| Author |
Message |
Robert Morein
Guest
|
 Posted:
Sat Oct 29, 2005 3:41 am Post subject:
NTSC-->DV-->MPEG ? |
|
|
I convert from VHS to DVD by the use of a hardware transcoder to DV, and a
software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
|
|
| Back to top |
|
 |
Richard Crowley
Guest
|
| Posted:
Sat Oct 29, 2005 4:41 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Robert Morein" wrote ...
| Quote: | I convert from VHS to DVD by the use of a hardware transcoder to DV, and a
software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
|
Unlikely (but remotely possible).
Reason: DV is significantly higher quality than either DV or MPEG. |
|
| Back to top |
|
|
Robert Morein
Guest
|
| Posted:
Sat Oct 29, 2005 4:41 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Richard Crowley" <richard.7.crowley@intel.com> wrote in message
news:djud6u$a0n$1@news01.intel.com...
| Quote: | "Robert Morein" wrote ...
I convert from VHS to DVD by the use of a hardware transcoder to DV, and
a
software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
Unlikely (but remotely possible).
Reason: DV is significantly higher quality than either DV or MPEG.
You mean, than either VHS or MPEG. |
What if the source were High-8 or SVHS ?
Has anyone critically examined this? |
|
| Back to top |
|
|
Mike Kujbida
Guest
|
| Posted:
Sat Oct 29, 2005 4:41 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
Robert Morein wrote:
| Quote: | "Richard Crowley" <richard.7.crowley@intel.com> wrote in message
news:djud6u$a0n$1@news01.intel.com...
"Robert Morein" wrote ...
I convert from VHS to DVD by the use of a hardware transcoder to
DV, and a software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
Unlikely (but remotely possible).
Reason: DV is significantly higher quality than either DV or MPEG.
You mean, than either VHS or MPEG.
What if the source were High-8 or SVHS ?
Has anyone critically examined this?
|
Assuming comparable camcorders (single chip consumer models), in a well-lit
environment, DV is far superior to either SVHS or hi-8. In a low light
situation, give me either of the "old" formats any day.
I was helping a student edit a project today (shot on hi-8) that was shot in
the late afternoon and early evening. Even when the sun was almost gone,
images were still visible and noise was acceptable. A comparable miniDV
camcorder would've been all noise with the light levels so low.
Mike |
|
| Back to top |
|
|
Richard Crowley
Guest
|
| Posted:
Sat Oct 29, 2005 4:41 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Mike Kujbida" wrote ...
| Quote: | Robert Morein wrote:
"Richard Crowley" wrote ...
"Robert Morein" wrote ...
I convert from VHS to DVD by the use of a hardware transcoder to
DV, and a software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
Unlikely (but remotely possible).
Reason: DV is significantly higher quality than either DV or MPEG.
You mean, than either VHS or MPEG.
|
Yes, you got me. Darned fingers want to think for themselves.
| Quote: | What if the source were High-8 or SVHS ?
Has anyone critically examined this?
|
According to Adam Wilt's "arrogant" rank-list...
http://www.adamwilt.com/DV-FAQ-tech.html
10 D-5 (10-bit uncompressed dgital)
9.9 D-1 (8-bit uncompressed digital)
9.7 Digital Betacam, Ampex DCT
9.6 D-9 (Digital-S), DVCPRO50
9 DV, DVCAM, DVCPRO (D-7), Digital8
8.9 MII, Betacam SP
8.7 1" Type C
6.5 3/4" SP
5.5 3/4", Hi8, SVHS
4 Video 8, Betamax
3.5 VHS
1 EIAJ Type 1, Fisher-Price Pixelvision
Mr. Wilt may think of it as "arrogant", but I don't know
anybody who seriously disputes his ranking and ratings.
| Quote: | Assuming comparable camcorders (single chip consumer
models), in a well-lit environment, DV is far superior to
either SVHS or hi-8. In a low light situation, give me
either of the "old" formats any day.
|
But that has everything to do with the camera parts of
the camcorders, and NOTHING to do with the tape format.
Modern camcorders are packed with all sorts of junk the
marketing people think will sell a few more cameras. I
think they are intentionally "dumbing-down" the camera
parts of DV camcorders to avoid competing with their
prosumer/industrial/broadcast lines.
| Quote: | I was helping a student
edit a project today (shot on hi-8) that was shot in the
late afternoon and early evening. Even when the sun was
almost gone, images were still visible and noise was
acceptable. A comparable miniDV camcorder would've
been all noise with the light levels so low.
|
Yup. But it would have been able to take multi-megapixel
stills and all sorts of other "features" that detract from the
primary mission. |
|
| Back to top |
|
|
Robert Morein
Guest
|
| Posted:
Sat Oct 29, 2005 4:41 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Mike Kujbida" <kujfamNoSpam@xplornet.com> wrote in message
news:3sg4u8Fnip2oU1@individual.net...
| Quote: |
Robert Morein wrote:
"Richard Crowley" <richard.7.crowley@intel.com> wrote in message
news:djud6u$a0n$1@news01.intel.com...
"Robert Morein" wrote ...
I convert from VHS to DVD by the use of a hardware transcoder to
DV, and a software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
Unlikely (but remotely possible).
Reason: DV is significantly higher quality than either DV or MPEG.
You mean, than either VHS or MPEG.
What if the source were High-8 or SVHS ?
Has anyone critically examined this?
Assuming comparable camcorders (single chip consumer models), in a
well-lit
environment, DV is far superior to either SVHS or hi-8. In a low light
situation, give me either of the "old" formats any day.
I was helping a student edit a project today (shot on hi-8) that was shot
in
the late afternoon and early evening. Even when the sun was almost gone,
images were still visible and noise was acceptable. A comparable miniDV
camcorder would've been all noise with the light levels so low.
Mike
No doubt. But the question is about the effect of concatenating codecs. In |
other media, two codecs applied serially sometimes produce degradation far
in excess of addition. |
|
| Back to top |
|
|
Ken Maltby
Guest
|
| Posted:
Sat Oct 29, 2005 4:41 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Richard Crowley" <rcrowley@xpr7t.net> wrote in message
news:11m5ouv991fnscf@corp.supernews.com...
| Quote: | "Mike Kujbida" wrote ...
Robert Morein wrote:
"Richard Crowley" wrote ...
"Robert Morein" wrote ...
I convert from VHS to DVD by the use of a hardware transcoder to
DV, and a software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
Unlikely (but remotely possible).
Reason: DV is significantly higher quality than either DV or MPEG.
You mean, than either VHS or MPEG.
Yes, you got me. Darned fingers want to think for themselves.
What if the source were High-8 or SVHS ?
Has anyone critically examined this?
According to Adam Wilt's "arrogant" rank-list...
http://www.adamwilt.com/DV-FAQ-tech.html
10 D-5 (10-bit uncompressed dgital)
9.9 D-1 (8-bit uncompressed digital)
9.7 Digital Betacam, Ampex DCT
9.6 D-9 (Digital-S), DVCPRO50
9 DV, DVCAM, DVCPRO (D-7), Digital8
8.9 MII, Betacam SP
8.7 1" Type C
6.5 3/4" SP
5.5 3/4", Hi8, SVHS
4 Video 8, Betamax
3.5 VHS
1 EIAJ Type 1, Fisher-Price Pixelvision
Mr. Wilt may think of it as "arrogant", but I don't know
anybody who seriously disputes his ranking and ratings.
Assuming comparable camcorders (single chip consumer models), in a
well-lit environment, DV is far superior to either SVHS or hi-8. In a
low light situation, give me either of the "old" formats any day.
But that has everything to do with the camera parts of
the camcorders, and NOTHING to do with the tape format.
Modern camcorders are packed with all sorts of junk the
marketing people think will sell a few more cameras. I
think they are intentionally "dumbing-down" the camera
parts of DV camcorders to avoid competing with their
prosumer/industrial/broadcast lines.
I was helping a student edit a project today (shot on hi-8) that was shot
in the late afternoon and early evening. Even when the sun was
almost gone, images were still visible and noise was acceptable. A
comparable miniDV camcorder would've been all noise with the light levels
so low.
Yup. But it would have been able to take multi-megapixel
stills and all sorts of other "features" that detract from the
primary mission.
|
It's not a matter of the "camera parts" but of the imaging
parts. You see the same effect in the low light level
performance of digital cameras, when compared to that of film.
An analog imaging is closer to film, in this regard, than a digital
one.
Luck;
Ken |
|
| Back to top |
|
|
Ken Maltby
Guest
|
| Posted:
Sat Oct 29, 2005 4:41 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Robert Morein" <nowhere@nowhere.com> wrote in message
news:AsudnRQKIemvNP_enZ2dnUVZ_smdnZ2d@giganews.com...
| Quote: | I convert from VHS to DVD by the use of a hardware transcoder to DV, and a
software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
|
Time, the speed of the process is greatly "degraded" by
that totally unnecessary step.
Luck;
Ken |
|
| Back to top |
|
|
Richard Crowley
Guest
|
| Posted:
Sun Oct 30, 2005 10:33 pm Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Ken Maltby" wrote ...
| Quote: | It's not a matter of the "camera parts" but of the imaging
parts.
|
Sorry, I don't see how substituting the phrase "imaging parts"
for "camera parts" changes anything? Can you explain?
| Quote: | You see the same effect in the low light level performance
of digital cameras, when compared to that of film.
|
All video cameras are analog. The analog signal just gets
digitized in a digital camera (and not in an "analog" camera).
In the same manner, all integrated circuits are analog/linear.
We simply design the circuits to be stable at only a couple
of values (which we arbitrarily name "0" and "1").
Yes, the electronic noise of a video camera is different than
the film noise (they call it "grain"), but IME my DSR-300
camera is far more sensitive and useful in low-light than any
film cine camera.
| Quote: | An analog imaging is closer to film, in this regard, than
a digital one.
|
You may be confusing the artifacts of digitization and data
compression with any differences between electronic (video)
imaging vs. chemical (film) imaging. These artifacts
disappear in the higher-end (less compressed) formats. |
|
| Back to top |
|
|
codecpage
Guest
|
| Posted:
Mon Oct 31, 2005 1:02 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
On Fri, 28 Oct 2005 18:41:56 -0400, "Robert Morein"
<nowhere@nowhere.com> wrote:
| Quote: | I convert from VHS to DVD by the use of a hardware transcoder to DV, and a
software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
It's the best you could do. Any degradation with this would be |
practically invisible. Capturing directly to MPEG2, both in software
or hardware, is still inferior.
Codecpage
http://www.codecpage.com |
|
| Back to top |
|
|
Ken Maltby
Guest
|
| Posted:
Mon Oct 31, 2005 2:34 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"codecpage" <removethis.codecpage@andthis.codecpage.com> wrote in message
news:9p5am198654n3kfs9mi95jvi465giajsqj@4ax.com...
| Quote: | On Fri, 28 Oct 2005 18:41:56 -0400, "Robert Morein"
nowhere@nowhere.com> wrote:
I convert from VHS to DVD by the use of a hardware transcoder to DV, and a
software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
It's the best you could do. Any degradation with this would be
practically invisible. Capturing directly to MPEG2, both in software
or hardware, is still inferior.
Codecpage
http://www.codecpage.com
|
Ok, let's say, for the sake of argument, that you have this
VHS quality video captured with absolutely no "degradation"
what so ever. But it is in DV. Now with this DV you can
perform some editing that is not yet available in the MPEG
format; but other than that what can you do with it? Do you
view your 26GB 2hr video files of VHS quality video on
your HD TVs? What is the benefit?
You will have to convert/encode/(or as PT insists transcode)
this DV into a useable format, most likely DVD compliant
MPEG. Now you will say: But then I can use an overnight
multi-pass encoding with filters and everything, creating a
"higher quality" MPEG. You will insist that these results are
vastly superior to any real-time single pass encoding. I say
POPPYCOCK!!! It's no problem to set a bit rate in any
good real-time encoder that will allow the capture of all the
quality available from VHS.
So unless you have some extensive processing to do to
your VHS video, such that each frame will be effected,
and you don't mind tying up your computer for many hours
per hour of video; capturing directly to DVD compliant
MPEG is a better choice.
Luck;
Ken |
|
| Back to top |
|
|
Ken Maltby
Guest
|
| Posted:
Mon Oct 31, 2005 3:27 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Richard Crowley" <rcrowley@xpr7t.net> wrote in message
news:11m9tf6rrhohm77@corp.supernews.com...
| Quote: | "Ken Maltby" wrote ...
It's not a matter of the "camera parts" but of the imaging
parts.
Sorry, I don't see how substituting the phrase "imaging parts" for "camera
parts" changes anything? Can you explain?
You see the same effect in the low light level performance of digital
cameras, when compared to that of film.
Snip |
| Quote: | Yes, the electronic noise of a video camera is different than the film
noise (they call it "grain"), but IME my DSR-300 camera is far more
sensitive and useful in low-light than any film cine camera.
Noise has nothing to do with it. |
| Quote: | An analog imaging is closer to film, in this regard, than
a digital one.
You may be confusing the artifacts of digitization and data
compression with any differences between electronic (video) imaging vs.
chemical (film) imaging. These artifacts disappear in the higher-end
(less compressed) formats.
|
I'm not talking about "artifacts of digitization and data
compression". It is a matter of "differences between electronic
(video) imaging vs. chemical (film) imaging". Also, to a limited
extent the similar, if of a much lesser degree, differences
between analog and digital video formats.
Quoting from a review:
"When digital cameras first appeared on the market, the resolution
was pretty coarse. Digital cameras have improved tremendously in
that department and high-end digital cameras rival 35mm film for
image detail. The one area where digital camera sensors are left
behind is tonal range. Chemical film can have a range of tones that
covers hundreds or even thousands of times more variation from
dark to light.
In an image where film can capture shadow details while maintaining
highlights, a digital camera will show only black blotches for the
shadows and blown out bright spots.
Digital photography is great for a scene with nice bright and flat
lighting, but when it comes to the tough situations, film wins hands
down. A subject in the shade of a tree with bright sky in the
background will either have lots of the shadow detail lost, or the blue
sky will turn white in the digital rendition. Film can capture this scene
if photographed carefully. Where I still use a film camera is for
underwater photography. Catching the eel hiding under the
overhanging coral would be lost on a digital sensor. With film, you
can pull old Mr. Eel out of the darkness."
Now while this is for still camera work, the principles are
the same. It is most likely that, at least to some small extent,
the analog signal off a CCD retains some little bit more of the
tonal range than what is there after digitization. So it could be
"An analog imaging is closer to film, in this regard, than a digital
one", as I said.
Luck;
Ken |
|
| Back to top |
|
|
codecpage
Guest
|
| Posted:
Mon Oct 31, 2005 4:25 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
On Sun, 30 Oct 2005 14:34:00 -0600, "Ken Maltby"
<kmaltby@sbcglobal.net> wrote:
| Quote: |
"codecpage" <removethis.codecpage@andthis.codecpage.com> wrote in message
news:9p5am198654n3kfs9mi95jvi465giajsqj@4ax.com...
On Fri, 28 Oct 2005 18:41:56 -0400, "Robert Morein"
nowhere@nowhere.com> wrote:
I convert from VHS to DVD by the use of a hardware transcoder to DV, and a
software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
It's the best you could do. Any degradation with this would be
practically invisible. Capturing directly to MPEG2, both in software
or hardware, is still inferior.
Codecpage
http://www.codecpage.com
Ok, let's say, for the sake of argument, that you have this
VHS quality video captured with absolutely no "degradation"
what so ever. But it is in DV. Now with this DV you can
perform some editing that is not yet available in the MPEG
format; but other than that what can you do with it? Do you
view your 26GB 2hr video files of VHS quality video on
your HD TVs? What is the benefit?
You will have to convert/encode/(or as PT insists transcode)
this DV into a useable format, most likely DVD compliant
MPEG. Now you will say: But then I can use an overnight
multi-pass encoding with filters and everything, creating a
"higher quality" MPEG. You will insist that these results are
vastly superior to any real-time single pass encoding. I say
POPPYCOCK!!! It's no problem to set a bit rate in any
good real-time encoder that will allow the capture of all the
quality available from VHS.
So unless you have some extensive processing to do to
your VHS video, such that each frame will be effected,
and you don't mind tying up your computer for many hours
per hour of video; capturing directly to DVD compliant
MPEG is a better choice.
Luck;
Ken
OK, so what amount of crap are you willing to accept ? |
Any of those real time soft MPEG capturers cheat with lots of things
that dump quality in favor of speed. Some even drop frames without any
notice or reduce effective resolution. They are very far from anything
you could really call 'quality'.
VHS is already bad, yes, but do we have to add more to it ?
CCE can encode almost real time but not capture. Yet it is written
entirely in assembler, i.e. it's the fastest possible and even then
and even with a 3 Gig CPU the additional load of capturing would make
it impossible to do without dropped frames (actually, they are
offering a pro soulution that rquires 2 such CPUs).
So if bitrate isn't crucial, why not capture in DV (these DV hard
encoders also have several goodies to stabilize frames etc.) and then
encode to MPEG in almost real time ?
Yes, I would usually add some filters to eliminate that noise and
flickering of VHS, but that's not mandatory.
Cheers
Codecpage
http://www.codecpage.com |
|
| Back to top |
|
|
Ken Maltby
Guest
|
| Posted:
Mon Oct 31, 2005 5:34 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"codecpage" <removethis.codecpage@andthis.codecpage.com> wrote in message
news:d0ham15vjld1et17nj3uq9l6e7kgi8dbn7@4ax.com...
| Quote: | On Sun, 30 Oct 2005 14:34:00 -0600, "Ken Maltby"
kmaltby@sbcglobal.net> wrote:
"codecpage" <removethis.codecpage@andthis.codecpage.com> wrote in message
news:9p5am198654n3kfs9mi95jvi465giajsqj@4ax.com...
On Fri, 28 Oct 2005 18:41:56 -0400, "Robert Morein"
nowhere@nowhere.com> wrote:
I convert from VHS to DVD by the use of a hardware transcoder to DV, and
a
software mpeg compressor.
This puts one superflous codec, DV, in the mddle.
Has anyone noticed any degradation from this approach?
It's the best you could do. Any degradation with this would be
practically invisible. Capturing directly to MPEG2, both in software
or hardware, is still inferior.
Codecpage
http://www.codecpage.com
Ok, let's say, for the sake of argument, that you have this
VHS quality video captured with absolutely no "degradation"
what so ever. But it is in DV. Now with this DV you can
perform some editing that is not yet available in the MPEG
format; but other than that what can you do with it? Do you
view your 26GB 2hr video files of VHS quality video on
your HD TVs? What is the benefit?
You will have to convert/encode/(or as PT insists transcode)
this DV into a useable format, most likely DVD compliant
MPEG. Now you will say: But then I can use an overnight
multi-pass encoding with filters and everything, creating a
"higher quality" MPEG. You will insist that these results are
vastly superior to any real-time single pass encoding. I say
POPPYCOCK!!! It's no problem to set a bit rate in any
good real-time encoder that will allow the capture of all the
quality available from VHS.
So unless you have some extensive processing to do to
your VHS video, such that each frame will be effected,
and you don't mind tying up your computer for many hours
per hour of video; capturing directly to DVD compliant
MPEG is a better choice.
Luck;
Ken
OK, so what amount of crap are you willing to accept ?
Any of those real time soft MPEG capturers cheat with lots of things
that dump quality in favor of speed. Some even drop frames without any
notice or reduce effective resolution. They are very far from anything
you could really call 'quality'.
|
Actually, I use an old hardware MPEG encoder (Broadcom's
Kfir-II, BCM7040) and my post wasn't specific to software
capture to MPEG. I would suggest that the OP would be better
off replacing that hardware DV encoder with a good hardware
MPEG encoder. That said; there have been some who claim
good results for their software MPEG captures of consumer
level video (normal TV, VHS, LD, ect...).
To further define a more practical approach for VHS tapes,
I use a DVD Recorder with a built-in TBC and NR circuitry.
Using a DVD RW disk, I can bring the results to my PC for
any further processing and/or proper authoring to DVD+R.
This puts the time consuming real-time capture "off-line", so
that neither I nor my computer are tied up. It only takes about
a half hour to author a DVD with neat menus & all, that way.
| Quote: | VHS is already bad, yes, but do we have to add more to it ?
|
There is no additional "bad" involved, in fact the signal
processing applied before the encoding can correct some of
the VHS "bad". If you are referring to the compressed MPEG
video vs. the less compressed DV, that is beside the point, unless
you are watching 26GB movies. It's what you have in the end,
which will be DVD compliant MPEG, whether it's ever been DV
or not.
| Quote: | CCE can encode almost real time but not capture. Yet it is written
entirely in assembler, i.e. it's the fastest possible and even then
and even with a 3 Gig CPU the additional load of capturing would make
it impossible to do without dropped frames (actually, they are
offering a pro soulution that rquires 2 such CPUs).
|
You neglect to mention that the "pro solution" is for HDV.
Actually CCE can transcode at better than real-time. What is
this "additional load of capturing", all the encoder needs are a
video and an audio stream. That is provided by whatever
capture hardware/vivo/video card is connected to the source.
There is far more "load" involved when your DV is being
encoded off the timeline of your favorite NLE.
| Quote: | So if bitrate isn't crucial, why not capture in DV (these DV hard
encoders also have several goodies to stabilize frames etc.) and then
encode to MPEG in almost real time ?
Yes, I would usually add some filters to eliminate that noise and
flickering of VHS, but that's not mandatory.
Cheers
Codecpage
http://www.codecpage.com |
|
|
| Back to top |
|
|
Richard Crowley
Guest
|
| Posted:
Mon Oct 31, 2005 5:41 am Post subject:
Re: NTSC-->DV-->MPEG ? |
|
|
"Ken Maltby" wrote ...
| Quote: | Quoting from a review:
"When digital cameras first appeared on the market, the resolution
was pretty coarse. Digital cameras have improved tremendously in
that department and high-end digital cameras rival 35mm film for
image detail. The one area where digital camera sensors are left
behind is tonal range. Chemical film can have a range of tones that
covers hundreds or even thousands of times more variation from
dark to light.
|
We were discussing video cameras. If you want to change the
topic to digital still cameras, that is your privelege. But not as
much of that rhetoric applies to video as you seem to believe.
Would you care to reveal the source so we can judge it in its
context (and consider its source)?
| Quote: | Now while this is for still camera work, the principles are
the same. It is most likely that, at least to some small extent,
the analog signal off a CCD retains some little bit more of the
tonal range than what is there after digitization. So it could be
"An analog imaging is closer to film, in this regard, than a digital
one", as I said.
|
The world of digital video is quite a bit broader than MPEG. |
|
| Back to top |
|
|
|
|
|
|
Profile
Log in to check your private messages
Log in