Problem with BT878 output in compar. with DC30

[Inc]

Hi there,

until now I used my Pinnacle DC30 via WDM driver to capture in Win2000 using VirtualVCR. I used the onboard hardware mjpg codec and it came out with a nice quality. Sound was captured via my 8rda+ Mainboard Sound device.

The only problem was that this card/driver is not supported by VirtualDubs capture engine and the driver of the card only supports 384x288, 384x576 and 768x576 (in PAL mode).

Now I got a BT878a (a video only card model) for cheap money just to do some tests using Vdubs capture engine and also do do tests using software Mpeg2 capturing via WinDVR.

I installed the btwincap.sourceforge driver using its automatic detection, and the card was well recognised in Vdub and WINDVR.

Now the Problem:
In comparison to the DC30 card output this Bt878a card gives me much more noise! And also this noise is strange cause it besides comes also with a horizontal-noise "effect".

Here's an example direct output of the bt878a in YUY2 via S-Video conector (I use both cards via S-VIDEO cable where an external S-VHS VCR is used as tuner device):

http://mitglied.lycos.de/incredibleweb/BT878.jpg

And here the example of the DC30 card(same S-Video connection etc. as used in case of the bt878a card):

http://mitglied.lycos.de/incredibleweb/DC30.jpg

Both images are not treated with avisynth or something else, just direct card outputs.
But you can see my problem that in comparison to the Dc30 the 878a comes with this "more noise" and "horizontal-noise" phenomenom.

Can someone give me an advice what's wrong with the bt878a card?
Is it just that 878a cards do provide a worse quality than a philips/Zoran chip based DC30 card?
Maybe wrong driver settings of the bt878a or should I change it at the store where I bought it??

It can't be a connection problem cause both cards get exactly the same signal via S-VHS connection of the same s-VHS-VCR device.

Thanks a lot for suggestions!

[avih]

this seems like signal interferences, and not aquisition problems. there could be two reasons for this:
1. your DC board has better amplifiers/filters than your BT board.
2. the signal quality was different (i.e. different tape, different channel, different atmospherical state, etc).
3. there really is a difference in the quality of capture by the chip iteself (internally).

to make a better comparision, compare exactly the same source (i.e. DVD, at the same point in time, same tv/cable channel, same VHS tape, same location of the computer, cables and source, etc). then put 2 images of the same source, and we (and you) can have a look at them, and reach more percise conclusions.

cheers
avih.

[Inc]

Originally posted by avih
[B]this seems like signal interferences, and not aquisition problems. there could be two reasons for this:
1. your DC board has better amplifiers/filters than your BT board.Yes this Could be



2. the signal quality was different (i.e. different tape, different channel, different atmospherical state, etc).And here's the point. I did some tests with exactly same sources, said the same channels said the same original noise the broadcast comes with. No matter which source channel I always get these strange card specific noise effects in comparision to the DC30 (but I know what you intended to say/explain in your post above refering to diff. channel inputs even via the same connection/signal



3. there really is a difference in the quality of capture by the chip iteself (internally).I've heard about this but I never thought that there could be such a high quality difference, thats why I started the thread. ;-)
to make a better comparision, compare exactly the same source (i.e. DVD, at the same point in time, same tv/cable channel, same VHS tape, same location of the computer, cables and source, etc). then put 2 images of the same source, and we (and you) can have a look at them, and reach more percise conclusions.Anyhow what I now said above, youre right and I'll do a really 1:1 comparison by capturing a s-Video input signal from my DVD standalone to get a real comparison. So I'll post the pics later.

Thank you very much for your fast response!

[trevlac]

The philips chip has a 'noise' filter. The BT is not filtering the noise for you. It will be intesting to see the 'DVD' results. You can test the filter by looking at a resolution test pattern. Here (http://forum.doom9.org/showthread.php?s=&threadid=67287) is a discussion of some of this with a test pattern link.

My guess is that the DVD picture will look better on the BT if there is high detail (like credits) in the picture, especially thru s-video.

But that's just a guess. :) Anyway, if you have noise in the source the Philips does the work for you. You'd have to do the filter post for the BT.

[Inc]

@ Trevlac,

I don't think the philips chip can treat such noise that good :-)
And I don't think the Philips chip of the old DC30 comes with an internal noise reductor.But I donno exactly ;-)
It seems that I got an low signal problem which will be amplified by the dc30 ??!?


@ Avih,

Now I did the test recommended by you.

Source PAL DVD standalone directly connected via S-Video cable to the tested card.

BOTH cards got the same connection (I just switched the s-video connection from one card to the other). Same Appl. VirtualVCR. Both Card settings set to standard settings and captured at 768x576 PAL.
Pics safed in jpeg mode at 80% Quality using Photoshop.

And the reult is: :eek:

Here comes the bt878:

http://mitglied.lycos.de/incredibleweb/bt878dvd.jpg


And here comes the DC30.

http://mitglied.lycos.de/incredibleweb/dc30dvd.jpg

ok, the dc30 is a touch sharper in its default settings, but thats not the problem I had ;-)


So what could be the problem with the VCR-Svideo-Connection in case of the bt878a ???? Cause the Dc30 doesn't have such problems.

This test was done under the exact same workout condition logic.
Said: I just switched the Connection from one card to the other.

Does that mean in case of Tv signals via s-VHS VCR connection, the signal could be too low which in case of the DC30 could be pre-amplyfied by the dc30 internally??
Cause no matter what channel or what kind of broadcasted quality - the 878a always end up with this horizontal errors and also more noise in general.

Could a Antenna Signal Amplifyer solve the problem??

Well in my eyes the 878a card itself now isn't the problem anymore :D

[trevlac]

Originally posted by incredible
@ Trevlac,

I don't think the philips chip can treat such noise that good :-)
And I don't think the Philips chip of the old DC30 comes with an internal noise reductor.But I donno exactly ;-)
It seems that I got an low signal problem which will be amplified by the dc30 ??!?



Well, most ADCs from Philip do have an anti-aliasing filter. THis might be your chip: http://www.semiconductors.philips.com/acrobat/datasheets/SAA7110_A_1.pdf

But I'd say your DC30 has a filter on the card, based upon the spec above. Hard to say unless you cap a test pattern. The sample you posted look to have the same resolution to me. The brightness is higher on the DC30. This maybe why "it's better". A histogram should show the difference. Contrast and Brightness have a big effect on PQ.

Anyhow, back to your original problem. A signal amplifier may help. Why does the philips look better without these? I'm going with avih's #1.

Hey, but what do I know? I'm just a newbie at this stuff. :D

[Inc]

Thanks for your replies trevlac! :)

Well it seems that it isn't an amplifier needed - problem.

I did another test:

The bt878 is still connected to the s-VHS VCR via the mentioned S-VIdeo cable. BUT now I disconnected the antenna signal and shure... a blank blue screen appears, BUT still these horizontal errors included :eek:

Here's an AVI sample of the "disconnected antenna"-state in 668kb captured in YUY2 mjpeg (http://home.arcor.de/the.incredible/blank.avi)

Ok, I guess you're right that the DC30 and its Philips chip does filter out this problem :confused:

But what about other users who own such a bt878a video only card??
Do they end up also with such a problem since the bt878 doesn't got that type of filtering.

[avih]

i have a bt878, and it depends. if i give it an RF signal, then i sually see these patterns, to some extent at least. on a clean composite or s-video inputs, i don't get these (and you didn't get them either with your DVD as well). i think it's an amplifier or filer issue. but anyhow, since u have both, why don't u just capture with the one that's best for you? ;). just make a note which is better for you in various scenarios, and use it. you have the proviledge to have both ;)

cheers mate


ok, i watched your clip. i see the pattern, and still i can't have a definite answer. since you plugged the antenna cable off the VCR, and the blue image is generated by the vcr itself (i know this since i see the VCR output of the channel #), still, the VCR signal may be too weak, or too noisy for the BT card. if u have one handy, i'll try an amplifier. also, you may try to move the BT card to another PCI slot, as the noise may be due to a specific physical location on the mobo itself.
good luck

[Inc]

Thanks avih,

I connected a different VCR to the Card and the same problem resulted.

I found an old Signal amplifier and connected it, but the signal got even worse :eek: But maybe the amplifier isn't ok, I also heared about problems if the signal is too much boosted so this could give on the other hand problems too. Well, ... the last thing I'll try is to pluf the cart into another PCI slot, but I've to watch out for INT/IRQ assignments.

[vhelp]

Hi incredible :)

Funny you brining this up all the sudden.

I've suffered from this (what I call, Line Noise) for ages. I first brought
this up on vcdhelp over a year ago I think. I've never quite found the
answer. Except that after countless testings, I narrowed it down to the

* Mobo/CPU/Chipset.

Here is the thread that started it all, for your reference, so you know
where I'm comming from too:

* Line Noise - mostly from VCR sources...not quite solved yet. (http://www.dvdrhelp.com/forum/viewtopic.phpt=107796&highlight=line+noise)

And guess what. Like you, I stumbled accross this same exact issue as
you, w/ my DC10+ card :eek: :eek:

However, my only answer was a comprimise. I got the ADVC-100 LOL.
But, I'm still seeking the answer to this VERY LONG quest, just as you
are beginning to now :)

My recommend, is to honestly try another brand Motherboard.
Some months ago, I did some digging around my CDR's I had encded some
MPEG samples to. I found one that I did w/ my ATI-TV Wonder (LOL) and
to my surprise, thre were no Line Noise in that (and others) sample :eek:
The card used to be inside my old K7T Pro2-A (MS-6330) - Athlon 900mhz w/PC133 128MB
Now, I have currently an ESC K7S5A w/ XP 1700+ and 1800+ (two identical
Mobo's) and both exhibit Line Noise w/ any BT type cards. Only my
DC10+ and ADVC-100 exhibit ZERO LINE NOISE.
The sample was a Winny The Poo cartoon holloween special. Cartoons are
easy to pick up Noise in.., but there were none. So, my theory about
if being:

* Mobo/CPU/Chipset.

would seem logical after all :rolleyes:
-vhelp

[Inc]

Vhelp, I apreciated very much you help!
Thats sounds like a nightmare, so I will stuck with my DC30 and its well working WDM Drivers under Win2000. It would had just be a funny thing to get another card working which runs under Vdub and WinDVR just for tsting, cause the driver of my dc30 (a modified dc10 driver) won't be recognised by those apps.even if using the wrapper only VirtualVCR does a good job with the dc30.

So I gonna sell my bt878 as it seems that it works correct but not with my actual running hardware combination :(

Thanks to all for the replies!

Inc.

[vhelp]

@ incredible..

great.

Listen, I heard some rumor that the DC10 can do 720 (vs it's 640) w/
a modified DC30 :eek: driver. I've never ben able to try it, nor find
any other info on it, though I still hear about it. If you should come
up against it, hold on to it, and point me in that direction :D

I like my clean DC10 captures. They're the best, but 640 will
not help my proper Aspect Ratios when I want to process as full 720
DVD compliancy. I don't want to "stretch" my source just to.. if you
know what I mean :cool:

Thanks,
-vhelp

[Inc]

Well my friend if I do see something,
you'll be directly informed! (I think you are also a Mod at KVCD like me thats why I already "know you" ;-)

The thing is that there are NO authentic WDM DC30 drivers for Win2000/XP.
All these versions of drivers do got their roots on modified DC10 drivers, cause the dc30 got's the same chip ... and that's why you also won't be able to do sound capturing via the dc30.
But anyhow its better to do the audio capturing in case of the dc30 via a good mainboard sound device cause the dc30 Soundchip got a bug on the right channel.

And as I do use a modified dc10 driver I have never heard about the possibility to capture in 720x480 px :(

And you're right, the quality in case of tv captures is awesome.
The TV source samples above are samples of noisy brodcasts each, just to show the difference of noise of the two cards.

I did a capture of a commercial and have put it last summer in a "DVB" thread at doom9.de just to show the capabilities of that "analog" card and its Zoran/Philips chip. First they didn't belive me there :D

Go to the Capture section of KVCD, there I posted a link to a DC30 Win2000/XP driver, ... maybe that one already gives you the choice to directly capture in 720x480?? Just try.

Another nice thing would be, if I could deactivate the onboard mjpeg compression, just to play a bit with other capturing codecs.

[Arachnotron]

Isn't the DC30 using a SAA7110/7111 for AD conversion?

How does the quality compare to cards using the SAA7133/34?

(Just curious)