I have a fairly simple question. Or at least I thought I would be simple to solve but couldn't find any answers online.
Anyway
I used this example form MSDN to play a file using DirectShow:
How To Play a File.
It's really simple with only a few lines of code and it works.
Then after some research I managed to create a VMR9 filter and add it to the graph. This also worked.
There's just one thing
When I play a video file using the VMR9 filter the quality looks worse in VMR9.
I tried changing it with IVMRMixerControl9::SetMixingPrefs but nothings seems to change. IVMRMixerControl9::SetMixingPrefs does actually return S_OK.
dwPrefs &= ~MixerPref9_FilteringMask;
dwPrefs |= MixerPref9_BiLinearFiltering;
Or am I using the wrong filter?
edit: problem solved
I just did a comparison with Media Player Classic by putting it on VMR9 (windowed) mode. It gave me the same quality. So if I want better quality am gonna have to use EVR (enhanced video rendering) instead of VMR9 (Video Mixing Renderer 9).
VMR-7/VMR-9 quality issues are a long standing problem:
Poor Picture Quality with VMR9 as renderer
VMR9 scaling issues on Vista
EVR is suggested to be used instead, to get proper/nicer scaling and visual picture quality.
In Windows Vista and later, applications should use the EVR if the hardware supports it. Otherwise, fall back to the VMR-9 or VMR-7. The EVR offers better performance and better video quality than previous renderers. Also, it is designed to work with the Desktop Window Manager (DWM).
Better performance is questionable, and EVR is sadly having its own issues though, but when output quality is in question EVR is the answer.
Related
So I have a programming project that I have to do for my school. What I have to do is setup a 2 player dice game. I could have gone the easy way and just display the number of the 2 die, but I was thinking of using images that I made in photoshop instead. However, the problem is that I do not know how to change images in an efficient way.
My first option is using the visibility tag on several images laid on top of eachother and change it accordingly as such
image1.visible = false
image2.visible = true
However, I do not think that is very efficient. Images also do not support changing the image with code from my research.
Secondly, I could use a PictureBox instead, which do support changing the image as the program is running. However, it does not support transparency, and the die images are transparent. Plus it gives me the invalid image file error, I guess due to the transparency in the gif files.
There is also the cheap workaround of me making the background of the images the same as the form background.
So is there a more efficient way I am missing out? I know that the cheap workaround would be the best option for this case, but I would like to have this knowledge for future use like semi-transparent pixels that blend in and such.
And before you ask, no, I cannot use another programming language as visual basic 6 is what my school teaches. Thankfully they are changing it soon, but I am stuck with this for now.
Turns out you CAN change the pictures of Images, while keeping transparency and stretch. I am going to properly show it:
Image1.Picture = LoadPicture("YOURPATHHERE.gif")
This is what I get for believing what I've seen on some forum.
Also, the error of invalid image file was due to the images being corrupted for some reason.
Does they have good algorithm to detect porn via detecting skin Colors frame by frame or they have human intervention ?
There exists a Porn Video Classifier/Editor with Caffe which uses Deep Learning, is Open Source and seems well documented.
I want to add a video to my game in MASM32. I looked it up on the web, and got zero results. How do I add a video in MASM32 and display it out on screen?
Short answer... It can be done... but it's not going to be easy. You'll probably want to use ffdshow (very good compatibility with all popular video formats), then learn about COM (common-object-model) programming (in MASM32) first. Then read this post to try Yves's approach.
Link to the project
Hey guys. I'm a longtime lurker, and usually like learning about this stuff on my own, but I figured I would see if someone (especially someone involved with the active open source project or its dependencies) would happen to know something I am really curious about.
If you compile the open source project, Audacity, for Windows and export any amount of silence, the resultant Wav has noise in it. By this, I mean (what seems at first glance) to be random noise appears in the resultant file. Everytime. And this noise varies from export to export - which I did by exporting the same content twice in a row without doing anything inbetween.
So far, the only other thing I have tested is Steinberg's Nuendo 4, and the same methodology produces nice, clean Wav files full of 00 every time.
Since this is my first foray into having the awesomeness of an actively maintained project, I figured somebody would have some knowledge about it, or at least let me know where else to post such a question. I am going to crosspost this to the Audacity forums also, and start there.
I managed to get this built in VS2010, and with DirectSound support, so can provide all the information to do so if anyone wants to try it out.
Cheers!
The default settings for (export) quality have dithering activated, which adds a little bit of random noise to improve the sound quality of a downconversion to 16bit WAV.
If you turn off dithering you get all 0 on an export of total silence, but for real world audio the sound quality is better with dithering on.
I'm running Audacity in German language, so I can't give you the exact menu terms, but you should be able to find it like this:
Open the settings via the Edit/Settings Menu.
Quality is in option 4 and there should be a setting for high quality conversion where you can select the kind of dithering you want. Here you can turn it off.
I rely on ImageCR3 for a number of my sites. However, I've come up against a variety of limitations with it over the past couple years (single-threaded, no crop anchor, etc) and all of my emails to the support addresses have been ignored. So I'm looking for an alternative.
My first thought was CFImage, but it seems to produce far too low of quality for the same image size, and seems excessively slow. Is there any other tool out there that can do what ImageCR does, as efficiently as ImageCR does it, that I could use instead? Or am I best off loading the jpg in CFImage, cropping and saving as PNG, then loading the PNG in ImageCR for the remainder of editing?
I'm using ColdFusion MX 7 and ColdFusion MX 9 (soon all to be migrated to the latter).
ImageMagick? Seems to work like a charm.