Tested the new Conexant CX2388 chip

[Anonymous65]

I have eagerly waited for a new simple AVI capure card that has the video quality of the high end capture devices like the ADVC from Canopus. I got hold of the new card from Prolink using the new chip, and gave it a test. However, I was quite disapointed...

Check out the test images, http://steve.kittelsen.com/cx2388/

[trevlac]

Steve,

I too have tested a CX based card (MSI) and was a bit let down. I compared it to my Aver BT878.

The comb filter was the bigest disappointment. I expected much better results. However, I did find that adjusting the driver settings made a difference. I never completed my tests, but I did conclude that the BTwincap default settings were much better than the default MSI settings.

Also, I found that the 8bit vs 10bit difference to be very minor. It can be best seen in a 6.75Mhz Circle pattern.

My biggest surprize was the tuner in my Aver card. It seems to apply a smoothing filter that is quite good for CATV. The loss of resolution is minor, and the results are quite pleasing.

My conclusions have been that the BT878 preforms very well with CATV source and worse. Hi8 may show it's limits.

Have you done a compare of a DV device to your BT878? Do you expect them to be better? Why?

Did you compare s-video connections?

Thanks


Edit:

I just found your writeup on the ADVC-50. Did you feel that the composite input to the DV device did better than the CX card? It's hard to tell from the test pattern.

I'm not sure why there is so much black on each side of the picture. My BT8x8 capture does not give the black borders on each side. I assume the ADVC-50 captures more of the video, and that is why I see more black.

The reason for the black is that (assuming NTSC) analog from a dvd player is about 52.666 microseconds. A dv device captures a width of 53.333 microseconds because it follows the digital standard (ITU-601). The BT878 card (depending on your driver) captures about 51.55 microseconds and streatchs it to 53.333 ms if you ask for 720 pixels wide (like you get from the dv device).

I've been reading alot about this lately. If you compare your test clips, you'll notice that the big circle is not quite round on the CX or the BT. The DV device looks round and should be.

Most people don't seem to be aware of this strange quirk.

[Anonymous65]

Nice information about the black sides on the DV captures. The reason I'm looking for a new AVI capture card is that I know the video can look better. With the ADVC I get much better detail, which means the ADVC use a very good video decoder chip (I don't know which one). I only use composite video, and not S-video. Most of my sources are composite, so that is why I stick to it. Obviously, the comb filter or notch filter is not used if you're using the S-video input on the video decoder.

I was hoping that the new capture card would be on-par with the ADVC. Since I have no need for the firewire connection and the hardware DV encoder, the ADVC is not my best option. So, I wonder if anybody will make a simple PCI capture card with a *high* quality video decoder chip on it?

[vhelp]

hay skittelsen.. good to see you around

yeah, I have a question about the pattern posted on your website (as well
as others I seen around)

* My quesiton is, HOW are you using it in your testing of sources ??

*I'm confused. Are you shooting a video of a huge picture on a wall or
something ??

I'd like to use this pattern myself in my tests.. since i have about 6 capture
devices in all my pc's. I'd be curious to see what results I get. Is that at
all possible for me ??

Any help would be much appreciated, thanks.
-vhelp 2130

[andie41]

I would be very interested in how a DV test pattern looked. I have a PCTV Pro (BT878/ BTwincap) and have been quite pleased with it - capture 704x480 huffy with no drop frames and considering the quality? of NTSC video, I have been quite pleased.

Recently I purchased a used Sony Digital 8 camera and now have the option of using it to convert my Hi8 video to DV. I have been trying to see which method will give me the best results, thus far I can't really see much difference. Again, just trying to do a visual on my NTSC monitor - none of my test clips look great, all look ok. Certainly, the DV capture interface is much easier to use.

andie