Component signal out?

[Martin Munthe]

Is it possible to override the mpeg2 tape compression by connecting the JVC to a workstation and capture using a DeckLink HD or CineWave straight to disk? Is it sending the signal component out all the time? I hope it's a HD D/A from the DSP. I've done the same thing with a PD150P to uncompressed SD via Y/C with good results (over riding DV25 compression).

[Eric Bilodeau]

That would certainly be a major quality boost. I do not know how the signal is transfered out in the component port though. It would certainly be worth trying. If the image is sent directly uncompressed it would make HDV a real good HD format indeed, a bit teedious to work with but who knows...

[Steve Mullen]

Connect the component output to an AJA Analog HD-to-HD-SDI converter. From there you can input to several different HD NLE's.

[Kevin Lee]

Hi Martin.

Can you describe your "PD150P to uncompressed SD via Y/C" setup and give me a rundown on your experience? Pictures perhaps too. Is there a significant boost in quality that is worth the effort in bypassing the DV25 compression? (very interested)

P.S. I am a fan of your work... if that helps. Haha.

Thanks in Advance.

[Steve Mullen]

Can't do.

The NTT encoder is half-duplex in HD.

[Alex Raskin]

Martin, you can't use Component out for uncompressed HD capture because these particular cameras (Hd1/HD10) only output 480p in *Record* mode, and NOT 720p.

Of course you can play the tape with the already recorded signal and capture it via Component, now at full 720p resolution... but your question was whether you can bypass mpeg2 compression during recording... so playing the already-compressed signal off of the tape won't provide you with any solution...

The cams do output full 720p during recording on their iLink ports - but this signal is mpeg2-compressed Transport Stream, which defeats your purpose.

Also, even if the cams did output 720p during Recording (which they don't), how do you know that the signal did not derive from internal mpeg2 decoder AFTER the compression/decompression? I for one am not sure that any prosumer camcorders can actually output uncompressed Component at this time.

[Martin Munthe]

Kevin: Color resolution doesn't gain anything from going Y/C. But image resolution does. Getting out of 5:1 compression is a good thing. DV25's heavy compression produces a lot of stair stepping artifacts in vertical lines that improves a lot by going Y/C. Perhaps I get some time to post a few examples later some day.

Steve: Ah! Forgot obout that HD SDI issue. Could you explain half duplex NTT?

I haven't seen the camera IRL but from what I heard the PAL version was setup, power on, connected to a HDTV displaying the HD signal from the camera (switched on). Did I misunderstand this? Or how do they get the signal out from the camera if the component interface doesn't deliver a HDTV signal but only 480p/576p? Is it only sending signal when the tape is rolling?

Alex: My hope is that someone that owns both the camera and a HD editing setup would try it out. Theres obviously a D/A interface in the camera. Usually the D/A sits before the hardware codec that compresses to tape. That's the case with all Sony miniDV cameras.

[Paul Mogg]

The PAL version doesn't do HD at all (for some obscure reason it was built that way), and the component outs on the HD1OU only output 720p during plaback, not during recording. The Firewire out DOES output MPEG2 TS compressed 720p during recording and playback.

Hope this clears up the confusion.

[Darren Kelly]

I am capturing 720p off a recorded image via component through converters to the KONA HD card.

Image quality is very good. Caution those4 though, you need very fast drives and lots of them to do this.

DBK

[Alex Raskin]

Darren and all who capture via Component: why?

The signal is already recorded on tape with mpeg2 compression in digital form.

On playback, camera (among other things) uses an internal D/A converter to output decompressed mpeg2 via analog Component (almost in real time). You then capture this via AJA (or whatever card you use) and use your card's A/D converter to write the digital output to your hard drive.

Inevitably, D/A + A/D conversions introduce their own noise/artifacts etc. Quality components may introduce very little problems - but why risk them at all, when you can simply convert DIGITAL (unaltered) m2t files (as output by cam's iLink) into DIGITAL AVIs simply by using HDTVtoMPEG2, then DVD2AVI software?

It's fast, and there's no super-expensive Component capture cards involved.

Because software does NOT convert in real time, even if you are saving Uncompressed AVIs, you won't need any super-fast/expensive RAIDs: software data stream is slower than real-time hard drive write required in case of hardware capture.

I personally do not know whether software conversion can introduce its own artifacts, but so far I have not noticed any by using the software-based HD file conversion method.

Anyone who can make an actual, instrument-measured comparison between the QUALITY of hardware- and software-based HD capture, based on signal analysis?

Please pitch in and let us know what's the real story here!

[Barry Green]

Alex is spot-on. Recapturing compressed/uncompressed data via Analog and then digitizing, is only going to lower quality (although very slightly). The highest quality you're going to get from this camera is through the firewire stream, capturing the direct digital data. There is no pre-compression analog feed out of it.

[Steve Mullen]

<<<-- Originally posted by Alex Raskin : An playback, camera (among other things) uses an internal D/A converter to decompress mpeg2 and then output analog Component (almost in real time). You then capture this via AJA (or whatever card you use) and use your card's A/D converter to write the digital output to your hard drive.

1) D/A and A/D do not have anything to to with compression/decompression.

2) The rest of the HD world uses HD-SDI input of uncompressed HD. This world knows nothing of FW. Of course you can ask them to capture via FW and software decompress.

Or, you can hook up a little black box to the camera and give them realtime uncompressed.

Now suppose you are paying them $500/hour for all your source footage.

Darren's right on. And, I've added this workflow to 4HDV. Sometimes it's easier to go with the flow than argue about D/As.

[Alex Raskin]

Steve, I respectfully disagree.

"rest of the world" using HD via SDI: they *do not* use $3K cameras as acquisition source.

Hello! :) This group is for the folks with $3K cameras! We need solutions for THESE cameras, not CineAltas or even SD cams that start at $25K etc... etc...

And we do not get paid $500/hr normally...

So what we need is *practical*, yet quality solutions that let us actually use the HD footage from HD10 cams and view/distribute the finished movies.

Correct me if I'm wrong, but software-based solutions are afforded by anyone on this forum (free downloads), introduce no known quality degradation, AND allow to capture/edit/output our footage NOW.

Or, we can discuss inferior quality hardware solutions where capture card is half of the camera cost, and more or less mobile RAID systems cost *starts* at double the camera price.

I vote for Practical solutions!

[Darren Kelly]

Alex,

Let's see if I can help you here.

First, I appreciate the comments from a budget minded user. I too am working on solutions that offer consumers and prosumer level users the ability to work with the HDV format. That doesn't mean the use of a capture card is out of the range.

First, I've been in the business for over 25 years (actually, I think it's about to become 30), so I know a little about this stuff. The use of a capture card provides the following:

1. 100% realtime capture at no compression. That means no loss of image quality.

2. The ability to see what I'm doing on a real HD monitor in Realtime. This means I can make an accurate judge of color and contrast. I can select graphics that look right and more. There are no software solutions that offer this today. There will be in the future as the speed of computers continue and the ability of faster drivers that are larger and more powerful. Remember it was FCP 4 and an Apple computer that first showed realtime DV editing with an RT output. That little accomplishment only happened a few short months ago, with a demo at NAB 2003 in Vegas and a product release in Mid June of this year. While it is acceptable for the consumer and prosumer to judge color and contrast, etc on the computer screen, it is not for working professionals.

3. Time is money. It's that simple a concept. What ever I can do to save some time in my world equals money. Doing software conversions are fine when the product is your own or your passion. It is not so when you are making a living with this stuff. Then , your ability to get a project in and out and off the drives is the single most important thing, allowing you to move on to that next big project, or perhaps have a weekend off, which too few small producers can do.

4. I do get paid several hundred an hour to make decisions in an edit session, or produce a corporate video, commercial etc. Part of the reason I can charge what I do is the level of professional gear I operate. My edit suite costs more in hardware and software than many peoples homes costs across North America. It certainly costs more than almost any car or truck you want to name. The reason..... it's reliable in it's results and performance, it's redundent in the key areas, it operates as fast as it can operate. Lastly, the monkey that pushes the buttons has almost 30 years of experience storytelling using these over priced toys. The latter is the most valuable, but the customer sadly thinks sometimes it's all the toys in the equipment room.


You made a quote that guys at my level don't use $3,000.00 cameras. Wrong. I have owned and still do cameras like the VX1000, PD150 and now the HD10U. I also own some more expensive camera gear too. The HD10U will allow me to sell a niche product to my customers who need it. Because I know how to best use the camera, and because I am a great story teller, I can take the image off that $3K camera and make it look much better. I can package an HD production now cheaper than I could last year at this time, when 2 days rental of an HD camera would cost the same as the HD10U.

We all want "Practical, yet quality solutions" as you put it. Steve has developed a plugin that gives you that, or you can download your choice of shareware or freeware or a combination of both. You might consider a capture card in the future, The least expensive for HD is now less than $2K, and I'll bet that a drive solution will be available in a few weeks or months that would be affordable too.

My first capture card was $18,000.00 and I bought 20 minutes of compressed stoarage(read 2 4.3gig SCSIHD) for an additional $5,000.00 Forget the computer and the single 17 inch monitor. Today, A card that does more than my first one sells for less than $1K, and a hard drive system to run uncompressed runs less than $500.00 and it holds a few hours of storage.

What I am saying is there are tools for every level and every requirement, and you should stay open to those options.

We'd all like to see it for less, but sometimes software tools are not the best option.

You wrote: Or, we can discuss inferior quality hardware solutions where capture card is half of the camera cost, and more or less mobile RAID systems cost *starts* at double the camera price" - The HD capture solutions available today are in every way but price superior to software. You're gonna have to trust me on that.

Cheers

[Martin Munthe]

Yes. And the most practical solution would be to capture from the camera via an SDI bridge to a $2K DeckLink HD card. Add a few disks and your still way under $10K for a complete HD editing system (including a HD monitor). With the FireWire solution you can't even view the footage out to a monitor as I understand it. What's the point of that? It eliminates all possibilities of showing your work to the world (if you're interrested in image quality). Or are we only going to show our work to the world in 85Hz Windows Gamma? Why even call it 720p then?

Just looking for some feedback :)