Articles Technology Blog News Forum Company

Troubleshooting

"Can't process with sound" problem.

Why do I get low quality?

File size is too big!

Video and audio go out of sync.

"Pins are not connected" problem.

Preview in filters doesn't work.

XviD compression doesn't work.

The program eats up too much memory (hundreds of MB).

 

General questions

What is super-resolution?

How to use Video Enhancer?

How to remove mosaic from video?

How to remove pixellation from video?

Are there any objective quality measurements of different upsize methods?

How to upsize interlaced video?

What filters to use with SR?

What file formats does Video Enhancer support?

How to enhance the resolution 4 times?

How about capturing and frame serving in Video Enhancer?

What codecs to use?

Multithreading: using VirtualDub filters on dual core and multi CPU.

 

Troubleshooting

"Can't process with sound" problem

Versions of Video Enhancer prior to 1.5 used standard Microsoft AVI muxer which cannot work with VBR (variable bitrate) audio. Unfortunately, a lot of video now has VBR audio streams. That caused Video Enhancer to refuse to write audio to result file. Now, starting from version 1.5, it uses its own AVI writing mechanism which allows different audio streams, including popular VBR MP3. If you still have problems with copying audio stream to result file, try choosing an audio codec, this usually helps.

File size is too big!

Usually the video you work with is compressed with a video codec. When Video Enhancer works with video, it decompresses it using a codec installed in your system. Then it uses codec you chose for video compression to compress result and stores it to an AVI. When you run VE, there's no video codec already chosen for compression. If you forget to choose one, the video will be stored uncompressed which takes a huge amount of disk space. So, just don't forget to choose a video codec and configure it for proper bitrate.

Video and audio go out of sync.

There are three known cases when this can happen.
1. In versions prior to 1.5 when processing some WMV files video could change speed. This was due to inconsistency between Microsoft WMV decoder and Microsoft AVI muxer in terms of timestamps for video. Starting with Video Enhancer version 1.5 it uses its own AVI writer that tries to keep proper video speed even in case of incorrect timestamps and missing frames (also happens with WMVs).
2. If you don't compress output video your hard disk may be not fast enough to read video with proper speed when playing. This will lead to video being slower than sound. The solution will be to compress video.
3. If you compress the video but when playing decoder can't handle the amount of work (usually consuming 100% of CPU), this leads to the same effect of slowing video down. The solution is to lower output bitrate or use another codec.

"Pins are not connected" problem.

In some cases when you start processing a video, Video Enhancer refuses to process it saying something about pins not connected. This usually means a problem with decoder. Video Enhancer uses codecs installed in your system. If there's no proper codec found for your video or the codec cannot decode to RGB or refuses to function properly, mentioned problem might occur. Try to find anther codec for this type of video.

Preview in filters doesn't work.

This function is just not implemented. You can see filter's results during processing.

XviD compression doesn't work.

In XviD version 1.1 to get it working one needed to change one option in Xvid's parameters (this needs to be done just once). When you choose Xvid for compression in Video Enhancer, press "Configure", then "Advanced options", "Debug" and switch off "Display encoding status". After you've done it once, Xvid will work in VE.

The program eats up too much memory (hundreds of MB).

This was a bug in version 1.5.0 published 05/13/07, the bug was fixed couple of days later, just
download the new version.

 

General questions

How to upsize interlaced video?

Super-resolution method in Video Enhancer works only with progressive video. If your original video is interlaced (like DV), you've got two options: deinterlace it before SR, or keep proper interlacing.
1. To deinterlace video before upsizing switch to Advanced mode, add one of deinterlacing VD filters (I personally like Deinterlace "Muksun") and then add SR filter. The output video will be progressive.
2. You can keep interlacing by using filter ViewFields before SR and UnViewFields after SR. These filters decompose interlaced frame into two separate pictures (fields) one under another and then combine them back. So each field will be upsized independently.
All necessary filters can be easily installed at once in
VirtualDub Filter Pack.

What filters to use with SR?

In some cases some pre- and post-filtering can really help. For example, if you deal with a strongly compressed MJPG video (which most digital photocameras produce today in video mode) the video has blockiness - divided into squares 8x8 pixels. There are some deblocking filters that can improve quality of such video. Another case is noisy video. Video Enhancer thinks of noise as a part of picture and resizes the noise as well, making it more annoying. And since in each frame the noise is different it makes harder to get right information from other frames to resize current one. So denoising filters can really help perform SR better. All these filters need to be used before SR pass. Sometimes applying sharpen filter after SR can improve visual quality.

Some filters, like deinterlacers and deblockers, must be used before super resolution. Some, like denoising and color correction, should be used before SR. And some, like sharpening, are worth using any time. When you know what filters do, it's easy to decide their order. And sometimes this is not obvious and only practice, trial and error can tell what's best.


What file formats does Video Enhancer support?

Video Enhancer uses basically the same technology for opening video as Windows Media Player. That means if you are able to play your video in WMP then most probably you'll be able to process it in Video Enhancer. By default it understands AVI, WMV, ASF and MPEG. It can open MOV files if
Quicktime Alternative is installed, and RM and RMVB files if Real Alternative is installed.

The output is always stored in AVI format.


How to enhance the resolution to 4x?

Each pass of Super Resolution in Video Enhancer can resize video to any size, but what is really happening is applying Super Resolution to double the picture size and then resizing the result to desired size using an image scaling method. That means if you try to increase resolution 4 times, then single pass will apply SR to increase it 2 times and then a simpler in-frame method to increase it 2 times more. In this case a chain of two SR filters each increasing resolution 2 times will do better but much slower of course because each pass of SR requires estimating motion and doing lots of computations.

If you tell SR filter to resize to the same or smaller size as the original, then no actual Super Resolution is done, just simple in-frame resizing. This means repeating SR without increasing the size will not enhance the quality. And also this means you can use SR filter to decrease picture size, this is done quite fast and with good quality (Lanczos filter is used internally).


How about capturing and frame serving in Video Enhancer?

Capturing video is not implemented nor even planned.

Frame serving is a convenient feature and ability to frame serve is planned for near future versions.


What codecs to use?

MSU codecs are lossless, that means decompressed video is identical to the original one. Lossless compression cannot be very strong, so the file size is so big. Also, MSU Lossless Video Codec is slow, it can't play high resolution video in realtime. Its main purpose is maximum lossless compression for storing.

For most tasks usually lossy video codecs are used. They compress much better by taking out some detail, degrading video quality. They differ by quality/compression ratio.

I think the best lossy video codec today is x264. It is an open source implementation of H.264 video compression standard. But x264 itself provides only compression, you need to have a decompressor either, which can be found in ffdshow.

If this is too complicated, I recommend you to use MPEG4 codecs: DivX and XviD. They are both implementations of MPEG4 standard, the first one is commercial and the second one is free and open source.