How should I know my mini dv cam is shooting a true 16:9

[joy]

I have read someplace that most mini dv cameras don't shoot a true 16:9 video. I wonder if mine which is an older sony model trv-17 is one of them.

The manual states that it can and I have tried shooting it in this mode but reading that article about those fake 16:9 made me think again. The 16:9 video which I took plays fine on the widescreen tv but looks tall on a 4:3 tv.

Any opinion will be greatly appreciated. Thanks!

[gshelley61]

If the video looks tall and skinny on a 4:3 display, then you do have an anamorphic 16:9 image recorded on the tape.

Some older camcorders used to record widescreen by simply masking the 4:3 frame with letterbox bars (resulting in a non-anamorphic, lower resolution widescreen image).

The other issue surrounding "true 16:9" camcorders has to do with how the CCD pixels are used to acquire the widescreen image. In many cases, less CCD area (fewer pixels) is used for widescreen images. However, I think there are several newer camcorders that have higher resolution CCD's, eliminating this problem.

You can also purchase a special anamorphic lens for many camcorders. This allows you to shoot 16:9 anamorphic images optically (using all your camcorder's CCD pixels). This is how it is done with film cameras.

[edDV]

I have read someplace that most mini dv cameras don't shoot a true 16:9 video. I wonder if mine which is an older sony model trv-17 is one of them.

The manual states that it can and I have tried shooting it in this mode but reading that article about those fake 16:9 made me think again. The 16:9 video which I took plays fine on the widescreen tv but looks tall on a 4:3 tv.

Any opinion will be greatly appreciated. Thanks!

The TRV-17 is an early 1CCD consumer cousin of the 3CCD TRV950. The same CCD usage issues apply but the single CCD models must share the 690Kpixels for filtered RG&B. After processing, the remaining resolution should be full height squeezed inside a 720x480 frame.

This picture of CCD mapping should explain the tradeoffs. The PDX10 represents the 3CCD prosumer models of the day. 3CCD uses all the resolution per color component.

[joy]

Thanks for the reply!

So you guys implying that this camera can shoot a true 16:9 video but at a less than perfect quality because of lesser resolution.

I have done a visual comparison of 2 ten second video clips taken by this camera, 1 in 4:3 mode and the other in 16:9 mode. The videos were of one corner of my living room. I played the video one at a time and tried comparing them. It's hard to tell the difference except for the aspect ratio of the clips. I wonder if this method of comparison is a proper way to tell if the other clip was better or not.

[edDV]

Thanks for the reply!

So you guys implying that this camera can shoot a true 16:9 video but at a less than perfect quality because of lesser resolution.

I have done a visual comparison of 2 ten second video clips taken by this camera, 1 in 4:3 mode and the other in 16:9 mode. The videos were of one corner of my living room. I played the video one at a time and tried comparing them. It's hard to tell the difference except for the aspect ratio of the clips. I wonder if this method of comparison is a proper way to tell if the other clip was better or not.

It depends on the quality of your monitor and the method of connection.

A EDTV/HDTV connected S-Video has enough resolution to compare.

[vhelp]

joy,

I have the Sony TRV-22 cam, while yours is the lower cousin. But, it is my
belief that your cam is outputing an Anamorphic 16x9 image inside an 720 x 480
pixel image.

When you import this inside vdub, it will *not* display the image in it's
proper aspect ratio, or un-squeeze it. But, you can still view it as a
16x9 widescreen w/out boarders inside vdub as long as you have the version
v1.5.10 (may even or higher ver's as well) by:

** right-clicking on the image window, and
** select, 16:9 frame ratio (HTDV)

This will expand it and you will have to adjust your window to show the extra
image area.

( Note, this will have no effect on the frameserving aspects. It will *still*
return your MPEG Encoder sofware with a 4:3 image. Its up to you to know how to
process Anamorphic 1.78:1 AR 16x9 720x480 formated source. )

Now, as to the proper way of encoding this strange image format from these
cheap cams..

using (vdub and TMPGenc) as the frameserver scenario:

** open/import your AVI source file.
** setup to frameserve, and into your MPEG encoder of choice.
** frameserve into TMPGenc.
** now, inside TMPGenc do:
** a) setting\video\aspect ratio: [16:9 display ]
** b) setting\advanced\field order: [bottom field first (field B) ]
** c) setting\advanced\source aspect ratio: [4:3 display]
** d) setting\advanced\video arrange method: [Full Screen ]

Note 1 - when working with MPEG-2 video's, its important to know that Aspect Ratios
only work properly when the resolution is respected according to the dvd standard.
The resolution should always be 720 x 480 and no less, otherwise, your dvd player
will not present your videos properly. It must be 720 x 480 to respect this
presentation.

Note 2 - ** b) setting\advanced\field order: [bottom field first (field B) ]
Because your source type was DV, it is BFF. Even if you used your cam as a
pass-through, via VCR; Cable box; Satellite; etc., you would still have set
your MPEG to BFF, because that is what DV image is set to during the internals
of the DV machanics. It you don't follow this rule, your vidoes will play as
jittery or bobbing or something similar, per your given stand-alone dvd player.

-vhelp 3926

[joy]

It depends on the quality of your monitor and the method of connection.

I'm using a 19" flat lcd (1280 x 1024 native resolution).

A EDTV/HDTV connected S-Video has enough resolution to compare.

I have transferred the video clips to the HDD so these must be identical to the source..

joy,
I have the Sony TRV-22 cam, while yours is the lower cousin. But, it is my
belief that your cam is outputing an Anamorphic 16x9 image inside an 720 x 480
pixel image.

When you import this inside vdub, it will *not* display the image in it's
proper aspect ratio, or un-squeeze it. But, you can still view it as a
16x9 widescreen w/out boarders inside vdub as long as you have the version
v1.5.10 (may even or higher ver's as well) by:

** right-clicking on the image window, and
** select, 16:9 frame ratio (HTDV)

This will expand it and you will have to adjust your window to show the extra
image area.

( Note, this will have no effect on the frameserving aspects. It will *still*
return your MPEG Encoder sofware with a 4:3 image. Its up to you to know how to
process Anamorphic 1.78:1 AR 16x9 720x480 formated source. )

Now, as to the proper way of encoding this strange image format from these
cheap cams..

using (vdub and TMPGenc) as the frameserver scenario:

** open/import your AVI source file.
** setup to frameserve, and into your MPEG encoder of choice.
** frameserve into TMPGenc.
** now, inside TMPGenc do:
** a) setting\video\aspect ratio: [16:9 display ]
** b) setting\advanced\field order: [bottom field first (field B) ]
** c) setting\advanced\source aspect ratio: [4:3 display]
** d) setting\advanced\video arrange method: [Full Screen ]

Note 1 - when working with MPEG-2 video's, its important to know that Aspect Ratios
only work properly when the resolution is respected according to the dvd standard.
The resolution should always be 720 x 480 and no less, otherwise, your dvd player
will not present your videos properly. It must be 720 x 480 to respect this
presentation.

Note 2 - ** b) setting\advanced\field order: [bottom field first (field B) ]
Because your source type was DV, it is BFF. Even if you used your cam as a
pass-through, via VCR; Cable box; Satellite; etc., you would still have set
your MPEG to BFF, because that is what DV image is set to during the internals
of the DV machanics. It you don't follow this rule, your vidoes will play as
jittery or bobbing or something similar, per your given stand-alone dvd player.

-vhelp 3926

Thanks for all these info vhelp. Very much appreciated.

[edDV]

Here is how the wide picture is generated in a TRV-17 and similar Sony consumer cams.

4:3 uses 690k pixels (shared RGB)
16:9 uses 3/4 the height for 517k pixels (shared RGB)
The reduced resolution is in the vertical dimension only.

After processing the picture is output 480 high and 720 squeezed horizontally.
The picture is displayed 16:9 wide. The horizontal stretch is done in the TV.

A EDTV/HDTV connected S-Video has enough resolution to compare.

I have transferred the video clips to the HDD so these must be identical to the source.

Is this a TV (interlaced) or a computer monitor (progressive)? If the latter, how are you playing it? What method of deinterlace?

A computer player would stretch 720 to 853 in horizontal. Vertical may be 480 interpolated or just stretched from a 720x240 field.

[joy]

Thanks again edDV!

Here is how the wide picture is generated in a TRV-17 and similar Sony consumer cams.

4:3 uses 690k pixels (shared RGB)
16:9 uses 3/4 the height for 517k pixels (shared RGB)
The reduced resolution is in the vertical dimension only.

After processing the picture is output 480 high and 720 squeezed horizontally.
The picture is displayed 16:9 wide. The horizontal stretch is done in the TV.

So actually in laymans term quality is not lost within the area bounded by the 16x9 frame. The reduced resolution(517k) is due only to the lost top and bottom portion within the original 4x3 frame.

Is this a TV (interlaced) or a computer monitor (progressive)? If the latter, how are you playing it? What method of deinterlace?

I have used my pc using WMP to play the transferred file(The monitor supposed to be progressive). I don't think deinteracing was involved. WMP displayed squashed picture horizontally for the 4:3 clip while displaying ok with the 16:9 clip. I also connected the camera to the 4:3 tv(picture look tall) and the 16:9 hdtv(picture look fine).

So with all these information I would like to ask the same question again, do you think this "aging" baby is shooting a true 16:9 video?

[vhelp]

Here's what I think is going on, based on my Sony TRV-22 cam..

(not counting every pixel laid out on the cam's grid) the cam is using the
full grid for the 4:3 image. But..

Section A:

** It is cropping (cutting) the 60 pixels from the Top and Bottom of the image.
** Then, it resizes the image [720x360] to 720 x 480, Anamorpically to 1.78:1 AR.

So, in theory, it is using the full CCD, but not in the most efficieant way.
I mean, these low-budget cams do a poor job of the above process. The evidence
is in the image.. when you view the image, you will see artifacts along the
Horizontals, depending on your magnifications and/or focused image in view.

I believe that your cam is doing the same thing as mine, the TRV-22 cam.

( personally, for my widescreen presentation videos, I like and prefer to shoot
footage in 4:3 mode, and then tweak it to 16x9 anamorphically, using 2.35:1AR.
And the reason for this, is because, at on my cam, there is no suffering of
artifacts when shooting 4:3 mode video. It seems (theory) that for Section A,
the process is not done properly or efficiently, hence the artifacts and the
reason I shoot 4:3 -> 16:9 projects. )

-vhelp 3934

[edDV]

Here's what I think is going on, based on my Sony TRV-22 cam..

(not counting every pixel laid out on the cam's grid) the cam is using the
full grid for the 4:3 image. But..

Section A:

** It is cropping (cutting) the 60 pixels from the Top and Bottom of the image.
** Then, it resizes the image [720x360] to 720 x 480, Anamorpically to 1.78:1 AR.

So, in theory, it is using the full CCD, but not in the most efficieant way.
I mean, these low-budget cams do a poor job of the above process. The evidence
is in the image.. when you view the image, you will see artifacts along the
Horizontals, depending on your magnifications and/or focused image in view.

I believe that your cam is doing the same thing as mine, the TRV-22 cam.

( personally, for my widescreen presentation videos, I like and prefer to shoot
footage in 4:3 mode, and then tweak it to 16x9 anamorphically, using 2.35:1AR.
And the reason for this, is because, at on my cam, there is no suffering of
artifacts when shooting 4:3 mode video. It seems (theory) that for Section A,
the process is not done properly or efficiently, hence the artifacts and the
reason I shoot 4:3 -> 16:9 projects. )

-vhelp 3934

Not for that class of camcorder. They were not offering full height CCD widescreen then. They just cropped the 4:3 CCD to 16:9 and stretched it back to 480 vertical lines in the processing stage.

The DV video that it outputs is 720x480 horizonally squashed which is standard for DV 16:9. In other words, the vertical resolution is interpolated. The horizontal resolution is also interpolated due to the nature of single CCD RGB filtering.

When presented to the player , the player will again stretch the video horizonally to 16:9 dimensions if it responds to the wide screen flag.

So with all these information I would like to ask the same question again, do you think this "aging" baby is shooting a true 16:9 video?

Depends what you mean by "true 16:9 video". It is 16:9 if the player detects it as such and it is video. The H and V will be stretched to fit.

To get close to CCD pixel to 720x480 wide DV mapping, you would need a 3CCD camcorder on the level of the VX-2100. The VX-2100 would have more than 720 horizontal pixels in each component. Those would be filtered to 720,180,180 Y,Cb,Cr for tape or firewire transfer and then the player would stretch them back to 16:9 for display (~853x480).

[joy]

I think I have read enough from you guys replies (thanks edDV and vhelp) and from this interesting article I found by chance:
http://docs.info.apple.com/article.html?artnum=75079 to have me convinced that I should just shoot my next video using 4:3 AR even though both ways will give not give what I badly need, a "perfect" 16:9 image on my 16:9 hdtv

Thanks a lot again!!

[edDV]

I think I have read enough from you guys replies (thanks edDV and vhelp) and from this interesting article I found by chance:
http://docs.info.apple.com/article.html?artnum=75079 to have me convinced that I should just shoot my next video using 4:3 AR even though both ways will give not give what I badly need, a "perfect" 16:9 image on my 16:9 hdtv

Thanks a lot again!!

A few comments on that article.

1. They are assuming you are using a high end 3 CCD VX or PDA class camcorder (or above) that actually will use the full CCD vertical resolution when shooting 16:9. Older single CCD camcorders vertically letterbox the CCD thus chopping 25% out of the 4:3 CCD resolution. The higher end camcorders acquire more than 720 horizontal pixels but filter these down to 720 for recording or distribution. This gives higher quality than simple 1:1 720 mapping, but a single CCD only uses about half to 3/4 of the 720 for luminance so the effective horizontal CCD resolution is more like 352-528x360 that then must be stretched out to 720x480 for output and then to 16:9 for display.

2. Computer geeks like to use the term "anamorphic" for this linear horizontally squished video. I guess this makes them feel inside and cool. Horizontally squished CCIR-601 16:9 video has been around since the mid 80's but nobody used the term anamorphic for that until the Quicktime yuppie wannabees appeared in the early to mid 90's. Inside the industry, anamorphic meant a special nonlinear lens (including spherical distortion) that was used to shoot the video and non-linear reverse curvature algorithms that were needed to recover a square grid. The geeks have highjacked the term so now I guess we need to say optical anamorphic to get around the terminology polution.
http://www.widescreenmuseum.com/Widescreen/apertures.htm

[whenloverageswild]

Hi everybody:

I get into this topic a little late, but I have teh same question.

I have a Digital 8 Sony TRV-520 (bought in 2000 in L.A.). I have no reason yet to replace it, since it record digital very similar to a DV cam. I'am waitng for new gen. HDef Cams, progressive format, supporting high capacity Memory sticks or a Hard disk big enough to do not have to run every 2 horus of recording to download to PC.
So said, I was recording with my Digi 8 cam in 4:3, up to the minute I discovered it allowed a 16:9 format.
So now, I record everything in 16:9.
To tell you the truth, I see no degradation at all. I am concerned, by the way, about the degradation made by converted DV to DVD -that got me nuts!!!).
My only concern on the 16:9 recording is that when I transfer this to DVD in 16:9 and will be displayed in a Widscreen LCD 40" or larger TV set it will remain 16:9 or not? (Obviously I do not have a widescreen LCD TV yet).

Thank in advance for you opinions.

[JohnnyMalaria]

Simplest way:

Set the camcorder up to view something in 4:3 - place two vertical objects at the extreme left and right of the view. Now switch to 16:9 mode. If the vertical objects move inwards, then you camcorder is using a wider range of the CCD than for 4:3 - therefore, you are shooting true 16:9. If the vertical objects stay where they are, more than likely it is faking it (and throwing away detail). And if two black bands appear top and bottom, it's just letterboxing it and it is still 4:3.