Kingston SSD with Watchout 6.1 installed - Watchout ShowsPath pointed to a Samsung SM951 M.2 PCie 3.0 x 4 with CrystalDiskMark seq. read speed of 2,252MB/s - AMD Firepro W4100 with 2GB VRAM and it will run smoothly: 1 x 7680x1080 HAPQ 0 @30fps over 4 1920x1080 outputs 1 x 4096x1080 H264 @30fps scaled to 4 1920x1080 outputs (I have no ProRes large resolution file to test; I only have one 1920x1080 ProRes file for these tests) Also running smoothly/without stutter are - 4 x 4 HAPQ 0 1920x1080 files @50fps ('4 x 4' means 4 files with 20-30sec cross-dissolve to another 4 files) 4 x 4 H264 1920x1080 files @50fps (If I have my SM951 M.2 on a PCIe 2.0 x 4 slot with Seq READ speed of 1,570MB/s, these H264 will stutter during the x-dissolve) 1 x 1 ProRes 422 1920x1080 files or 2 instances of one ProRes file with staggered start (eg 2 secs apart), or 4 x 4 instances of 1 ProRes file started simultaneously out to 4 displays. Thomas (BTW, these old eyes mis-read your Intel DC S3500 SSD as the P series of NVMe drives).

Duh!...missed the "Intel@ SSD DC S3500 Series 480GB" in your equip details. That's where Watchout should be installed, or 'Shows' folder directed to.

Well, your Xeon E5-1630 V3 seems almost equiv to the i7_5820K except for having 2 cores less. I have neither, nor do I have the HAPQ quality 0 files to test, so it is difficult to say the cpu is lacking in this case. However, in my opinion, for what you have invested in your Display PC, the more capable i7_5960X would have been "more fitting". Here is a shot in the dark, if you have not already done so : Ensure Watchout is installed on the RAID array, not on the single 1xSSD Samsung 850 Pro. Either uninstall Watchout from the 850 Pro, and re-install on the RAID, or Refer: http://forum.dataton.com/topic/1986-show-storage-location/to direct shows to the RAID. Additionally, you have 16GB RAM. Are these in 4GB modules such that you are using the quad channel specs for your motherboard? If in 8GB modules, you'd only be on dual-channel and this may affect your performance versus quad-channel RAM. For the future, if is matters to you...there are 2 Ultra M.2 slots in that mobo, and I would have gone for 1 or 2 of the Samsung SM951 512GB, or the new Samsung 950 Pro M.2 instead of the SSD RAID. In my opinion, the PCIe Gen3 x4 interface is faster than the SATA 6GB/s interface even in RAID mode. Thomas Leong Addendum: upgrade to Watchout 6.1. It has solved some HAP problems, more blue-screen HAP than hiccups on cross-dissolve. But it may still help.

From the X-keys website, seems that they have 2 versions: TruCom or Virtual Com. Both would essentially communicate with a Watchout PC via serial port (RS232). So you would need a RS232-TCP/IP converter installed in the Watchout PC which would convert any serial input signals to TCP/IP which Watchout can understand. I have used the single port freeware version of HW-VSP3 from http://www.hw-group.com/products/hw_vsp/index_en.htmlnot with X-Keys but another serial keypad device by Genovation. Just direct the serial port for the X-Keys to TCP port 3040 of your Production PC. (hint: You have to 'login' with the default password to make any changes to the Virtual Port). In my opinion, with today's PCs/Laptops, the Virtual Com version with USB connection would be more convenient. The TruCom version would need a RS232-USB adapter. Thomas Leong

Good question...but in this case, it does not seem to make a difference. '-ShowPath' and '-ShowsPath' will both work. Bear in mind that the option is only available to Display PCs. From one of my earlier posts, it did not work for me because I tried it on a Production PC shortcut. My boo-boo, then. Wiser now . Thomas Leong

This might help, from John Huntington's blog on Jan 7, 2014 - Controlling an ETC Ion Console Over an IP Network and from Dataton Watchout(If the link above does not work, go to January 2014 in his blog) Thomas Leong

Thanks, Fredrik. However, I have found that CTRL-R is necessary (and Ctrl-D to update the Displays) to refresh the media after any change to the 'Video Standard' in the Live Video window, else nothing happens and it looks like the change does not work!

JJ, that is a brilliant idea! The Corner tween works as a simple keystone adjustment applied either on the virtual displays or the media content itself. Thomas

I just tried your problem with the same Korg NanoKontrol2 and you are correct. Solution is to enable 'Jump to Time 0' in the same run cue. Works!

JFK has kindly offered to host my images on his server, so here they are - used to test capture from Decklink Intensity (old version) and Decklink Studio 2 (also old version). Source in all cases was my Windows 10 laptop via HDMI. SDI capture was also tested with DataVideo DAC-70 and Decimator to convert the HDMI source to SDI. Step 1: Step 2: Step 3: Step 4: Watchout Preferences Step 5: Watchout Production PC (PS: Post Modified with added Screen Captures...fingers too fast last night!) Thomas

I'm not familiar with the Blackmagic converters, but do they output a known and acceptable resolution to the Intensity Pro and hence to Watchout? I'm not able to post any images here, so shall pm you with some screen capture procedures I've done to see if it works for you. Thomas

Found another workaround for the moment: Re-sizing the Geometry window (hit-and-miss) until the aspect ratio is approximately there. Thomas

Well, looking at the template submitted, despite his 'request' there does not seem to be any keystoning required. If there are small bits here and there, it could be fulfilled with a mask(s) from within the respective Watchout Displays.

If you are still having problems capturing to Watchout, then I'm quite sure a Decimator MX-HD Cross Converter (or Datavideo DAC-70 - more expensive) will solve the problem for you. Thomas Leong

Not possible to see if there is any distortion of the mask with v5.5.2 since the Geometry tab in this version does not super the grid over the media. A simple Circle mask with 'Suppress Rendering' enabled illustrates the distortion clearly in the Geometry tab. Happens only in the Production PC. Does not distort in the Display PC. But then, if the mask is distorted, how would one go about accurately correcting geometric points with the Production PC? My workaround for the moment is to not use the 'Suppress Rendering' feature and have instead 2 masks of the same object, one the negative of the other. Thomas Leong

Should be do-able with Watchout 6's Virtual Displays where - - Virtual Displays 1 & 2 and possibly also 3,4,5 & 11 are assigned to one Actual Display (If so, then may need a mask because of the 'overlap' with 6&7) - Virtual Displays 6,7,8,9,10 & 12 to a 2nd Actual Display (otherwise, it is 3-12 to this 2nd Actual Display; a bit much depending...) - Virtual Displays 13 & 14 to a 3rd Actual Display ...and so on, ending up with 5 Actual Displays, 6 if 13 and 14 are assigned to an Actual Display each instead of sharing. How 3-12, and 15-24 are grouped and split will probably be critical depending on the hardware capabilities, and the respective resolution of each video. Alternatively, if the width of #11 (and #16) could be shrunk down to meet the vertical edge of 4 (#23), this would simplify things a lot as 1-5 & 11 (16 & 22-26) could share an Actual Display with no complications. Thomas Leong

Watchout v6.02. I have a mask (of a car's outline; black car on white background) with 'Suppress Rendering' enabled and Image Blend Mode = Normal; on Layer 2 in the Timeline. In the layer below (Layer 1), I have a red solid as background with a setting for 'Masked by Layer Above: Luma Mask' or 'Luma Mask Inverted' as the case may be. Both are 1920x1080 resolutions and positioned to fill a Stage 2D Display of 1920x1080. When 'Suppress Rendering' is enabled, the Geometry tab shows a distorted mask (the black car) regardless of Geometry being None, Perspective, Horizontal, Vertical, or Full. The distortion can be described as taller, less wide, and Y position moved down by about -162 pixels (owing to the distortion; i.e. instead of the black car's wheels at the lower bottom of screen, the black car has been displaced by about -162 pixels vertically such that the wheels are no longer visible in the Geometry window). The Mask tab, however, shows the mask in its proper proportion with no distortion. Toggling between Geometry and Mask tabs, it seems that the aspect ratio of the Display (and the media contents in the Display) changes under Geometry. Is this a bug? Have not tried earlier versions or v5.5.2 yet. Thomas Leong

In WO Production, if 1080i2997 does not work, try 1080p2997. Also, always CTRL+R to refresh/update the Media whenever you make any change to 'Video Standard' in the Live Video window of Production.

Sydney, for the latency test, my friend connected the Decimator direct to his laptop's HDMI output. Likewise with the DAC-70. Watchout was not involved. If used in a blend setting, using different branded converters may introduce more problems than necessary. So we have not tried a blend scenario yet - don't have enough Decimators...yet. Looking into getting more.

To put an end/solution to this thread, 3 units of Watchpax were delivered to site last week, and after some initial troubles getting a proper connection to the LinSn Sending cards (set at 1920x1080@30Hz), it was finally discovered that only PASSIVE mDP-DVI adapters worked between Watchpax and the Sending Cards, with Watchout's Preference set at 30fps. The Sapphire active mDP-DVI adapters were unable to get a fix on the 1920x1080 resolution required by the LinSn Sending Cards.

Bluetonesblue said: "I like DataVideo's DAC-70 which can take an HDMI input, and spit out SDI and HDMI at the same rate/resolution. I like this, as the graphics card sees a static handshake, while the SDI can be 100 m / 333 ft without loss, and you can connect/disconnect as much as you need to without resetting the graphics card and screwing up the settings." The Decimator MX-HD Cross converter is similar. Seems to have an 'EDID Manager" built-in. Disconnect the output end, and Windows does not even blink an eye. Decimator is also about half the price of the DAC-70. A friend/Watchout user recently did a latency test between the DAC-70 and the Decimator. If I recall right, the DAC-70 had 2 frames of delay from original, and the Decimator 3 frames. Thomas Leong

One alternative is to have the panel of buttons built for you. I know an engineer friend in London who builds panels, switches for RS-232 out (which would work with Watchout). If you want to discuss this further direct with him, let me know and I'll pass on his email to you.