TMPGEnc - Difference between CQ and CQ_VBR?

[Hulk]

I don't understand the difference between Constant Quality and Automatic Constant Quality. I would think that by definition CQ would have to be VBR to maintain constant quality.


Also, is there a tool that allows you to graphically look at the bit rate allocation for a VBR MPEG-2 video stream?

[Dr Mordrid]

Constant Quality (CQ) is a single pass encoding that varies the bitrate, but attempts to keep the perceived quality of the final video the same

Automatic VBR (CQ_VBR) is also a single pass encoding that is used which automatically adjusts the video bitrate up during high motion sequences and down during low motion sequences.

The key here is that both try to maintain constant quality but in slightly different ways. CQ_VBR is best used in videos where the motion covers a wide range while CQ is better for talking heads and other relatively low motion sequences. If the motion changes get too tough for CQ_VBR then it's time to bring on 2-pass_VBR.

In 2-pass_VBR the first pass is used to create a table of the complexity of each frame. In this table frames with a high complexity are allocateded more bits. The second pass does optimized encoding of each frame using the tables info.

There is supposed to be a "server" version of TMPGEnc in the works that can do up to 20 passes, which is very much a pro feature.

The tool you're looking for is called BitRate Viewer.

The free version will show you the bitrate allocations and quality setting over time using a graphed timeline along with most of the MPEG's settings & properties in a side panel.

The full verserion (E29.95/~$29.00 USD) adds the following features (their bad English );

GOP structure informations
Seconds informations
Printing ability
Simple I frame trimming for elementary streams
Can be save the once read stream's datas and reload that later, you do not need the original stream anymore.
Can modify some basic parameters of MPEG stream.
Scene change detection and Variable GOP Pattern can be detected
Bitrate Calculation tool
Frame compression/quantization scale view
AVI file reading if the compression is standard MPEG
User support what is not available for the freeware version users.

Excellent tool. You can get BitRate Viewer here:

http://tecoltd.com/bitratev.htm

Dr. Mordrid

[Hulk]

Interesting. Thanks for the info. I'm currently compressing a beach patrol video with lots of waves/ocean, tough for the encoder. I'm experimenting with various settings, your advice will help me hone in on the best setting.

[Hulk]

To what does the DC precision value refer?

[Dr Mordrid]

DC precision controls the accuracy of the discrete cosine transform values used for image compression. Generally speaking a higher value results in better quality.

DC precision in the better encoders can be set from 8 to 11.

For MPEG-1 DC is restricted to 8.

For MPEG-2 encoded as Main Profile of either Low or Main Level or 4:2:2 Profile/Main Level (aka: "4:2:2 Studio Profile") you can use a DC value of 8, 9 or 10. Most times you would stick to 9 or 10. I always use 10.

A DC of 11 is normally only used for High Profile/High Level and High Profile/High-1440 encoding, but these profiles can also use 8, 9 or 10 if desired.

Dr. Mordrid