At the last Videomaker Expo East, we gathered a panel of experts from leading capture card manufactures. Here’s what transpired.
Chuck Peters
(Videomaker)
The quality of digitizing capture cards has really improved in the last few years, but there still remains a fear in the consumer world, [about] having to buy a card and install it. I d like each of you to address the subject of installation. What can be done to make it easier for us to configure our computers for nonlinear editing?
Rob Hawranko
(Pinnacle Systems)
I think we ve seen "out of box" experiences really improve. It’s easier because PCs have really grown up. I think the PC manufacturers, because they have more resources for us, have made [the installation process] somewhat easier. I think it has gotten a lot better. There really aren t the horror stories of the past that people associate with installing a card in a PC.
Jim Holland
(FAST Multimedia)
It’s really important to FAST that, whenever you purchase one of our products, you have a good experience out of the box. And, I think that a lot of people have run into problems, not only with FAST products, but with a lot of products. And the reason for that is simply you re dealing with complicated machines. And as a result of that, there are going to be, well, complications. One of the things that FAST does is we extensively update our Web site with hardware compatibility. We work very closely with any hardware manufacturer out there to make sure that we re compatible. And we encourage users to use the software that comes bundled with their hardware. That is, software that we re most closely testing to make sure that our customers have a good "out of box" experience.
We’ve certainly seen the number of complications go down significantly. And we ll continue to work to try to get them down to nothing. But again, I encourage you, visit our Web site and go with the bundled software. Those are the two sure-fire ways to make sure you ll have a good experience right from the get-go.
Ralph Messana
(NewTek)
At NewTek, we re a little bit spoiled. We actually came from the Amiga platform. That’s the only computer I know that you can just stick a card in and it works. And, there really weren t any configuration problems. If you re taking software from manufacturer "A," and a sound card from Manufacturer "B," and a capture card from manufacturer "C," and a different motherboard, there are always going to be some difficulties.
It’s getting better. Is it ever going to be perfect, where you just stick it in and it works? Maybe someday. I think the key is to try to buy from a dealer who specializes in nonlinear systems or PC Desktop systems, who knows the product and knows how to put it in. Or even buy a turnkey system. You can buy a computer that you can just plug in. They already dealt with the headaches. They got the sound card to work with the capture card, and everything’s all done for you. And to me, I d rather spend my time editing video and making money or having fun, than spending two weeks trying to get my computer up and running.
Stephen Davies
(Matrox)
The thing to remember, I think, in integrating a PC is that PCs are not made for video. PCs, at least 99.9% of them, are made for office automation. Computer manufacturers are building their machines for the largest market they can. And that’s not video. What we, as hardware manufacturers are trying to do is take our square peg and fit it in that round hole. So that means there’s always going to be some level of complexity. It goes with the territory. It’s getting better.
Microsoft is making it better. But, on Microsoft’s side, high performance operating systems, by definition, are complex. They try to get a lot of the extraneous software out of the way, so the computer can use all of its computing power for just that, computing. If you put layers of simple "guide me" steps into how to integrate a system, the computer the operating system has to make basic assumptions for you. And then it puts restrictions in the way. In video editing, we can t survive with those restrictions. We need the system to operate at its peak performance.
Again, that makes a system complex to put together. That puts a lot of pressure on us, on you or your reseller to put together the system properly. You have to know a lot about computers and follow the recommendations of the manufacturers, be it for recommended motherboards, recommended hard drives, what have you, to the letter. Or, the easiest thing, and what we try to encourage every user to do, is buy your system integrated from a reseller. You don t make money at integrating PCs. You make money at editing video. The reseller makes money building PCs. That’s what they do. That’s what they know.
Chuck Peters
(Videomaker)
Some of us are making the move to the digital formats: Mini DV and Digital8. Others have not done that yet. Some of us are halfway [there]. We re still shooting S-VHS and Hi8, as well as DV. Can you talk about products that allow you to work with both? Do I need a separate card to capture my analog video?
Ralph Messana
(NewTek)
I don t want to shoot down the "Holy Grail" of FireWire, but I think FireWire is a little bit over-hyped. I mean, it’s a great, great feature, really good if you re going from a DV deck to a DV camcorder, or from a DV deck to a DV capture card. FireWire is just a cable. FireWire is not better quality than an S cable or better quality than a component cable. The reason you get better quality going through FireWire is if you re going from a DV camera to a DV capture card, you re saving the step of uncompressing, converting to analog and recompressing. So, when you see that your FireWire footage looks better than S-VHS, it’s not that FireWire is better than S-VHS. It’s that you didn t uncompress, convert to analog and recompress. FireWire is great, it’s just not the "Holy Grail" that everyone thinks it is.
Stephen Davies
(Matrox)
I would tend to agree with Ralph. I think it was at this show last year where I had an editor [who said] "I want DV. I want DV. I have to have DV! If you don t have DV, I don t want it!" And I asked him, "Why do you want DV?" And he pointed right over to the booth beside me and he said, "Look at that camera. Look at the output of that. That’s a DV camera. It looks phenomenal!" I said, "Yeah, but what are you looking at?" [He said,] "Well, look on the monitor!" I said, "Is that a DV monitor? What you re doing is actually monitoring the output of S-VHS output. If it looks great on the monitor, it’s going to look great to the capture card, as well." 90% of the people are buying into DV. DV cameras are great. They re small. They re portable. They re relatively inexpensive. They re wonderful! But DV, as a video format, is not, like Ralph said, the "Holy Grail." There are limitations. It’s 5:1 compression by default. It’s 4:1:1. You don t have the color depth that even S-VHS decks have. So, if you try to do a color key, it’s problematic. It has a lot of advantages. As an editor, you really have to evaluate what you re doing with that video. If you re just going to cut it up, then that’s fine. If you re going to do a lot of keying with it, then it may not be what you re really looking for.
Jeff Bierly
(Digital Origin)
I happen to like DV a lot. A lot of camcorders now come with analog/digital conversion built in. All your DV decks do. Sony makes a little media converter. You can put analog into DV, if you just want to work that way, as well. We support all these different things.
Rob Hawranko
(Pinnacle Systems)
I think it’s simply a matter of qualifying your customers needs. Obviously, there are a lot of reasons to go to DV today. And there are a lot of misperceptions on what DV has to offer. It really boils down to the customer’s needs and the price points. We have two card solutions, you can do at a very inexpensive price point; one analog and one digital. Or, the DC1000 offers actually both on the same card for a very reasonable price point, as well. It just depends on what your needs are, what your expectations are. Everybody is migrating to DV. It’s the format of choice going forward, but you have to be real careful not to abandon the analog and legacy footage that’s out there. The great majority of content you have to work with today still resides in the analog world. So don t forget about the analog and just immediately flip the switch to digital, because you really have to co-exist for a while yet.
Chuck Peters
(Videomaker)
Let’s talk about render time. Just say that word and everyone kind of shivers. Rendering has always been an issue with video. There’s a myth that nonlinear is fast. You plug your footage in, you make your edits and you re done, right? No. Not until you press Render, then you go away for a few hours. Later you come back to see if your finished product rendered out how it should. If not, you make an adjustment and render again. We ve all been there. Can you speak to render time? Is there a solution that you have, or at least, something that we can go and do while we re waiting? And what about the latest industry jargon? Can you explain terms like "real time" and "faster than real time?
Stephen Davies
(Matrox)
First of all, I think you brought up a good point. Nonlinear editing is considered fast. But, for many editors out there, nonlinear editing is not even appropriate. If you shoot your material in sequence that it will be presented and all you need to do is cut, then don t buy a nonlinear editor. You re going to spend more time capturing your material than it would take you to produce your entire project in an analog domain. Having said that, Matrox has a number of "real time" nonlinear products. By real time, I mean, they can play back two streams of video at the same time. Because anytime you need to do any kind of transition, short of a cut, you need to be playing two streams of video at the same time, you have to deliver two video streams to the switcher simultaneously. We can do all of our transitions in real time, all of our dissolves in real time, 2D effects, picture in picture in real time. 2D transitions, squeezes, wipes, that type of thing in real time. We have five keyers on board that are all real time: two chromakeyers, two luminance keyers and an alpha keyer available in real time. Color correction available in real time. Color effects such as posterization and mosaic and tinting effects available in real time. The key thing with the DigiSuite family is all the effects are available all the time on all layers simultaneously in real time, with no rendering. So if you do one, you re not negating the other. You can do all the effects and stack them up and maintain real time. There are going to be times you re going to render. Nobody is going to do everything in real time. For those times, it simply boils down to computer power. The faster your computer, the faster you re going to get it done.
Rob Hawranko
(Pinnacle Systems)
Well, in the days of old, rendering used to be a really bad word. It used to be a dirty word. It’s the one problem that all of the manufacturers here have really worked hard to solve, because if you get your project laid out and it takes two evenings to render, then you save no time going to a nonlinear application. Essentially, it depends on your hardware. We have real-time capability where we funnel two streams of video and enable you to do effects between those two streams without rendering.
If you can t afford that dual-stream product today and you just want to get started, our DC30 product has a feature called miroINSTANT Video, that actually only renders changes or transitions. So it doesn t re-render things that don t need to be rendered. It’s not real time, but it’s as close to real time as you can get without a dual-stream card. The other option we have, for any product we have out there in the Pinnacle line, is a product we just announced, called Free FX; essentially, it’s a 3D effects generator, and it’s near real time. It essentially not relying on our hardware on our board as much as it relies on your processor. But you can get 3D effects in near real time, with just a software application. So again, it boils down to price point and functionality, and how much you can invest in your system to get to take the rendering thing away.
Ralph Messana
(NewTek)
As Rob said, back in the old days, it took probably a minute to render a single dissolve, and that was not very useful. The thing that always bothered me is, I can handle waiting a minute or 15 seconds to render a dissolve. It just bothered me that every time I made a change to that dissolve, I had to re-render to see how it looked. Our Video Toaster system in 1.0 provides real-time previews. As soon as I put in a dissolve or transition, I can scrub through that area and see what it looks like. Once that’s done, I can move on to the next area and do what I have to do and at the end of my project, render just the transitions. That’s in version 1.0. As I said before, our system is capable of four streams. Probably sooner than most of us think, we re going to be going to dual streams, with real time DVEs; LightWave generated DVEs that can have motion blur, depth of field, all kinds of stuff that nobody has in a DVE engine. Because we re the first product that’s using the computer to do everything. It’s the first product that’s using the CPU to do the effects, effects that you normally see being done with dedicated hardware. So even though we re going to be dual stream very shortly, our render time now is very quick, because a lot of times when you re rendering, the intense part of the rendering is uncompressing the video, moving it to RGB color space, manipulating it, and then recompressing it.
Our system’s already uncompressed. So we don t have to uncompress two frames, convert them to RGB, manipulate them, and then recompress them back to YUV. It’s all uncompressed already. And next year at NAB we ll be shipping a 16-input live switcher. As Stephen said, some projects are not right for nonlinear. If you ve got a four camera shoot of a play, do that linear. It doesn t make sense to put four hours or eight hours of footage onto your hard drive to do a play or something that was shot multi-camera.
Jim Holland
(FAST Multimedia)
Everyone has mentioned, and I think all will agree, this is much less of an issue than even a year ago. Certainly, much less of an issue than it was three years ago. Every time I get a new computer and I do a project, and I hit the Render button, I m surprised at how little time I m looking at. There’s a couple of things I recommend. One of them is use the CODEC that’s assisted by the hardware. That’s always going to help you, of course. And keep the size of your projects reasonable. As far as what FAST is doing to address render times, we re recommending you go with the best hardware you can get. It’s all about the speed of the processor.
Jeff Bierly
(Digital Origin)
Digital Origin is constantly optimizing. We re a Software CODEC, so we do everything in software. The beauty is, this time last year, CPUs were maxing out at like 350MHz, where now you re hearing 600MHz. So Moore’s law, real fun in computers. The faster you get, the faster you get to real time and beyond. Also, processor technology is like Apple calling theirs Velocity Engine; Pentium III with the MMX architecture. When you plug into those types of vector rendering assistance types, you re going to see things accelerate quite a lot. The nice thing about software is that as soon as you put it in a faster computer, it runs faster. What a concept!