I made a better remote-live-view page [OLD]

[bp2008]

None of the camera cycles are configured in UI2 - that is entirely done in Blue Iris. After that, in UI2 you just need to choose a cycle in that dropdown list under Current Camera. If it cycles at the wrong speed, it is because Blue Iris is changing between views at the wrong speed, and it has nothing to do with UI2. Reloading the clip list won't help -- not sure where you got this idea from. By the way, you reload the clip list by clicking either of the two little buttons below the word Clips:

Anyway I'm on the latest version of Blue Iris and I can confirm the cycle timing is correct in this version.

 

[dashony]

None of the camera cycles are configured in UI2 - that is entirely done in Blue Iris. After that, in UI2 you just need to choose a cycle in that dropdown list under Current Camera. If it cycles at the wrong speed, it is because Blue Iris is changing between views at the wrong speed, and it has nothing to do with UI2. Reloading the clip list won't help -- not sure where you got this idea from. By the way, you reload the clip list by clicking either of the two little buttons below the word Clips:

Anyway I'm on the latest version of Blue Iris and I can confirm the cycle timing is correct in this version.

Thanks bp2008,

I thought the setting I needed to change for cameras cycle timing was located here . But after your reply, I've dug a bit deeper in the actual Blue Iris and I've found where to change this: by clicking on the cog and then changing the seconds in the box from the screenshot attached.
Thanks for confirming this setting can't be adjusted via UI2 so I stopped looking there and found the setting in the right spot.

 

[J0akim93]

I just made an new user and it don´t work. Locally it works..

I´ll test that combine or cut again. Because last I tested it I thought that it did work normally in blueiris locally. But I'll test and report back.

Now I tested this again. so here's an example. One camera records from 03.49.58 to 03.54.50 so that's 4minutes 52 seconds. And locally in blue iris this clip is 22 seconds long. And in the default.htm page it is 22 seconds long. But somehow the ui2 page makes this 4minutes 52 seconds long. And that 22 second long clip runs again and again until 4min and 52sec has gone.. Do you understand what I mean?

I Would love to use the ui2 page but that's quite frustraiting.. Doesn't anyone else have this issue?

 

[bp2008]

Now I tested this again. so here's an example. One camera records from 03.49.58 to 03.54.50 so that's 4minutes 52 seconds. And locally in blue iris this clip is 22 seconds long. And in the default.htm page it is 22 seconds long. But somehow the ui2 page makes this 4minutes 52 seconds long. And that 22 second long clip runs again and again until 4min and 52sec has gone.. Do you understand what I mean?

I Would love to use the ui2 page but that's quite frustraiting.. Doesn't anyone else have this issue?

This sounds like you are letting Blue Iris combine multiple alerts via the Combine or Cut option. I believe Blue Iris reports the duration of these clip as the full time between first alert and last alert. So even though there is only 22 seconds of video, UI2 thinks it is 4 minutes and 52 seconds because that is what it was told. The way Blue Iris's web server works, when you ask for video beyond the end of a clip, it just starts over. And that is why it loops.

I will try and remember to test this again myself and maybe ask Ken (the developer of Blue Iris) to fix it.

 

[J0akim93]

This sounds like you are letting Blue Iris combine multiple alerts via the Combine or Cut option. I believe Blue Iris reports the duration of these clip as the full time between first alert and last alert. So even though there is only 22 seconds of video, UI2 thinks it is 4 minutes and 52 seconds because that is what it was told. The way Blue Iris's web server works, when you ask for video beyond the end of a clip, it just starts over. And that is why it loops.

I will try and remember to test this again myself and maybe ask Ken (the developer of Blue Iris) to fix it.

Ah okay that makes sense. That would be great because I like ui2 very much but this detail makes it really hard to use. I tried to have blueiris to make clips that are as long as the movement and then autoplay with ui2.. But that wasn't so smooth. So i'll use the default page.

 

[bp2008]

Ah okay that makes sense. That would be great because I like ui2 very much but this detail makes it really hard to use. I tried to have blueiris to make clips that are as long as the movement and then autoplay with ui2.. But that wasn't so smooth. So i'll use the default page.

I just tested Combine or Cut, and the duration is correctly reported so the clip does not repeat. What version of Blue Iris are you running? If you are really out of date, that could explain why you are seeing this bug.

 

[erkme73]

BP, is it possible the UI2.htm is causing the BI server to crash? I've got Ken logged in to my server with Teamviewer right now, and I noticed that when I refresh the UI page, the server throws up a crash screen. Could just be a symptom of a bigger problem (with x.x.9.4), too...

 

[bp2008]

Possible, certainly. I've known it to happen on some systems when the clip preview animation in UI2 is enabled and you mouse over the clips really fast, causing a bunch of animations to try to start and it overloads BI and causes a crash.

 

[erkme73]

Interesting. It's happened twice now, when I reload the page, it kicks the server to the curb. Ken is still logged in via TV, but I haven't said anything to him yet.

I tried what you described, but simply switching from alerts to clips caused the . Hmmpf... I didn't have this before the latest upgrade. I wonder if it isn't something that he changed on his end that makes the UI a bit less compatible.

 

[bp2008]

Hmm. I didn't update to the latest BI minor revision because I read it caused crashes often for someone.

 

[erkme73]

Yes, and I had to roll back to 9.0 (thanks to your dropbox link, btw). Ken has been logged in to my server all day with BI trying to hunt down the bug. But, I'm now able to cause these crashes pretty consistently by toggling views within UI. What is confusing (to me at least) is that I've never had UI2 cause crashes. It's only after the update on Ken's side. I let him know just as another clue to the puzzle. For now, he's disabled the webserver completely to see if the crashes stop. I'm quite surprised to see how painstakingly slow the diagnostic process is for Ken. He's literally turning one camera on/off to see if the crashing stops. I would have thought there was some master black box voo-doo magic dohickey that he massages a certain way, and it tells him what the problem is. Guess not.

 

[bp2008]

Remote debugging is difficult. He can't very well set up his development environment on your machine and actually step through code to see what is wrong

 

[erkme73]

That's true. He did take my reg export and is putting that on his local machine for testing.

 

[bp2008]

I just installed 4.3.9.4 for testing and I can't make it crash. Granted it has only been a few minutes, but it looks just as stable as ever for me.

 

[erkme73]

Well, that makes sense. If there's a problem to be had, I usually take it for the team.

 

[nejakejnick]

I can crash 4.3.9.4 and earlier easily by just hovering over clips for preview, I already reported that 2 weeks ago, Ken said that is should be fixed in 4.3.9.5

 

[erkme73]

I can crash 4.3.9.4 and earlier easily by just hovering over clips for preview, I already reported that 2 weeks ago, Ken said that is should be fixed in 4.3.9.5

Correct. It was crashing on 9.3 and 9.4. I was unable to crash it by hovering on 9.0 or earlier. The 9.5 did fix it. No crashes after 24 hours. And I've been riding it extra hard to make sure.

 

[bp2008]

Correct. It was crashing on 9.3 and 9.4. I was unable to crash it by hovering on 9.0 or earlier. The 9.5 did fix it. No crashes after 24 hours. And I've been riding it extra hard to make sure.

Good to hear!

 

[nejakejnick]

I have a problem caused by high load time of an initial clip image. I store everything but DB on Synology NAS DS215j with WD Purple 4TB, and the PC is connected via 100MBit Cisco router (unfortunate part of a cow milking robot solution - I am working on getting a newer gigabit model which is actually way cheaper :mad-new:, and also my notebook is quite slow). So the system has not a perfect performance and the result is this:



The first image takes way too long to load, and then it is immediately replaced by a second one, however the latency is not taken into the account so it can easily skip a half of the 10s clip, forcing me to restart the playback to see what actually happened.

I think it would be better to load the first image and then start the playback, not the other way around.
Also notice that the first image has not time 0 (not always).

Thanks.

 

[bp2008]

What an unexpected problem! Normally the first frame doesn't take significantly longer to load than any other. Try turning off "Clip preview animation" under UI2 options > UI Behavior as that can reduce the burden on Blue Iris and it may improve the response time when it really matters.

I can't properly "load the first image and then start the playback" because when the first frame is loaded, it is immediately time to start the second frame. I must know what time value to ask for. Normally, this is simply the previous frame's time value plus the time it took to load the previous frame, but if I can't rely on the first frame's loading time being representative of the second frame's loading time, then what do I do? I can limit the time difference between the first and second frames, but then people with slow network connections would experience clips that play back at an inconsistent rate for the first few frames. Admittedly that would be a minor issue, but I think it would affect far more people than your issue.

The reason the time of the first image is not always 0 is that there is quite a bit of script being run between the moment when the playback clock is zeroed and the moment when the first frame is actually requested. I can certainly move the logic around a little to mostly eliminate the oddity, though it won't have much real-world effect.