[Released] Anilip 2 [Commercial]

12346

Comments

  • DobitDobit Posts: 151

    Thank you for the interest. But there is no news about an Anilip 2 mac version. We don't want to make any promises here at the moment.

  • his xhis x Posts: 728

    Perhaps you would consider a 3.0 Roadmap containing what, not when.

  • DobitDobit Posts: 151

    We are currently testing a free update for all Anilip 2 customers. We have improved the error handling, support more characters and an optimized, even more natural lip animation. We will report more on this when the release is available. Version 3 will take some time. We are still investigating the feasibility of possible features and do not want to promise too much at the moment.

    But maybe there are suggestions for a version 3 here in the forum, which we would be happy to take into account if possible.

  • his xhis x Posts: 728
    Dobit said:

    We are currently testing a free update for all Anilip 2 customers. We have improved the error handling, support more characters and an optimized, even more natural lip animation. We will report more on this when the release is available. Version 3 will take some time. We are still investigating the feasibility of possible features and do not want to promise too much at the moment.

    But maybe there are suggestions for a version 3 here in the forum, which we would be happy to take into account if possible.

    Thank you for your response. I would like to offer two suggestions.

    First I would add automatic facial gestures. It adds a lot of credibility and is far faster than adding them manually. Second, I would provide an Intensity Level slide switch. Of course, either one would be easy to disable.

    These are the two main things Mimic has that Anilip does not.

     

  • DobitDobit Posts: 151
    his x said:
    Dobit said:

    We are currently testing a free update for all Anilip 2 customers. We have improved the error handling, support more characters and an optimized, even more natural lip animation. We will report more on this when the release is available. Version 3 will take some time. We are still investigating the feasibility of possible features and do not want to promise too much at the moment.

    But maybe there are suggestions for a version 3 here in the forum, which we would be happy to take into account if possible.

    Thank you for your response. I would like to offer two suggestions.

    First I would add automatic facial gestures. It adds a lot of credibility and is far faster than adding them manually. Second, I would provide an Intensity Level slide switch. Of course, either one would be easy to disable.

    These are the two main things Mimic has that Anilip does not.

     

     I think I know what you mean. Our aniHead product may be a step in this direction. But we will definitely think about further improvements.

  • his xhis x Posts: 728
    Dobit said:
    his x said:
    Dobit said:

    We are currently testing a free update for all Anilip 2 customers. We have improved the error handling, support more characters and an optimized, even more natural lip animation. We will report more on this when the release is available. Version 3 will take some time. We are still investigating the feasibility of possible features and do not want to promise too much at the moment.

    But maybe there are suggestions for a version 3 here in the forum, which we would be happy to take into account if possible.

    Thank you for your response. I would like to offer two suggestions.

    First I would add automatic facial gestures. It adds a lot of credibility and is far faster than adding them manually. Second, I would provide an Intensity Level slide switch. Of course, either one would be easy to disable.

    These are the two main things Mimic has that Anilip does not.

     

     I think I know what you mean. Our aniHead product may be a step in this direction. But we will definitely think about further improvements.

    Personally I am unlikely to use Anilip until it is on a level with Mimic.

  • gitika1gitika1 Posts: 792
    edited July 2020

    I am recording the voice over in Adobe Audition.   Noise reduction and silence (in breaks of speech) are employed.  Yet, bringing in even the simplest of 16 bit wav files to Anilip prompts warnings of too loud and the messages similar to these here:

     

    Suggestions?

    Anilip1.PNG
    526 x 129 - 7K
    Anilip2.PNG
    522 x 187 - 13K
    Anilip3.PNG
    529 x 136 - 8K
    Post edited by gitika1 on
  • DobitDobit Posts: 151
    gitika1 said:

    I am recording the voice over in Adobe Audition.   Noise reduction and silence (in breaks of speech) are employed.  Yet, bringing in even the simplest of 16 bit wav files to Anilip prompts warnings of too loud and the messages similar to these here:

    Suggestions?

    These are just warnings from Windows speech recognition. You don't have to take them too seriously. We will soon post an free update in which you can optionally suppress the warnings. The goal is not 100% speech recognition, but a list of phonemes. If one or the other word is not recognized correctly, you will hardly notice it in the lip synchronization.

  • Makumba666Makumba666 Posts: 23
    edited July 2020

    I was considering buying this, cause I really need lip sync for daz. Good I tried it before (a friend of mine was stupid enough to buy it and he showed me).. Not only it can not recognize audio, but it can not lipsync even plain text. Text says one thing - lips say absolutely different. I even tougth may be it translates text into some other language - chinese for example :) - before syncing. Personally I would not pay a dime for this.

    Post edited by Makumba666 on
  • DobitDobit Posts: 151

    I was considering buying this, cause I really need lip sync for daz. Good I tried it before (a friend of mine was stupid enough to buy it and he showed me).. Not only it can not recognize audio, but it can not lipsync even plain text. Text says one thing - lips say absolutely different. I even tougth may be it translates text into some other language - chinese for example :) - before syncing. Personally I would not pay a dime for this.

    I'm sorry you're having problems with Anilip 2. But it would be fair if you gave us the chance to track down the problems. Anilip has been around for a few years now and so far we have got most of the user problems under control. Let's take care of your Text To Speech problems first. Maybe you can help us with the analysis. Which Windows language do you have? Which DAZ figure do you test with? What text did you enter for LipSync?

  • gitika1gitika1 Posts: 792
    Dobit said:
    gitika1 said:

    I am recording the voice over in Adobe Audition.   Noise reduction and silence (in breaks of speech) are employed.  Yet, bringing in even the simplest of 16 bit wav files to Anilip prompts warnings of too loud and the messages similar to these here:

    Suggestions?

    These are just warnings from Windows speech recognition. You don't have to take them too seriously. We will soon post an free update in which you can optionally suppress the warnings. The goal is not 100% speech recognition, but a list of phonemes. If one or the other word is not recognized correctly, you will hardly notice it in the lip synchronization.

    I have found that if there is more than one or two warnings, the frames are created but there is no lip animation.  Even in the text window, it deletes all but the first word that I typed in.  Single sentences work.  Two or three sentences at the same time yields too many warnings and no animation.

  • DobitDobit Posts: 151
    gitika1 said:
    I have found that if there is more than one or two warnings, the frames are created but there is no lip animation.  Even in the text window, it deletes all but the first word that I typed in.  Single sentences work.  Two or three sentences at the same time yields too many warnings and no animation.

    If you want, you can send us examples of your recordings to our support address. Maybe we can give you some advice.

  • gitika1gitika1 Posts: 792
    Dobit said:
    gitika1 said:
    I have found that if there is more than one or two warnings, the frames are created but there is no lip animation.  Even in the text window, it deletes all but the first word that I typed in.  Single sentences work.  Two or three sentences at the same time yields too many warnings and no animation.

    If you want, you can send us examples of your recordings to our support address. Maybe we can give you some advice.

    Thanks!  That's very kind of you.

    I discovered (trial and error) that if I took the voice over and coverted it to Mono 16 bit (with only 1 channel being used), that I had much better results with Anilip 2.  In Audition, the default for stereo to mono mixes half of each channel.  I was literally amazed when Anilip processed all 5 sentences with no errors.  It was the first time ever, to not struggle with Anilip 2; and  I was able to move on to learning about the setting and parameters.

     

  • Dobit, where do I go to delete all of the keyframes? How do I start back at square 1 where character is no longer talking because in the next take I need her to not talk?? Lot of issues here! Also one major problem I'm hitting is that my character doesn't ever actually fully close her mouth on fricatives. So while there is something going on there its still not right.

  • zmodiszmodis Posts: 6

    Is there any way to increase the _S  or pause between words using ... or . or something like this in the text to speech box?  Changing start time individually is not really an option. It can be done in a spreadsheet of course, but this is a little bit more time consuming I think.  It would be nice to just enter [...] and have it add 100 miliseconds per dot when it renders the animation...

  • DobitDobit Posts: 151
    edited August 2020
    zmodis said:

    Is there any way to increase the _S  or pause between words using ... or . or something like this in the text to speech box?  Changing start time individually is not really an option. It can be done in a spreadsheet of course, but this is a little bit more time consuming I think.  It would be nice to just enter [...] and have it add 100 miliseconds per dot when it renders the animation...

    Not yet, but we think about those options. In the mean time you can use a spreadsheet. It is very easy. We tested it with Excel:

    1) Open Tab "Viseme Animation"

    2) Select all rows (ctrl a)

    3) Press the "Copy Rows" button

    4) Paste the content to the spreadsheet

    5) Make your changes

    6) Select the rows and colums and copy

    7) In Daz Studio press the paste button

    Hope this helps.

     

    Post edited by Dobit on
  • DobitDobit Posts: 151
    edited August 2020

    Dobit, where do I go to delete all of the keyframes?

    In tab "Viseme Animation" press button "Remove all rows..." and press "Write to timeline"

     

    Also one major problem I'm hitting is that my character doesn't ever actually fully close her mouth on fricatives.

    This can be changed in the configuration file. It is located in \ Program Files\DAZ 3D\DAZStudio4\plugins\dobitAnilip2\dobitAnilip2.dsx. You need admin rights to edit it.

          <configuration name="Genesis8">
            <viseme id="V0">
              <property category="Neutral" value="0.6" id="eCTRLvS0"/>
            </viseme>
    ...
          </configuration>
    
    

    Here you can adjust the "Neutral" position.

     

    How do I start back at square 1 where character is no longer talking because in the next take I need her to not talk??

    Anilip is not yet designed for creating dialogues. We would recommend doing a project for each take and saving the animation as a pose preset. Then use the saved presets in the actual scene. We explained how to save a pose preset in the thread above.

    Post edited by Dobit on
  • Hello Again, I have installed and reinstalled Anilip 2 10 times.  When i add the audio file to speech recognition it does not make the lips move.  I have loaded the genesis 8 characters and it still will not make the lips move.  What am i doing wrong? 

  • DobitDobit Posts: 151
    donte4011 said:

    Hello Again, I have installed and reinstalled Anilip 2 10 times.  When i add the audio file to speech recognition it does not make the lips move.  I have loaded the genesis 8 characters and it still will not make the lips move.  What am i doing wrong? 

    Does the "Text To Speech" Option work?  Do you get any error messages in your log file ? (Help/Troubleshooting/View Logfile). You can also send us your audio file for inspection.

  • No, the text to speech option does not work either. That is why I feel it is not my audio file because i have used several exports of audio and played around with the export settings and have tried loading it in anilip 2 speech recognition and it does not animate the lips. I even tried text to speech and still nothing.
  • DobitDobit Posts: 151
    donte4011 said:
    No, the text to speech option does not work either. That is why I feel it is not my audio file because i have used several exports of audio and played around with the export settings and have tried loading it in anilip 2 speech recognition and it does not animate the lips. I even tried text to speech and still nothing.

    Ok, you are right. It is not tthe audio file. Which Windows Display Language do you use? Can you send us your log file (Help/Troubleshooting/View Logfile)?

  • I am using anilip 2 and I am very satisfied but I would like to change the value of the default parameters: overall strength (80%) and possibly jaw strength and lips strength. I tried in viseme style settings without result. Is there a possibility?

  • DobitDobit Posts: 151
    edited September 2020

    I am using anilip 2 and I am very satisfied but I would like to change the value of the default parameters: overall strength (80%) and possibly jaw strength and lips strength. I tried in viseme style settings without result. Is there a possibility?

    In order for the presets to take effect, you must:
    1) either make the default settings first and then generate the phonemes.
    2) Or mark all phonemes. If you then change the default settings, they are adopted for all marked lines.

    The changes have to be written back to the timeline. Therefore press the red flag so that the changes take effect.

    I hope, this answers your question.

     

    2020-09-12 13_27_01-DAZ Studio 4.12 Pro Publishing Build.png
    1332 x 645 - 144K
    Post edited by Dobit on
  • I do have one question.  If I were to make the voice sound more happy and not monotone, how could I do this?  The one problem I have with AniLip2 is that I get a monotone voice generated every single time.

  • DobitDobit Posts: 151

    I do have one question.  If I were to make the voice sound more happy and not monotone, how could I do this?  The one problem I have with AniLip2 is that I get a monotone voice generated every single time.

    Anilip 2´s integrated voices are dependent on the SAPI 5 voices from Microsoft and therefore the voice quality is tied to what Microsoft provides. However, you have the option to use Amazon AWS (Amazon Polly) voices instead. How to implement them is documented in the anilip 2 tutorials.

  • EightiesIsEnoughEightiesIsEnough Posts: 894
    edited October 2020

    I do have one question regarding Amazon Polly.  I noticed there is an option to pay as little as 2¢ and as much as 8¢ for animated avatars.  How does that work?  Do I pay per usage of under 3.5 minutes of speech using the Polly voices?  This sounds very interesting.

    I have another question: What does "neural voice" refer to?

    Post edited by EightiesIsEnough on
  • It mentions something about downloading JSON files.  I am having a hard time finding JSON files for voices on Amazon Polly.  Where do I have to go to download them?

  • DobitDobit Posts: 151

    I do have one question regarding Amazon Polly.  I noticed there is an option to pay as little as 2¢ and as much as 8¢ for animated avatars.  How does that work?  Do I pay per usage of under 3.5 minutes of speech using the Polly voices?  This sounds very interesting.

    I have another question: What does "neural voice" refer to?

    We only support the Amazon Polly voices and not the animated avatars. Here is a link to  the pricing options for this service: https://aws.amazon.com/polly/pricing/?nc1=h_ls Neural voices here refer to artificial voices that are made using AI technology from real voice samples provided by an actor. A visual guide on how to use Amazon Polly in anilip 2 can be found in your "hints" Tab in the anilip 2 pane. The video tutorial is called "Text to Speech".

  • EightiesIsEnoughEightiesIsEnough Posts: 894
    edited November 2020

    I'd like to ask, is there a way for me to generate the visemes with just the audio in AniLip2?  Reason I ask is because I am wondering if I could use audio files generated in Speechelo?  Speechelo has some voices in friendly, normal and serious tones.

    Or, could I generate speech marks in Speechelo?

    Also, regarding Amazon Polly, do they accept Visa Debit?  I tried to add my debit card number, but it registered as a credit card.  I have since deleted my account.

    Post edited by EightiesIsEnough on
  • DobitDobit Posts: 151

    Speechelo voices have a really amazing quality. You can use the speech recognition feature of Anilip 2 to take advantage of Speechelo voices. We tested this briefly and it works quite well.

    You must convert the Speechelo mp3 export to wav file format.

    The payment methods of AWS you find here.

Sign In or Register to comment.