Canopus ADVC-100 better than TV capture cards?

[erratic]

I read this in a newsgroup (rec.video.desktop). Nomen Nescio wrote: (http://groups.google.com/groups?&selm=5eeac71db1f397b698c5e85a03b693f2%40dizum.com)I've had a Canopus ADVC-100 for over a year now and it kicks total ass. It is better than any analog capture card I've used in its price range.

With many cheap analog capture cards (e.g., Brooktree chipset based, ATI All-In-Wonder), they may use cheap filters/parts and I've found colors come out not correct and the images don't seem clear. I guess some people can tweak the captures thru filters, etc. to clean them up, and then sometimes it still doesn't work. You've also got potential problems with audio/video synch, especially on long captures, which I guess can be somewhat solved if you use the proper capture software. Also, some PC motherboards with poor PCI bandwidth (e.g., VIA chipset-based AMD) have problems capturing at high-resolution/frame rate because of too much PCI traffic.

None of these problems exist with a Canopus ADVC-100.

Some of the very high end analog capture cards (e.g., Canopus DVStorm2, Matrox RT.X10) have better quality than a Canopus ADVC-100 if you capture at HuffYUV or high-bit rate MJPEG.

However, if you are capturing VHS quality material, these will give you no better quality than a Canopus ADVC-100, and they cost hundreds more.

There's a reason nearly everyone at dvdrhelp.com says it's great. That's cuz it is. I wholeheartedly recommend it.
Most people seem to agree, but not everyone. Brent Geery wrote: (http://groups.google.com/groups?&selm=qhm54vcpb6rbnfsbqkslold49iijv5r1ud%404ax.com)
As you are already comfortable with using HuffYUV, and off-line MPEG encoding, I don't see what the ADVC-100 DV transcoder has to offer you! It's quality is not better then the current (second generation) of TV tuner/capture cards (it just uses the same phillips chip, found in current TV tuner cards,) with the disadvantage of adding an extra layer of compression artifacts, not to mention the cost. I'd suggest a $70-$100 TV tuner/capture card using the new CX23880 or CX23883 chipsets, capture with VirtualVCR, using huffYUV compression, clean up the video with AVISynth filters, then edit in the editor or your choice, then encode to mpeg-2 with CCE SP. You will not get better quality. If your video material is really poor (card dropping frames because of it,) also pick up a $270 TBC-1000 time base corrector.I wonder what the capture experts here would recommend: a tv capture card, maybe with a Philips chip, or an ADVC-100 with an external tuner connected to it. And by the way, if you connect an external tuner (VCR for example) to an ADVC-100 or similar device I suppose s-video is better than composite, but then you would need a VCR with s-video support. They aren't cheap, are they?

I assume an analog->DV converter is more user friendly but you have more control with a tv capture card. However, an internal tuner can suffer from interference caused by other hardware in your pc, which wouldn't happen with an external tuner. What I really want to know is: do internal tuners generally add more noise to the video than external tuners connected to an analog->DV converter?

Has anyone ever compared both possibilities (ADVC-100 versus tv capture card) and is there a website with more information?

[SiliconSoul]

ive heard DataVideo DAC-100 is like the canopus ADVC-100 but maybe even better

im not sure if all those statments are valid.

It's quality is not better then the current (second generation) of TV tuner/capture cards (it just uses the same phillips chip, found in current TV tuner cards,) with the disadvantage of adding an extra layer of compression artifacts, not to mention the cost.

so maybe the phillips chipset is not so great but it should not have interference that the internal cards are subject to.

and the compression should be DV, which ive found to be really good. (using it with a canon GL1/ZR30 on passthrough and a GL1 on taped footage)

well with firewire integrated in my motherboard and the external solution not that much more im pressed to see a that much higher cost...

TV tuner/capture card using the new CX23880 or CX23883
chipsets
I not seen any of these in cards yet.. who makes a card with these chipsets? are they better?


ive love to see what the experts around have to say! :-)

[fingernailX]

Its not the difference between the devices (ADVC vs. capture card) as it is the difference between the codecs. I have used a BT8x8 based capture card and iuvcr to capture with the M$ DV codec, Sony, Panasonic, Canopus and Main Concept. All provide essentially the same quality capture. I have also used a Sony DSR20 (from work) to record to dvcam from the tuner on my JVC SVHS recorder. It does an excellent job as well. Neither however look as good as a huffyuv capture, either straight from the tuner or from the SVHS deck. I have also tried out a ADVC-100 and a Hollywood Bridge. Both again do an excellent job. YUV is lossless DV is not. The other advantage for capturing TV is using the scheduler function for unattended capturing although I suppose a ADVC100 in conjunction with a vcr could be setup to schedule record. Personally, I'd rather spend the money on a big hard drive, pick my favorite flavor of PVR software and record to the lossless codec for better quality.

[vhelp]

First, and foremost, the end result is pendant upon the user and
his/her energy level (skills and techniques)

Now.., that in mind..

I have the ADVC-100. I use it 99% of the time.. if not, 101% anyways..

I also have the DC10+ and another card, Osprey-210 (I paid too much for
it, but then again, it gives me zero framedrops - good driver support
on my windows 98 setup)
.
.
It's ben a while since I last compared each output quality and encodes
from such (each) output quality. So, I won't go their :)

But, I will say this till I die or retire. Nothing can beat a good
Analog capture device vs. a DV device.

If the Analog is properly setup and configured in one's given system,
and together, they make beautiful music or conchertos, then there is
no way a DV device will output better quality. Especially, if you
Analog capture in "uncompressed" mode. You can't compare that to a
DV project.
.
.
BUT, on the other hand, if you have one of those setups that just don't
work w/ your Analog capture card, then to spare you the greaf, the DV
would be the way to go. ..no frame drops, and no audio sync issues
w/ DV (advc-100 in my view)

For other odd sources like VHS, I would recommend an Analog capture
device vs. a DV one. But, since it is nearly imposible to work w/
these Analog device and their multiple issues, the DV route is the
way to go, plus the quality ain't so bad. In fact, imo, I feel that
(not comparing aginst Analog devices) DV is great quality. At least
from my ADVC-100's point-of-view, the quality is fantastic! Well, that
all depends on your source too. Oh, and your skills levels too!
.
.
You could take a great device like the ADVC-100 and put it in the hands
of an amature, and he/she will turn out crap in your hands. Yes, the
final ingredient is the end user. And, if that user has some scares,
aka, special techniques and things, the sky is only the limit.
Just remember that DV has set limits. Be aware of this. Analog capture
do not (in so much)

VHS...
I would much rather use an Analog capture device in my VHS projects,
but I too still have some issues (though minor) w/ this route, and as
such, I would rather not put up w/ it, and so, I go the DV route, for
flexibility, and still good quality - - compremises are our daily
chore. anyways..
So, I use my ADVC-100 for these. Only, like the above analigy, its up
to the "end user" (me for instance) to make or break this project. I
have the necessary skills to work w/ VHS sources. I've also aquired
(through numerous trial-and-error stages) to develop the skills to
incorporate "filter chaines" for these source projects. Just like the
chef, we all have our "special" ingredients that stands out from the
rest. Can you taste the sweet-savory.. anyways..

Those PROS on the ADVC-100 weren't false. They were correct. Some
people just can't seem to break away from the old (though I still like
to maintain a close grip - speaking of Analog devices) So, you'll naturally
have those that CONS about this device or medium.

The ADVC0-100 rules (two years straight thus far) :D

-vhelp

[SiliconSoul]

okay is there anyway to make the DV less compressed as it comes in? would that make it look better?

my point of view is if DV is good enough for them to make movies with its good enough to backup tapes and capture tv.
plus it can still only look as good as your source.

i currently use a leadtek winfast tv2000 with huffyuv but im going to switch money permitting.

ive used the canon camra when i can use them from school :-)

i think it is worth the time to make it easier by using DV and DV editing programs...

should be able to use any encoder you want either way :-)

[SiliconSoul]

is there any device that would let you connect it to firewire and capture with huffyuv?


soemthing like this?
http://www.datavideo-tek.com/navigation_frames/dac-10/entire.htm

or any of these?

http://www.datavideo-tek.com/navigation_frames/dv_format_converters/entire.htm

[vhelp]

hi SiliconSoul,

but you still have to understand something - important.

Remeber, DV is a compression (lossy) and, say you capture something
w/ your Analog device, and you now want to archive it to a DV format,
you will be compresing it, though DV, you will be compression
it, hence lossy.
.
.
So, now you have TWO compressions or conversions, then you have to add
to that, when you decide to encode it, yet another (3rd) compression
or conversion.

So, you have to watch out for this pit-fall.
You could think or look at DV as if it were MPEG'isized :D

At the very least, I would capture uncompressed, to get a first-generation
like look/feel, and then archive it to DV (if that's your end goal)
.
.
But, don't capture to a "lossy" codec (cause of the compression/conversion)
being done on the source being captured.

Or, the other alternative is to just record your sources (cable or
satellite) to miniDV tapes, and then work w/ them at your leasure. At
best, these should last for many years - probably better than storing
these sources on DVDR disks. I hear that they don't reall last that
long anyways. Real shame :( why bother w/ DVD is they're gonna cause
you greaf two years later.

-vhelp

[SiliconSoul]

currently im capturing analog cable so my source is okay... but i get to watch what i miss :-)

with my card and sytem
the analog way with tv cards gets a bit too much interference for my taste. on top of the already okay source.

so im looking to get something else


end result would be
TV source to DV then edit then encode to svcd or dvdr.
ultimately i want a good svcd or a few episodes per dvdr

okay i understand DV is compression.

where does the compression happen?
from the camera/converter box?
firewire card?
codec?

can i filter DV just like when i use huffyuv?
with avs scripts and all that?

[vhelp]

with my card and sytem
the analog way with tv cards gets a bit too much interference for my taste. on top of the already okay source.

so im looking to get something else

The majority of Analog capture cards exhibit Noise to some degree. Some, more to than others. Still, there is a combination of causes..
* Mobo, CPU and Chipsets
* capture card

These above is everyone's main cause of Noise in their final capture
source .avi files. I've suffered from Noise (what I refer to as Line
Noise) for many years. Except for when I had a slower system, 233MHz
for instance. I think there is a sweet spot in this phase of mobo's.
AMD w/ XP's 1400 - 2000 exhibit the worse Noise. Slower systems don't
have any. And, I've read rumor that faster pc's ie, P4 has none.
.
.
Riding on this wave, the DC10+ and the ADVC-100 both DO NOT exhibit any
Noise, hence your cleaner source .avi files :)
But, don't discount your Cable. Cable signal is Noisy. Satellite, is
NOT :D
.
.
All's I can say is, if you want no issues, then the ADVC-100 is it.
No Noise, frame drops, and no audio sync issues. What more can anyone
ask for.

where does the compression happen? from the camera/converter box? firewire card? codec?

The conversion (I believe) is on the fly, via the DV chip inside the
cam or dv device.

can i filter DV just like when i use huffyuv?
with avs scripts and all that?

Yes, you can :)

-vhelp

[SiliconSoul]

first off sorry to overtake this thread... but this is info ive been wanting to know about for a while now!


The majority of Analog capture cards exhibit Noise to some degree. Some, more to than others. Still, there is a combination of causes..
* Mobo, CPU and Chipsets
* capture card

These above is everyone's main cause of Noise in their final capture
source .avi files. I've suffered from Noise (what I refer to as Line
Noise) for many years. Except for when I had a slower system, 233MHz
for instance. I think there is a sweet spot in this phase of mobo's.
AMD w/ XP's 1400 - 2000 exhibit the worse Noise. Slower systems don't
have any. And, I've read rumor that faster pc's ie, P4 has none.

YES i know you are right on this... i had a athlon 1700+ with a epox 8KHA+ and the noise was so bad that it was not FUNNY! with the same cables the tv or vcr did not get the same problems... ive recently upgraded to a athlon 3200+ and gigabyte GA-7NNXP and little to no noise issues.

I think overall DV is easier to edit and smaller. and the quality difference is not so great as to be that much of an issue.

[Erik_Osterholm]

Originally posted by SiliconSoul

my point of view is if DV is good enough for them to make movies with its good enough to backup tapes and capture tv.
plus it can still only look as good as your source.



I just wanted to jump in and point out a slight fallacy here...to compare consumer DV to "filming" movies digitally (as in Attack of the Clones) is a big mistake. They are working with extremely high resolution video streams. If you don't believe me, you can just think of this little example:
Movies on DVD are 720x480 (for this example, we'll assume NTSC). A movie theater screen is about 50 feet wide (I suspect this would be about average). 50 feet wide, handling 720 pixels comes out to 1.2 pixels per inch. That's pixels that are very nearly an inch long. If you saw a DVD projected onto a screen this size, believe me, you'd notice it.

Now for your purposes, having a little pixellation may be fine, and obviously you won't be projecting this onto the big screen. But realize that Lucas didn't shoot Star Wars Episode 2 on a little $500 Sony DV camera.

Also, that the encode can only be as good as your source is only partially true. If you have a bad source (mosquito noise, etc.) it's even more important that you not lose any information (lossy encode, such as DV) until you've had a chance to clean it up. Noise like this can cause really bad artifacting that would be difficult to remove in later processing, however you can usually clean up the video a bit if you've captured it losslessly.

For me, that extra encoding step isn't worth the artifacts I'll get, particularly since my source isn't as clean as I'd like it in the first place.

[Arachnotron]

A bit of topic, but perhaps nice to mention in regards to the noise issue:

I saw a japanese page somwhere (forgot where) where someone had modified his TV card by completely removing the tuner, replacing the capacitors for high grade ones and shielding the card by building a metal cage around it.

If I can find the time, I want to try this myself. TV cards are dead cheap now anyway. :D

[SiliconSoul]

@Erik_Osterholm
i do not use the ZR series camera any more... that was when i first started with video production. Now im using mostly canon GL1 and sometimes XL1s.. so the cameras are not the best but very good and no where near the $500 ones! It nice that my school has spent the money to get good equiptment

:)


id like to see a dvd projected on a movie theatre screen and see what happwn or how bad it is!
:p

[mustardman]

All's I can say is, if you want no issues, then the ADVC-100 is it.
No Noise, frame drops, and no audio sync issues. What more can anyone
ask for.

Not exactly true.

A source such as videotape with a good signal going to no signal (blank piece of tape) going to a good signal can give you repeated frames.

A handycam source can give you dropped fields at scene change boundaries (about 4 frames after the change). I have found this (and can proove it) with an ADVC100.

However, since it is an external device, it has no way of reporting dropped/duplicate frames to the computer. So you are blissfully unaware of them.

This is my only complaint about an external device - the advantages outweigh this though.

[SiliconSoul]

external devices can drop frames and reports it to the software when it happens. im not speaking from experience with the canopus but with Canon ZR and Gl and XL series cameras are external and drop frames and reports when it drops a frame.... the ZR series is by far the worst at dropping frames! :(

[trevlac]

Originally posted by vhelp
The majority of Analog capture cards exhibit Noise to some degree. Some, more to than others. Still, there is a combination of causes..
* Mobo, CPU and Chipsets
* capture card

These above is everyone's main cause of Noise in their final capture
source .avi files.

I'm confused. I mainly use a CATV source. I would expect the sources of noise to be:
- Signal (most)
- RG6 Cable/splitter/etc. (maybe a little)
- Tuner in card (a bit more depending on the channel)
- Internal PC radiation (more, but not much)
- Once we get to digital, loss of info is the problem

Your quote and the Japanese project mentioned, imply the Internal PC radiation is the big problem. Maybe you are talking about very clean source. But clean source does not come from a tuner. I would think a tuner would have the bigest problem with PC radiation.


So as a Test:

I could isolate the Tuner by using a VCR as a tuner set away 1 meter from the PC and run an RCA to the Card. If this had the same amount, I would say the Card Tuner was not to blame.

I could isolate the Entire card by using the VCR as a tuner and record to my miniDV cam, across the room with the PC off. I could then transfer via firewire. If the noise is not less, I would say it is not the card.

I could buy good cables to start the test. That way I don't have to suspect the cables.

If it's not the Tuner or the Card or the Cables, I'd say it was in the source.

How's My Logic? I'm going to try this test. Any guess as to the source of the noise? I have an MSI Master in a Dual Athlon XP 1900+ machine. :)

I just thought of a quicker test. Maybe not as thurough or valid. I could run s-video from a DVD player to the card. If this looks as good as using PowerDVD from the harddrive, I'd say maybe the Tuner, but not the card. Is this a valid test?

[SiliconSoul]

what i think he was saying is that some motherboards, chipsets, and CPU introduce noise into the capture.
and maybe some do not...
but the majority having problems with a good source that is clean are having the noise come into play else where

[mustardman]

@SiliconSoul

That's very interesting. If there is only the IEE1394 (firewire) cable going to the computer, there must be a field in the DV data frame that can tell the computer some stats about it - such as whether it is dropped or duplicated.

In your setup, is there only one cable (ie: firewire)?

What software are you using that reports such mishaps?

All software I've used reports 0 frames dropped, all the time. Also, there may be a difference between cameras and the ADVC, since it is just a converter, not an optical system as well? Maybe cannopus just skimped on the spec?

Thanks,
MustardMan

[SiliconSoul]

im using camera to firewall cable to firewire card on pc

video vegas and adobe premiere

maybe there is a difference between camera and ADVC...

maybe the ADVC just never drops frames...

im not sure where the frame drop detection is...
software knows it because of missing frame?
hardware tells the software a frame drops?

[Arachnotron]

@ trevlac:

a source of interference often is the PC itself, but not internally (directly into the capture card), but outside through the cables or connected devices. When the case is open or incorrectly shielded, noise is picked by the cables or the camera/vcr/dvd player which are often not shielded at all.

A lot of people have at one time or another removed the metal shielding plates in front of the drivebays, but forgotten to put them back when moving a drive to another bay.

[vhelp]

I don't know where you guys are getting dropped frame from the advc,
but, perhaps its because of the following (of which I have encountered
on my own:

* fw card
* fw card, mobo, and chipset
* fw card and OS
* fw cable is loose
* fw cable is not properly plugged in or snug
...if loose, you need to replace w/ a newer and tighter cable that
...will fit into the holes w/ a little dificulty.
* poor driver install (even if your setup says perfect)
* driver issues. ie, on my main pc, I have one dv driver (dah) and its
...not the same as my 2nd pcs's (for other reasons) and when I connect my
...dv cam (cannon zr-10) it works perfectly (due to the older style dv
...circitry) but when I connect my NEW Sony TRV-22 to it, it does notwork,
...nor the my advc.
...So, it could just be the unit vs. the driver (oh, and posibly vs. the OS)
...Just to note, I'm under Windows 98 Gold version. But, on my 2nd pc, and
...using only DVIO, it's pure heaven. No madness.. zero drops.

Also, I'd like to point out, that in other expeirences encounting any such
frame drops, were due to bad VHS tape spots or issues, of which are very
rare in deed. These only happen when I'm tinkering in my hobby.

If you're dropping frames, its definately one of the above itemized reasons.
AFA I'm concirned, I never drop frames w/ the advc-100.. never! But...
.
.
Check out DVIO. This app (though monitor not inluded) is the worlds
most simpliest DV transfer app.. (AFAIK :) ) and it also reports frame drops.
There are times when I will drop frames.
ie, when I don't have the fw cable connected all the way in. Sometimes,
it finds its way out partially, even when I feel it's in just right.
You have to jiggle it. I admit, one shouldn have to jiggle hehe..
But, sometimes, if you use fw quite heavily (as I do) and disconnect
and reconnect (as I do) lots and lots of times (I have other dv devices)
you'll run into this problem. Perhaps, its time for a dv router ??
Anyways..

Hope the above helps.
-vhelp

[trevlac]

Originally posted by Arachnotron
@ trevlac:

a source of interference often is the PC itself, but not internally (directly into the capture card), but outside through the cables or connected devices. When the case is open or incorrectly shielded, noise is picked by the cables or the camera/vcr/dvd player which are often not shielded at all.

A lot of people have at one time or another removed the metal shielding plates in front of the drivebays, but forgotten to put them back when moving a drive to another bay.

If noise does not go into the card, why would you build a cage around it?

I wasn't busting on VHelp. I am doing some testing. I do have a dual XP machine and my MSI card is terrible in it. My Aver BT card from CATV source looks HD in comparison. :) So far, I have nothing to report. I'm just confused myself by being unsystematic. I'm back to step 1 with the cables.

[mustardman]

@vhelp

Hi there!

I agree, the ADVC is a very good card, beats the .. out of a DC30 I also have. However, the frame drops I am talking about are not because of loose connectors or dodgey drivers...

1] Complete frame drop - looks like a normal frame drop from an analog capture card, the last good frame is simply repeated.
2] Very subtle field drop - only picked this up by going through a clip field by field (used AVIsynth to split the full frames into their respective fields). I noticed a field was missing from the movement of trees past the window of a car. Since I doubt the car can speed up and slow back down again in 1/25th of a second, its pretty surely a field dropped.

Both these captures are from a very bad VHS copy of camcorder footage. The dropped field happens about 8 fields after a scene change where the field order reverses (because of the way analog camcorders work, this happens on average 50% of the time). I think type [2] field skips would occur with any footage that has field order reversals in it.

I can email clips if you like, but I have no web space to post them for all to see.

By the way, on the entire sequence, the ADVC does [1] four times. The DC30 drops so many frames it's not even funny (I'm talking hundreds here).
The ADVC is awesome! Who needs a TBC!

BTW, I found the most awesome DV capture (and playback) software to be WinDV. Check it out, I think you'll like. Unfortunately the author has never responded to any of my emails.

MustardMan.

[Arachnotron]

@trevlac

If noise does not go into the card, why would you build a cage around it?

As you noted yourself, there are several sources for noise. A cage could perhaps help against internal noise; how much a card is susceptible to that depends a lot on design (multilayer PCP's with proper grounding planes, something a lot of manufacturers try to cut corders on).

I merely stated I had seen it done somewhere and wanted to try it sometimes.

What I did note myself is that noise levels (wavy interference patterns in large areas with an uniform color in the capture) increase when I leave the cover off my PC (P4 byt he way).
From this I conclude that interference generated by the inside of PC, but "cought" by leads or devices outside the PC is also a problem.

So paying attention to the correct shielding supplied by the PC case is also important, especially since shielding a plastic VCR may not be so easy to do.

edit
I noticed some issues with my hauppauge cx23881 based card in an athlon 1800+ bord. See horizontal green stripes in my captures, a sign there is not enough data captured to calculate the pixels for a complete line I think. Card worked file in a P4....

edit2
I wasn't busting on VHelp.

I wasn't suggesting you were :confused: (I'm not a native speaker, so if I did it was unintentionally), merely trying to be helpfull.

[vhelp]

hi guys.

@ mustardman,

It looks to me like you're from PAL land. Perhaps there are issues w/
PAL versions w/ certain sources ie, your given source and it's issues
which happens to be age and ware'n tare :( Either your source is soo
bad, or some of the above (earlier post I made) or who knows..
has something to do w/ your drops in frames.
.
.
Like I said, in my hand (NTSC land) a frame drop is unheard of.
.
.
But, you probably have a unique situation going on there in the Outback :)

I admit, when I used my DC10+ card on VHS (even brand new store bought
movies on VHS) I dropped frames. It was very frustrating because these
were crip sources (store bought ones) and my installations of the DC10+
and it's drivers (and IRQ's) were free from issues) but still would
drop frames.
.
.
After dealing this this card and many others, I found that this met w/ the
above "Canopus ADVC-100 better than TV capture cards?"
a reality. Stop the madness I said. Yes, frame drops and Line Noise
were my biggist grip w/ Analog capture cards. I had zillions of VHS
tapes (you know, fun projects) to archive to CD, and I was waisting lots
of time wrestling w/ Line Noise and frame drops. The best thing about
the DC10+ card was that there were zero Line Noise. That was a plus
for this card. But for VHS projects, there were not getting around
the frame drops. Mind you, the drops were minimal.. about 20 to 50 in
an hour, but that was too much for me. One drop is too much. It's bad
enough to work out the IVTC on these (those that were Movies) and have
issues as such. Well, now I have a better IVTC (w/out using plug-ins,
hehe) but other than that, I don't know what else to say, but that the
DC10+ does a great job on my other sources, be it Cable or DVD
or Satellite. No drops in frames and zero Line Noise :)

But, w/ the ADVC-100, I can concentrate on refining my "filtering" for
VHS noise (not Line Noise) and as such, has proven to be a challenge.

I still keep my Analog capture cards in my pc because:
a) they work w/out issues as they are (in both pc's) and
b) I am forever doing experiments or testings, and I need them where they are.

Out of curisity, have you tried the ATI-TV Wonder pci card on VHS ?
Not the VE, mind you. This card seems to match my ADVC-100's output,
which makes me wonder about some "common" things that they both do or
perhaps share. Anyways..

On another note...
I was hoping (planning) on doing a comparison test w/ my ADVC and an
Analog capture card (of my choice) to see which of the two yields the
best output quality. I haven't met up w/ all the proper ingrediants
as of yet. But, if I get everything all set up (time aloud) I'll run
a few tests and see what I come up with.
.
.
DV has one problem. Resizing and color degration (due to resizing)
The 411 filter (thanks to FredThompson's help on this issue) helps,
but IMO, you can't use it more than once (you'll begin to see vertical
lines if you do) - - so, resizing has to be performed w/ awarence of
that fact :rolleyes: Since I'm aware of this, I avoid such nonsense,
and all is well. To my knowledge, Analog capures (non DV conversion)
does not suffer from this phenomina (spelling) Anyways..

Still, I'm sorry you are having trouble w/ your home-sources.

-vhelp

[SiliconSoul]

where can we find this 411 filter? im going to need that in the future! :)
possibly today!

when do we need to use it?

i understand the why from the DV to DVD thread i found on the search.

i have a 30 min long captured DV that im going to convert to dvd
do i use it everytime (once per source/encode) to be sure?
only when resize?

[vhelp]

Hi siliconsoul,

These are the filters I D/L'ed, but I can't exactly remember where
they came from, but..

* (Filter) - (vdub) - (DV 411) - (by Xesdeeni) - 411Helper.zip
* (Filter) - (AVIsynth) - (DV 411) - (by xxx) - ReInterpolate411.zip

Just look/search for those names that are bolded. Sorry that
I don't recall the URL links.

-vhelp

[vhelp]

.
.
i have a 30 min long captured DV that im going to convert to dvd
do i use it everytime (once per source/encode) to be sure?
only when resize?

Yeah, I did ask about where to place (or use) the filter in a given
AVS script or inside vdub, but I got no replay. Basically, it was
expected of me to figure it out for myself. In any case, since I
was not given the recommend, and since i had to learn and figure these
nonsense things out for myself, I'll share w/ you (and others) what
little I've learned, and you can go from there :)

Anytime I "resize" or "alter" my source .avi, be it in vdub or AVS,
I insert the 411 filter as the LAST command in the chain. I have found
that after you do and of the above, things get mangeld in the color
space. Not by much, but if you look in an area covered in red, you
might notice some things are blocky or pixelated incorrectly. Once
this filter is use (last command) you should see a slightly smoother
looking color (red in this example)
.
.
I've taken the liberty of using this any time I alter my DV sources
as a standard practice.

Note, if you don't alter your DV source in any way ie, resize or
filtering like spacials and temporals and smoother etc, then you do
not need to use this. As you source is not tampered with. At
least this has ben my experience. If untouched, leave 411 out. If
touched, ad 411 in :)

EDIT: oops, after reading this, it looks like I'm vearing off the subject
topic :eek: I'm warning myself.
.
.
You can post quesions related to DV in the DV forum :)

-vhelp

[SiliconSoul]

this sucks i can not search for dv to dvd because the "DV" is too short!
any way around this?

ohh shoot ill move all dv questions over there sorry!
:)

[mustardman]

@SiliconSoul

Yeah, I've got a DC30 which is really giving me the craps. Like your DC10, it drops frames like crazy, even on good source. It is for this reason that I would never reccommend Pinnacle hardware. (As an aside, I tried studio8 when I bought a IEE1394 card, and it was the biggest pile of ... I wouldn't recommend their software either!)

Anyway, if I can get a couple of short grabs, I will email you them so you can see what I mean (if you're interested that is!). The field drop is my only real concern, the frame drop I spoke of... well the source is so bad I'm not surprised it happened!

Oh, and yes, I am from PAL land. Does your NTSC conversions (Analog to DV) have rather wide 'front porch' and 'back porch' (thats the bits to the left & right of the picture itself)?

@vhelp

The 411helper sounds intruiging. I've downloaded it, but would like pointers on when to use it - you say when resizing (horiz, vert, both?) but it only works once??? That sounds pretty weird. Any clues?

The other prog - reinterpolate411 - seems from the blurb that it is very specific. Do you use it?

Thanks,
MustardMan

[trevlac]

@Arachnotron

I just wanted to wrap up my comments about noise. The largest amount of noise seems to come from (into) the cables and external connections. (Like you suggested.) I was concerned that my CX card in my AMD machine was picking up lots of noise. Here are my unscientific results:

- My AMD machine creates more signal noise.
- Noise is picked up from External devices (mainly cables)
- My MSI tuner is not more effected by noise than an external VCR
- My Aver card tuner blurs noise out. I'd call it some sort of low pass filter (maybe it's just the bandwidth of the tuner)
- The RCA in and/or the s-video in on the BT card from the VCR show the same amount of line noise as the MSI card with good cables
- Some noise can be filtered out of the RF line using cheap line filters

Conclusions:
- Line filters can help
- Strong signal matters (powered line amp)
- Good cables / tight connections make the biggest difference
- Interference from the computer is limited if you use shielded cables



@eratic

What have you decided to do? Is DV cap worth the price? With what kind of source?

[Arachnotron]

@ trevlac

Another thing I want to try:

An electronics engineer at my workplace suggested using an ferrite ring in / around the VCR powercord as close as possible to the VCR, since a lot of (PC) interference comes in through that route.

[erratic]

Originally posted by trevlac
What have you decided to do? Is DV cap worth the price? With what kind of source? I haven't decided anything yet. I'd like to read some reviews of the ADVC300 first. The noise reduction will have to be very good before I pay that much money for it.

[Blackout]

i have an ADVC-100. if youre dropping frames here and there when recording, its probably an old VHS tape, and you will need a TBC - Time Base Corrector. Do some research on them...thats your prob.

i suggest the ACE Video convertor, cheap, great...on the input of the ADVC. thats what ive got,...never dropped a single frame since.

Blackout

[trevlac]

Originally posted by Blackout
i suggest the ACE Video convertor, cheap, great...on the input of the ADVC. thats what ive got,...never dropped a single frame since.


Can you post a link to some info about this specific device? I like your description "TBC..great..cheap"

Thanks

[Arachnotron]

I just looked them up here (http://www.gthelectronics.com).

It's a video processor for about 300 uk pounds. According to an on-site review:

Inside the ACE, all of the circuitry is on two neatly-designed PCBs. The first is the main board, which contains most of the circuitry, controls and socketry. The second board, which is mounted on the front panel, contains another row of controls. On the underside of the main PCB are the surface-mounted video ADC and DAC, which are sourced from Philips. Both of these chips are commonly encountered on video capture cards, of the type used for PC-based non-linear video editing.

So if your VCR already has a TBC I don't expect much advantage compared to post processing. Or by an ADVC 300 with TBC build-in.

Fun thing: one of the developers is over at Neuron's forum (http://neuron2.net/ipw-web/bulletin/bb/viewtopic.php?t=583) right now.

[trevlac]

Well, 300 pounds is not that cheap. To me :D

That's about $600. You can get quite a few TBCs in the states for sub $600. Not to mention Ebay. But of course your milage may vary ... this device may be fantastic. There is a fully analog video processor they sell here (not a TBC) the BVP4 Plus. Can get them on ebay for ~ $200. They say great things about them.

It'd be nice to have a bunch of control on the analog side, but at some point how much xtra are you getting that you can't do in digital for most source.

BTW: I still think you could do a TBC in software, if you could get the sync signal. Regardless of what any crazy Croats say. :)

[Arachnotron]

BTW: I still think you could do a TBC in software, if you could get the sync signal. Regardless of what any crazy Croats say.

I know you can, that is how TBC's work....:D

(by digitizing the whole signal and messing around with that I mean)

[racerxnet]

I had seen this same subject come up in the Canopus forums and the response is that you have a bottleneck somewhere on the PC. This could be hardware related or software. As was my case, I was using a software based raid array, specifically a PDC20265 chipset. The IRQ interupt was causing dropped frames when monitoring the system. I'm not suggesting your problem is as mine, but the possibility exists that it is hardware related on the PC.

I have been using ferrite cores to wrap my power cables on the 3 Bryston amps as well as other items on the pc. It is also used on my NEC Pc-VCR regarding the RS232 port wiring. This has been done for some time in the electronics field. Great idea and works well! Do you see the problem when playing the tape to the TV from the VCR?

Good to see all still frequenting the site.

MAK

[leonid_makarovsky]

I've already posted it to rec.video.desktop a while ago. I just wanted to mention that I did some test captures and found out that:

1. TV Tuner card with Philips chipset with Huffyuv codec is the best.
2. Pinnacle DV500 with Philips ADC and with DV capture is pretty good but not as sharp as TV Tuner card with Philips chipset.
3. BT878 based cards with Huffyuv codecs are the worst. The AD conversion looks pretty blurry. Pinnacle DV500 is much closer to Philips based TV Tuner card than BT878 based TV Tuner card to Pinnacle.

I have snapshots of all the captures for those who's interested.

--Leonid

[trevlac]

Originally posted by leonid_makarovsky
3. BT878 based cards with Huffyuv codecs are the worst. The AD conversion looks pretty blurry. Pinnacle DV500 is much closer to Philips based TV Tuner card than BT878 based TV Tuner card to Pinnacle.


I don't agree with this characterization of the BT878. :)

1st I wanted to note that the entire process (source to display) is much more important than a given chip. VHS source has zero practical benefit from an 8bit adc vs say a 10bit adc.

Far from blurry ... I'd say the BT resizer tends to artificially sharpen. So, when you say the BT is blurry ... I have to say Untrue. Especially if you don't give more details on the process/test.

http://trevlac.us/pics/Res.jpg

That's a link to a test I was doing to compare DV capture devices. The Laird is a pro level device. Looking at the TVL wedge, the BT does quite well. This is for 6.75MHz source, which no one would really capture anyway. Also notice that there are 2 BT tests. On the 1st, I had some setting wrong and it came out much worse.


@MAK

Good to see you here. Have not talked in a while, have we?

[leonid_makarovsky]

Originally posted by trevlac
I don't agree with this characterization of the BT878.

Well, take a look yourself. Same VHS, same VCR (SECAM), same frame:
Pinnacle DV500:
http://us.geocities.com/leonid_makarovsky/WinTV/Pinnacle_Composite.jpg
FlyVideo 3000FM (Philips based):
http://us.geocities.com/leonid_makarovsky/WinTV/Fly2000.jpg

Hauppauge WinTV (Brooktree based):
http://us.geocities.com/leonid_makarovsky/WinTV/WinTV_Composite.jpg

Now the 3rd one does look like sh*t.

Originally posted by trevlac
Good to see you here. Have not talked in a while, have we?

Are you talking to me? If so, sorry, but I'm not recognizing you.

--Leonid

[Sergei_Esenin]

Originally posted by leonid_makarovsky
Now the 3rd one does look like sh*t.

But this might not be representative of the 878's usual performance. Firstly, because SECAM is increasingly rare these days and possibly not as well-supported/tested as mainstream PAL and NTSC signals, and secondly because it looks like the card really is having trouble getting a clean lock on the signal. Look how straight objects like the scorebox aren't straight in the 878 cap--it probably isn't locking on to that signal as well as the other two cards.

[scharfis_brain]

I would never ever go for a DV-Devide for capping my analogue stuff.

Its (at least for PAL) vertical reduced Chroma is horrible.

I always do some kind of heavy pixel manipulation on my analogue Handycam captures (deshaking, denoising, maybe deinterlacing for Film-look)

for all of this, interlaced-YV12 is more than pain.

[Arachnotron]

On the website of Fly 2000 TV (and elsewhere) I remember reading the performance of the BT878 is extermely poor with SECAM, especially when compared to the SAA7134.

When dealing with PAL or NTSC the difference is much smaller.

[trevlac]

Originally posted by leonid_makarovsky

Are you talking to me? If so, sorry, but I'm not recognizing you.

--Leonid

Sorry ... no. After replying to you, I was trying to sneak a hello in to MAK the poster just before you. He and I had some nice conversations a while back at vcdhelp.


As far as SECAM ... I have no doubt that your BT experience has been poor. But this does go back to what I was trying to say. One needs to cover the entire process/system to say if something is any good. My BT experience with NTSC has been very good. I have an Aver card with a Philips tuner, and for NTSC B-Cast, my BT does a fantastic job. It is a little softer than the CX (MSI) card I have, but that is what I need for my source because there is so much noise that comes thru on the CX that I'd have to do a post filter. If I want to do VHS, the colors on the BT are also better than the CX, and the BT has more than enough "bandwitdth" to get all the details in VHS. In fact, the BT is better than some of the higher end DV stuff I have.

[leonid_makarovsky]

Originally posted by Arachnotron
When dealing with PAL or NTSC the difference is much smaller.

There's still a difference.

--Leonid

[Arachnotron]

There's still a difference.Oh, certainly there is.

But having tested a philips saa7134, a BT878 and a CX23881 I have come to the conclusion that the differences in amount of detail are so small that when you look back the result on a TV you cannot see the difference.

That is providing you have functioning drivers, good capping applications etc. etc. and the source is ok. For example, there are some truly crap CX23881 drivers out there that also mess up brightness settings. Also, there are differences in exactly which sharpness filters are enabled, and how far. Especially for conexant drivers this can make a difference.

But when you are comparing full res caps with all settings neutral (that is no additional sharpening filters) the results are about the same as far as detail is concerned. The differences are more a matter of taste.

The main area of difference lies in the comb filters. I would expect there to be a larger difference between the chips for high bandwidth composite NTSC sources like analog TV broadcast (not VHS) since there is more of an overlap between the chroma subcarrier frequency and the high frequency luma than there is for PAL or SECAM. An illustration of this is the recent release of a NTSC only TV card that uses a separate NEC combfilter chip in front of a philips SAA713x chip, just to improve NTSC TV capturing.

Personally, I use the SAA7134 most since it has the best PAL-60 support and gives me nice luma ranges which I rarely have to adjust. The drivers are stable and simply work. I have had no end of troubles with my WinTV CX23881 card, and my WinTV BT878 card also has had it's problems. There comes a point when you are finished looking for special drivers and tweakers tools and want something that simply works. :)

But having said that, I have just registered fly 2000 TV and am playing with the special SAA7134 tweak options it offers. :D

[trevlac]

Originally posted by leonid_makarovsky
There's still a difference.


Sure it's different ... some say the BT is better. :)

Anyway ... it is important to know where the differences are, why, and what really matters.

For example: Sound like the BT is poor for SECAM. On the other hand, the bit depth of the original samples (8 vs 9 vs 10 bit), in my opinion provides zero advantage to 10 or 9 bit chips. The final results are 8 bit.

Another area of difference is the color subsampling. As pointed out, DV is inferiour to some forms of analog, where as a non-DV device allows you to grab more info.

I don't want to be argumenative here ... so I'll just get back to my main point. When comparing various devices, you have to specify what is happening and what is of concern each step of the way. In my experience, my BT based card produces better results on things like VHS or B-Case NTSC than my CX card or various DV devices I have.

The Doom9 analog capture guide has a bunch of examples from various chips (CX/BT/ATI/Philips) ... check out section 5.

http://www.doom9.org/index.html?/capture/sizes_advanced.html

But if you capture at full frame ... none of them are really blurry.

[leonid_makarovsky]

Unfortunately I don't have WinTV (BT 878) card any longer and I don't think I have any captures left. I believe as far as I remember it gave me blurer image than FlyVideo 3000FM for NTSC VHS. Both drivers had default settings. WinTV had Hauppauge drivers.

My friend now has this card. If he ever installs it, I'll just get him to do a test capture from DVD.

One thing that comes to mind is the following. If both chips give about the same quality, why was it only Philips ADC used in professional boards such as Fast AV Master, Pinnacle Pro-One, Pinnacle DV500. Some even say Canopus used Philips ADCs. Why BTs were only used in PCI TV Tuner cards exclusively?

--Leonid

[Arachnotron]

One thing that comes to mind is the following. If both chips give about the same quality, why was it only Philips ADC used in professional boards such as Fast AV Master, Pinnacle Pro-One, Pinnacle DV500. Some even say Canopus used Philips ADCs. Why BTs were only used in PCI TV Tuner cards exclusively?

Could simply be price, features, driver support, support in general, availability of good reference designs, easier interfacing to other hardware, availability of custom features (there is an enormous range of video-related chips from Philips) etc. etc.

There have been various improvements for working with ACPI, improved FIFO sizes, better combfilters, better interfaces to hardware MPEG2 encoders, better sound decoding and whatnot since the bt848 and bt878. These improvements matter. However, they are not intended to improve the amount of detail simply because the BT878 already had enough to deal with the maximum amount of detail available in TV broadcasts.

A good example of this is the fusion 878: it is the same as the old BT878, only the interface has been brought up to date to deal with PCI 2.2 and ACPI.

The point is this: the BT878 already had a high enough sampling rate to deal with capturing PAL TV (5.5 MHz bandwidth) at 768x576. The CX kept the same raw PAL sampling rate as the BT878 (35.5 MHz), the Philips series is actually a bit lower. (27 MHz). There simply is not that much detail available in broadcast and certainly not in VHS (only 3 MHz luma bandwidth), and the BT878 already reached the point where it could deal with all of it long ago. Newer chips have improvements on many other points, but not there.

[Morbo]

I use a Canopus ADVC-100,its macrovision proof,never drops frames and if you can de-interlace via avisynth like the back of your hand,it looks like what you feed it......bar none.

I had a Dazzle Mpeg2(Spawns site sold me a long time ago),but the audio synch and dead color sent me looking for uncompressed goodness.

If you know the usual tricks to cleaning up DTV captures anyway,its a rock solid choice,even if you use it like TiVo with movie maker in XP.

Id buy another one,mainly because its firewire,and my laptop will even use it.

Cheers! and Good Luck!

[Sergei_Esenin]

Originally posted by Morbo
but the audio synch and dead color sent me looking for uncompressed goodness.

Just bear in mind that it isn't uncompressed, it's compressed with DV. It's by no means a bad choice and gives better PQ than hardware MPEG-2 encoders, but it's not quite as good as capping an uncompressed feed to Picvideo MJPEG @ level 18 or 19, which in turn isn't quite as good as capturing uncompressed video to a lossless codec like huffyuv.