Qusetion on : capture DV to Computer using FireWire...

[rover0man]

Thanks for your time for my questions... and I am new to this forum.

What i need:
Capture the best video quality(professional to prosumer standard) from a DV tape or TV source and write into DVD disk.

p/s : I always belive that "Best VIDEO QUALITY" does not necessary to mean most expensive products but the best way.


Facts that i know:
1) DV is a 5:1 compression about 4.2Mb per second or 15Gb per hour of data...
2) DVD is about 4.5 Gb per hour, so DVDV = 1:3.33? DVD is only 1/3rd quality of DV quality thru MPEG compression?
3) VCD is about 650Mb per hour, so VCDV = 1:23? VCD is only 1/23rd quality or DV quality?


My qusetion is:
A) will there be any different if i capture DV into computer using
1) On board FireWire port...
2) PCI FireWire add on card...
3) A Video capture card that with a ForeWire input on the card itself (software encoding)...
4) Or a Video capture card that with a ForeWire input on the card itself (hardware encoding)...

B) Does DV camera (PAL) record video in 25 frames or 50 fields (interlaced or non-interlaced)?

C) Do not know what to do as analog video are recorded in 50 fields (interlaced) but computer capture and played in 25 frames (progressive)... So???
1) Should i capture video into computer using 25 frames or 50 fields (if i am asked to choose)?
2) After edit, should i out put back to tape using the 25 frames video? or should i change it to 50 fields before out put?
3) I experience interlaced problem on computer screen but not when i link it to TV screen, is there any software that can have WYSIWYG (what you see is what you get) on both the computer screed and Tv screen. I can find progressive scan video player in the market but preview from converter does not provide this feature...


As i am looking for 1) a TV capture card and 2) method to capture DV video using firewire and produce DVD as a end product...

So i need to know does different method of capture video produce different result? Does a onboard firewire port with a 2.8GHz processor produce good reault? or i should atleast get a PCI add on card with firewire port (some chips on the PCI card will help to improve the video quality)? or i should go all the way to get a video capture card (even for DV tapes) that will enhance the quality by both hardware and software. How much the different will be?


Can anyone kindly recommand some TV capture card that can capture TV source into high quality Mpeg2 to burn into a DVD...
iS Leadtek Winfast DV2000 good? what about WinTV 350 with hardware encoder and decoder?

THANKS...

[underwurlde]

Interesting...

You base quality on compression ratios!!?!?!?!?

So if you ZIP up a program to say 10:1 compression you will only get back 1/10th of the original program which will therefore not work?

THINK about it.....

OK, ZIPPING a file is LOSSLESS but the fact that any DIGITAL representation of our analogue world (beit audio or video) is inherently lossy - i.e. digital is not infinite.

Once transfered into a computer readable digital domian we can then apply digital compression algorithms - some of which are lossy, some of which are not.

The video & audio stored on a DV tape is compressed 5:1 - this is not 'lossy': a small compression is used: Small because it is fast, compressed because if not, then storage tapes would be 5x bigger and cost 5x more.

DVD is as good a quality as you'll ever need & uses MPEG2 compression algorithms.

To transfer video (that will be DV compressed) from camera to PC I use a firewire port - it does NOT matter where or how this is connected - I use a PCI card (unless its a poxy PCMCIA laptop card): Firewire is a STANDARD.

The captured DV is interlaced and I De-interlace and store as progressive (because my PC is not a TV!) and then burn onto a DVD - interlacing the signal for TV is taken care of by the DVD player electronics.

A harware capture card with MPEG compression just does (in REALTIME) what I do on my PC when I compress in software to MPEG - just takes longer.

The MPEG compressed video file will be 99% quality of original. i.e. you won't notice the difference!

Andy.

[trevlac]

@rover0man

I'll try to answer a few of your questions. But you had too many for 1 post.

A) DV Firewire capture is a file copy. If it works, it works. If not, the PC just can't find the cam. Some cards/software don't work together. 99% probably work.

B) 50 Fields interlaced

C) If you are going to watch on TV, use interlaced. PC and TV monitors are different. There is no perfect solution for both. If the source started as interlaced (such as DV cam) the best is to keep it interlaced and have the progressive monitor de-interlace as you play it. Film based source started as progressive. For this, the best answer is a little different. Depends on the details of what you are watching (like DVD or xvid or etc.)

Here is my little tutorial on interlaced video:

- Interlace or Progressive is created when the scene is filmed.
- For Film, a 24 pictures are taken every second.
- For video, 50 pictures are taken every second.
- Strange thing of video, is that every other line of the 50 pictures is thrown away

The point is that the 2 sets of pictures hold different information. If there was something moving, interlace will have 2x as much time detail but only 1/2 line detail. An ideal conversion would properly create the missing info for the given format. But this is near impossible to do. Most solutions kinda fake it. The problem with faking it is that either the motion is not as smooth as it should be and you notice, or the picture is not as sharp as it should be. It is blurred.

Hope this helps a little. There is a lot to learn about video. It takes time.

[mikesbytes]

First step is getting the DV data from the camera to the computer. As underwurlde mentioned, you are not capturing the video, you are actually transferring the digital data that is the video. Hence it is unlikely that the type of firewire port selected will make a difference.

What is important is that your computer doesn't get behind while the data is being sent from the camera, when this happens, the computer will skip a frame to catch up. The speed of the CPU has little effect on the transfer, a Pentium3 450Mhz is more than adequate.

What will impact it is what software you use and the performance of your hard drive. I use WinDV to transfer and this software will buffer the video data being received, so if the hard drive gets behind, WinDV will save the frames in memory, rather than drop them. The next most important thing, is the Hard drive. If you only have one drive, that drive will be interrupted every so often by the operating system, potententially causing dropped frames. The most common trap, is not having DMA activiatd on you existing hard drive. If your budget will stretch, buy a new second drive 7200Rpm, DMA100 or better. Pretty much all new hard drives meet the spec nowdays.

Having got the video to the computer, your journey is just begining. You have a lot to learn about video quality. Spend the next 3 months reading this site and others like it.

[trevlac]

The video & audio stored on a DV tape is compressed 5:1 - this is not 'lossy': a small compression is used: Small because it is fast, compressed because if not, then storage tapes would be 5x bigger and cost 5x more.

DV is actually lossy. DV uses most of the same stuff as MPEG. The main difference is that MPEG uses frame sameness to improve compression. DV only applies to a single frame.

The captured DV is interlaced and I De-interlace and store as progressive (because my PC is not a TV!) and then burn onto a DVD - interlacing the signal for TV is taken care of by the DVD player electronics.

If your source is something like a TV show, you are loosing quality by de-interlacing. If the source was 'filmed' as interlaced, you should try to keep the interlace if you are making dvd.

A harware capture card with MPEG compression just does (in REALTIME) what I do on my PC when I compress in software to MPEG - just takes longer.

Yep. My experience is that hardware tends to be easier to use, but software is much more flexable.

The MPEG compressed video file will be 99% quality of original. i.e. you won't notice the difference!

Well, this depends upon how much you compress it. If you only want to stick an hour of DV on a DVD you are set. If you want 3 hrs, you are going to have to work at it to get reasonable quality.

Noone can sit thru more than an 30 minutes of my home movies, so I'm set.

[underwurlde]

Points taken! In an effort to keep things simple I kind of generalised!



Andy

[trevlac]

Points taken! In an effort to keep things simple I kind of generalised!



Andy

Yeah, I know what you mean. Sometimes it is hard to know where to start ... and how much to go on about ....

[rover0man]

thanks alot guys...

i think i need sometime to study those informations...

thank you "underwurlde", "trevlac", and "mikesbytes"... will pm you all...

[rover0man]

once again thank you guys for the informations...

You base quality on compression ratios!!?!?!?!?

Yes that is what i think as for video compression but not for data compression... I always think that there is no free lunch... so in order to save video data storage space and reduce processing speed, we need to compress the video file by
1) throw away unnecessary data (that is what Jpeg and Mpeg does in their high res compression) to achive not noticable quality loss...
2) compromise in quality loss by reduce the image size from 720 x 576 to 352 x 288 (1/4th of the area) or change the frame rate from 25 frames to 12 frames per second...
I do notice that different codec will give different quality for the some compression rate, but in general i still think that compression rate = quality...

DVD is as good a quality as you'll ever need & uses MPEG2 compression algorithms.

Yes i agree with you that the quality of DVD is good... but still wonder how DVD keep the high quality when it is only 1/75th (4.32Gb : 75.6Gb) of uncompressed video and 1/3rd (4.32Gb : 15Gb) of DV video...

Well, this depends upon how much you compress it. If you only want to stick an hour of DV on a DVD you are set. If you want 3 hrs, you are going to have to work at it to get reasonable quality.

but i notice even you keep it to one hour per DVD disk... we still can choose different bits rate like 6Mbps to 15Mbps... so what is "best"?
1) video bits rate ( 6 - 15Mbps) ?
2) audio bits rate ( something like 96 - 386??) ?
3) CBR or VBR ? (i once experience capture video into computer sing VBR, but the quality is bad) ?
Quality is my concern... As long as our eyes and ears (in professional or prosumer level) can't detect the different between the original and DVD video, i will consider as a good quality...

Once transfered into a computer readable digital domian we can then apply digital compression algorithms - some of which are lossy, some of which are not.

First step is getting the DV data from the camera to the computer. As underwurlde mentioned, you are not capturing the video, you are actually transferring the digital data that is the video. Hence it is unlikely that the type of firewire port selected will make a difference.

I'll try to answer a few of your questions. But you had too many for 1 post.
A) DV Firewire capture is a file copy. If it works, it works. If not, the PC just can't find the cam. Some cards/software don't work together. 99% probably work.

DV firewire capture is a file copy? Yes video signal in DV tapes are digital. But when DV cam play the video thru firewire into computer... Computer does not just copy it, does it? Because the captured/transfered file are either in AVI, MOV, OMF/AIF (AVID version) or Mpeg2... there are some re-computation done, isn't it? File copy should be 1 to 1 copy and no calculation done... and although is the some AVI file (for example). If captured/transfered by software "A", don't mean that editing software or player software "B" can reconize that "AVI". Simply say the file is not reconized... So as long as CODEC and convertion involved i will consider as capture... sorry just my knowledge and understanding towards DV and computer world... Please correct me if i am wrong... thanks...
p/s : If is file copy... how nice if we can just key in the in and out point and computer can transfer the video/data in seconds... and not in real time "batch capture/digitize".

[rover0man]

I'll try to answer a few of your questions. But you had too many for 1 post.

Ya, you are right... i myself get confused too... I wanted to buy a TV capture card and burn video into DVD, that is why all the questions pop out together.. ha ha...

Ok can we narrow down to this topic first and go on to other questions later...

First Topic is : DV TAPES / FIREWIRE
Things to disscus under this topic is

1A) DV TAPES
...data stored on tape is digital?
...video is interlaced fields?
...does the vieo data stored in DV tapes are Mpeg compressed or in DV own digital format?
...how does DV achive the 5:1 compressoin?

1B) FIREWIRE port...
...is a form of data transfer only, so it does NOT matter where or how this is connected?


Keep the following topic till we finished with topic 1.

Second Topic is : TRANSFER / CAPTURE / DIGITIZE INTO COMPUTER
Things to disscus under this topic is

2A) COMPUTER (receiving end)...
...Does computer just copy the data, no calculations done?
...Does computer apply different CODEC from different software?
...what does CODEC does at this point?
...Most P3 or P4 with 512Mb RAM and ok hard disk can handel the 4Mb data transfer, so in DV/FireWire technology do we still need to buy a DV capture card? How does these high ended capture made a different in quality or they don't?

I notice there are alot professional out there (people from Videohelp.com too)... Please join in and don't worry that we might not be able to understand what you said (we will PM you and ask you). Give your professional view and we will decide what to take in... what to ack blur... ha ha...

Forum is a good place to learn, learn thru other's problem. But it will be good someone can consolidate all information into one help file or web page... because i face a problem is that normally i need to view thru a numbers of topic before find answer or hints to my question untill i need to start one here... and there are alot repeated or similar topic in the forum. These information no longer is some company secret... let share it in a more effective way. Thanks, just my point of view...

[guns1inger]

If you take the original DV footage from the tape, and store it on the hard drive as a DV type avi, then it is a file copy exercise. Remember that avi is a container for digital content. If the content is DV taken from a camera, it is a bit for bit copy of the original. If you elect to re-encode the DV footage on the fly, as some capture programs allow for, then you will change the data and may impact the quality.

It is not a fair or correct call to compare the compression vs quality equation across codecs, only within a single codec. The way DV compresses is sufficiently different to the way mpeg-2 compression is applied to make comparisons purely on file size irrevevent. On the other hand, there is a relationship between compression rates within a codec. Squeeze mpeg-2 hard, quality drops. The art is finding a happy medium that allows for the right bit rate at the right places.

Codecs aren't only about compression, either. They are about how the data is represented. As Trevlac mentioned, mpeg compresses across frames, as well as within frames. DV compresses only within frames. This makes DV a much better codec for editing in, as you can be frame accurate with little effort. With mpeg-2 editing is more difficult at the frame level because not every frame contains all the data required to recreate it.

So, to answer 2A,

If you save the file as a DV format avi, it is a direct copy of the original data. As DV tapes play at a single speed, capture is real-time. 1 hour of footage takes 1 hour to transfer.

At this point, a DV codec is required purely to decode the data for presentation, either as a preview during transfer, or for playback through your media player. It does not alter the content of the file while doing this.

Generally, the codec will be the ms DV codec. There are better ones available. Sony has a good one that ships with Vegas, and that Vegas uses when rendering new footage in the DV format. A good DV codec can go several generations with little or no dicernable loss of quality.

I have processed DV on a celeron 500 with 768mb ram and 7200rpm disks. For best results - fast disk, not a system disk, formated as NTFS, defragged before transfering the dv files across.

[mikesbytes]

rover0man, pickup a cheap firewire card and start playing. You can learn as you go.

[trevlac]

First Topic is : DV TAPES / FIREWIRE
Things to disscus under this topic is

1A) DV TAPES
...data stored on tape is digital?
...video is interlaced fields?
...does the vieo data stored in DV tapes are Mpeg compressed or in DV own digital format?
...how does DV achive the 5:1 compressoin?

1B) FIREWIRE port...
...is a form of data transfer only, so it does NOT matter where or how this is connected?

1A) yes, yes, DV compression, it is slightly lossy

1B) yes ... but since a tape plays at a constant rate, the pc has to keep up. Most pcs are fast enough.


BTW: I work for a bank in a job that has nothing to do with video. I just learned by reading and mostly doing. (Like a lot of people around here.) I've also collected a bunch of hardware and software over the years. As people have said... dive in and join the fun.