Why do 16x9 in Post? (vx2000 and pd150) at DVinfo.net

Basem Elsokary · September 22nd, 2003, 04:22 PM

it seems people always recommend going with a 4x3 shot, framed for 16x9, and adjust it in post production...

I have been using the built in widescreen setting on the PD150, and it seems to work just fine, and is less of a problem in terms of trying to reframe in Post if you didn't correctly do so during shooting...

the argument that you are losing resolution seems to be flawed here. The argument usually goes like this, that when shooting in 4x3, you are using more of the CCD and thus get higher resolution...in post however, all you are doing is masking out that extra resolution. When you use the camera's built in widescreen setting, it puts the mask up for you, and encodes the video as widescreen, thus saving time in post for conversion...You also don't have framing problems as you can frame on the spot by looking at the LCD or in the viewfinder. And you're still using the same amount of the CCD's resolution in the end.

So if you do the math, lets say the CCD captures 500 lines of resolution...if you shoot full 4x3, you get full 500 lines. In post, you mask the top and bottom 15% of the image...so now you're using own 70% of those lines, or 350 lines.

Now lets use the built in widescreen setting. In this setting, the CCD doesn't use the top and bottom 15% of the CCD to capture the video image...so you lose the same 150 lines of resolution, resulting in 350 lines of resolution.

so if you ask me, the only thing I gain by shooting full frame is if I change my mind and decide to go full-frame, otherwise, I'm better off framing it correctly the first time...is there something wrong with my analysis of this situation?

Mike Rehmus · September 22nd, 2003, 06:05 PM

There is at least one counter argument to your position, Basem.

When you capture 4:3, you get a choice in Post to correct your vertical framing and, of course, you get the choice to change your mind about the aspect ratio of the finished piece.

The mind set of most of us is to try to not limit our options when we tape.

You can use the 150's overlay feature to make yourself a bold 16:9 mask for use in framing.

Sean R Allen · September 22nd, 2003, 06:14 PM

I thought there was also an issue of the 5:1 DV compression, that you are more likely to get artifacts and that it isn't as efficient when using in camera 16x9, or so I've heard. So for best quality shoot 4:3 and crop in post. You also get the added benefit of total flexibility.

Barry Green · September 22nd, 2003, 07:34 PM

The reason people advise against using the in-camera 16:9 mode on the VX2000 is because it does a fairly poor job. Boyd Ostroff did some resolution chart comparisons here:

http://www.greenmist.com/pdx10/mode/03.jpeg

He compared cropping and stretching in post, vs. in-camera, and there's just no contest, stretching in post yields a much higher resolution picture.

Some cameras do okay with 16:9 -- Canon XL1's are supposed to be quite good at it. Sony's VX2000 is not.

Also, your results will vary depending on what post-production solution you use to stretch the footage. Boyd used PhotoShop, but my testing showed Premiere 6.0 to be not so good at it. Vegas 4.0 is phenomenal at it.

Boyd Ostroff · September 22nd, 2003, 08:29 PM

<<<-- Originally posted by Barry Green : Boyd used PhotoShop, but my testing showed Premiere 6.0 to be not so good at it. Vegas 4.0 is phenomenal at it. -->>>

Excellent point, software may make a significant difference and motion video isn't the same as still frames. Run your own tests, using your own camera, lighting, exposure settings and software. Then decide for yourself which you like better. You can download a nice test pattern at bealecorner

Basem Elsokary · September 23rd, 2003, 10:50 AM

well,...I'm working a project now that I shot in 4x3, and will crop to 16x9 in Vegas,....so hopefully I'll see for my own eyes if there's a difference...

again,...the whole mathematics of the two different solutions didn't make sense to me...technically speaking, it just seems to be that both methods would yield exactly the same quality image in the end...

as I mentioned and was mentioned by others...i do realize there are benefits to shooting 4x3 and reframing in post...

Barry Green · September 23rd, 2003, 11:28 AM

The math is the same. You do lose the same number of lines. But the quality of the work done by the in-camera algorithm is just lousy compared to what can be done in post by a good re-sizing algorithm, such as PhotoShop or Vegas.

Seeing is believing. Look at Boyd's picture, or shoot your own, and you'll come to the same conclusion.

Basem Elsokary · September 23rd, 2003, 11:58 AM

...when you're talking about algorithms in the NLE software,...what exactly are you talking about?

What I planned on doing in Vegas for example was just creating the 16x9 mask on the top layer...and then using the widescreen template to render to dvd...where in this process do different programs use different algorithms...again, this process is new to me, so I wasn't too sure what you meant by the widescreen algorithms used by the NLE

Barry Green · September 23rd, 2003, 12:08 PM

Let's first clarify if you're talking about letterbox footage or 16:9 footage.

Letterbox footage is 4:3 footage with a black mask applied on the top and bottom. This is appropriate for making a VHS tape of widescreen material, for example.

16:9 footage is actually anamorphically stretched footage, different from letterbox. If displayed on a 4:3 TV, 16:9 footage will look "squeezed". If displayed on a 16:9 TV it will look "proper". Widescreen DVD's have anamorphic 16:9 footage on them.

Using in-camera 16:9 is one way to get anamorphic 16:9 footage, but on the VX2000 it's not a good way to get it.

So the footage needs to be stretched in post. It is that stretching algorithm that is at the root of the debate: a good stretching algorithm can yield good results, a bad stretching algorithm will ruin your footage. Vegas has a great stretching algorithm. Premiere 6.0 has a lousy one. I have heard mixed reports on FCP but I don't have access to it so I can't test.

With Vegas you use the crop/pan filter, apply the 16:9 mask and make sure that it's set to stretch to fill the frame. You should be quite happy with the results.

Basem Elsokary · September 23rd, 2003, 12:22 PM

ahhhhhh! I'm beginning to understand :)

Yes...I do plan on creating anamorphic content for use on a dvd...I was under the wrong impression of how to go about accomplishing this...I'll definitely give your method a try in vegas!

a second question, based on what was stated below...how do I use a mask within the pd150 for framing my shot without actually recording the mask to tape?

again, thanks for all the feedback on this issue...it has been something I did not understand fully for quite some time!

Timothy Lab · September 26th, 2003, 04:48 PM

Hello Basem,

I would like to offer this information on shooting widescreen. The BBC uses the VX-2000 for widescreen shooting and they offer free online courses and guides for shooting for television with the VX-2000. There is one giude, in particular, that you should read. Again, here is the address;

http://www.bbctraining.co.uk/television.asp

Look under - "Online Courses and Guides"

Select: DV Usage Guides

Then Open Guide #2; How to shoot widescreen on small cameras

You may want to look at the other information they offer, too. Such as post production and camera modifications (anamorphic) and their list of what cameras do well with widescreen and which one doesn't.

Good luck,

Tim

September 22nd, 2003, 04:22 PM	#1
Basem Elsokary Regular Crew Join Date: Mar 2003 Posts: 33	Why do 16x9 in Post? (vx2000 and pd150) it seems people always recommend going with a 4x3 shot, framed for 16x9, and adjust it in post production... I have been using the built in widescreen setting on the PD150, and it seems to work just fine, and is less of a problem in terms of trying to reframe in Post if you didn't correctly do so during shooting... the argument that you are losing resolution seems to be flawed here. The argument usually goes like this, that when shooting in 4x3, you are using more of the CCD and thus get higher resolution...in post however, all you are doing is masking out that extra resolution. When you use the camera's built in widescreen setting, it puts the mask up for you, and encodes the video as widescreen, thus saving time in post for conversion...You also don't have framing problems as you can frame on the spot by looking at the LCD or in the viewfinder. And you're still using the same amount of the CCD's resolution in the end. So if you do the math, lets say the CCD captures 500 lines of resolution...if you shoot full 4x3, you get full 500 lines. In post, you mask the top and bottom 15% of the image...so now you're using own 70% of those lines, or 350 lines. Now lets use the built in widescreen setting. In this setting, the CCD doesn't use the top and bottom 15% of the CCD to capture the video image...so you lose the same 150 lines of resolution, resulting in 350 lines of resolution. so if you ask me, the only thing I gain by shooting full frame is if I change my mind and decide to go full-frame, otherwise, I'm better off framing it correctly the first time...is there something wrong with my analysis of this situation?

September 22nd, 2003, 06:05 PM	#2
Mike Rehmus Wrangler Join Date: May 2002 Location: Vallejo, California Posts: 4,049	There is at least one counter argument to your position, Basem. When you capture 4:3, you get a choice in Post to correct your vertical framing and, of course, you get the choice to change your mind about the aspect ratio of the finished piece. The mind set of most of us is to try to not limit our options when we tape. You can use the 150's overlay feature to make yourself a bold 16:9 mask for use in framing. __________________ Mike Rehmus Hey, I can see the carrot at the end of the tunnel!

September 22nd, 2003, 06:14 PM	#3
Sean R Allen Regular Crew Join Date: Jul 2003 Location: RTP, NC Posts: 66	I thought there was also an issue of the 5:1 DV compression, that you are more likely to get artifacts and that it isn't as efficient when using in camera 16x9, or so I've heard. So for best quality shoot 4:3 and crop in post. You also get the added benefit of total flexibility.

September 22nd, 2003, 07:34 PM	#4
Barry Green Barry Wan Kenobi Join Date: Jul 2003 Location: North Carolina Posts: 3,863	The reason people advise against using the in-camera 16:9 mode on the VX2000 is because it does a fairly poor job. Boyd Ostroff did some resolution chart comparisons here: http://www.greenmist.com/pdx10/mode/03.jpeg He compared cropping and stretching in post, vs. in-camera, and there's just no contest, stretching in post yields a much higher resolution picture. Some cameras do okay with 16:9 -- Canon XL1's are supposed to be quite good at it. Sony's VX2000 is not. Also, your results will vary depending on what post-production solution you use to stretch the footage. Boyd used PhotoShop, but my testing showed Premiere 6.0 to be not so good at it. Vegas 4.0 is phenomenal at it.

September 22nd, 2003, 08:29 PM	#5
Boyd Ostroff Wrangler Join Date: Dec 2002 Location: Mays Landing, NJ Posts: 11,802	<<<-- Originally posted by Barry Green : Boyd used PhotoShop, but my testing showed Premiere 6.0 to be not so good at it. Vegas 4.0 is phenomenal at it. -->>> Excellent point, software may make a significant difference and motion video isn't the same as still frames. Run your own tests, using your own camera, lighting, exposure settings and software. Then decide for yourself which you like better. You can download a nice test pattern at bealecorner

September 23rd, 2003, 10:50 AM	#6
Basem Elsokary Regular Crew Join Date: Mar 2003 Posts: 33	well,...I'm working a project now that I shot in 4x3, and will crop to 16x9 in Vegas,....so hopefully I'll see for my own eyes if there's a difference... again,...the whole mathematics of the two different solutions didn't make sense to me...technically speaking, it just seems to be that both methods would yield exactly the same quality image in the end... as I mentioned and was mentioned by others...i do realize there are benefits to shooting 4x3 and reframing in post...

September 23rd, 2003, 11:28 AM	#7
Barry Green Barry Wan Kenobi Join Date: Jul 2003 Location: North Carolina Posts: 3,863	The math is the same. You do lose the same number of lines. But the quality of the work done by the in-camera algorithm is just lousy compared to what can be done in post by a good re-sizing algorithm, such as PhotoShop or Vegas. Seeing is believing. Look at Boyd's picture, or shoot your own, and you'll come to the same conclusion.

September 23rd, 2003, 11:58 AM	#8
Basem Elsokary Regular Crew Join Date: Mar 2003 Posts: 33	...when you're talking about algorithms in the NLE software,...what exactly are you talking about? What I planned on doing in Vegas for example was just creating the 16x9 mask on the top layer...and then using the widescreen template to render to dvd...where in this process do different programs use different algorithms...again, this process is new to me, so I wasn't too sure what you meant by the widescreen algorithms used by the NLE

September 23rd, 2003, 12:08 PM	#9
Barry Green Barry Wan Kenobi Join Date: Jul 2003 Location: North Carolina Posts: 3,863	Let's first clarify if you're talking about letterbox footage or 16:9 footage. Letterbox footage is 4:3 footage with a black mask applied on the top and bottom. This is appropriate for making a VHS tape of widescreen material, for example. 16:9 footage is actually anamorphically stretched footage, different from letterbox. If displayed on a 4:3 TV, 16:9 footage will look "squeezed". If displayed on a 16:9 TV it will look "proper". Widescreen DVD's have anamorphic 16:9 footage on them. Using in-camera 16:9 is one way to get anamorphic 16:9 footage, but on the VX2000 it's not a good way to get it. So the footage needs to be stretched in post. It is that stretching algorithm that is at the root of the debate: a good stretching algorithm can yield good results, a bad stretching algorithm will ruin your footage. Vegas has a great stretching algorithm. Premiere 6.0 has a lousy one. I have heard mixed reports on FCP but I don't have access to it so I can't test. With Vegas you use the crop/pan filter, apply the 16:9 mask and make sure that it's set to stretch to fill the frame. You should be quite happy with the results.

September 23rd, 2003, 12:22 PM	#10
Basem Elsokary Regular Crew Join Date: Mar 2003 Posts: 33	ahhhhhh! I'm beginning to understand :) Yes...I do plan on creating anamorphic content for use on a dvd...I was under the wrong impression of how to go about accomplishing this...I'll definitely give your method a try in vegas! a second question, based on what was stated below...how do I use a mask within the pd150 for framing my shot without actually recording the mask to tape? again, thanks for all the feedback on this issue...it has been something I did not understand fully for quite some time!

September 26th, 2003, 04:48 PM	#11
Timothy Lab New Boot Join Date: Aug 2003 Location: Owosso, MI Posts: 13	Hello Basem, I would like to offer this information on shooting widescreen. The BBC uses the VX-2000 for widescreen shooting and they offer free online courses and guides for shooting for television with the VX-2000. There is one giude, in particular, that you should read. Again, here is the address; http://www.bbctraining.co.uk/television.asp Look under - "Online Courses and Guides" Select: DV Usage Guides Then Open Guide #2; How to shoot widescreen on small cameras You may want to look at the other information they offer, too. Such as post production and camera modifications (anamorphic) and their list of what cameras do well with widescreen and which one doesn't. Good luck, Tim