ExportController - Plugin loading.log & argument from Flex

February 11, 2013, 11:20 pm

≫ Next: Why is Src and Dst memory not pinned?

Hello guys

I'm having a problem porting my ExportController to cross platform and it recently refuses to get loaded.

Here is Plugin loading.log:

Loading C:\Program Files\Adobe\Common\Plug-ins\CS6\MediaCore\ExportControllerSDK.prm
Loading from disk...
No loaders recognized this plugin, so the plugin is set to Ignore.

I found CS6_ServiceManager.log but I'm not sure if it's related (altho I changed some IO precisely before it broke)

2013-02-12 01:13:50 : ERROR vcfoundation::system::VCSysError: WriteFile() failed <232: The pipe is being closed.>
race:    1 0X0049199F
   2 0X0048E7A3
   3 0X0048E91E
   4 0X00488C16
   5 0X00491CF8
   6 0X00490C6F
   7 0X0048ED72
   8 0X004884BB
   9 0X0049BAF5
10 0X0049BB7F
11 0X759F8543
12 0X7791AC69
13 0X7791AC3C

Where else can I find more information on my problem or can I enable debug launch logs? It compile and builds fine and the code has been tested externally of PPro so I'm having an hard time figuring it out.. Any help or advice appreciated.

Also is it possible to pass an argument from a flex plugin to the ExportController, i.e ExportTimeline in the snippet below:

DllExport PREMPLUGENTRY xExportControllerEntry (
          int                                        selector, ecStdParms                    *stdParms, void                              *param1)
{
          prMALError result = exportReturn_Unsupported;
            switch (selector)
          {
...
                    }
          case exportController_ExportTimeline:
                              result = ecExportTimeline(stdParms, reinterpret_cast<ExportControllerTimelineInfoRec*>(param1));
                              break;
...

Call is currently made from JSX function (invoked by the flex) but if I could propagate an argument it would solve all my problem!!!

Thanks in advance,

↧

Why is Src and Dst memory not pinned?

January 9, 2014, 11:34 am

≫ Next: Is it possible to create an Importer which automatically can provide the needed resolution?

≪ Previous: ExportController - Plugin loading.log & argument from Flex

I just completed my first video filter that is CUDA accelerated using the Premiere Pro SDK for CS6 and CC7. If suitable CUDA hardware is available, it use it, but if not then it uses a multithreaded software implementation. Both work extremely well. However, the CUDA implementation could be a lot faster if the source and destination memory buffers were "pinned" memory. Since they are not, I must copy the source and destination memory to a pinned buffer, and then asynchronously copy that to CUDA device memory and back. The overhead to copy from the source/destination memory to pinned memory is significant. Without the copy to pinned memory the CUDA on my laptop is fast enought to process 130 fps for a 1920 by 1080 HD video. However with the pinned memory I only get about 45 fps.

If I do the exact same filter in DirectShow, the source and destination buffer pools are always pinned and the filter runs much faster than it does on Premiere Pro. I noticed that the new GPU filter example uses and AE interface, but it does allow access to pinned memory. However, I have not mastered the AE interface, and I am reluctant to giving up 13 years of learning curve on the Premiere SDK.

Is there any good reason why the source and destination buffer pool in the Premiere Pro SDK is not pinned memory?

Gene

↧

Is it possible to create an Importer which automatically can provide the needed resolution?

October 17, 2014, 2:17 am

≫ Next: Get sequence start from Transmit API

≪ Previous: Why is Src and Dst memory not pinned?

Hi,

I am creating an synthetic importer to be used in Premiere and Media Encoder.

The synthetic importer is file based. The file contains information about the contents to be rendered.

Since the code can automatically generate frames and arbitrary resolution and framerate I want the user to just add the file and produce a sequence or to a media encoder convert.

But the Importer API wants me to set a definitive resolution and framerate on the imGetInfo8 callback.

To give a fixed framerate does not seem to be a problem because if I set "canDoContinuousTime" during imInit I get all calls for every frame of the sequence even if does not match the framerate given in imGetInfo8.

But the fixed resolution is a problem because I don't know the resolution of the target material.

When importing a clip without a file (File -> New -> Custom Importer) a "New Synthetic" dialog is shown to configure framerate and resolution. These values are defaulted to the values of the current sequence.

The SDKFileInfo8->vidInfo on the imGetInfo8 callback is pre filled with the values entered in the dialog.

The same workflow for importing files would be very helpful. (but without the "New Synthetic" dialog)

Especially because the Media Encoder has no option to set a synthetic as source.

tl;dr

Is there a way to implement a "match destination" like the H.264 exporter offers a "match source"?

Message was edited by: Bjoern Gdanietz

↧

Get sequence start from Transmit API

August 3, 2012, 10:30 am

≫ Next: Three questions about the CS4 SDK Exporter

≪ Previous: Is it possible to create an Importer which automatically can provide the needed resolution?

I'm using the Transmit API to get the timecode of the sequence as it plays. I'm using the inTime and working out the number of seconds from the ticks.

This gets the number of seconds from the start, which is fine if the sequence begins at 00:00:00:00, but if it begins at 01:00:00:00 or some other value then all of the timecode values are wrong. I could not figure out a way of getting the sequence start position - am I correct in assuming that such a function doesn't exist?

↧

Three questions about the CS4 SDK Exporter

April 9, 2010, 3:43 am

≫ Next: loading content into project/timeline from http listener

≪ Previous: Get sequence start from Transmit API

Hello. I'm working to make an Exporter Plug-In that supports "Pro CS4" and "Elements 8.0" single binary.

Please give me any advices to complete it. I show you the questions as follows:

How can I get the slider screen without checkbox (in parameter GUI)
How to enable an Exporter Plug-In at "Elements 8.0".
What is the difference between ver.4.0 and ver.4.2 in exExpoterInfoRec.

Details below.

1st:

I tried PrSDKExportParamSuite::AddParam() with exParamFlag_slider and exParamFlag_optional.

Both cases parameter GUI shows slider with checkbox. Please teach me a method to use simple slider (no checkbox) in parameter GUI.

BTW, Standard plugins are seem to use "Premiere Export Private Statndar Param Suite". Are you planning to publish it?

2nd:

The Premiere Elements 8.0 never shows any additional Exporter Plug-in, including SDK sample (ExporterSDK.prm).

Please tell me if I need special installation procedures.

3rd:

I found that the exExporterInfoRec is difference between 4.0 and 4.2. The exExporterInfoRec[ver. 4.2] is defined in CS4SDK_r1. But I can't find any definition of exExporterInfoRec[ver. 4.0] in the SDK.

An Exporter Plug-In is build with CS4SDK_r1 never work on the Premiere Pro CS4 (4.0) / Media Encoder CS4 (4.0). I need definition of exExporterInfoRec[ver. 4.0] to support CS4 (ver. 4.0).

Thanks.

↧

loading content into project/timeline from http listener

May 10, 2010, 7:39 am

≫ Next: Importing multiple files in a sequence

≪ Previous: Three questions about the CS4 SDK Exporter

I have an application which helps producers/researchers browse and search for video assets. Once they have located one or more assets, I would like to be able to "push" them over to Premiere Pro.

The browser is written in Silverlight, so the only mechanism it has to communicate with Premiere Pro would be via http. I think what I need is a plug in which has an HttpListener listening on a specific port, so that the silverilght application can post information (in edl/xml format) to the port the plugin is listening on. The plugin will then load the assets (mms video streams or files from uris) into the project and/or timeline.

I would like to write this either using Flex or the Premiere Pro SDK. I need it to run on both Windows and the Mac.

If anyone can provide any pointers, I would be very grateful.

Thanks

Nigel

↧

Importing multiple files in a sequence

October 21, 2011, 3:57 am

≫ Next: Unexplained problem with PF_TRANSFER_RECT in Premiere

≪ Previous: loading content into project/timeline from http listener

Hi -

I am new to this forum - excuse me if this has already discussed eslewhere. I am trying to see if there is a way of importing clips from an excel file which has the folder name, file name and the time stamp from which the video shd play and import all of them by reading the rows in excel file into premier pro. Any pointers will be helpful. Thanks for the help.

vimmy

↧

Unexplained problem with PF_TRANSFER_RECT in Premiere

June 1, 2012, 5:36 am

≫ Next: Pixel iteration when footage smaller than sequence size?

≪ Previous: Importing multiple files in a sequence

This is cross-posted from After Effects SDK forum, since I could not get help there:

http://forums.adobe.com/message/4456832

I am trying to perform a simple operation using AE SDK for an effect in Premiere:

Create a new world (mask), fill it with color and alpha, and then blend it with a selected composite mode to the footage.

I have made it working in After Effects perfectly. It uses WorldTransformSuite1 transfer_rect, which is not supported by Premiere. However, its counterpart, PF_TRANSFER_RECT is supposedly supported.

The results in Premiere, regardless of whether it's YUV or RGB, are really almost random depending on the color that I select. Sometimes I get a blue cast all over the picture regardless of alpha channel, sometimes I get the alpha channel transferred, even though I have the rgb_only flag set to true, sometimes seemingly nothing happens at all.

The mask is drawn correctly. If I simply copy it to the output world, everything is fine. Clearly it is the PF_TRANSFER_RECT that is to blame for improper transfer.

Can somebody help me? I tried to debug this thing, but I can't seem to find any sense in the behavior.

The callout of PF_TRANSFER_RECT is pretty generic:

     ERR(PF_TRANSFER_RECT(                          in_data->quality,                         PF_MF_Alpha_STRAIGHT,                          in_data->field,                          &output->extent_hint,          // rectangle                          &mask,                    // source world                          &composite_mode,          // comp mode                          &mask_world,               // mask world                          0,                         // dest_x                          0,                         // dest_y                          &temp) );                    //dst world

temp is already populated with &params[PLUGIN_LAYER]->u.ld

mask_world is defined as the mask world for mask and PF_MaskFlag_NONE flag.

I have been fighting with it for past week, and I am totally frustrated.

↧

Pixel iteration when footage smaller than sequence size?

August 18, 2012, 3:01 pm

≫ Next: Creating C++ objects inside an exporter plugin (CS6 SDK)

≪ Previous: Unexplained problem with PF_TRANSFER_RECT in Premiere

Hi everyone,

I am having a problem with my plugins when the footage is smaller than the sequence.

For some reason Premiere (and only Premiere, AE is fine in this regard) iterates pixels not from the clip origin (let's say 200,100), but from sequence origin (0,0). Since it only iterates the amount of pixels that is proper clip size (correct), it results in part of the layer being totally cut off (in this example 200 pixels from the right, and 100 pixels from below).

For what it's worth, if the clip has proper aspect ratio, and scale to sequence size is clicked, or if the bigger clip is scaled down, the effect is applied properly.

I have tried using various offsets, but I could not get it to work. Obviously I'm missing something. I am using AE SDK.

Thanks for any help.

Bart

↧

Creating C++ objects inside an exporter plugin (CS6 SDK)

February 20, 2013, 11:01 pm

≫ Next: how can i rename item of rootitem in premiere project and move this item to a bin by actionscript

≪ Previous: Pixel iteration when footage smaller than sequence size?

I have downloaded the Adobe CS6 SDK. In the exporter example, inside exSelBeginInstance(), there are calls to "memory suite" API. This grabs some memory (at runtime) for a C-struct/array.

In my case, I need to call the new() operator on a derived C++ class. The idea is that I'll create the object during exSelBeginInstance(), then delete it during exSelEndInstance.

If I use the new() operator, then the memory-suite won't know about the object at all. Alternatively, I could call the memory-API to grab a chunk of memory the size of "sizeof(my_class)" But this doesn't work correctly either, because the object itself may need to create instances of other classes.

So is there a way to safely use dynamically created C++ objects inside an exporter plugin? Or do I have re-architect everything to avoid calling the new() constructor?

↧

how can i rename item of rootitem in premiere project and move this item to a bin by actionscript

October 17, 2014, 3:49 am

≫ Next: SDK_ProcAmp_GPU: documentation or example to interact with CUDA

≪ Previous: Creating C++ objects inside an exporter plugin (CS6 SDK)

I use this code in my button:

var prj:Project = Premiere.app.project;
var arr:Array = new Array();
arr[0] = "c:\abc.mp4"
prj.importFiles(arr);

Premiere project add an item named "abc.mp4" but i want display this item with name like "somethingiwant.mp4".

and then i create bin by:

prj.rootItem.createBin("tempbin");

i want move the clip into this bin.

I can not do these, help me please !!

↧

SDK_ProcAmp_GPU: documentation or example to interact with CUDA

April 20, 2015, 12:46 pm

≫ Next: QE: What is this rabbithole?

≪ Previous: how can i rename item of rootitem in premiere project and move this item to a bin by actionscript

Hi- I see the SDK_ProcAmpGPU example uses OpenCL and casts the memory info as cl_mem for GPU processing:

void* frameData = 0;

mGPUDeviceSuite->GetGPUPPixData(*outFrame, &frameData);

cl_mem buffer = (cl_mem)frameData;

When using CUDA, what is the proper struct/class to cast the frameData for use with CUDA?

Thanks,

John

↧

QE: What is this rabbithole?

October 12, 2015, 5:17 pm

≫ Next: Is there a way to export EDLs from the Premiere api?

≪ Previous: SDK_ProcAmp_GPU: documentation or example to interact with CUDA

Soooo as the title asks: What is QE? There's absolutely no documentation on it, and it appears to house most of the information that we want to access, albeit not reliably.

Most importantly out of anything that I'm looking for - why is there no documentation on not just QE, but the rest of everything within Premiere's capabilities of extendscript?

A single piece of code on Github is not sufficient for a documentation substitute.

As far as the seeming instability of QE:

Consider the following code for Premiere Pro CC 2015:

app.project.activeSequence.exportAsFinalCutProXML("C:\\Users\\user\\Desktop\\XML_Testing.xml")


app.enableQE()
var mySeq = qe.project.getActiveSequence(0)
var myTrack = mySeq.getVideoTrackAt(0)
var myClip = myTrack.getItemAt(0)
myClip.name

If I have a project open in Premiere and execute this using ExtendScript Toolkit, I receive back a myTrack object with a name of 'Wildlife.wmv'.
If I then switch to a new project in Premiere(does not contain 'Wildlife.wmv') and re-execute that same code in ExtendScript Toolkit, the myTrack object stays the same, and I don't receive a new object.

This seems hazardous at best.

Thank you.

↧

Is there a way to export EDLs from the Premiere api?

November 4, 2015, 9:04 am

≫ Next: custom transmitter plugin not push video on time

≪ Previous: QE: What is this rabbithole?

Hello friendly Premiere people,

I see options to export FCPXML, AFF, etc, but nothing for EDL. Maybe the exportTimeline() function? If so, how do I use that? I tried these two:

app.project.activeSequence.exportTimeline("edl");

app.project.activeSequence.exportTimeline("EXL");

If there's nothing available I might create a fcpxml -> edl converter. Any ideas about a better way to go?

Cheers,

Raphael

↧

custom transmitter plugin not push video on time

November 22, 2015, 11:21 pm

≫ Next: I want to create new sequence from a movie clip using script.

≪ Previous: Is there a way to export EDLs from the Premiere api?

my custom transmitter plugin not push video on time, when start play, I called "clockCallback " function 11 times but just push video 6 times, so my plugin in will have not enough frame to play, after several frames, It works well, I make sure timeElapsed is enough, why?

please help me , thanks in advance

↧

I want to create new sequence from a movie clip using script.

December 28, 2015, 2:59 am

≫ Next: Downloading file with NodeJS in CEP extension

≪ Previous: custom transmitter plugin not push video on time

I want to create new sequence from a movie clip using script.

↧

Downloading file with NodeJS in CEP extension

December 29, 2015, 2:37 pm

≫ Next: Media types supported by ProjectItem.refreshMedia()

≪ Previous: I want to create new sequence from a movie clip using script.

I'm trying to get file downloading with NodeJS working in my CEP extension for Premiere Pro CC 2015. I have the following script:

var http = require('http');
var fs = require('fs');
var file = fs.createWriteStream("file.jpg");
var request = http.get("http://i3.ytimg.com/vi/J---aiyznGQ/mqdefault.jpg", function(response) {    response.pipe(file);
});

The script works perfectly well if I run it on it's own. For example, if I put it in a file called test.js, and run node test.js in terminal, it successfully saves the image to the same directory the script is in.

When I include it in my CEP extension, however, nothing happens. The script executes without errors, but no file is downloaded anywhere. Even if I try:

try {     response.pipe(file);     alert("File downloaded");
} catch(e) {     alert(e);
}

The "File downloaded" alert pops up. So the lack-of-download doesn't seem to be from an error in the code, but perhaps in the way CEP works itself? I'm at a loss. Any suggestions?

Thanks!

↧

Media types supported by ProjectItem.refreshMedia()

January 27, 2016, 11:23 am

≫ Next: Where can I find documentation for the Premiere SDK?

≪ Previous: Downloading file with NodeJS in CEP extension

Hi,

Can any type of media be refreshed using the ProjectItem.refreshMedia() function, or does this only work for OP1A MXF growing files?

I have tried calling this function on ProRes files on a NAS and they do not refresh unless Premiere is tabbed out and back into again.

↧

Where can I find documentation for the Premiere SDK?

May 12, 2016, 5:08 am

≫ Next: Log Premiere startup info re: extensions

≪ Previous: Media types supported by ProjectItem.refreshMedia()

I've already asked several questions in the forums, but I still feel lost. For every feature I want to implement, I have to come here to ask if it's even possible.

So, what I would like to be my last question: Is there any documentation around?

↧

Log Premiere startup info re: extensions

May 12, 2016, 2:17 pm

≫ Next: Video Filter Crash on Mac

≪ Previous: Where can I find documentation for the Premiere SDK?

I have an extension bundle that loads when Premiere starts up and it stopped loading today and can't figure out why.
I'm on CC 2015.2, Windows 7.
I'm hoping there's a way I can get Premiere to print out running information to the console when calling the .exe. I already tried looking in the .log files generated in ~/AppData/Local/Temp but nothing useful is output there.

My hope is there's a command-line arg that I can pass the Premiere.exe to print out information. I see it can be done on OSX and was curious about the same capability on Windows.

↧