4:3 mode

[Keith Mottram]

whats the resolution in squeeze mode is it still hdv or can you only use 4:3 in dvcam mode?

 

cheers,

 

Keith

[Phil Rhodes]

Hi,

 

You get true 16:9 in DVCAM mode. Excellent feature.

 

Phil

[Keith Mottram]

Hi,

 

You get true 16:9 in DVCAM mode. Excellent feature.

 

Phil

<{POST_SNAPBACK}>

 

 

Cheers for the reply Phil,

 

But what I really need to know is in 4:3 mode specifically with the squeeze function are you can you still record to HDV or only DVCAM? hope this clarifies,

 

Keith

[David Mullen ASC]

HD is a non-squeezed, native 16x9 format, either 1920 x 1080 pixels of 1280 x 720 pixels. You can't record 4x3 in HD; you can only frame 16x9 HD for matting on the sides to 4x3.

 

In 4x3 SD mode (if there is one), I'm sure the camera crops the 16x9 CCD image to 4x3.

 

16x9 is only squeezed in SD. In HD, the pixel ratio is naturally 16x9 so it isn't necessary.

[Keith Mottram]

HD is a non-squeezed, native 16x9 format, either 1920 x 1080 pixels of 1280 x 720 pixels.  You can't record 4x3 in HD; you can only frame 16x9 HD for matting on the sides to 4x3.

 

In 4x3 SD mode (if there is one), I'm sure the camera crops the 16x9 CCD image to 4x3.

 

16x9 is only squeezed in SD. In HD, the pixel ratio is naturally 16x9 so it isn't necessary.

<{POST_SNAPBACK}>

 

I know that, when I talk about squeeze mode I was under the assumption that when shooting 4:3 the 4:3 image was squeezed to fill 16:9, thus holding more resolution. My question is basically what resolution 4:3 mode is recorded on the Z1 or whether it is just an option for SD output. To simplify is it possible to to have a 4:3 image squashed to the 16:9 of HDV- then in post you could unsqueeze and have a very sharp 4:3 image. Hope this clarifies.

 

Keith

[David Mullen ASC]

I know that, when I talk about squeeze mode I was under the assumption that when shooting 4:3 the 4:3 image was squeezed to fill 16:9, thus holding more resolution.  My question is basically what resolution 4:3 mode is recorded on the Z1 or whether it is just an option for SD output. To simplify is it possible to to have a 4:3 image squashed to the 16:9 of HDV- then in post you could unsqueeze and have a very sharp 4:3 image.

<{POST_SNAPBACK}>

 

That would only make sense if there were such a thing as 4x3 HD broadcasting.

 

No HD camera stretches 4x3 out to fill 16x9 HD (except maybe the Viper by use of its sub-pixel arrangement in its unusual CCD's, just like how it can create 2.35 on full-frame 16x9 HD).

 

For 4x3 SD, you shoot 16x9 HD and crop to get 4x3 SD in post.

 

Or if the HDV camera has a 4x3 SD recording mode, then it would use a 4x3 area from the center of the 16x9 CCD and record that straight to 4x3 SD.

 

There's no reason for the HD camera to stretch 4x3 to fill 16x9 HD since a 4x3 crop of HD is still HIGHER in resolution than SD, so it's still oversampling to get to a 4x3 SD recording.

 

On a 1920 x 1080 HD recording, cropping the sides to 4x3 gets you a 1437 x 1080 pixel image which is then downconverted to 720 x 480 (NTSC.)

 

So there isn't much advantage for an HD camera to first crop its 16x9 CCD to 4x3 and then STRETCH that back out to 16x9 for HD recording, then SQUEEZE that AND SHRINK that down to 4x3 SD. You aren't really gaining any true resolution because you had to crop the 16x9 CCD to get 4x3 anyway (if you optically stretched 4x3 to fill 16x9, maybe -- it's the same issues as using a 4x3 SD camera to get 16x9 only reversed.)

 

The only advantage of a camera stretching 4x3 out to fill 16x9 HD would be for a transfer to 35mm film IF the final projection format was 1.33 -- but since almost all projection is either 1.85 or 2.35, that really isn't necessary either.

[Keith Mottram]

This might seem daft but the reason I'm asking is because I've used a p+S technic with scope lenses and I thought about trying it out with the Z1 (which I have access to) so if there was a way of squeezing the 4:3 onto HDV there could be some advantages.

 

Keith

[David Mullen ASC]

This might seem daft but the reason I'm asking is because I've used a p+S technic with scope lenses  and I thought about trying it out with the Z1 (which I have access to) so if there was a way of squeezing the 4:3 onto HDV there could be some advantages.

 

Keith

<{POST_SNAPBACK}>

 

No.

 

I'm going through the same issues right now actually. You'll end up with a 3.56 : 1 image once you unsqueeze it, and you'll have to crop the sides to get to 2.39.

 

Even in 35mm, the negative is only about 1.20 : 1 (more square than 4x3) when using a 2X anamorphic lens for a final projected image that is 2.40 : 1.

 

 

I actually had this discussion with Arri about the D20 because unlike the Genesis or Dalsa, it has a 4x3 sensor, but it crops it to 1.78 for recording to an HDCAM-SR deck. So I said, "why can't you stretch the whole 4x3 sensor image to HD for recording?"

 

And they said "why would you do that?"

 

"Because then I could shoot with standard 2x anamorphic lenses on the Arri D20, record to 16x9 HDCAM-SR and end up with a 1.33X electronic squeeze with 2.35 filling the recording, just like the Viper does in 2.35 mode" (or the Genesis with a 1.34X anamorphic lens if they finish building one.)

 

But Arri really does not believe anamorphic photography has a future and have no interest in accomodating that for an HD recording.

 

They said in the future, you could record the whole 4x3 sensor area to 2K data, but they haven't worked the bugs out of that. But when sending out a dual-link 4:4:4 HD signal to HDCAM-SR, it crops the sensor to 16x9. But this is an unusual case because the Arri D20 is the only HD/2K camera with a 4x3 sensor.

 

Panavision is building a set of anamorphic lenses with a 1.34X squeeze to fit 2.39 onto 16x9 for their Genesis, but it could also do the same thing on a 3-perf camera.

[Keith Mottram]

thanks for the reply David, it does seem strange that all these companies want to reject what would seem on paper a small modification, but then again a small modification for a small amount of users. I was just hoping that the way you discussed with arri could aply to the Z1 as it has a 4:3 outout option, I assumed that it would not use the whole of the sensor anyway, but would at least have a higher resolution than sd. tis a shame as they say. I knew that there would be some cropping but i had assumed a final aspect of 2.66:1 - stretched over the 16:9, you could then unsqueeze the image in shake, for example, color correct and resize for a 2k output . I had some nice results using the p+s with scope lenses on a canon with this method (though this was 4:3 dv)- so it would have been great if this had worked. rereading the bumph on the sony site it seems to allude that the 4:3 option is an SD downconvert which I assume spits out the centre of the image- so I might still do a test and see what happens. It might be useful as an effect for SD producitons.

 

once again thanks for your patience and it will be interesting to see if panavisions developments will sway arri at all into rethinking.

 

Keith

Edited June 23, 2005 by keith mottram

[David Mullen ASC]

If your project is just for downconversion to SD anyway, cropping spherical 16x9 HD to 2.39 or putting a 2X anamorphic lens and then cropping 3.56 : 1 to 2.39, is all probably going to still give you more resolution than SD.

 

On a 1920 x 1080 recording with a 2X squeeze, you'd be using only 1290 x 1080 for a 2.39 image (1.4MP) -- if you just cropped top & bottom to get 2.39, you'd be using 1920 x 800 (1.5MP). Either way, that's more resolution than the final letterboxed SD image.