Speech Analysis as layer markers - How?

this page states:
"If you import files with a speech metadata into After Effects, each word appears as a layer marker on layers based on these footage items."
I have tried various methods of achieving this without any success. Can someone please explain exactly how its done. I have the metadata appearing in Premiere's clip metadata panel so the analysis is complete. I can get the same result in Soundbooth, but no layer markers appear in After FX when I import the files.

I just tried it, and it worked for me. The temporal XMP metadata stored in a movie exported from Premiere Pro was converted to layer markers in After Effects.
You have to have the Create Layer Markers From Footage XMP Metadata preference turned on. This tutorial shows all of the steps.

Similar Messages

Speech Analysis to Clip markers?

In an hour interview, I usually manually make cuts and rename the clips giving me a basic description what the person is talking about when creating a string-out.
Im curious, is there anyway to get the Speech Analysis meta data in every detected word into a marker within the audio clip?
Just like in After effects or other Adobe authorizing software once you import it.
-Thanks!

I just tried it, and it worked for me. The temporal XMP metadata stored in a movie exported from Premiere Pro was converted to layer markers in After Effects.
You have to have the Create Layer Markers From Footage XMP Metadata preference turned on. This tutorial shows all of the steps.

How can I export speech analysis text with imbedded timecodes for the creation of subtitle files?

I found a script that allows you to export all your text layers with time codes as a .srt subtitle file. However, instead of multiple text layers, I have only one: essentially the meta data from Premiere speech analysis that with the aid of a different script, is being used at a "subtitle layer". Essentially, I'm trying to use this data to create a .srt file so I can create actual subtitles, not this hybrid animation/text layer.
Here is the script I used to export the text layer to .srt:
Subtitle Exporter v01
Script by Philipp Grolle
www.display-artists.de
Instructions:
Select all the Text Layers you want to export, run the script, and save the file as .srt
Attention: The Script has not been debugged or testet. Try it on your own risk!
// get all the selected Layers
var theComp = app.project.activeItem;
var alleLayer = theComp.selectedLayers;
// count how many layers are selected
var anzahlLayer = alleLayer.length;
// Enable the line below if you want to change the order of your text (ascending / descending)
/* alleLayer.reverse(); */
   // prompt to save file
   var theFile = File.saveDialog("Save the text file.", "untitled.txt", "TEXT txt");
   // if user didn't cancel...
   if (theFile != null) {
      // open file for "w"riting,
      theFile.open("w","TEXT","????");
         // Do it for all the selected Layers
         for (x = 0; x < anzahlLayer; x++)
             // get In- and Outpoint of the current Layer and convert to timecode (00:00:00:00)
            var timecodeIn = timeToCurrentFormat(alleLayer[x].inPoint, 25);
            var timecodeOut = timeToCurrentFormat(alleLayer[x].outPoint, 25);
           // split the value of the inPoint; converting last 2 characters (frames) to milliseconds, and finally join it again with comma instead of colon
            var str_in= timecodeIn;
            var timeCodeAnfang_in=str_in.slice(0,8);
            var timeCodeEnde_in=str_in.slice(9,11);
            var millisek_in = timeCodeEnde_in*40; // 1000 milliseconds divided by 25 fps = 40
            var zusammen_in = timeCodeAnfang_in +"," + millisek_in;
            //split the value of the outPoint; converting last 2 characters (frames) to milliseconds, and finally join it again with comma instead of colon
            var str_out= timecodeOut;
            var timeCodeAnfang_out=str_out.slice(0,8);
            var timeCodeEnde_out=str_out.slice(9,11);
            var millisek_out = timeCodeEnde_out*40;
            var zusammen_out = timeCodeAnfang_out +"," + millisek_out;
            //counter for the number above the timecode (in the results)
            var zaehler = x+1;
            //writing the results to file
            theFile.write(zaehler);
            theFile.write("\r\n");
            theFile.write(zusammen_in);
            theFile.write(" --> ");
            theFile.write(zusammen_out);
            theFile.write("\r\n");
            theFile.write(alleLayer[x].property("Source Text").value);
            theFile.write("\r\n\n");
      // close the text file
      theFile.close();
      // open text file in default app
      theFile.execute();
As I am only using one text layer, it shows me only one continious subtitle in the .srt file. Does anyone have any suggestions?

I should add that I myself am not particularly familiar with scripts or scripting, I'm merely trying all possible solutions to my problem and this seems to be a good approach. Any help would be much appreciated.

How do I clear or delete text in Speech Analysis?

I tried "Speech Analysis" using a transcript and the results were not good. How can I clear this text from each clip's metadata?

Hi EPC,
See this response from Premiere Pro engineering: https://forums.adobe.com/message/5752095#5752095
Thanks,
Kevin

How do you search speech analysis metadata?

I'm new to Premiere - I used the Analyze function to create a metadata element of 'Analysis Text' - which is great. How can I search this for keywords.
I have been given about 15 hours of footage and I need to find a few moments where certain things are mentioned. The analysis didn't come back with the most accurate results even though the audio is extremely clear (a speaker on a stage with the microphone as the audio source) but I'm hoping it's good enough to pull out what I need ...
if i can only search!

Very surprised no one else is having this issue. Seems like half the point of speech analysis is then to be able to search the text....

How do I install Speech Analysis Models for PPro CC?

I downloaded the International English speech analysis model from here:
http://www.adobe.com/uk/products/premiere/extend.displayTab3.html
Which says they're for CC as well as CS6.
But when I install it, I don't get the option in CC to analyze clips using anything but US English.
The models are installed here:
C:\Program Files (x86)\Common Files\Adobe\SpeechAnalysisModels\CS6 so I tried creating a CC model, copying the CS6 content across, and restarting Premiere. No joy. What else can I try?
Cheers
Mark

joshweiland wrote:
To clarify, you created a new "CC" folder inside C:\Program Files (x86)\Common Files\Adobe\SpeechAnalysisModels\ and put the files there?
Exactly that. Also tried naming the folder CS7 and CS7.0
joshweiland wrote:
If that doesn't work, perhaps try putting them in the 64-bit Program Files folder, like the following:
C:\Program Files\Common Files\Adobe\SpeechAnalysisModels\CC\
There is no SpeechAnalysisModels folder inside the 64-bit Program Files folder.
But I tried your suggestion, and tried the different names for the folders as above.
No success.
Thanks for your suggestions, though.

Speech analysis troubles

First time trying speech analysis, and yes ... I've looked through Adobe and haven't found anything useful. Sigh.
Adobe CC on everything, btw ...
Have a clip in P-Pro source monitor, tried the speech analysis tool whatever. MAJOR fail ...
1) though the clip is over 6 minutes long, it decided to take a cut of that clip from an earlier sub-clip, and so "analyzed" only about 1:15 or so (one minute 15 seconds or something like that) ... and did it VERY poorly.
Searching here and Adobe in general, I got a bit from this forum where someone also complaining of a partially analyzed clip was told by staff that if they simply re-analyzed the clip it would replace the original too-short version, as there is no designed way to delete an analysis once done. I've tried that ... 5 times now. Nope, it still chooses the exact same sub-segment, even though this full clip is on it's own sequence and all in/out markers removed on this sequence. How do I get rid of this useless crap and get it to analyze the whole clip?
2) Near to half the words need editing ... even though it's a talkie-head of two women with no other noise going on. Very clear to listen to. Standard US English, we're from Oregon, so ... no mid-west twang, no Dakota-near Canuckian, no southern drawl or any of that. Some of it isn't even close to what was said. How the hell do you edit the text? I haven't found one thing about doing that ... I have just loaded "Story" ... and that might be able to help from what little I understand of it, BUT ... I'd need a "full" copy first even if it can 'edit' the analyzed script. Which ... see problem no. 1.
3) My preferences in Adobe Media Encoder for output is on my active projects drive ... but it keeps placing files on the C: drive, in the whole damn Windows Users/morass. That C:drive is an SSD, it's not unlimited, and I do my best to keep unwanted CRAP off it. I've got most of my other Adobe junk from putting extra crap there, and using the D: drive which is planned for all that. How the Hades do I get AME to stop putting multiple copies of everything on the Users section of my C drive?
Rather ... frustrated. And yes, I've spent a good bit of time searching on these things ...
Neil

Thanks, I updated Premiere Pro CC to 7.2.2 & downloaded English & French Speech models from Exchange, but the Speech options are still greyed out. I would have thought English would be installed by default anyway? I'm sure it was in the original PPCC (and I had downloaded a number of other languages previously. I was looking for them in the various Premiere folders but didn't find them.)
Seems there's no solution for this on the Using Speech Analysis page (no "Analyze" button in the metadata panel).

EXPRESSION help,layer markers

Hi
I've got a comp with one camera and 100 3d layers distributed randomly in 3 space.Each layer has got a different z-Rotation value.
The camera animates from one layer to the other via script (the script generates
automatically layer markers and keyframes for each animation).So,there are 101 layer markers.
Example:
The camera moves between layer marker 2+3 from layer 2 to 3 for example.
Layer 2 has got a z-rotation value of -14° and layer 3=>45°.
basically,i need a camera (z orientation) expression so the camera animates
from -14° to 45° between these two layer markers and so on.
THX

Yeah, what part of the equation do you not understand? The first part simply checks whether markers exist at all, then counts them down until there are none left after a given time. If you want to blend between the different values, you'd simply need to plug the marker times into the remapping. Something like this:
markNow=marker.nearestKey(time).index;
markNext=marker.nearestKey(time).index+1;
timeNow=marker(markNow).time;
timeNext=marker(markNext).time;
rotationNow=thisLayer.transform.zRotation.valueAtTime(timeNow);
rotationNext=thisComp.layer(index+1).transform.zRotation.valueAtTime(timeNow);
rotationFinal=linear(time,timeNow.timeNext,rotationNow,rotationNext)
You only need to include this in the curly braces or return n outside the function to make use of this. Haven't tested it, as I'm not in front of AE, but on an abstract level this is how it should work. Hope it helps to get yoou started.
Mylenium

Clearing out speech analysis

Is there a way to clear the speech analysis / text analysis metadata so all the words son't show up as markers when I bring into after effects (I do need the clip markers - but want to get rid of the speech and hoping I don't haveto do that one word at a time.
Thanks

Won't be able to check that now because I'm (unfortunately) in the middle of breaking and trying to fix things.
Here is a solution that was posted on Creative Cow - which I think worked, though I've already blundered into new problems.
http://forums.creativecow.net/thread/3/928644 (Involves setting After effects preferences not to use the metadata for the markers from speech)
Now, so I can get back on track, is there a way to terminate a dynamic link that I started in Premiere Pro, because I want to go back and start in in After Effects. I though things got terminated automatically when you shut the programs. But link seems to be persisting, because any changes made in AFX are reflected in Premiere.

Using Speech Analysis to create subtitles

Does anyone know if its possible to take the results of a Speech Analysis transcription in Premiere Pro and export it as a Subtitle file to be used in Encore?
Mainly, I'm thinking of two possible uses for the Speech Analysis function, that I can use in our workflow.
1) Creating subtitles with the click of a button. We have a script for our video, which is narration only - with the help of a reference script, Speech Analysis reproduces the script nearly 100%, and now has timing associated with each word. Is there any way to take this timing and turn it into something that is useful for subtitle purposes?
2) Creating a transcription for clients. If we had an interview, and - without a script - did a Speech Analysis, [which I imagine would have varying amounts of success, but may be good enough] - how can we export that analysis, with some timing for reference, for the client to use.

Hi,
I'm not sure if this thread is really about the accuracy of the speech-to-text process, but because Frank_xrx asked, I thought I might give my own answer. I've been working with this workflow for the last four weeks now, and have finally decided that it is not possible to do serious work with the tools as they are at the moment.
All the comments from Adobe about the quality of the voice-recording affecting the accuracy of the text analysis are true, but, as others have noted, when your sound-track has spontaneous speech you cannot always separate one voice from the background, nor have recording-studio quality all the time. In such cases the text that is produced will simply not be accurate enough to use, whether you have attached an accurate script or not. Also, this speech analysis process does not like long pauses between speakers (e.g. if you have any action in your video, this will really mess up the timing of text that is produced).
However, above and beyond all this, the main flaw in the whole process, from my point of view, is that even with a word-perfect script typed into Story, attached to video clips in On Location, and then "speech-analyzed" in Premiere... if the resulting text has omissions or mistakes, there is almost nothing you can do to edit them. OK, I know there is an editing facility but it is a sham. If you add missing words into the text, then these words appear in the text, but they never get a time stamp, so they never appear in the exported xml-file, and they can never be used to "go to" a particular point in the video. If, like me, you want to find out where and how many times certain words are used in your video, then none of these newly added words will be found! Also, if you look closely, you'll find that you cannot add anything before the first word of the analyzed text that is produced in Premiere, nor can you add anything after the last word of that text. It is just not possible (right click on the first and last words to see the greyed out items). I've had several instances where the first speaker in a scene is completely missed out of the analyzed text, and also, in my experience, the analysis process always (i.e. always) misses off the last word in each scene, regardless of sound-track quality or the attachment of a full script. When these things happen this is a major problem for me. For me, the feature of searching for and playing back individual words or phrases in a video was the main attraction of the Production Premium Suite; this was why I bought it. I'm trying to produce spoken and visual "concordances", as a step up from the text-only versions now available (sorry if that's a bit esoteric).
I know that a lot of editing can be done to the analyzed text if you use After Effects. And this is great if you want to show subtitles. But anything you add, or change in AE is still beyond the reach of the xml files exported by Premiere or Encore, so, again, no changes you make in AE will be searchable (or findable) in the final video. There is another route to producing a text-searchable video, which involves putting physical cue points into the video whilst in AE, and then writing a Flash program to use those cue points, instead of using the xml-file data. I haven't tried that yet, as I kept on hoping that the (frequently) advertised speech-to-text workflow would actually work. But after four, increasingly frustrating weeks I've given up on it, simply because I can't get even a small proof-of-concept project to work exactly as advertised. Starting a serious project, with the current tools, would be unwise.
I realize that there must be work-arounds for some of the problems, but at the price of this suite (even more expensive for those of us who don't live in the US), I feel that things should work as advertised. This is not Version 1.0 from a start up company.
So, I hope this wasn't too verbose. I have specific requirements of the speech-to-text workflow which I've tried to explain above. If your requirements are not the same, then maybe you can get away with the work-arounds.. and... this thread was originally about subtitles, and there is some hope for those, using AE. I thought I'd still write this, though, as I need a bit of an outlet for all these weeks of frustration :o)

CS5 Speech Analysis Tool For Transcription, Acceptable Media Types?

Hi all,
I've been fiddling with the oh so handy, speech analysis toolin CS5 PPro, and can really see the potential for cutting down on transciption time. I spent a bit of time poking around the CS5 PPro help document, as well as tutorials via lynda.com etc. to gain a bit more insight into the file types that the speech analysis will accept and am drawing a blank other than the observations I have made doing a bit of experimenting with the feature. So far, I can not get the tool to recognize 2 of our 3 cameras, native media. Currently it will analyze only raw media shot via SDHC Class 6 cards on a JVC HGMY700 in 1080i60 format. The other two camera/media types in question are Panasonic P2 MXF files and Ikegami HDS-V10 MXF files, both are also 1080i60 format.
Is there perhaps a codec/plug-in etc. available to access the speech analysis tool with these other types of media? I have found it works well with quicktime .mov files, but I might end up spending as much time waiting to convert the file to an .mov as I would just transcribing the old fashioned way.
Any insight on how I might best utilize this tool would be greatly appreciated. Thank you!
On a separate random note, if anyone has any useful info, links or otherwise, in regards to media management using CatDV and Cache-A for archive, it would also be appreciated. That's one of many projects looming...
Lindsay

Believe it or not, it has nothing to do with the media type--well, specifically, it doesn't have anything to do with the format or codec.
So, what's the problem? Annoyingly, Panasonic P2 structures are write-protected--that is, they're marked as "Read-only." While this isn't much of a problem in normal operation, when you try to do things like speech transcription (and other metadata entry, if I recall correctly), you'll effectively be shut out. The files in the structure are writeable, but the structure itself is not.
Yep. Weird.
What to do? In Explorer, simply right-click the CONTENTS folder (you might even be able to do this at the top level, if you save your P2 offloads in that way), click Properties, and clear the Read-only checkbox in Attributes. Once the Read-only attribute is cleared, you should be able to run Speech Analysis on the P2 clips. You'll probably want to change the Read-only flag with Premiere Pro closed.
Now--that might not solve it. There seems to be a latent bug in Pr that prevents you from running Speech Analysis on a single P2 MXF clip. If this seems to be the case, try selecting two or more clips in the Project Panel, and then check to see if the Analyze button or Analyze Content menu command is available. Pretty silly, but at least its a reasonably painless workaround.
Hope that gets things working for you.

Keyframes control with layer markers

Hello,
I have an animation of a line that has two sets of keyframes of two different properties in CS4.
Basically a line gets generated in the first phase then it gets cleared out in the second phase.
This is done using two slider controls that act on the horizontal size of two rectangles and using merge paths (using the second rectangle as a matte for the first to clear it out)
I've put two layer markers on my layer hoping to link the keyframes to each of them.
Then I have found this script on the web and I put it on the first property :
L= thisComp.layer("ShapeLayer");
n = 0;
if (L.marker.numKeys > 0){
n = L.marker.nearestKey(time).index;
if (L.marker.key(n).time > time){
n--;
}if (n == 0){
valueAtTime(0);
}else{
t = time - L.marker.key(n).time;
valueAtTime(t)
This indeed lets me move the whole animation with the layermarker (which is very good) but I can't find a way to link the second set of keyframes to do the output.
I will add a link with the project file: WeTransfer
Please help!
Thanks

First, thank you for taking the time to solve this.
I feel like a handicapped... and I am when it comes to scripting. (I'm not really a handicapped person (not that there's anything wrong with that))
It has been now over two hours of tweaking and replacing values...no go.
I can follow the logic behind the script - telling to use the index or the name of the marker instead of the nearest one but I'm a script retard.
So, like a child, I'm asking what does this do:
t = time - L.marker.key(n).time; ?
and
should there be a dot between L.marker and (1) like this L.marker.(1) ?
and there are many other questions (very stupid ones) that I will not mention here.
If you (me) don't have some basic knowledge about scripting, you (you) can give me everything word by word and I still wouldn't know how to assemble it in a script.
By now I think I've already made a fool of myself so the variable shame is not of the importance anymore.
I know...I shouldn't even be in this forum with my knowledge regarding the scripts in AE, but here I am.
I'm a motion graphic designer with over 15 years of experience with AE, and so far I've managed to elude the tricky part of learning scripting. (my bad)
If anybody has a question about motion graphics, or just graphics (ex. Oh my God, how did they do that in AfterEffects or Cinema 4D, Maya, Photoshop, Illustrator ?(or What do you think about this font, should I use it on my wedding invitation ?)) I will be glad to help!
Thank you everyone!
You are a great community!
exit.(stage(left))
P.S. (post.scriptum) I'm sorry if this turned into some kind of written stand up comedy but I didn't know how to approach this situation otherwise. I'm sorry about the retards too.(me included)

Installing speech analysis in Premiere Pro

What are the specific steps in installing speech analysis (premiere_cs5_speechanalysis_mac_ie) in Premiere Pro after I download it?

Hi Vinay Dwivedi,
I Download and install the speech analysis, but it seems to be installed in the CS6 folder, I'm running CC, and just keep Encore CS6 to have DVD Authoring.
I don't know how I can say to Premiere CC where is the file for speech analysis.
Thanx for your help

Speech Analysis in 5.5

Where do I find the Speech Analysis in CS 5.5? The Help file says it's under Metadata, but I don't see it there. In Soundbooth, it was found in Metadata.
Thanks!

This is the same question I'd love to see answered.
Steve and Lone are correct, the instructions do not describe how the metadata panel actually looks or behaves in for Speech Analysis.
I am running PPro 5.5 on a MAC OS trying to convert a speech-only MP3 audio file to text - the speech analysis panel is completely blank when the file is selected (even after running "Content Analysis" on the file or clip)
I note the variances in bold:
The help link for PPro 5.5 converting speech to text is here:
http://help.adobe.com/en_US/premierepro/cs/using/WS29F692D2-7A49-44db-A109-AE016B766767.ht ml
it says:
"Analyze speech to create text metadata
Select a file or clip.
At the bottom of the Metadata panel, click Analyze Speech, or Analyze (Adobe Premiere Pro). 5.5 actually says: "Speech Analysis"
Set the Language and Quality options, No such settings or selections appear.
and select Identify Speakers if you want to create separate speech metadata for each person. No such settings or selections appear.
Note: Speech Search can use any of several language-specific and dialect-specific libraries, such as libraries for Spanish and UK English. No such settings or selections appear.
Click OK. There is nothing to click
The spoken words appear in the Speech Analysis section.
To retain the speech metadata, save the project. "
Have tried this on both a file and sub-clip with no luck. The functionality described just is not there.
There must be some missing step. Would downloading Soundbooth help?
Thanks for any help with this.

Speech Analysis in multicam mode

Does anybody know if you can speech analysis with multicam sequences? We usually shoot our interviews with two cams, then put the footage to one timeline to cut it with a multicam sequence. Once you put it to the preview monitor there is no option to use speech analysis. Would be very helpful havng this tool in multicam sequences too. What I figured out yet was that you can only trnascribe the original footage but you cannot read the metafiles when the footage is nested in a new sequence.

Good idea, I tried to place markers to the original footage. unfortunately they do not show up in the multicams sequence. Did you try this before? Maybe I did not place them correctly.

Speech Analysis as layer markers - How?

Similar Messages

Maybe you are looking for