Miro M-JPEG file --> MPEG Conversion Problems
David Longley
I have been experimenting with the XING MPEG encoder. However:
It will only handle a max of 640 x 480, which means half
resolution, single field (ie DC1 size).
When I try it, and feed the resulting AVI file from capture
(presumably in M-JPEG format), to the MPEG encoder (mir/xing),
the sound is out of sync. How do I prevent that? Is it something
to do with the capture /data rate I am specifying at capture? Is
it the frame rate???
Also, there is noise on the sound capture....why does that
happen? If I capture at 11Khz and 16bit mono, the raw playback in
Premiere is fine, but once converted by MPEG maker, it's swaful.
I get better results if captured at 44KHz and 16 bit mono and
then let the MPEG maker do its stuff, but I don't appreciate the
logic of this..
--
David Longley
John Anderson
In article <834924...@longley.demon.co.uk>, David Longley <Da...@longley.demon.co.uk> says:
>Also, there is noise on the sound capture....why does that
>happen? If I capture at 11Khz and 16bit mono, the raw playback in
>Premiere is fine, but once converted by MPEG maker, it's swaful.
>I get better results if captured at 44KHz and 16 bit mono and
>then let the MPEG maker do its stuff, but I don't appreciate the
>logic of this..
You don't say what bitrate and sample rate you're using for the MPEG
audio. MPEG audio can only be 32, 44.1, and 48KHz. The MPEG audio
encoder must resample the 11KHz audio to one of these frequencies. It
may not do a good job of resampling.
John Anderson
David Fowler
I am going to upgrade my vesa 1meg vidio card to a pci 2meg vidio card.
I do some games but not a lot. Is the difference between vram and dram
going to be that big of a deal. Also am interested in possibly mpeg
support since I am seeing this more and more and seems to be used on
some games.
Thanks in advance
Dave, fow...@ix.netcom.com
Tristan RAMBO Savatier
This is correct (for MPEG-1 audio). You can use SoX to do the
rate conversion.
I am not an MPEG audio expert, but I believe that
MPEG-2 audio supports lower frequencies in its NBC mode
(NBC = non backward compatible with MPEG-1).
--
Regards,
-- Tristan Savatier (email:tri...@creative.net - http://www.bok.net/~tristan)
"My karma ran over my dogma"
David Longley
One of the problems has been a lack of sound understanding on my
part.
1. I am capturing PAL from a source which is VHS and non HI-FI
audio. The video is basically "talking heads".
2. Using the Miro DC20 and Premiere 4.2 on a Pentium 166 with
32MB EDO RAM and two 2GB drives (a 7200rpm Seagate barracuda for
capture and a Western Digital drive for rendering.
3. I have set the capture to 15 fps and the compression/data rate
to 300kbs as I am assuming this will give me maximum recording
time for the disk space available (2GB maximum under W95).
4. Audio is being captured at 16 bits 22Khz. If I reduce to 8
bits I get sound like an old radio.
5. When I then convert with the objective of rendering for dual
speed CD, I am using CINEPAK at 30% with and audio/video
interleave of 1 frame with audio at 11Khz and 16 bits. Data rate
is set as maximum at 300kbs.
6. The problem with the Xing MPEG encoder I have used is that it
does not let me edit the fps from 25 for PAL, which means I have
to origibally capture at 25cps too. If I do this, and then use
Xing to convert to MPEG, the audio and video tend to go out of
sync.....is this avoidable with DC20 generated 1/2 PAL?
With these factors in mind, do you have any suggestions for
improvements?
--
David Longley
Paul Masters
David,
I e-mailed Xing and they said ring Miro. Miro said ring Xing !
If you discover a way of encoding please let me know at PMas...@Cellnet.co.uk
I'll return the complement if I discover it first.
Regards,
Paul.
Thomas Borowski
fow...@ix.netcom.com(David Fowler ) wrote:
Be sure to get VRAM, because it's faster than DRAM. OK, you'll have to
pay a couple of bucks more, but if you want quality...pay for it.
thomas borowski
spac...@tuebingen.netsurf.de
...
Errare humanum est.
But you need a computer to really mess things up.
Gregg E Economou
Yes, there is a difference.
The only mail difference between dram and vram is this:
dram requires a refresh every so often, because it stores charge in
capacitors which decay after a while.
When you read a cell of DRAM, it is refreshed at the same time by the nature
of a DRAM read.
Video memory is read very frequently and unfiormly anyways, because the display
is updated many times a second.
So, manufacturers can save money by removing the refresh mechanism
from the video memory.
chip density goes a bit up,
power consumption goes a bit down,
and the main plus: its cheaper to make.
speedwise, there is no difference between vram and dram.
you can use dram in place of vram in most cases
but you cant use vram in place of dram in most cases.
Theres nothign magic about either one.
Isildur
Andy Wilson
true, but the other salient fact about video memory, apart from the fact
that you read it out frequently, is that you want to be able to write it
while it's being read out. therefore VRAM is dual ported; therefore
it's actually a little more expensive to make a VRAM of a certain size
than to make a DRAM.
In theory EDO DRAM, which, like VRAM, can read out extended lines of
data without requiring extra wait states, is nearly as fast as VRAM for
many graphics applications. Still -- personal opinion -- if I were in
the market for the absolute hottest graphics card for my PC and didn't
mind paying a little extra, I would look for VRAM.
cheers
atw
not speaking for intel....
Robert Cook
Gregg E Economou (gee...@pitt.edu) wrote:
: Yes, there is a difference.
: The only mail difference between dram and vram is this:
: dram requires a refresh every so often, because it stores charge in
: capacitors which decay after a while.
: When you read a cell of DRAM, it is refreshed at the same time by the nature
: of a DRAM read.
: Video memory is read very frequently and unfiormly anyways, because the display
: is updated many times a second.
: So, manufacturers can save money by removing the refresh mechanism
: from the video memory.
: chip density goes a bit up,
: power consumption goes a bit down,
: and the main plus: its cheaper to make.
: speedwise, there is no difference between vram and dram.
: you can use dram in place of vram in most cases
: but you cant use vram in place of dram in most cases.
: Theres nothign magic about either one.
: Isildur
I believe you are mis-informed. Some things you say are true: DRAM is
cheaper, DRAM required refreshing, reading the DRAM refreshes it. You have
, however, only addressed the output portion of the display. If all you want
to do is draw a static display, then DRAM is the most economical way to go.
Speedwise, there is a big difference. It has to do with writing to the memory,
and available bandwidth. If 100% of the memory's bandwidth is consumed in
reading it to produce the display, then you can't write to it while it draws.
If reading the memory to draw the display consumes 50% of the memory's
bandwith, then you might have to wait a while to write to the display. Dual
ported memory can be written to and read from at the same time, and for a
display system, this is what you want...VRAM... No waiting while the display
is being drawn.
Regards,
Robert Cook
Stephen Frearson
In article <54jj8l$l...@usenet.srv.cis.pitt.edu>, Gregg E Economou
<gee...@pitt.edu> writes
>Yes, there is a difference.
> The only mail difference between dram and vram is this:
>dram requires a refresh every so often, because it stores charge in
>capacitors which decay after a while.
>When you read a cell of DRAM, it is refreshed at the same time by the nature
>of a DRAM read.
>Video memory is read very frequently and unfiormly anyways, because the display
>is updated many times a second.
>So, manufacturers can save money by removing the refresh mechanism
>from the video memory.
>chip density goes a bit up,
>power consumption goes a bit down,
> and the main plus: its cheaper to make.
>speedwise, there is no difference between vram and dram.
>
>you can use dram in place of vram in most cases
>but you cant use vram in place of dram in most cases.
>
>Theres nothign magic about either one.
>
>Isildur
>
DRAM, and is duel ported - meaning it can be read and written to at the
same time! DRAM cannot. VRAM is more expensive than DRAM.
--
Stephen Frearson
Joseph H Allen
Oh boy, you people are clueless. Rarely have I seen so much misinformation.
Both VRAM and DRAM require refresh. The screen refresh does not usually
cycle through the ram fast enough to refresh it (this goes for both VRAM and
DRAM), although it may be possible in certain architectures. Ram is usually
arranged as 512 rows by 512 columns by 4 - 16 bits. Each of the 512 rows
must be accessed in 8ms to refresh the VRAM. Screen update takes 16.67ms
for a 60Hz system. A single row could correspond to 1/2 to 4 screen lines
depending on the architecture, so it may or may not be able to do it
(usually not). In any case it doesn't matter because explicit refresh takes
an insignificant amount of time.
VRAM and DRAM are completely incompatible unless the board manufacture made
special arrangements for accepting either kind (which is not usually the
case). VRAM is dual ported, which implies that an entirely different design
is needed to use it compared to DRAM.
VRAM has about 2.5 times the bandwidth of fast-page-mode DRAM of the same
speed and width, but the video port on the VRAM is limited to sequential
access within each row (which is just fine for video screen refresh). VRAM
has the added benefit that the address bus is not critical (you only need a
single clock line to operate the video port, not an external address
pointer). This means that is more likely that you'd be able to actually get
the maximum speed of the video port, compared to the maximum speed of a
fast-page-mode data port.
The actual bandwidth for a 64-bit bus are 176 MB/s for DRAM and 440MB/s for
VRAM. A 1280x1024 256 color display needs 100 MB/s, and whatever is left
over can be used by the computer.
The cool thing about VRAM is that internally it transfers an entire row in
one 140ns cycle to the video port. That's a bandwidth of 256 - 1024 bytes /
140ns which comes out to 1.8 - 7.3 billion bytes per second! Unfortunately,
once the data is in the video port it can be read out of the chip only at 16
- 66 million bytes per second (33MHz multiplied by the word size). However,
this vast untapped bandwidth leads to other architectures. Synchronous DRAM
can operate up to 100MHz (100 - 200 MB/s). RamBUS DRAM can operate up to
500MHz, but I'm not sure if it can transfer an entire row in one 500MHz
burst (would be cool if it did). Still, none of these come close to the
7.3GB/s that's available. It would be very cool to see a DRAM and some kind
of 2d or 3d accelerator engine on the same die. This may be comming because
NEC (I think) is going to offer DRAM as an ASIC library part.
--
/* jha...@world.std.com (192.74.137.5) */ /* Joseph H. Allen */
int a[1817];main(z,p,q,r){for(p=80;q+p-80;p-=2*a[p])for(z=9;z--;)q=3&(r=time(0)
+r*57)/7,q=q?q-1?q-2?1-p%79?-1:0:p%79-77?1:0:p<1659?79:0:p>158?-79:0,q?!a[p+q*2
]?a[p+=a[p+=q]=q]=q:0:0;for(;q++-1817;)printf(q%79?"%c":"%c\n"," #"[!a[q-1]]);}
Peter Choi
Gregg E Economou <gee...@pitt.edu> wrote in article
<54jj8l$l...@usenet.srv.cis.pitt.edu>...
> Yes, there is a difference.
> The only mail difference between dram and vram is this:
> dram requires a refresh every so often, because it stores charge in
> capacitors which decay after a while.
> When you read a cell of DRAM, it is refreshed at the same time by the
nature
> of a DRAM read.
> Video memory is read very frequently and unfiormly anyways, because the
display
> is updated many times a second.
> So, manufacturers can save money by removing the refresh mechanism
> from the video memory.
> chip density goes a bit up,
> power consumption goes a bit down,
> and the main plus: its cheaper to make.
> speedwise, there is no difference between vram and dram.
>
> you can use dram in place of vram in most cases
> but you cant use vram in place of dram in most cases.
>
> Theres nothign magic about either one.
>
> Isildur
>
i always thought vram was faster than dram, if they're the same, why is
vram a lot more than dram? what about wram and sgram? what's the story on
those? i know wram is just dual ported dram, but what does "dualported"
actually mean?
Yok pang Ngo - INEN/W94
Robert Cook (rc...@skopen.dseg.ti.com) wrote:
: Gregg E Economou (gee...@pitt.edu) wrote:
: : Yes, there is a difference.
: : dram requires a refresh every so often, because it stores charge in
: : capacitors which decay after a while.
: : When you read a cell of DRAM, it is refreshed at the same time by the nature
: : of a DRAM read.
: : Video memory is read very frequently and unfiormly anyways, because the display
: : is updated many times a second.
: : So, manufacturers can save money by removing the refresh mechanism
: : from the video memory.
: : chip density goes a bit up,
: : power consumption goes a bit down,
: : and the main plus: its cheaper to make.
: : speedwise, there is no difference between vram and dram.
: : you can use dram in place of vram in most cases
: : but you cant use vram in place of dram in most cases.
: : Theres nothign magic about either one.
: : Isildur
: I believe you are mis-informed. Some things you say are true: DRAM is
:
Mark Anderson
In article Joseph H Allen <jha...@world.std.com> wrote:
>VRAM is dual ported, which implies that an entirely different design
>is needed to use it compared to DRAM.
You forgot to explain "dual ported" which is the major difference.
Dual ported simply means that your hardware can read and write the
memory at the same time. This is important for video monitors because
the display hardware must constantly read the memory for each pixel
it fires on the screen. In order to change what's on the the screen,
the hardware must write to the video memory. VRAM simplifies the board
design and display logic by being able to write on a separate bus.
A DRAM design, which is either read or written to, requires some
kind of scheduling so that writes do not interfere with the reads that
operate in real time. That raster putting dots on your screen never
stops or slows down.
Hope this helps somewhat. There's probably a lot more video specific
stuff that they put into VRAMs nowadays. I haven't followed that market
in years.
Joseph H Allen
In article <54ole4$a...@xochi.tezcat.com>, Mark Anderson <m...@tezcat.com> wrote:
>In article Joseph H Allen <jha...@world.std.com> wrote:
>>VRAM is dual ported, which implies that an entirely different design
>>is needed to use it compared to DRAM.
>You forgot to explain "dual ported" which is the major difference.
>Dual ported simply means that your hardware can read and write the
>memory at the same time.
It would be more accurate to say that two things can access the memory at
once (they don't necessarily have to be read and write, but that's what
counts for video card frame memory).
>A DRAM design, which is either read or written to, requires some
>kind of scheduling so that writes do not interfere with the reads that
>operate in real time. That raster putting dots on your screen never
>stops or slows down.
Well, both DRAM and VRAM are going to require scheduling. The transfer of a
row to the video port requires a command on the DRAM side. Once this
transfer occurs, then both sides can be accessed simultaneously until
another row is needed.
A VRAM design may also be simpler because you can probably get away without
having a line buffer in the video controller chip. Complex video
controllers (VGA) may not be able to get away with this though.
Mike
Stephen Frearson wrote:
>
> In article <54jj8l$l...@usenet.srv.cis.pitt.edu>, Gregg E Economou
> <gee...@pitt.edu> writes
> > The only mail difference between dram and vram is this:
> >dram requires a refresh every so often, because it stores charge in
> >capacitors which decay after a while.
> >When you read a cell of DRAM, it is refreshed at the same time by the nature
> >of a DRAM read.
> >Video memory is read very frequently and unfiormly anyways, because the display
> >is updated many times a second.
> >So, manufacturers can save money by removing the refresh mechanism
> >from the video memory.
> >chip density goes a bit up,
> >power consumption goes a bit down,
> > and the main plus: its cheaper to make.
> >speedwise, there is no difference between vram and dram.
> >
> >you can use dram in place of vram in most cases
> >but you cant use vram in place of dram in most cases.
> >
> >Theres nothign magic about either one.
> >
> >Isildur
> >
> DRAM, and is duel ported - meaning it can be read and written to at the
> same time! DRAM cannot. VRAM is more expensive than DRAM.
> --
> Stephen Frearson
VRAM is definately better...obviously, if it's more expensive, right?
Don't forget the new, insidious WRAM, which can read a write more than
once simultaneously.
mad