Posted by Ken Maltby on 01/10/27 11:53

This thread is a hoot!


OK, to address the largest number of posts first;
MPEG2 was created to properly deal with interlaced
broadcast video. MPEG1 with only 240 lines couldn't
do it, MPEG2 with 480 could. As long as they were
at it, they also implemented VBR to greatly improve
compression potential. MPEG1 is CBR progressive
only. MPEG2 can do VBR and interlaced video.
( This is the main reason that if you use only 240lines
with MPEG2 you should make it progressive.)

Now the OP's DVR: The encoding options that are
offered by such consumer devices vary a great deal.
There is nothing with the two formats that require the
limited settings available with DVRs or PVRs or TiVos
or whatever.

You were right you could just pick an image size and
adjust the file size by the bitrate setting. (If your DVR
were to provide you access to such a control)

It is all a matter of what the manufacturers feel the
largest number of consumers will tolerate in terms of
how complex it is to use their product. It is not a
matter of the hardware. For instance the TiVo Series 2
uses the Broadcom BCM7040 Encoding chip. I have
that same chip powering my PCI capture card and I can
set a wide range of parameters for the encoding. Even to
the extent that I can capture to "I" frame only GOP and
frames per GOP up to 30. I doubt any such adjustments
are available on the TiVo. To address your questions I
can also adjust the image size and bitrate directly.

As to why the switch to a smaller image size when
lowering the bitrate sometimes, it is in an attempt to keep
the "bits per pixel" in an acceptable range.

You could look at it this way:

Your bitrate says how many bits of image data can be put
into your file per second.

Your frame rate tells you how many frames there will be
each second. (Since that stays the same, we can look at
it as if it were 1 frame)

Now it stands to reason that, the more bits you can use to
paint the frame the better it well look. Consider also, how
much area there is to cover with the bits.

If you are trying to have a smaller file all you need do is lower
the number of bits being put into your file every second. Now
if you do that, at some point you will start to not have enough
bits to paint the frame, with same amount of detail. So there
comes a time when it is a good idea to have a smaller frame to
paint, and not need as many bits. ( Of course the smaller frame
can't hold as much detail as the larger one, but if the bitrate drops
too low on the larger frame it can lose detail and make a mess of
things, as well.)

[ There are other technical factors like Quantization and the effort
involved in motion estimation; that also benefit from having a
smaller frame at low bitrates.]

So the manufacturers decided that most consumers would not
have an interest in learning how to address such issues and they
just provide those settings they thought most people would
like/want. They figured that they would be in trouble if they
made their product any more complex than a VCR.

So, you could consider yourself lucky if you have more
settings than SP, LP and EP.

Luck;
Ken

[Back to original message]