On Tue, 10 Jun 2003, Steven M. Schultz wrote:
Hi -
From: Selva Nair [EMAIL PROTECTED]
As the source is progressive ppm frames, better keep it that way.
Use -F 24000:1001 and encode with pull down flag:
Well, as it turns out it seems that the encoder is doing the right
On Tue, 10 Jun 2003, Steven M. Schultz wrote:
Hi -
From: Selva Nair [EMAIL PROTECTED]
Forgot to mention that if this is meant to be 4:3 aspect, you should
scale to 704x480 and then pad to 720x480. You can avoid vertical scaling
using 640x480 ---(scale)-- 704x480 ---pad-- 720x480.
Hi -
From: Selva Nair [EMAIL PROTECTED]
But why do you care about top_field first when the source is progressive?
I don't care which order the fields come in - could have been the DV
standard's bottom first and I'd have been just as happy.
If you use 24000:1001 and pulldown
Hi -
From: Selva Nair [EMAIL PROTECTED]
The general advice is to avoid vertical scaling mainly because vertical
scalers should be interlacing-aware -- also for interlaced frames the
and y4mscaler is, I believe, interlacing aware ;)
Downscaling feels more correct. Removing
Here's is my 5 cents on various aspects of this thread:
From: Selva Nair [EMAIL PROTECTED]
...
The general advice is to avoid vertical scaling mainly because vertical
scalers should be interlacing-aware -- also for interlaced frames the
vertical samples per field is only half as much. For
Hi there,
On Wed, 11 Jun 2003, Steven M. Schultz wrote:
If you use 24000:1001 and pulldown instead of 3:1001, the encoder will
keep switching between top_field_first = true/false to generate the
pulldown sequence.
Now it's my turn to ask why would I want to do a pulldown when
From: Steven M. Schultz [EMAIL PROTECTED]
...
Downscaling feels more correct. Removing information is preferable
to synthesizing it.
(But in this case you may end up removing more information than you would
have started with by taking the synthesis route.)
may not matter
On Wed, 11 Jun 2003, Steven M. Schultz wrote:
Hi -
From: Selva Nair [EMAIL PROTECTED]
The general advice is to avoid vertical scaling mainly because vertical
scalers should be interlacing-aware -- also for interlaced frames the
and y4mscaler is, I believe, interlacing aware ;)
Hi -
From: Matto Marjanovic [EMAIL PROTECTED]
I guess Martin Sitter read a lot of books on Photoshop, instead of reading
books on video engineering. NTSC TV's use 10:11 (as referenced to the
industry standard square pixel aspect ratio).
That might be - and I thought he was making
Anyone used ppmtoy4m to create composited from PPM images
(for example from a scanner/camera) for a DVD?
It's working - but what I'm ending up with is 720x480 progressive
and that's not, as far as I know, legal for use in making a DVD.
What I'm doing so
Hi -
From: Selva Nair [EMAIL PROTECTED]
As the source is progressive ppm frames, better keep it that way.
Use -F 24000:1001 and encode with pull down flag:
Well, as it turns out it seems that the encoder is doing the right
thing for me (despite my being confused ;)).
Hi -
From: Selva Nair [EMAIL PROTECTED]
Forgot to mention that if this is meant to be 4:3 aspect, you should
scale to 704x480 and then pad to 720x480. You can avoid vertical scaling
using 640x480 ---(scale)-- 704x480 ---pad-- 720x480.
Hmmm, I was told (at one time) to avoid
12 matches
Mail list logo