Messages in this topic - RSS

Home ? Tech Help ? lip synch off after rendering

the topic is closed
09/07/2013 03:31:52

synovia5
synovia5
Posts: 3
just finished my first project. a 7 minute, two person dialogue. created tts .wav files for each text component using ivona and then pieced it together into one .wav file using audacity. set alternating speak/hushes and it seemed to work right when i ran it by timeline, but after i rendered it, the lip synchs didn't switch properly, getting increasingly off after the first minute. any ideas?

gb
permalink
09/07/2013 04:18:28

urbanlamb
urbanlamb
(Account inactive)
Posts: 1796
Hi yes this is a known issue it tends to get worse the longer the clip is and if you start altering the clip half way through the timeline after you do the shush/talk part. My only suggestion is try to leave the shush/talk till last and when your working on that part always start from the beginning of the timeline. Unfortunately even that doesnt always work well...so I find I have to redo dialogue inside a video editor the timing is okay its simply the sound becomes disjointed. So if you have saved your dialogue in a seperate file then you can redo it afterwards if you have a video editor. This is on muvizu's "todo list" to try to optimize this a bit better
permalink
09/07/2013 13:29:20

NeilExperimental userMuvizu staff
Neil
Posts: 396
I think this has been fixed in our next release, which should happen in the next few weeks.
permalink
10/07/2013 00:10:51

synovia5
synovia5
Posts: 3
thanks to both of you for responding so quickly. saved me hours of probably unproductive retooling. i love the product and can't wait for the next version.

i learned a few things on this project so far:

1) with that many talk/shush alternations, setting cues on the time line and then synching to the them while editing turned out to be very helpful.

2) this time, i started out by making many (like 57) wav files for each alternating voice that had to be pieced together. (there was another tts program that offered "round robin" but the voice quality was not as good as ivona.) i think a better strategy would have been to have ivona read my script document first in "voice one" and then in "voice two." this would have output 2 medium sized wav files. then, the two files could have been read into audacity as separate tracks in the same file. audacity makes it very easy to piece chunks together using cut and paste. there seemed to be enough space at the end of paragraphs to make the distinctions clear. the zoom feature is helpful for this.

things i still need to know:

1) is there an optimal screen resolution/compression scheme?

2) even a small screen resolution avi was 1.3 gigs. will that be a problem for uploading to a personal web site, your web site, or youtube?

3) does the pay version improve quality or just remove the watermark? I would be happy to pay even with the watermark - you guys deserve free advertising!!!




gb
permalink
10/07/2013 01:16:10

urbanlamb
urbanlamb
(Account inactive)
Posts: 1796
for your avi files these are basic codecs this is why the size is so massive if you want a better compression you will need to either install some codecs or edit it and rerender it in a video editor. Muvizu play has no price tag hence no fancy codecs or they would have to pay a license fee per user . For the moment there is no pay version also the only pay version is the option to remove the watermark via the interface for making the video if you open it up there is a no watermark you can purchase by the minute. They have mentioned a few times that they will eventually be releasing a pro version (paid version) but for the moment the only option to remove the watermark is to pay per the minute which unfortunately for many projects is not practical.

anyhow if you want to install some codecs everyone here uses klite
http://www.free-codecs.com/download/K_Lite_Codec_Pack.htm

make sure to use the one that goes with the version of muvizu you chose (32bit or 64bit) or there are some strange results although somewhere i read that if you have 64bit you need to install both 32 and 64bit codecs but then i found some contradicting posts on that one.. so take this last piece of advice with a grain of salt .. obviously though you dont want 64bit codecs on a 32bit system
permalink
10/07/2013 07:06:16

ukBertyMuvizu mogulExperimental user
ukBerty
Posts: 976
Synovia,

I can report that the sync does seem a lot better in the up coming version.

On your other points I know that a couple of us have avoided the whole codec issue by just outputting in uncompressed format. You do get enormous files but I never upload directly from Muvizu as I always want to use an editor (in my case Sony Vegas). I then output to mp4 from Vegas to have a more reasonable upload size.

A 7 minute talking head piece in one shot could be a little dry. If you split this up by using some single head shots then this would enable you to resync the whole thing at that point, plus your individual files would be more manageable.
permalink
10/07/2013 13:37:16

WozToonsExperimental user
WozToons
Posts: 494
There is an easier way to do dialogue without using talk/shush that works very well.
Let's say you have two voices for simplicity. We'll call them voice1 and voice2.

Record the conversation with both voices, load into Audacity. Duplicate the tracks twice, one for each voice and name them voice1 and voice2 ( or whatever you like ).

In Audacity silence all of voice2 in voice1 and all of voice1 in voice2. Don't cut just select each piece of dialogue and use "silence audio".

Export both voice1 and voice2 and import into Muvizu. Set each track as dialogue for the appropriate character and set both characters initial state to talk.

Problem solved.

Hope this helps.
permalink
23/07/2013 22:19:12

synovia5
synovia5
Posts: 3
thank you all.

the audio strategy worked fine. made the mistake of trying to save it before splitting the tracks (it automatically combines tracks on the save) so had to do it a second time. but after that, worked perfectly even after the render.

loaded dvix and it automatically installed into muvizu. how cool is that? after playing with various screen sizes/compressions, i was able to get something working at a reasonable file size - between 2.5 and 5 megs.



i have temporarily posted onto a blog page until i figure out my bigger strategy. will use the watermark removal if ever go commercial. if anyone wants to take a look and critique it, i would be happy to hear your comments. if you think it would be of interest, i will put it up on your site.


http://berkaholic.blogspot.com/2013/07/blog-post_20.html


gb

permalink
24/07/2013 08:08:35

Mike_Num_5
Mike_Num_5
Posts: 165
Remember, you can upload your videos on to our site as well for other members of the Muvizu community to view also!
permalink
24/07/2013 10:05:35

MrDrWho13Muvizu mogulExperimental user
MrDrWho13
Posts: 2221
Yeah, it's a good film, you should put it in the gallery!
permalink
the topic is closed

Home ? Tech Help ? lip synch off after rendering