Shooting progressive with HC1?

[Douglas Spotted Eagle]

Thomas,
I didn't say the images were "perfect 24p." I said they're not jerky. The cadence doesn't "feel" like the perfect 24p, but it is not "jerky."
In real-world use of the cam with CineForm vs measurebating by reading text on a web page, I'd submit I'm at least marginally qualified to hold the opinion that I hold. Additionally, CF24 samples both with and without CineForm processing have been available and comparable on my website for nearly 2 years now. In other words, I've viewed them side by side on many an occasion, and am quite familiar with the differences.

[Thomas Smet]

I also happen to own a HC1 and to me it looks jerky. A non perfect 24p does mean it is slightly jerky. If it was that good then why do more people not use it on a mainstream level. I love my HC1 to death and think for the money it does a great job but I can tell you it stobes a little bit more than true 24p footage does. Maybe that doesn't fit your definition of jerky but to me it does. I recently used my HC1 to work on a documentary in South Africa with my wife that is from there. We spent 5 weeks in many different areas and I shot well over 8 hours of HDV with my HC1 that I am very happy with but I'm not sure if I would use the CF mode. I have done some visual effects tests and it looked nice but it did strobe more so then other 24p footage I have worked with.

I am not measurbating here but going off my own personal experience as a visual effects artist using the camera. Besides in the case of measurbating, 95% of Dvinfo users do it and the other 5% lie about it.

I'm not sure why you have to get so negative with me. I'm also perfectly qualified to know what the camera does and I do not appreciate you indicating that I only know things from reading text on a website.

[Douglas Spotted Eagle]

Unless something changed in later HC1's, the Cinema Effect mode isn't the same as the CineFrame 24 mode of the Z1/A1.
"Strobes a little bit more" doesn't equate to jerky for me. Personally, I'm very impressed as to how CineForm removes the "pulldown" that CineFrame 24 embues. No, it's not perfect, but given the opportunity it provides for artistic creativity, it's useful in the hands of someone who has made informed decisions. I've seen several very impressive pieces of work using the CF24mode and the CineForm HDI.
Reading information from the CineForm website vs experience with the actual Cineframe modes doesn't equate to personal experience, but rather forming an opinion based on written word.
I don't own an HC1, given that I have several A1's, but based on my short experience with the HC1 prior to its introduction, and information from Sony; it's not the same. If your HC1 has exactly the same CF24 mode as the A1/Z1, I'll stand corrected.

[Mikko Lopponen]

Originally Posted by Douglas Spotted Eagle

And of course there are interlace lines, at any shutter speed. It's *always* an interlaced output from the HC1.

Output is interlaced? Sameway like the output of some toshiba hd-dvd players is interlaced even though the material is progressive? That doesn't matter if the material is progressive. You can't just add lines everywhere, because the lines are created from the frames that are encoded together.

You miss how interlace works. It works by combining two frames into one to create one frame with fields. Those fields are played back lower field first (dv) to separate the fields from the frame.

Now if you combine twoo frames that are exactly the same frame (like for example shutter speeds 1/30 etc) then there will be no interlace lines. Because the frames are the same (combine a and b that are exactly the same, you will not get lines), there cannot be any lines. Those lines are different frames encoded into one, there's nothing magical about it.

So how come hc1 has interlace lines in 1/30th mode? Because the fields are flipped. Not flipped as in upper or lower but flipped in time. Frame A is displayed in the wrong place than frame B. They encode the wrong frames together. I'll make some captures and study them more this week.

[Douglas Spotted Eagle]

you're apparently right, Mikko. I guess I don't know how interlaced or progressive works. I look forward to reading your whitepaper on how interlacing and the HC1 operate.

[Mark Donnell]

Mikko, either you are having trouble saying what you want to say, or you have real misunderstandings regarding interlaced versus progressive. Shutter speed has nothing to do with whether two interlaced fields are identical. The only time two interlaced fields are identical is when there has been no motion of anything being recorded during the time between the recording of the fields. Thus, interlaced video of a perfectly motionless scene could be deinterlaced electronically to frames that would be identical to video recorded by a progressive frame camera. Similarly, true progressive frames can be interlaced into fields and recorded, as is done in some modes on the Panasonic HVX-200 camera.

[Thomas Smet]

Originally Posted by Mikko Lopponen

Output is interlaced? Sameway like the output of some toshiba hd-dvd players is interlaced even though the material is progressive? That doesn't matter if the material is progressive. You can't just add lines everywhere, because the lines are created from the frames that are encoded together.

You miss how interlace works. It works by combining two frames into one to create one frame with fields. Those fields are played back lower field first (dv) to separate the fields from the frame.

Now if you combine twoo frames that are exactly the same frame (like for example shutter speeds 1/30 etc) then there will be no interlace lines. Because the frames are the same (combine a and b that are exactly the same, you will not get lines), there cannot be any lines. Those lines are different frames encoded into one, there's nothing magical about it.

So how come hc1 has interlace lines in 1/30th mode? Because the fields are flipped. Not flipped as in upper or lower but flipped in time. Frame A is displayed in the wrong place than frame B. They encode the wrong frames together. I'll make some captures and study them more this week.

I may not always agree with Spot but trust me when I say I am pretty sure he knows the difference between interlaced and progressive.

A shutter speed of 30 is interlaced on the SONY HC1. I have one and have tried pretty much everything with this camera to fool it into doing progressive. Trust me it does not work.

I'm not sure what the HC1 is doing with the shutter but it is not the same as other cameras with a 30 shutter. I have captured uncompressed component video by using a shutter of 30 and even putting the camera in photo mode and all of the frames are interlaced. They may look different but they are interlaced. The only way and I mean the only way to get progressive frames from the HC1 is to use the Cinema mode which does work fairly well if you give it a chance. It isn't perfect but neither is using a $1,500 camera.

[Graham Hickling]

My understanding has always been that for 60i-based cameras, when using shutter speeds of less than 1/60th, the image is recorded to a buffer and then written to the appropriate number of 1/60th second fields. So at, say, a 1/15th shutter, the buffer writes the identical images to 4 successive fields, and then refreshes and repeats.

I think Mikko believes the same thing, and hence he (and I) would expect successive pairs of fields in 1/30th footage to be identical.

[Mauritius Seeger]

Originally Posted by Douglas Spotted Eagle

you're apparently right, Mikko. I guess I don't know how interlaced or progressive works.
I look forward to reading your whitepaper on how interlacing and the HC1 operate.

that's a bit uncalled for. i think that mikkos explanation makes alot of sense to me.

the way i see it, when people say interlaced 'lines' what they mean is that you can see discontinuities between fields a and b when you combine both fields into a frame. if fields a and b were recorded at the same time then one would not expect to see those lines - interlacing or not!

it's also hard to see how when the shutter stays open for 1/30th of second and the camera derives two fields from that one exposure how the resulting two fields can be different enough to show up 'interlacing lines' - because there are only 30 exposures during one second yet the camera needs to derive 60 frames from that. similarly if you shoot footage of a static object (camera on a tripod) you would also not expect to see interlacing lines.

[Douglas Spotted Eagle]

whether you see interlacing or not, it's there. It's an interlaced stream regardless of what your eyes might tell you.
No way around it, it's interlaced. However, according to Mikko, I don't understand interlacing vs progressive. It's not a subject I'm going to reduce myself to debating. The camera is interlaced. Mikko claims it's not. So, I look forward to seeing how he demonstrates that it is not interlaced at 1/30 when shooting anything other than a static object with the camera on a tripod.

[Peter Ferling]

Oh good grief, if I pan or zoom then everything sucks.

[Graham Hickling]

Give him a break for language and terminology. He says this: "... if you combine twoo frames that are exactly the same frame (like for example shutter speeds 1/30 etc) then there will be no interlace lines"

Now he means two fields not two frames when he says that. He's saying when you combine two identical fields you should not SEE any interlace lines, even though the image is indeed interlaced.

I agree with him on that point.

[Douglas Spotted Eagle]

Regardless of language, terminology, or any other point to be nit-picked, it's an interlaced stream. Period. Now, whether your NLE correctly interprets those two fields and does or does not display interlacing is up to your NLE. But it is still, and always will be, an interlaced image regardless of what your eyes see, or what you *think* it should be.
From my perspective, shooting a non-moving subject with a non-moving camera for purposes of being able to deinterlace an interlaced stream that doesn't have horizontal artifacting somewhat goes against the grain of producing motion pictures, wouldn't you agree?
More importantly, it's simply dishonest to tell others that the HC1 is progressive at any state of it's shooting function.
Further, from what we have shot with the A1, Sony does not invert the fields, HDV is upper field first. Our NLE's (Vegas and Avid XPress) interpret the fields in correct order.
I'm not at all trying to create chaos, but rather settle it out. Mikko claims the camcorder has progressive qualities, it does not. He claims Sony deliberately inverts the fields; they do not. Mikko claims I don't understand interlacing vs progressive scan formats, I do.
either way, it's not something worth debate; the camcorder and its functions are quite well documented.
If you want a deinterlaced image of a non-moving object with a non-moving camera, the HC1 has a great still photo mode, which would be the better (and higher resolution) image.

[Graham Hickling]

Yes, fine to all of that.

But the question that kicked off this thread still stands.....I fail to understand how interlace lines could be VISIBLE in footage shot at 1/30.

[Greg Boston]

I think the important distinction to make here is the 'appearance' of interlaced or progressive vs. the 'aquisition mode' of interlaced or progressive.

What DSE is saying is that no matter what it 'appears' to be, the camera 'aquires' the image using an interlaced scan in all modes of the camera's operation.

That doesn't mean that under certain conditions that an interlaced image comprised of two fields can't 'appear' to be progressive. In fact, that's the magic of 'frame' mode on the XLH1 and its two newer siblings.

-gb-