XL1s direct to SCSI

[Sam Brandt]

I'm just moving from 13 years of print and publishing to adding video to my studio. I'm new at this and I'm sure most of you know this by rote, so I'm asking everybody and anybody.

1. Is it posssible to capture direct to disk without recording to the DV tape in the XL1s?
2. What port and pinouts would I use?
3. Are there any flaming hoops I'd have to jump through to make it work?
4. Are there any neat tricks to make this work smoothly?
5 If not, are there any workarounds?

Thank You !!
Sam





P.S. I've had my XL1s for only about 2 months now, and I'm thrilled at the output. I know, it's not fully professional level, but it's so much better than the usual fare that's available, it's astounding.
I also tried out the Panasonic DVX100 when I bought this. The deciding factor for me, was the Panasonic is all one piece like any cheap consumer camcorder, and I like the color space of the XL1s much better.

[Boyd Ostroff]

I think you mean firewire (ieee 1394) not SCSI?... Are you using a Mac? If so there's a nice little shareware program called BTV Pro than can do this. If on the PC then it's my understanding that Premier can also capture direct to disk. For that matter, I'd think almost any NLE software on either computer could capture directly from the camera without rolling tape. All you need is a standard firewire cable to do this.

Different camera handle color in different ways, but I think the term "color space" actually refers to the color sampling/data compression that is common to all DV cameras and is known as "4:1:1".

[Sam Brandt]

Thanks but no, I mean directly from the camera that's taking the live shot, TO the storage device (SCSI, SATA, ATA133 or 150, or whatever) bypassing the need to go to the DV tape.
It could be BY WAY OF Firewire or S connector or SDI... that's what I don't know. I need more exact information and not that IEEE 1394 might work or there's some software somewhere that could do the job.

And, by color space... I mean ''color space''... the range, depth and saturation of the colors that each camera ''sees'', like the difference between an RGB and a CMYK gamut. The Panasonic looks way cool compared to the XL1s. The Canon has cleaner reds and a warmer more fleshy ''real'' look. I don't know how each camera samples and compresses color but the final output of the Canon's range of color is much less muddy and is a better ''palette''.

I'm just real sorry I editorialized about the camera comparison. I really need to focus on and know about the direct to disk capture possibilities for the XL1s.

I'll take a look at BTV Pro. Thanks again.

[Sam Brandt]

By way of information:
The capture system, as it currently stands, is a SuperMicro X5DA8 with an e7505 chipset and three SCSIs in RAID 0 with an SATA as the system and audio drive.



Boyd,
Those stills from Madama Butterfly are REALLY stunning. They're GORGEOUS !!! (...and, like ''awesome'', I don't use that word very often!)

[Dean Sensui]

Sam...

I've been in the print business for 23 years now (newspapers) and have been in the video business for about half of that. I'm really enjoying doing ever more-demanding jobs in video.

Is there a compelling reason to avoid tape? If you can run both that would be a better solution in terms of reliability. Hard drives can go bad at the worst moments unless you run some sort of backup. RAID level 0 (striped pair) is fast but can suffer from potentially catastrophic problems. If one drive goes slightly bad you lose all the data. RAID level 1 (mirrored pair) is much more robust but not as fast, although for DV formats might well be fast enough.

As for color, I think you mean color fidelity and balance. Color space refers to the range of colors available for reproducing a particular scene. With NTSC it's pretty limited compared to Adobe RGB or even ColorMatch RGB, tho not nearly as restricted as generic CMYK (which is purely a print-based reflective color space). I can go on and on about geeky prepress stuff but won't right now...

Here's a nice portable hard drive solution that you might find useful: http://www.focusinfo.com/products/firestore/firestore.htm. It'll allow you to record directly to the hard drive and, if you want, record directly to tape as well. Firewire, by the way, is your best pathway. It's built-in to the XL1s and is essentially lossless in that it's a direct data transfer with no conversions. People might argue this point, but that's how DV cameras work. In fact, there's actually software out there that will turn a digital video camera into a data storage device!

Good luck with the venture into this medium. It's not better or worse than print. Just different. Challenging, too.

Dean Sensui
Base Two Productions

[Don Palomaki]

To sort of answer the basic question, yes it is possible to capture directly from at least the XL1 (and thus, most likely the XL1s) from either the IEEE 1394 (aka iLink or fiewire) port or the analog output provide your computer has the require input capability - either a IEEE1394 port or analog capture card and appropriate software drivers for the capture devices.

[Sam Brandt]

I found it... I stayed up a good deal of the night searching the net to find the site for the card that originally sparked my interest in this matter. I'd had it at one time and then couldn't find it again.
This is the technique it uses.

As you see it is a capture/controller that goes DIRECT and UNCOMPRESSED to storage.
http://www.ioindustries.com/dvrexpress/record.html

...and this is the PCI card: http://www.ioindustries.com/dvrexpress/overview.html

This would be the end of my search, but this card will not work with the XL1s. http://www.ioindustries.com/dvrexpress/cameras.asp

I've also seen on the web at some point, but can't find it now, a whitepaper written by a professional Hollywood cinematographer on a homebuilt Direct-to-Disk system specific for the XL1s.

As I recall bypassing the tape improved the image quality of the XL1s to near that of studio production cameras with MUCH larger RGB chips. (That's the reason I'm interested)

I'm sure it was built with off the shelf components, and used 160 or 320 SCSIs for storage.

Even if I find the whitepaper, it still won't give me the details I need to duplicate his unit. (Although I would have a chance to track down an eMail address at that point)

The only Pro Shooter I've found so far that has lots of great information is Ron Dexter at: http://www.rondexter.com/ . This guy is a treasure chest full of goodies, but I can't find anything regarding the direct to disk storage issue. (What I might do is eMail him and see if he knows anything specifically about the topic... although I'd hate to start another round of confusion and speculation.)

Thanks Dean, I guess you're right, what I meant was "color fidelity and balance", but graphics terms are what I'm used to.

And thanks for everyone's interest and responses... kind' makes a feller' feel (sniff) wanted... :)

[Sam Brandt]

This monster is designed just for direct to disk.
It's only about $100,000.00.

http://www.thomsongrassvalley.com/do...r/viper_br.pdf

[Jeff Donald]

Sam, what platform are you on? Most of the uncompressed capture cards are platform specific. Also what NLE software are you using.

The "White Paper" you're referring to may be something on Scott Billups site. His site is here. A couple of years ago he wrote about trying to get RGB directly off the chips before being converted to DV. I don't think his experiments met with the success he hoped for.

What type of work are you doing? Most people find the limitations of directly connecting their camera to their computer a bit limiting. Long cable lengths and the need for AC power preclude all but some studio work.

[Sam Brandt]

As I said before in my posts it's a SuperMicro X5DA8 w/e7505 chipset, 3 SCSIs in RAID 0 w/SATA as OS and audio drive.

I'm using an XL1s, VideoToaster[2] TEd, Lightwave(a little... it's new to me and complicated) and VegasVideo.

And yeah, yeah, yeah, I know, the VT2 will capture real time, uncompressed, but I'm looking for another MAYBE better way to do it. (anything for an edge)

And the whitepaper I'm referring to WAS a completed project that was used in production shoots and worked. Maybe I'll run into it again on the web someday

And, yes of course, I know about the limitations of packing a computer around with cable and needing power and all the obvious things, but (as a fer' instance) setting the system up for the stationary master shot in say, a wedding would be doable, and the potential increas in quality would be worth it just to offer to the client a better product that the competition down the street.

[Dean Sensui]

Sam....

Using a complex system like this for a wedding is a bit of overkill for what will be a marginal improvement. Not to mention the reliability problems that might crop up, and the loss of mobility.

If the cost goes too high, it might be worthwhile considering:

-- setting up ideal lighting (huge difference).
-- better glass. Get the 16x manual lens.
-- larger CCD (different camera).
-- larger CCD and better glass.
-- better format (DigiBeta).

Canon just shot a commercial using the XL1s and the manual lens. Their results were excellent. Again, part of it is how the scene is lighted. And using the manual lens helps, too. The manual lens is slightly sharper. But the difference between the manual and automatic IS II lens is the ability to maintain focus throughout the zoom range, maintain the same f stop throughout the zoom range, and being able to focus while zooming. The last capability is often needed in event videography.

The manual lens is definitely on my wish list, as soon as I get the Betacam deck paid off!

Dean Sensui
Base Two Productions

[Sam Brandt]

Good suggestions, thanks!

[Virginia Benedict]

Can anyone explain the differences between Firewire and Flexible Analogue Video I/O: Composit, component, Y/C?

[Juan P. Pertierra]

I've been wanting to ask this question for a while...seems this is the thread to do so:

i saw a couple of comparison screen grabs done by Scott Billups on his www.pixelmonger.com site that shows the quality of the XL1s being a little better when recorded to disk than when taken off tape. Has anyone experienced this with the camera? or is it just a difference in the way he encoded the screen grabs, etc. At the very least, the brightness looks different.

Of course, usually it doesn't make a difference because it is the same DV stream :) Just curious...

Juan

[Jeff Donald]

Can anyone explain the differences between FireWire and Flexible Analogue Video I/O: Composite, component, Y/C?

Composite video is a mixed video signal in which the luminance (black and white information), chrominance (color information), blanking, sync and color burst are carried on a single cable. The signal is coded using NTSC, PAL, SECAM, etc.

Y/C a set of video signals that contain a separate Y (luminance) and C (chrominance).

Component video is a set of video signals in which Y and C are sent as separate components. Component video signals maintain full bandwidth, unlike composite signals. Component can be both analog or digital. Analog component is used in M II and Betacam SP, digital components YCRCB in CCIR 601.

FireWire (IEEE 1394, I-Link) is a high speed serial data bus. FireWire offers high-speed communications and isochronous real-time data services.

FireWire can transfer data between a computer and its peripherals at 100, 200, 400 or 800 Mbps, with a planned increase to 2 Gbps. Cable length is limited to 4.5 m but up to 16 cables can be daisy-chained yielding a total length of 72 m.