Announcement: Our new CyberLink Feedback Forum has arrived! Please transfer to our new forum to provide your feedback or to start a new discussion. The content on this CyberLink Community forum is now read only, but will continue to be available as a user resource. Thanks!
CyberLink Community Forum
where the experts meet
| Advanced Search >
Project renders with wrong duration using SVRT - 11:11 instead of 11:54 .
Julien Pierre [Avatar]
Contributor Joined: Apr 14, 2011 01:34 Messages: 476 Offline
[Post New]
Cyberlink has been unable to reproduce the issue.

Indeed, I can only reproduce it on one of my systems, but not the other.

Packing the project and re-rendering it seems to work around and make the project work on the one computer that is experiencing the issue

It is still very weird.

I was able to reduce the test case to a smaller one.

-----------------------------------------------

I tried to reduce my project test case, and was successful.
I was able to reduce it to just 2 video clips, for a total of 338 MB.

In the shortened test, the normal "good" rendering should be 1min58sec, but when the bug is seen, the "bad" rendering is only 1:14 - 44 seconds short.
The shortened test case consists of just 3 files :

FrenchSuite6-svrt-short-problem2.pds
gigue-firsthalf.m2ts
gigue-secondhalf.m2ts

They were originally located in D:\Documents\Videos\Full French suite 6 harpsichord
along with the other PDS projects and other m2ts source files which aren't needed by this shorter project.

I then prepared to pack the project to upload this much smaller test case. I packed it on my desktop .
The packed project has the same 3 files, but the PDS file is modified.

I thought I would try to render the packed project from the desktop as well before uploading it to your ftp site.

And lo and behold - it rendered fine, without skipping a beat, for the full duration.
However, before the packing procedure, the smaller project still shows the issue.

I tried to do an XML diff between the unpacked/packed PDS projects, but couldn't really figure out what differences are making the non-packed project fail to render properly vs the packed project.
I do see that before packing, there are references to the project location on my D: drive, and after packing, these absolute paths are gone. I don't think these explain everything, however.

While packing is a workaround for the problem, it is not a real solution. I still would rather not have to do this in the first place.
I only ever edited the project file with PowerDirector - I never did any manual edit.
It got in that state somehow because PowerDirector wrote something corrupt in it.
It is possible that I moved, copied or renamed a PDS from one location to another with the Windows file manager. I would not expect that to cause problems, though.

So, I kept experimenting to try to find why the original project has this problem before being packed.

1) just copied the 3 files of the test case to a folder on the desktop on my C: drive, but without packing it. It also rendered incorrectly.
So, the packing process in PowerDirector is somehow fixing something that's wrong in the original project file.
2) after doing 1, I renamed the original project folder
D:\Documents\Videos\Full French suite 6 harpsichord
to
D:\Documents\Videos\Full French suite 6 harpsichord-renamed
3) I then rendered the test case located on the desktop again.
This time, it rendered properly, likely because it could no longer reference any source video files from
D:\Documents\Videos\Full French suite 6 harpsichord and instead picked them up from the folder on the desktop
4) but then, I opened the project in
D:\Documents\Videos\Full French suite 6 harpsichord-renamed and tried to render it.
The bug was still seen, even though the actual source file location had been renamed.
I did not expect this.
5) I then copied the 3 test case files from the desktop on the C: drive that were rendering correctly in step 3, to a new folder on the D: drive.
And I rendered again. And this time, the bug was seen in the rendering !
I don't really see why that is the case, but there it is.

This means PowerDirector is rendering the same project differently based on whether the same files are located on the C: or D: drive, even though in both cases
the location of the original source files, D:\Documents\Videos\Full French suite 6 harpsichord, has been renamed, so the PATH references in the PDS files are not good
anymore, which forces PowerDirector to look for them in the same directory as the PDS file.
6) no matter what I tried, either with the original large or with the smaller test case, I could only reproduce the issue on my main computer,
on which the project was originally created. Trying to render on another computer, even using the exact same original path location of
D:\Documents\Videos\Full French suite 6 harpsichord on its local disk, failed to reproduce the rendering problem on that system.
Since the using the same path location doesn't trigger the issue on the other machine, I'm thinking perhaps the original disk signature is stored somewhere in the PDS file,
and is being checked when the project is opened, triggering different behavior in PowerDirector if it doesn't match.

Because of this, I'm not really confident that you can reproduce my test case on another machine.
I would still love to see the issue truly solved, rather than just worked around with the packing feature, as the bug has caused me to waste so much time.

I have still uploaded the (unpacked) shortened good and bad projects. I'm in the process of uploading the good and bad renderings as well - this will take more time.
The location is ftp://ftp.gocyberlink.com/vcpriya/CS001513122/ .

Even if you can't reproduce the issue on your system just yet, at least you can watch the good and bad renderings and see what I have been complaining about.
Perhaps you or your developers will have an idea about why the issue occurred in the first place, and can identify the part in the packing process that causes it to
go away.

---------------------------------------------------

Here are some of my notes from the case :

bad.m2ts was generated from project FrenchSuite6-svrt-shortproblem2.pds

good.m2ts was generated from project FrenchSuite6-svrt-short-good

The only difference between these two projects is the title track .
The first project has the title track enabled, while the second does not.

Indeed, bad.m2ts and good.m2ts have different length. That's part of the bug and why I opened this case ! In the bad rendering, the video jumps too early to the second video clip, and then everything gets cut off at 1:11, when it should last until 1:58 in the good rendering.

I have done a screen capture to show the problem when rendering and play back both the bad clip and then the good clip second.

As I said, this bad rendering issue is only reproducible on my own computer. FrenchSuite6-svrt-shortproblem2.pds renders correctly with the proper length on another computer . So, I don't know if you will be able to reproduce it either. Even if you can't, the capture shows you that it's happening on my computer. Hopefully, product development engineers will have an idea of why that is the case.

I did a PD14 screen capture as you requested. I used voice over, but somehow the tool did not work properly and my voice is cut every second or so and is pretty much unintelligible, even after raising the levels 12 dB in post-processing.
However, you should still be able to see what's going on in PowerDirector. Also, my monitor is 2560x1600 and the capture can only go to 1920x1080 so it is not 1:1 .

In the screen capture, I did the following.
1. open the FrenchSuite6-svrt-shortproblem2.pds
2. render it to bad.m2ts
3. play back bad.m2ts and show that the video and audio are out of sync, and it cuts off early at 1:11.
4. disable title track in the project
5. render again to good.m2ts
6. play back good.m2ts and show that the video and audio are in sync, and it ends correctly at 1:58

This screen capture has been uploaded to your ftp site at :
ftp://cyberlinkcs:7J9CUMKG@ftp.gocyberlink.com/vcsoro/CS001513122
MSI X99A Raider
Intel i7-5820k @ 4.4 GHz
32GB DDR4 RAM
Gigabyte nVidia GTX 960 4GB
480 GB Patriot Ignite SSD (boot)
2 x 480 GB Sandisk Ultra II SSD (striped)
6 x 1 TB Samsung 860 SSD (striped)

2 x LG 32UD59-B 32" 4K
Asus PB238 23" HD (portrait)
Dafydd B [Avatar]
Senior Contributor Joined: Aug 26, 2006 08:20 Messages: 11973 Offline
[Post New]
Hi Julien,
Please contact online support, your issues are being looked into by CyberLink. Anything you message CyberLink, will land on the "right desk".
Dafydd
Julien Pierre [Avatar]
Contributor Joined: Apr 14, 2011 01:34 Messages: 476 Offline
[Post New]
I have, all my cases are in "open" state and I'm waiting for their update. I cannot send anything more on the cases - and I don't have anything more to send. I just hope there will be a proper diagnosis. I bought PD14 on september 17 so I'm still in my 30 day period to get a refund. All 6 cases were submitted during the first week of ownership. Given the lack of progress on all 6 cases, I'm considering asking for a refund on day 29 if there is no progress. MSI X99A Raider
Intel i7-5820k @ 4.4 GHz
32GB DDR4 RAM
Gigabyte nVidia GTX 960 4GB
480 GB Patriot Ignite SSD (boot)
2 x 480 GB Sandisk Ultra II SSD (striped)
6 x 1 TB Samsung 860 SSD (striped)

2 x LG 32UD59-B 32" 4K
Asus PB238 23" HD (portrait)
Dafydd B [Avatar]
Senior Contributor Joined: Aug 26, 2006 08:20 Messages: 11973 Offline
[Post New]
Hi Julien,
I'm sure you'll pick up on the new beta patch that's been released. It might be worth you checking it out. I'm hopeful it'll fix the issue I have.
http://forum.cyberlink.com/forum/posts/list/45816.page#236780
Dafydd
Julien Pierre [Avatar]
Contributor Joined: Apr 14, 2011 01:34 Messages: 476 Offline
[Post New]
Thanks, Dafydd, I downloaded and installed it already. Now I'm testing one of the bugs that Cyberlink supports emailed me about which they said was fixed . It is for CS001514573 , the "audio only" rendering issue. Rendering is still going on. MSI X99A Raider
Intel i7-5820k @ 4.4 GHz
32GB DDR4 RAM
Gigabyte nVidia GTX 960 4GB
480 GB Patriot Ignite SSD (boot)
2 x 480 GB Sandisk Ultra II SSD (striped)
6 x 1 TB Samsung 860 SSD (striped)

2 x LG 32UD59-B 32" 4K
Asus PB238 23" HD (portrait)
Julien Pierre [Avatar]
Contributor Joined: Apr 14, 2011 01:34 Messages: 476 Offline
[Post New]
I can confirm this particular issue is not fixed in the latest patch. I did not expect it to be, given that Cyberlink could not reproduce it on their end - and indeed, I can only reproduce it on one computer.

I do have a workaround for it though, which was pack the project and then copy the packed project file back over the original - that resolved the issue.

Nice to see the SVRT has been upgraded from 3 to 4 from PD14 release to this PD14 patch .

Guess they forgot to put the SVRT4 logo in the release they had rushed out the door .
MSI X99A Raider
Intel i7-5820k @ 4.4 GHz
32GB DDR4 RAM
Gigabyte nVidia GTX 960 4GB
480 GB Patriot Ignite SSD (boot)
2 x 480 GB Sandisk Ultra II SSD (striped)
6 x 1 TB Samsung 860 SSD (striped)

2 x LG 32UD59-B 32" 4K
Asus PB238 23" HD (portrait)
Powered by JForum 2.1.8 © JForum Team