Capturing DV vs. Analog Question

[johnny quest]

I have a Canapus ADVC 100. I have captured Hi 8 home movies from my Sony camcorder and have also captured some Tivo'd programs.

I am unhappy with the resulting dv avi and its quality of picture. In both instances I notice macroblocking (looks like poor quality mpg). My question is whether or not this is the best quality I can expect? I thought capturing in dv with the Canopus would result in high quality avi's given the size of the avi's. Is this normal, would I get better quality from a straight dv camcorder?

I capture via firewire using the Canopus and Arcsoft Showbiz software. My computer is Athlon 2000 (1.6 ghz) with 512 ddr ram. This should be sufficient to capture with no dropped frames, but I don't even know if it is possible to drop frames when capturing dv.

Can someone please advise me? Thanks

[bugster]

Are you encoding during capture, or do you capture to DV Avi and encode to mpeg later. In either case, what bitrate do you use?

[johnny quest]

I capture to the dv-avi, no encoding at all. The resulting avi is a type 1. I have also converted to type 2 but no change. Bitrate is whatever the standard is for dv-avi. I am viewing the avi quality.

The gist of my questions are that my dv-avi files look like low bitrate mpgs.

Any help or advice appreciated.

[bugster]

There are several possibilities here as to what is going wrong.

Firts thing, what are you using to view the .avi files with. If it is WMP, don't. Use a recent version of a software DVD player like PowerDvd or WinDvd. They both (I think! Certainly PowereDvd does) will de-interlace the video for display. DV is interlaced, your PC monitor is NOT. Encoding to a high bitrate mpeg, burn to DVD and viewing on your TV will also fix this to a large extent.

The other possibility is simply that this is home movies. Hand held cameras create camera shake, eating up the available bitrate of any encoder, even DV. This won't help the image quality but really shouldn't cause macroblocks in DV encodes. You may want to take a look at some of Lordsmurfs or Satstorms guides/posts with respect to applying filters to avi capture to help remove noise and other artifacts.

Hope this helps.

[johnny quest]

Thanks for the info. I have tried viewing in Power Dvd, same prob. I don't think its the home movies thing because it also happened with the Tivo programs I captured. Thanks for the response. Any other ideas?

[Tommyknocker]

Have you tried capturing with DVIO?

[BJ_M]

there may be a problem with your canopus ... some units had this problem, can you take it back to the dealer and have it checked? or call canopus tech support ...

[spiderman2k1]

My might want to try Dazzle 2 capture card analog MPEG 2 very good. Go to www.dazzlegeek.com to read more Hauppauge WinTV PVR-350 and ATI AIW are good. All these are real time MPEG 2.


johnny quest I capture all the 1996 johnny quest cartoon's to SVCD and they are top of the line. I have almost ever johnny quest cartoon LOL.

[Sulik]

Contray to popular belief, capturing via DV is usually worse than capturing via analog if you intend to make a MPEG file.

The reason is that DV is interlaced 4:1:1 video format, analog is 4:2:2, and MPEG is 4:2:0.
Converting from 4:2:2 -> 4:1:1 -> 4:2:0 (DV->MPEG) will result in losing half the chroma resolution compared to converting from 4:2:2 to 4:2:0 (Analog->MPEG).

[Blkout]

Ya know, I was about to create a topic this morning asking about this but it sounds like you may be experiencing the same problem I'm having.

I've been using my ADVC-100 for almost two weeks now and every capture I've done has produced noticable compression artifacts in the picture. On scenes with dark backgrounds its not as noticable but with lighter backgrounds its very noticeable. Looks a little like mosquito noise. I know DV compressed more than analog HuffyUV, but after reading the rave reviews on the Canopus, I'm a little dissapointed. I sold my Radeon 9700 Pro All-In-Wonder for the Canopus and as of right now, I can say without a doubt the ATI is superior.

Does anyone else notice DV compression artifacts or is my Canopus just faulty?

[trevlac]

Contray to popular belief, capturing via DV is usually worse than capturing via analog if you intend to make a MPEG file.

The reason is that DV is interlaced 4:1:1 video format, analog is 4:2:2, and MPEG is 4:2:0.
Converting from 4:2:2 -> 4:1:1 -> 4:2:0 (DV->MPEG) will result in losing half the chroma resolution compared to converting from 4:2:2 to 4:2:0 (Analog->MPEG).

I tried to bring this up in another thread. Hopfully people will check this out.

This problem may be mitigated a bit by the DV codec used to decompress the dv and upscale it 4:2:2 (or 4:4:4) before it goes to 4:2:0. This means the codec is important. Another attempt people have made to get around this problem is to scale up the entire frame and then scale it back down. I assume the way this works in direct show is that the chroma gets interpreted from 4:1:1 to 4:4:4, the editing tool would then scale up and down. This would smooth the colors back out. Then you go to 4:2:0.

I don't have one of these DV capture devices, so I can not truely test.

Edit:

PS. A transfer via firewire vs via s-video does have advantages. If NTSC Hi-8 has a better luma resolution than s-video can transfer you would loose luma resolution going hi-8 -> s-video capture. Hi-8 -> DV capture would keep more detail. See this link for some compares: http://www.bealecorner.com/trv900/copy/copy.html

I believe Hi-8 does not have better resolution than s-video can cary, and a CX capture card probably caps as much detail as a DV device. So dv is not better. Unfortunatly I am not speaking from experience. I do not have a CX card nor a Hi-8 cam to test. I may test a DV to s-video to BT8x card, because I am now interested.

Trev

[beingfunny]

I would try another program other than the ArcSoft Showbiz also. I was "unimpressed" with anything I did with that.

[bode]

I've done quite a bit of "capturing" with ADVC and I can't honestly fault it. I would ask the techie at the Canopus forums.

http://forum.canopus.com/postlist.php?Cat=&Board=GenADVC

[KrisS]

Here what I found on
http://www.cesnet.cz/doc/techzpravy/2001/12/

...
There are two important encoding schemes of color spaces used in video processing of interest for us (R, G, B shall mean gamma corrected values for reg, green and blue in range 0 to 255):

YUV
This is the basic color space used by PAL, NTSC and SECAM. It transmits luminance (or luma - denoted as Y) information stream and two color information streams (denoted B and V). This is connected to transition from B/W TV receivers to color ones. It allows B/W receiver to decode B/W information only (Y stream) and color receiver to decode complete color information.
Y = 0.299R + 0.587G + 0.114B
U = 0.147R + 0.289G + 0.436B = 0.492(B - Y)
V = 0.615R + 0.515G + 0.100B = 0.877(R - Y)
YCrCb
This is the format used in DV processing. It's modified (scaled and offset) version of YUV format. DV sampling schemes (e.g. 4:2:2, 4:1:1, 4:2:0) are defined with respect to this format.
Y = (77/256)R + (150/256)G + (29/256)B
Cb = - (44/256)R + (87/256)G + (131/256)B + 128
Cr = (131/256)R + (110/256)G - (21/256)B + 128
Following down-sampling modes are commonly used in DV:

4:2:2 means 2:1 horizontal down-sampling, no vertical down-sampling.
(Think 4 Y samples for every 2 Cb and 2 Cr samples in a scan-line.)
4:1:1 ought to mean 4:1 horizontal down-sampling, no vertical.
(Think 4 Y samples for every 1 Cb and 1 Cr samples in a scan-line.)
It is often misused to mean the same as 4:2:0
4:2:0 means 2:1 horizontal and 2:1 vertical down-sampling.
(Think 4 Y samples for every Cb and Cr samples in a scan-line.)
For consumer and prosumer DV-based standards only 4:2:0 (for PAL) and 4:1:1 (for NTSC) are used. 4:2:2 sampling is used in higher quality technologies (like Betacam SX).
...

Assuming that capture is done @ 720 x 480, then resized to 352 x 480 (MPEG2, interlaced) or 320 x 240 (MPEG4, progressive) -- can somebody suspect compensation of horizontal down-sampling?

My experiences with ADVC-100 capturing from cable TV are good with those settings...

What are your thoughts about that?

KrisS

[trevlac]

Assuming that capture is done @ 720 x 480, then resized to 352 x 480 (MPEG2, interlaced) or 320 x 240 (MPEG4, progressive) -- can somebody suspect compensation of horizontal down-sampling?

When you feed a 4:1:1 image in an mpeg encoder, the DV codec probably upsamples (interpolates) the 4:1:1 to 4:4:4 (or 4:2:2). This is then down sampled to 4:2:0 for the dvd by the mpeg encoder. The quality of this is dependant upon the interpolation of the DV codec.

If you down size the entire image, the editor is interpolating the pixels again. This may smooth rough color edges caused by the 4:1:1 capture. This depends upon the quality of the resize.

If you are trying to improve your process, you should try to create an example of the problem. You will then know if your process/software improves on the problem.

Bottom line, s-video gives more color and less luma than DV. Does it matter in either case? Probably depends upon your source and your process.

[sch_m5]

I’ve been using ADVC-100 for about year as well as been reading many threads about macro blocks, washed edges, and other artifacts in DV .avi captured by amazing ADVC-100. I’m almost sure that the issue being discussed caused by some differences between MS and Canopus DV codes. ADVC-100 captures video using Canopus HW DV codec but resulting DV .avi gets wrapped in MS DV .avi (unless you use Canopus firewire card and software). To playback such .avi a player uses MS DV codec (or something similar, like Main Concept if it’s installed). Download from canopus.com Canopus DV File Converter (free) and DV codec (playback is free but only playback is needed in this case). The converter does not decompress/compress video, just converts .avi format from MS to Canopus, so there is no extra loss. To playback such DV .avi file a player uses Canopus DV codec. It seems to me Canopus DV HW encoder and SW decoder better “understand” each other. I’ve made a lot of home video as well as TV captures using this method and do not have this artifact issue (w/ MS codec I do).

BTW, when you encode DV .avi in TMPGEnc or frameserve it to TMPGEnc from VirtualDub also try Canopus DV.

Hope this will help.