HVX200B? When?

It would, I'm just saying that it won't happen because the DVCPRO-HD compression isn't scalable. With AVC they could do it...
 
Barry, the thing is, though that SI-2k will be a single sensor 2/3" chip.

Versus HVX

three 1/3" sensors shifted with hydra to behave single sensor-ish. Also, with hydra, the ADAC is 14-bit not 12. Also because SI-2k records to cineform, recording is limited to 10bit for now. Not bad at all, but for HVX owners, going on specs, I think hydra would win, we just need to to see pics. Si- LOOKS fantastic.

A side by side comparison would need to be in order.
 
SI-2K will easily out perform a hydra modified HVX. The single sensor doesn't have overlaped pixels so it can resolve a higher resolution image, whereas pixel shift can't go as far. 3 x 1k x 0.5k pixels is less than one 2k x 1k sensor, so even if pixel-shift was equivalent, it couldn't match the SI-2K. Yes the compression is 10-bit, but is 10-bit LOG from the very clean 12-bit linear source, more bits will not help if they only contains noise. Compressing images from a Arri D20 or Red One, also only has 12-bit linear best compressed to 10-bit log. Viper Filmstream is 10-bit log. Film scans 10-bit log. You get the idea ;)
 
And budgets that big filmmakers are used to now will NEVER happen again. They really won't. The NY times did a huge article about how studios are aquiring financing with other studios, but are increasinly turning to movie-investment funds (like mutual funds for hollywood), the fund managers are pushing for lower budgets. That is #1 on their agenda.
Click to expand...

I don't think big budgets will end. They may lower overall...but there will always be people daring enough to spend more than has ever been spent before to try and make something new and exciting.
 
David Newman said:
SI-2K will easily out perform a hydra modified HVX.
Click to expand...
I would expect it to as well, but let's wait and see before making blanket proclamations like that, because you may end up being surprised.

The single sensor doesn't have overlaped pixels so it can resolve a higher resolution image
Click to expand...
Huh? Only in black and white! Yes it has more pixels, but they need to be overlapped or else you're only looking at a monochrome image. If you stripped out the bayer filter then yes you'd get a 1920x1080 image at full resolution, but because there's a bayer filter in place you have to combine pixels (in a process that's remarkably similar to pixel shift) in order to get color from the sensor.
whereas pixel shift can't go as far. 3 x 1k x 0.5k pixels is less than one 2k x 1k sensor, so even if pixel-shift was equivalent, it couldn't match the SI-2K.
Click to expand...
Again, that remains to be seen. If I'm reading the spec sheet properly, the SI-2K has a bayer'ed 1920x1080 single sensor, which means its actual pixel count is 960x540 blue and 960x540 red, and 960x1080 green, with 5-micron pixels. The HVX has 960x540 blue and 960x540 red and 960x540 green, with 5-micron pixels. So yes the green is higher, but the HVX also uses vertical spatial offset (something that AFAIK has never been done before) so ... it's too early to tell, but I wouldn't be surprised if the Hydra results are *extremely* competitive with the SI-2K. Soon enough the results will speak for themselves.

Considering the amazing results Juan got from a standard-def DVX, I think it would be unwise and premature to dismiss what he'll be able to pull from an HVX; we should let the results speak for themselves.
 
Unless there is a new HVX200 model that provides a perfect script and actors without egos, I am perfectly happy NOW ! Sometimes I think some people on this forum spend too much time fantasising and not enough time shooting.
Carpe Diem.
 
arrestthisman said:
Yes aculag, that's basically what I said. I was speaking about the individual pixels NOT the chip size or total resolution.
Click to expand...
Yeah, I just misunderstood. I'm not used to hearing people talk about pixel size instead of chip size or resolution round these parts. Maybe I just don't pay enough attention. Hehe.
 
ksteiger said:
Unless there is a new HVX200 model that provides a perfect script and actors without egos, I am perfectly happy NOW ! Sometimes I think some people on this forum spend too much time fantasising and not enough time shooting.
Carpe Diem.
Click to expand...
Then what are you doing here? :)

Just kidding.

Acutally I've been bed ridden with an infection so I've been hanging out here for a while, but I get back to work on Monday and then no one will have to deal with me here :)

Back to the SI2k debate: Anyone who has seen andromeda samples knows that Juan pulled such a clean image, it's almost hard to believe they came off a DVX. Given that the HVX has so much more to work with, I don't think it's a pipe dream to expect amazing results.

The SI-2k can produce some beautiful images, and now that Barry has identified it has having 5micron pixels I'm not as surprised at the results. But on look alone I think andro could go head to head with SI-2k any day (if si-2k was cropped for the same resolution), therefore I don't think it's impossible for the Hydra to come out on top visually speaking.

I think they are both amazing solutions. For me though, now knowing that hydra pulls sync sound, and being under the impression that SI-2k does not, I doubt the edge of SI-2ks picture, IF there turns out to be one, would sway me.

What it all comes down to is that IMO, I think they'll be in the same league quality-wise. I think they are both excellent options for people who aren't ready or don't need 4k.

And not that I'm a RED skeptic, because I'm not, but I just don't think they'll be shipping big numbers for a few years. I think they'll deliver everything promised, and THEN some, but it'll be wait after wait after wait. So unfortunately, to me it's not a viable solution to use unless I have it in my hands. But I'm glad it's there.

The more I spend looking at all of the digital cinema choices available, the more it makes me want start writing, and working again... because I think we've hit the time where digital filmmakers no longer have to justify themselves against film. The quality is now there. It's cheap, (compared to film), and it's a workflow my generation is far more comfortable with.

This is a great year for everyone because we can work, and know that we'll be able to achieve the results we want in the digital ream without worry or sacrifice. We're getting closer to the dream of creativity being the only barrier for film. Like writing. Pen and paper were not always so cheap; now with word processing, it's all in the imagination. We're getting close for film. 10 years and we might even get there!

Good times!
 
One area where the SI-2K will have an undeniable lead is in getting the shallow-DOF look; when using 2/3" lenses you'll get Super16-class DOF, whereas the HVX is going to deliver Super8-class DOF.
 
I’m pretty sure the SI2K will outperform a hydra HVX200. I know we need to wait and see but everything points to that direction. First the SI can use the best glass available while the HVX200 will always be limited to it’s fixed lens. I don’t think a $4,000 camera comes with a lens good enough to resolve 2K. That’s why it’s hard to believe you will get real 2K with hydra, even if it does wonders with the pixel shift.
Then the bigger 2/3” CCDs will surely have some advantages over the tiny 1/3” sensor in the HVX200.
Quite frankly I’m not even sure you would get much better results with hydra as opposed to just capturing uncompressed 4:2:2 DVCPROHD via the component out or uncompressed from the HD SDI out of the Canon XL-H1. Since with hydra you are linked to a computer no matte what, just get a Decklink and capture uncompressed. The decklink is surely much cheaper than hydra will probably be.
 
The real advantage of the Hydra, SI2K & Red is the RAW workflow. It really is amazing how much is lost in the DSP.
 
Exactly, and that's what's being missed here. Put an XLH1 through HD-SDI to a Decklink capturing through DVCPRO-HD and you're still getting 8-bit data. The Hydra is 14 bits.

I would expect that the SI should outperform a hydra'd HVX, but put another way: it had better, or else it'll end up looking pretty silly. But even so, I wouldn't bet against Juan and the Hydra'd HVX, at least until it's had a chance to prove itself. We haven't seen a single frame from the Hydra yet.
 
well based on how great andromeda looked, i imagine hydra to look quite remarkable.

watch that foggy beach clip on the website.... YOWSA
 
Yes, I know RAW is a big improvement, but that is basically a color improvement. I was talking resolution. Do you guys really think the HVX200 lens is good enough for 2K? I can't see that when true real HD lenses cost 4 times more than the HVX itself. If the lens can't record it it doesn't matter what the hydra pixel shift does.
I'm not sure the bit increase alone is worth the mod unless for effects. That would be most useful for effects work. To be an all around worthwhile mod it would have to deliver the promised 2k resolution. But I don't see how if it has to go through the fixed HVX lens. The SI2K can use Digiprimes if you want. That should beat the HVX lens by light years. No arguing in my opinion. No way hydra will come close to the SI2K IMO. Not ditching Juan's efforts. It's just logic.
 
Mediacre said:
Yes, I know RAW is a big improvement, but that is basically a color improvement. I was talking resolution. Do you guys really think the HVX200 lens is good enough for 2K? I can't see that when true real HD lenses cost 4 times more than the HVX itself. If the lens can't record it it doesn't matter what the hydra pixel shift does.
I'm not sure the bit increase alone is worth the mod unless for effects. That would be most useful for effects work. To be an all around worthwhile mod it would have to deliver the promised 2k resolution. But I don't see how if it has to go through the fixed HVX lens. The SI2K can use Digiprimes if you want. That should beat the HVX lens by light years. No arguing in my opinion. No way hydra will come close to the SI2K IMO. Not ditching Juan's efforts. It's just logic.
Click to expand...

That's not really logic, more like opinion mixed with information that is generally regarded as true.

Also, RAW is far more than just a color improvement, It enables greater dynamic range, and non-destructive color correction in post, depending on one's workflow.

We've already seen the results from andromeda, and like I said, put it side by side with SI-2k footage and it looks very similar, qualitatively speaking. Since we can only estimate that hydra will look better than andromeda, and at a higher resolution... well, just use your imagination.

Also, about the lens. The HVX is a 1080p camera, we can argue about the ins and outs of how it achieves the resolution, but the native res, coming off of the array is 1440 x 810, it's then expanded to 1920 x 1080, and resterwised to 1280 x 1080p for DVCPRO HD. The idea is that this camera was engineered to offer 1080p.

There is a bigger difference between 720p and 1080p than I was led to believe when I bought the camera. People were quoting a 10% improvment or so, and I think it looks about 20% at the very least.

In short, I don't believe the bottleneck is the lens. Especially since fixed lenses can get away with a bit more murder being that they are engineered specifically for their CCD arrays. Interchangable lenses have their own issues to overcome. Remember that the HPX500 has a special DSP system in place to deal with the abberation problems inherent to interchange lens systems.

Yes digiprime lenses are better, but they are not made to fit like a glove to any given camera.

Also, I had a good chat with a guy who does 35mm lens adapters for the HVX, and his theory, which made sense to me, was that since the fixed lens is zoomed past the center and the focus is on the ground glass, the interference of the fixed lens is almost nil. (The theory came after the results. Anybody who's seen footage of a properly mounted 35mm adapter on an HVX is stunned by the results and aside from the light loss is damned to find any degradation in the image.)

Back to Brass Tacks, at has to do with the final image. Period. Specs aside, it's all about what you see with a trained eye. The lens on the DVX is an SD-fixed lens. But andromeda certianly resolves at least 720p (and more if you look at the higher quality samples).

By YOUR logic that would be impossible because it's an SD lens.

My final point is simple that Juan has a track record, he said he'd do what most people thought impossible with andromeda, and delivered. He's saying he's gotten the same principal off the HVX, and I don't see any reason to doubt.

The only reason I compare hydra to the SI-2k, specifically the mini is because they are in similar price ranges, and are both RAW 2k cameras.

6K + 3.5K(mod) + 2.5K(laptop and storage) in $ = 12k for the hydra

SI-2k is 12 - 14 k for the mini + 2.2k for a digiprime + 2k for laptop

So they are both in the 12 to 18k price range to get up and running. Ironically, not THAT far off from the RED. price order = hydra/SI/RED. In like 6 grand increments. 12/18/24 to get up and running for 2k RAW cinema.

Then there's the other advantage with the HVX that you have the possibility of 2k cinema when tethered but also a compressed HD and SD camera for run and gun, unthered. Lot's of options in one little package. AND it includes the deck! It's a great value.

Again, I think the SI-2k is awesome. And it may pull a better image, and will certianly have DOF advantage (unless you put an M2 on the HVX, then the HVX has the advantage), but I could forsee a future where you have films using both cameras. They could easily match up because of the easy LUTs on their respective systems, and they are both RAW.

I think they will be very competitve, and SI-2k will have an edge for people ONLY involve with cinema, and Hydra will make more sense for people who need more options with a smaller crew, and also if you're like me, a deep DOF is a plus. f64 rules :)
 
But I don't see how if it has to go through the fixed HVX lens.
Click to expand...
Of course a DigiPrime will beat the HVX lens (or, at least, it certainly should).

But the biggest problem I have with your argument is that it's reliving the arguments of the past, where people made sweeping proclamations based on some specs, while disregarding the actual observable evidence we have in front of us. How can you say the HVX lens won't hold up? We have evidence that the DVX lens (the "standard-def" lens on the standard-def DVX) does a fine job of resolving a high-def image, delivering higher resolution images than the HD lenses on the HD100 and Z1, and by a decent margin. That's already been demonstrated by Andromeda, the footage already exists. So to assume that the HVX's even-better Leica Dicomar won't be up to the task seems presumptuous at best. Let's look at footage before you say the footage won't hold up!

I mean, I was one of the biggest arguers about lenses being the limiting factor for small HD cameras; it simply did not mathematically add up that we'd be getting decent-performing lenses on these small cameras. But you know what? The lens on a $925 Canon HV20 delivers a high-def image. There's no way it should, but it does. It doesn't matter that we don't understand how, all that really matters is that it does it. I have no doubts that the HVX lens will be PLENTY adequate to the task. And sooner or later we'll have actual images and photographic evidence that will answer the question definitively, so why stress about it? It either will or it won't, but if the last couple of years have taught us anything it's that "theory" gets its *** kicked by "reality" every single day.
 
HERE HERE Barry! That was exactly my point, of course, as always your explaination sums it up more quickly and pithy, but yeah.

It all comes down to the image, period. You can argue theory till you're blue in the face, but if you have an image, especially a charted image, there's no more talking, it's either there or not there.
 
You must log in or register to reply here.
Back
Top