1/2" versus 1/3"

A

arrestthisman

Well-known member
I did my searching, so please no scolding.

Can anyone shed some light on the qualitative difference between 1/2" and 1/3".

I know there is so much more to imaging than just the size, but if the HVX200, say, had 1/2" sensors, what would be different about the camera. Higher rez with the same 5 micron pixels?

I can't imagine the DOF would be THAT much different. Thanks for any info.
 
1/2 vs 1/3

1/2 vs 1/3

I hate to say this.
I am under the impression that 1/2 is alote better than the 1/3.

The same as a point and shoot digital camera vs a digital SLR. My Fuji S2pro digital SLR is 100 times better than my point and shoot.
The point and shoot has 9 mega pixels and the digital SLR is 6 mega pixels.

Bigger is better!
 
better yes, all other things being equal

but not $100 point and shoot versus nice SLR better!!
 
could be up to a stop faster -- and dynamic range.
Click to expand...
But only if they kept the pixel size the same. If they crammed more pixels onto it, then you'd get probably the same sensitivity and dynamic range, but with more pixels. They can use the bigger size to make bigger pixels, or they can use it to fit more pixels into the same area, but they'd have to choose which one to go with; they wouldn't likely have both.

A 1/2" sensor is going to be about 33% wider and 33% taller than a 1/3" sensor. That makes it about 1.78x larger in surface area. So they could go to 1280x720 and have the same pixel size as an HVX, or 1440x1080 with pixels a lot smaller than the HVX's. Even so, you can't just compare pixel size to pixel size because it matters as to whether the pixels are sourced from a progressive-scan CCD, from an interlaced CCD, or from a CMOS chip; each will affect sensitivity and noise performance.

There will be a difference in DOF, 1/2" will deliver somewhat shallower DOF. Not night-and-day different, not 35mm-style, but shallower. For example, at full telephoto on an HVX, 55mm, on a 1/2" cam you'd get the same angle of view from about 75mm.

Assuming a subject 20 feet away, the depth of field on the 1/3" camera (at f/2.8) would be about 30 inches, and on the 1/2" camera it'd be about 24 inches, so maybe 20% shallower at equivalent field of view.
 
lycrarob said:
My Fuji S2pro digital SLR is 100 times better than my point and shoot.
The point and shoot has 9 mega pixels and the digital SLR is 6 mega pixels.

Bigger is better!
Click to expand...
All other things being equal, bigger is better, but more pixels is rarely a fix-all, as you demonstrate. You're saying your 6-megapixel SLR is 100 times better than your 9-megapixel camera. And I'm sure you're right; but the point is, you can't just go counting pixels and declare that "more = better". Very few of us would be qualified to judge which would make a better picture than the other, just from reading specs. We should instead stick to looking at footage! :thumbsup:
 
Barry, do you just have all this info in your head? :) I love reading your perspective.
I must have 2-3 opinions dropped on me daily about size, density, sensitivity, sensor type etc. So many people just say because. It's nice to read your facts. That and "how does it look?". :)
Thanks!
 
It's not all in my head, but it comes from calculators like the Panavision DOF calculator. Tracking down the optics info is easy, that doesn't change regardless of what type of sensor you're using, so DOF and such is an easy one to get a grip on and to answer.

It's comparing the different types of chips that starts to get silly. Seriously, there's no way to compare differing technologies on paper; it's a mental exercise that won't result in any useful results. People can clamor that this chip has more pixels than that one so it must be sharper, but you know what? When we look at the footage, it's all basically the same. All these companies are employing brilliant engineers, and they're all delivering comparable performance within the particular engineering compromises that they each make. They all employ different means to get there, and that has serious repercussions on the features each camera can have, but the core imaging is going to be in the ballpark on each of 'em, and there are going to be compromises with each of 'em.

And then, even when we think we understand how something's going to perform, we can say that "the bigger pixels of an HVX will mean much better low light performance" -- but then an XHA1 comes out with 3x as many pixels crammed into the same surface area, and it delivers comparable sensitivity! Who saw that one coming? Certainly not the numbers-obsessed. You have to factor in that the A1 is using interlaced chips, and that doubles the effective sensitivity. So they get more pixels, and comparable sensitivity, so isn't that "better"? Well, you can't do variable frame rates on an interlaced chip (unless you're willing to throw away half your vertical res, like the XDCAM HD system does for frame rates from 31 to 60!) So because of the need to do variable frame rates, Panasonic chose progressive chips, which are good for some things and not as good at others (such as the few benefits interlace actually offers, which include lower noise with increased sensitivity because of the interlaced-scanning process). You end up with comparable sensitivity, comparable sharpness, comparable dynamic range; one has the bragging rights of having 3x as many pixels and a perceivable 5% increase in sharpness (when in 24F mode); the other does variable frame rates but gets picked on for "not having as many pixels" -- even though the footage is comparably sharp.

So what ends up mattering? Bragging rights? Internet fights over subjects that we non-camera engineers can't possibly be qualified to understand? Or just putting the footage to the test and seeing results that we *are* qualified to understand and judge! :)

The numbers are important, yes. But only to those who can properly interpret them! And very few of us short of Adam Wilt or Juan Pertierra would have the experience and engineering background to understand how a bayer-pattern 1920x1080 single-CMOS 1/2" chip will compare to a 960x540 spatial-offset three-CCD 1/3" chipset, and be able to confidently predict the outcome.

All the blather and clamor over these specs that we just don't understand -- it amounts to as much useful information as two fleas arguing over which one owns the dog. Look at the features, then look at the footage, and look at the price tag, and then pick the one that does the job you need done. There is no one "best" camera system, but there probably is one that's best for what you intend to use it for.
 
Last edited:
Barry, as always, your wisdom is plentiful and valuable!

Obviously it's the new sony 1/2" cams that had me wondering, but since sony isn't the only one with 1/2" sensors, I didn't want to turn into an XDCAM versus Panny topic.

My background is computers for 21 years, and videography for only 7. Even after this long I'm still engrained in the Moore's law mindset. In comps, newer really is better, there is no saving grace for old equipment (unless it has some very special niche). Also, in comps the specs almost always follows performance. It takes a lot to remind me that cameras, even in the digital, are quite different. I may want to do a movie on a comsumer camera because the look is "right" for that project.

I think I need to staple a board to my head that says, "there is no better, and there is no worse."

I'm hoping I'm not the only one who thinks this way, otherwise this would be a very self serving thread :)

Thanks all!
 
Well, there is better and there is worse, but it (as always) depends on your particular application. I think an HVX is much better for a filmmaker than, say, a DVC30 would be. And the DVC30 is probably a much better wedding cam than an HVX200.

So there's definitely better and worse, it all depends on the target use.

I think Sony is the only manufacturer currently offering 1/2" chips, aren't they?

If all other things were equal, 1/2" would definitely be preferable to 1/3", no argument at all. Problem is, things are rarely equal, so we have to see what the actual comparison will be. We still don't know if the new Sony will be CMOS or CCD, for example. Sony gets a great image out of their 1/4" CMOS chips in the V1U (provided you avoid the oil paint effect) so they could maybe do something really good with 1/2" CMOS. It all depends on price, features, and performance. And we can't really predict those things until we see the results.

I mean, can one really say with a straight face: 1/2" chips are better than 1/3" chips? Because, if so, then Sony is indicting their entire camera lineup. They would be admitting that other competitors' 1/3" products are naturally "better" than their own 1/4" V1U. And they'd be admitting that their competitors' 2/3" products are naturally "better" than their own 1/2" XDCAM products. So if they try to portray that their 1/2" is "better" than the competitors' 1/3", they had better be prepared to realize that that particular knife cuts both ways. Competitors have larger chips at much lower cost than Sony (the HPX500 is 2/3" and costs about half as much as the 1/2" F350; the XHA1 has 1/3" chips and costs about 25% less than Sony's 1/4" V1U).

Chip size is one factor to consider, but it's only one factor among many. Pixel count, chip type (CMOS or CCD or Foveon etc), # of chips, pixel size, interlaced or progressive scanning, rolling shutter or global shutter, FT or FIT, and the age of the technology all play factors in how well the equipment performs. (I mean, is anyone going to try to say that a 1/2" Betacam UVW100 outperforms a 1/3" XHA1? I'd laugh 'em out of the room if they tried to claim that!)

I say bring out the product and let's compare features to features, dollars to dollars, and performance to performance.
 
There are also other considerations, If I remember correctly, Sony 1/2" chips are not 4:2:2 and the compression is 50ms. So what's better, 1/3 4:2:2 AT 100MS or 1/2 4:1:0 at 50ms?... I think too much time is spent looking at numbers rather that looking at the final output or looking at how the chip algorithm handle contrast, sharpness, saturation etc...
 
That's even moving outside of the realm of the chips though. That's getting into recording formats, and then there's the bit depth of the DSP, and all sorts of stuff.

Which does bring home SomewhereinLA's point -- what good is having the "bigger" specifications, if the end footage doesn't deliver the better result? We aren't buying chips here, we're buying a finished production system. It seems to make much more sense to judge the systems based on the final output, than it does to base a decision on a bunch of numbers that frankly, we aren't qualified to understand or assess properly anyway! :)
 
It's not the chips that do the compression or set the color space. Chips are analog, monochrome devices. The compression and processing are done after the info is read from the chips.

Now, chips can be interlace or progressive, but that's independent from compression and color space.
 
I remember reading a very long article about still cameras and the "megapixel" factor. The article was very technical at time, but the whole point was that more pixels doesn't mean better picture, and that the megapixel factor became in fact a marketing gimmick. We do live in a society where more is too often perceived as better... I also think that too often we count on technology to deliver great things for us... Shooting in 4k won't do anything for you if you can't light a scene properly. I have seen stuff shot with a DVX that literally stunned me, like wowwwww... And I still use my one megapixel camera with great results....
 
I liken the "megapixel" race to the "digital zoom" race. Is having a 70x digital zoom better than having 30x? Er... sure, yeah, that's the ticket...
 
Barry_Green said:
I mean, can one really say with a straight face: 1/2" chips are better than 1/3" chips? Because, if so, then Sony is indicting their entire camera lineup. They would be admitting that other competitors' 1/3" products are naturally "better" than their own 1/4" V1U.
Click to expand...
If the case solely revolved around image quality, and with the caveat of "all else equal" then I certainly could say that with a straight face, and elsewhere in that post you yourself say "If all other things were equal, 1/2" would definitely be preferable to 1/3", no argument at all." It's simple physics, the laws of which can't be changed, and (all else equal!) factors such as diffraction due to the wavelength of light are as relevant as larger surface area (so higher sensitivity and/or higher resolution), to say nothing of more controllable depth of field.

So is it a one way street? Is Sony then indicting their own lineup? No, because that argument is solely based around image quality, and the ADVANTAGE of smaller chips has always been smaller and cheaper cameras can be built around them, with all the peripheral advantages that brings of lower power consumption etc. It's horses for courses - quality or convienience, but not both together.

What I personally find so relevant about the EX announcement is that whereas previously to get a 1/2" chip camera, it had to be of the size/weight/price of the 330 or 350, the EX seems to offer that chip size in a package comparable to that of the HVX200 or the Z1. That is the real significance of the announcement.
I say bring out the product and let's compare features to features, dollars to dollars, and performance to performance.
Click to expand...
Absolutely. All this is all speculation at this point, but..... is that not what forums like this are for? But 1/2" chips, true manual lens, much cheaper solid state recording format (for starters) - it's easy to see why it has caused a stir.
 
and the ADVANTAGE of smaller chips has always been smaller and cheaper cameras can be built around them,
Click to expand...
Yes, but that's not what they're doing. They're building more-expensive cameras using smaller chips. The 1/4" V1U is substantially more expensive than the 1/3" XHA1, and the 1/2" F350 is twice the cost of the 2/3" HPX500.

All I'm saying is, they can't have it both ways. Either bigger is better, or it's irrelevant. I say, with all other things being equal, bigger is better. But it's a moot point because it's a very rare case that all other things are equal!

All this is all speculation at this point, but..... is that not what forums like this are for?
Click to expand...
Not so much, actually. We went through a speculation phase, but now we're more interested in finding out what people are actually DOING with their gear, and helping them to do more and accomplish more. There was a time for speculation and "mine is better than yours" and whatnot, but now that each manufacturer is offering excellent tools, we'd rather focus on the business of accomplishing something useful rather than just endless tech blather that very few of us are actually qualified to even understand.
 
You must log in or register to reply here.
Back
Top