|
 |
 |
 |
 |
 |
 |
- What about command-line interface? Does MSU VQM support it?
- Your program reports that it failed to open input file.
- How can I obtain version for Linux?
- What values correspond to better quality for each metric?
- Where I can get more information about your own metrics (MSU Blocking/Blurring measure)?
- I recieve very strange results for AviSynth scripts. What's wrong?
- Why average value of the PSNR is not equal to arithmetic mean of the PSNR values for frames?
- Is it possible, to open video files, which are not supported by MSU VQMT directly (.mp4, .mpg, .mov etc.)??
- Results are very strange (too low) for one ore more codecs?
-
Q: What about command-line interface? Does MSU VQM support it?
A: Public version doesn't support command line. We have such feature in PRO version, but it's available only with
commercial license.
-
Q: Your program reports that it failed to open input file.
A: Our tool supports .avi, .yuv, .avs, .bmp files as input. Other filetypes
(.wmv, .vob, .mpeg, etc.) can be opened through AviSynth. Detailed information can be found in measure_files_with_directshow.avs and measure_mpeg1_mpeg2.avs in directory with the tool.
If you experience a problem with .avi, then it is possible that MSU VQM is not
compatible with this codec. Please send information about
this codec, or part of such avi file (1-2 megabytes) to our e-mail.
-
Q:How can I obtain version for Linux?
A: Sorry, our program is Windows-only now.
-
Q: What values correspond to better quality for each metric?
A:
|
Metric
|
Values for sequences
|
Interpretation
|
|
Blocking measure
|
1 value for 1 video
|
Lower values corresponds to lower blocking
|
|
Bluring measure
|
1 value for 1 vide, 1 visualisation video for 2 input sequences
|
Lower values corresponds to higher blurring.
|
|
Delta
|
1 value for 2 sequences
|
0 means, that frames are equal. Positive and negative values means deviation. Lower absolute values are better.
|
|
MSAD
|
1 value for 2 sequences
|
0 mean equal frames. Lower values are better.
|
|
MSE
|
1 value for 2 sequences
|
Lower values are better. 0 for equal frames.
|
|
SSIM
|
1 value for 2 sequences
|
Higher values are better, 1 for equal frames
|
|
VQM
|
1 value for 2 sequences
|
0 for equal frames, lower values are better
|
PSNR
|
1 value for 2 sequences
|
100 for equal frames, higher values are better
|
-
Q: Where I can find more information about your own metrics (MSU Blocking/Blurring measure)?
A: We plan to prepare some paper first and next publish this
measurement methods on our webpage.
-
Q: I recieve very strange results for AviSynth scripts. What's wrong?
A: AviSynth support is unstable, because we have no reference source code for working with AVS files.
If measurements are wrong try this:
- Check our recomendations for AVS scripts.
- Specify output colorspace using ConvertTo() instruction.
-
Q: Why average value of the PSNR is not equal to arithmetic mean of the PSNR values for frames?
A: The correct way to calculate average PSNR for the sequence is to calculate average MSE for all frames
(average MSE is arithmetic mean of the MSE values for frames) and after it calculate PSNR using ordinary equation for PSNR
)
-
Q: Is it possible to open video, which are not supported by MSU VQMT directly (.mp4, .mpg, .mov etc.)?
A: Our program supports AviSynth, which can be used to open such files. (it could be found in archive with program). Some formats (such as DVD) requires additional plugins for
AviSynth. See AviSynth website for more information.
-
Q: Results are very strange (too low) for one ore more codecs?
A: Some codecs (for example DivX 6.0) shift video file one or two frames
back, or make two first frames equal. Check for this shift in your video files after encoding.
You can open two encoded videos in two different VirtualDubs.
After it go to the 10th frame, for example. And at last switch between
them to see are 10th frames in two videos corresponded or not.
If shift really exists, use AviSynth scripts with DeleteFrame() and DuplicateFrame() to fix it.
| e-mail: |
 |
See also MSU Video Quality Metric
|
 |
 |
 |
 |
 |
 |
|