February 2008 - Pro-apps-dev

Re: Motion - weird handling of large images
by Darrin Cardani 07 Feb '08

07 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com On Feb 7, 2008, at 12:36 PM, Brian wrote: ... it asks the video card how much VRAM it has, and also what the size of the largest texture it can handle is. ... so we may estimate that you will be able to process less and we'll still downsample for you. Darrin, Is there any sort of Preference setting in Motion (and/or FCP) for a user to say that they want to use Software rendering as a fall-back, if there is insufficient resources to run a plug-in in Hardware? If the FxPlug-in is both software & hardware enabled, it seems that a good Software render is preferable to a bad Hardware render. I'm working with Film resolution mostly (2K-4K). So, I'll likely to be hitting the same problems as Steve is having. Darrin -- Darrin Cardani dcardani(a)apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… No, there's no such setting. The problem is that filters aren't the only thing which renders on the GPU. Almost all other objects do, too, such as shapes, particles, etc. So we would have to write the code to actually render those things in software before we could add such a switch. We are working on ways for the application to deal with this more gracefully in the future. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Building a generator FxPlug that reads from a device driver...
by Paul Schneider 07 Feb '08

07 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com Hi, Kris, Here's some further reading that might be of interest to you: Using C++ with Objective-C: http://developer.apple.com/documentation/Cocoa/Conceptual/ObjectiveC/Articl… Final Cut XML: http://developer.apple.com/documentation/AppleApplications/Reference/FinalC… QuickTime sample code: http://developer.apple.com/samplecode/QuickTime/idxCompressionDecompression… Hope that helps, - Paul On Feb 7, 2008, at 9:26 AM, Kris van Rens wrote: Hi there, I've built a device driver for MacOS X to be able to read and write bulk video data from/ to a USB device, this all works fine. Now I'd like to write a FxPlug plugin to acquire the video data from my USB device directly into Final Cut Pro. I have read some documentation and came to the conclusion that I will have to write a generator FxPlug plugin, this plugin should then communicate with the device driver. I checked out the examples that come with the FxPlug SDK but as far as I saw there was no example of a plugin that does direct communication with a device driver. Are there specific ways to go about dealing with this plugin to device driver communication? Should I implement the communication with the device driver exactly the way I would when building an user application? Would that have to be done in Objective-C or are there ways to partly implement the plugin in C++? I hope my problem is clear, if not; tell me. Thank you in advance for your time, I really appreciate your help. Greetings, Kris _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/pschneider%40apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… it sounds like you want to write a file importing plugin. FxPlug generators are not really the best choice for that. For one thing, there's no "native format" for generators - they are expected to be able to generate any frame size, frame rate, and field order, and support any duration. If your USB device is exposed as a filesystem to the Mac, the best thing to do would be to write a QuickTime component that allows the video data to be wrapped in a QuickTime movie. If the Mac doesn't have access to the device's filesystem, you could write a standalone application that pulls the video stream from the device and saves it as a QT movie on disk. Your application could then create some XML that references the movie and load that into Final Cut. If neither of those options are available, then you could try the FxPlug generator route. It's straightforward to call C++ code from Objective-C. The easiest thing to do is to write your plugin in Objective-C++. As far as talking to the device from a plugin, I haven't done that myself, so I don't know what obstacles you'll encounter. Your plugin will run inside of Final Cut, so I wouldn't expect it to be too different from the code you'd write in a standalone application, but it really depends on the specifics of your device. This email sent to pschneider(a)apple.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: Motion - weird handling of large images
by Brian 07 Feb '08

07 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=dk20050327; d=earthlink.net; b=RFm6mYD5STfzisaujbeliqEtkKAZjLM4OcK0m7r4IVwWOzn1ZcF4aOfhHGOs/5FS; h=Message-ID:Date:From:Reply-To:To:Subject:Cc:Mime-Version:Content-Type:Content-Transfer-Encoding:X-Mailer:X-ELNK-Trace:X-Originating-IP; > ... it asks the video card how much VRAM it has, > and also what the size of the largest texture it can handle is. > ... so we may estimate that you will be able to > process less and we'll still downsample for you. Darrin, Is there any sort of Preference setting in Motion (and/or FCP) for a user to say that they want to use Software rendering as a fall-back, if there is insufficient resources to run a plug-in in Hardware? If the FxPlug-in is both software & hardware enabled, it seems that a good Software render is preferable to a bad Hardware render. I'm working with Film resolution mostly (2K-4K). So, I'll likely to be hitting the same problems as Steve is having. -- Brian _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… This email sent to site_archiver(a)lists.apple.com

1 0

Building a generator FxPlug that reads from a device driver...
by Kris van Rens 07 Feb '08

07 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; bh=bhJHrGmOzlXk7dr3rik1gki1SP1QAKoMDcv3G+hj3aE=; b=c8WroQHTnLfVZ2CYmgJ4FsznnQGn24O8iK+jJDwX6FOFoIETDNJOlqIhfrOtgQAHf0GnH10SDpoHfV7uhivxFlWe3TnHOFiO9NrFSPO7jNHNPWYWzNXuKLHtRC6t7MYmHgF9BGhzKCxxLQSTb/Gxdc8is46xdzfYG/JqrAAqKsk= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=ODfZgKyXUMS6X+zPA8qYpEtWmB5OfvDzsez5cNtXQpKMU58Mdv6ZgHzE5g0pYG8QiSHONlQyzMpkxtr6UDZn8HoLQS7aE/NZN/K5XjzDHZ9OapoViHN6+P2mrwZjshmSJV4Zb55iqL+5ljp5i4AN7os82U//gSUCuEyfyxw4gyk= Hi there, I've built a device driver for MacOS X to be able to read and write bulk video data from/ to a USB device, this all works fine. Now I'd like to write a FxPlug plugin to acquire the video data from my USB device directly into Final Cut Pro. I have read some documentation and came to the conclusion that I will have to write a generator FxPlug plugin, this plugin should then communicate with the device driver. I checked out the examples that come with the FxPlug SDK but as far as I saw there was no example of a plugin that does direct communication with a device driver. Are there specific ways to go about dealing with this plugin to device driver communication? Should I implement the communication with the device driver exactly the way I would when building an user application? Would that have to be done in Objective-C or are there ways to partly implement the plugin in C++? I hope my problem is clear, if not; tell me. Thank you in advance for your time, I really appreciate your help. Greetings, Kris _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… This email sent to site_archiver(a)lists.apple.com

1 0

Recommendations for plugins that want to work internally in linear color
by Pete Warden 07 Feb '08

07 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Motion - weird handling of large images
by Steve Christensen 05 Feb '08

05 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com On Feb 5, 2008, at 8:29 AM, Darrin Cardani wrote: On Feb 4, 2008, at 10:18 PM, Steve Christensen wrote: rendering at kFxDepth_UINT8: rowCount = 2648 columnCount = 2276 inputRowBytes = 9104 outputRowBytes = 9104 rendering at kFxDepth_FLOAT32: rowCount = 2048 columnCount = 2048 inputRowBytes = 32768 outputRowBytes = 32768 glTexParameteri (GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE); glTexParameteri (GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE); steve _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… I have been testing some effects on both 8-bit and 32-bit (float) channel images and ran across some weird (to me) behavior. After launching Motion, I dragged a large (2276x2649) Photoshop file into the middle of the workspace and applied a filter. This worked fine with an 8-bit depth. I then opened the project properties window and changed to 32-bit float depth, then clicked the OK button. Motion popped up a sheet saying that the object is too large to render and that it would be cropped to 2048x2048. My filter was called to render, and then the final image looks like the edge pixels were repeated to fill out the original image size. (I didn't do this since the filter was passed a 2048x2048 image. Why am I seeing the difference, especially given that the original image is larger than 2048x2048? I'd expect the same message for the 8-bit case. Here are the numbers Motion is passing to my plugin: The reason you see this message is because Motion has to guess how much of your computer's VRAM your plugin is going to use up during processing. Since it can't know for sure how many pBuffers, etc. your plugin may allocate, it assumes that it will probably not use more than 4 times the size of the image at any one time. So it asks the video card how much VRAM it has, and also what the size of the largest texture it can handle is. Those things are controlled by a combination of your hardware and the video card drivers. Most modern cards can handle textures which are 4k on a side (4096x4096). However, when you switch to 32-bit per channel processing, that takes up 4 times as much memory per texture, so we may estimate that you will be able to process less and we'll still downsample for you. This can happen in 8-bit per channel mode, too, if you're on a card with less VRAM (such as a 9600 with 64 Megs and 2 monitors, for example). There are better ways of handling this situation and we're always trying to improve them, but that's the current situation. Are you working in hardware or software? It sounds like hardware. If that's the case, the clamping to edges is part of the OpenGL state and you can change it. I believe you need to call: You can use just GL_CLAMP to get the behavior you see right now, or GL_REPEAT to tile the image. (And you should probably first save the current state with a call to glGetTexParameteri () and then restore it after your processing.) Let me know if that helps at all or if you have any other questions. OK, good to know for the hardware rendering case, but this effect is running purely in software with no OpenGL calls. The plugin is returning software=YES, hardware=NO to the - [frameSetup:inputInfo:hardware:software:] method. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Motion - weird handling of large images
by Darrin Cardani 05 Feb '08

05 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com On Feb 4, 2008, at 10:18 PM, Steve Christensen wrote: rendering at kFxDepth_UINT8: rowCount = 2648 columnCount = 2276 inputRowBytes = 9104 outputRowBytes = 9104 rendering at kFxDepth_FLOAT32: rowCount = 2048 columnCount = 2048 inputRowBytes = 32768 outputRowBytes = 32768 glTexParameteri (GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE); glTexParameteri (GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE); Let me know if that helps at all or if you have any other questions. Darrin -- Darrin Cardani dcardani(a)apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… I have been testing some effects on both 8-bit and 32-bit (float) channel images and ran across some weird (to me) behavior. After launching Motion, I dragged a large (2276x2649) Photoshop file into the middle of the workspace and applied a filter. This worked fine with an 8-bit depth. I then opened the project properties window and changed to 32-bit float depth, then clicked the OK button. Motion popped up a sheet saying that the object is too large to render and that it would be cropped to 2048x2048. My filter was called to render, and then the final image looks like the edge pixels were repeated to fill out the original image size. (I didn't do this since the filter was passed a 2048x2048 image. Why am I seeing the difference, especially given that the original image is larger than 2048x2048? I'd expect the same message for the 8-bit case. Here are the numbers Motion is passing to my plugin: The reason you see this message is because Motion has to guess how much of your computer's VRAM your plugin is going to use up during processing. Since it can't know for sure how many pBuffers, etc. your plugin may allocate, it assumes that it will probably not use more than 4 times the size of the image at any one time. So it asks the video card how much VRAM it has, and also what the size of the largest texture it can handle is. Those things are controlled by a combination of your hardware and the video card drivers. Most modern cards can handle textures which are 4k on a side (4096x4096). However, when you switch to 32-bit per channel processing, that takes up 4 times as much memory per texture, so we may estimate that you will be able to process less and we'll still downsample for you. This can happen in 8- bit per channel mode, too, if you're on a card with less VRAM (such as a 9600 with 64 Megs and 2 monitors, for example). There are better ways of handling this situation and we're always trying to improve them, but that's the current situation. Are you working in hardware or software? It sounds like hardware. If that's the case, the clamping to edges is part of the OpenGL state and you can change it. I believe you need to call: You can use just GL_CLAMP to get the behavior you see right now, or GL_REPEAT to tile the image. (And you should probably first save the current state with a call to glGetTexParameteri () and then restore it after your processing.) This email sent to site_archiver(a)lists.apple.com

1 0

Motion - weird handling of large images
by Steve Christensen 05 Feb '08

05 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com rendering at kFxDepth_UINT8: rowCount = 2648 columnCount = 2276 inputRowBytes = 9104 outputRowBytes = 9104 rendering at kFxDepth_FLOAT32: rowCount = 2048 columnCount = 2048 inputRowBytes = 32768 outputRowBytes = 32768 _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… I have been testing some effects on both 8-bit and 32-bit (float) channel images and ran across some weird (to me) behavior. After launching Motion, I dragged a large (2276x2649) Photoshop file into the middle of the workspace and applied a filter. This worked fine with an 8-bit depth. I then opened the project properties window and changed to 32-bit float depth, then clicked the OK button. Motion popped up a sheet saying that the object is too large to render and that it would be cropped to 2048x2048. My filter was called to render, and then the final image looks like the edge pixels were repeated to fill out the original image size. (I didn't do this since the filter was passed a 2048x2048 image. Why am I seeing the difference, especially given that the original image is larger than 2048x2048? I'd expect the same message for the 8- bit case. Here are the numbers Motion is passing to my plugin: This email sent to site_archiver(a)lists.apple.com

1 0

Re: questions on pixel formats
by Darrin Cardani 05 Feb '08

05 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com On Feb 4, 2008, at 3:12 PM, Steve Christensen wrote: As of 3.x, yes. I think there were cases in 2.x where they were not. Darrin -- Darrin Cardani dcardani(a)apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… On Feb 4, 2008, at 8:21 AM, Darrin Cardani wrote: Correct. Input and output should always be the same bit depth and channel order. Note that there is a bug in Motion in which the output bitmap will claim to be RGBA, when it's really ARGB, just like the input (when working in software). It's safe to ignore it and output ARGB. 1. So the input bitmaps are advertised with the correct channel order? 2. Is this separate from what the formatsFloatRGBABitmapsAsARGB selector in FxHostCapabilities returns? Sort of. In Motion 2.x, we claimed that float bitmaps were RGBA, but they were really ARGB. At the time, I thought it only affected input bitmaps and bitmaps obtained from the parameter retrieval API or temporal API. I didn't realize it also affected output bitmaps. In Motion 3.x and later, you should no longer need to check that flag. It's only in 2.x that it's an issue. 3. Or in the Motion case are all bitmaps ARGB, independent of the pixel depth? I want to say that all bitmaps should be ARGB independent of pixel depth. Let me know if you find that not to be the case. This email sent to site_archiver(a)lists.apple.com

1 0

Re: questions on pixel formats
by Steve Christensen 04 Feb '08

04 Feb '08

site_archiver(a)lists.apple.com Delivered-To: pro-apps-dev(a)lists.apple.com On Feb 4, 2008, at 8:21 AM, Darrin Cardani wrote: On Feb 3, 2008, at 12:02 PM, Steve Christensen wrote: In FCP, I think it's this: In Motion: 1) From the "Edit" menu, select "Project Properties" 2) Change the "Bit Depth" popup from "8 Bit" to "32 Bit". 1. So the input bitmaps are advertised with the correct channel order? _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.a… I don't remember seeing these specifically addressed so I figured I'd ask so I can make sure I'm doing everything correctly. I'm currently interested in the software rendering case, but if there some differences in the hardware case, that'd be nice to know as well. 1. So far all the images or video I load into a project for testing use 32-bit ARGB/RGBA for the pixel format. If I've read the docs correctly, I should also expect to see 128-bit floating point ARGB/RGBA pixels in some circumstances. If that's so, what can I do to get that to happen so I can test that I'm handling that case correctly? 1) select your sequence in the browser, or click on the timeline window 2) From the "Sequence" menu, select "Settings..." 3) Click on the "Video Processing" tab 4) Click on the "Render all YUV material in high-precision YUV" Thanks. I've already started testing in Motion, and the FCP info will come in handy since I wouldn't necessarily associated that with floating-point pixels. 2. Can I assume that the input and output images will always have the same sized pixels, i.e., I shouldn't expect to see a 32-bit pixel in the input image(s) and a 128-bit pixel in the output image? Correct. Input and output should always be the same bit depth and channel order. Note that there is a bug in Motion in which the output bitmap will claim to be RGBA, when it's really ARGB, just like the input (when working in software). It's safe to ignore it and output ARGB. 2. Is this separate from what the formatsFloatRGBABitmapsAsARGB selector in FxHostCapabilities returns? 3. Or in the Motion case are all bitmaps ARGB, independent of the pixel depth? This email sent to site_archiver(a)lists.apple.com

1 0