EX1 Vs. HVX200 with Green Screen at DVinfo.net

Buzz Remde · May 15th, 2008, 09:43 PM

I have been talking with a client about shooting a series of "Green Screen's" for a TV series. He is getting info about the HVX200 outperforming the EX1 in this area. 4:2:2 vs. 4:2:0

Any tips?

You help is always appreciated.

Leonard Levy · May 15th, 2008, 10:07 PM

I can only tell you what I've read on the web and conversations I've had with people . The consensus is that the Ex-1 vastly outperforms the HVX200 for green screen.

The HVX was never that good as it had comparatively lower resolution and was very noisy. The 4:2:0 thing seems to be over valued. Its a ratio of chroma sampling to luminance and the the resolution of the green channel is so much higher on the Ex at 1080 and the number of pixels so much greater that it at least equals the HVX in chroma info . Also you can shoot 4:2:2 if you can deal with the SDI output and in a few months with a Matrox II that should be easy.

A number of people have said that it equals or actually outperforms an HDCAM on green screen. Adam Wilt told me that just in terms of specs it should. That's all heresay to but it convinced me. Go elsewhere on the web and search for Jim Arthurs whose tested this stuff extensively.

Lenny Levy

Steven Thomas · May 15th, 2008, 10:12 PM

The tip is - don't believe everything you hear.

The EX1 is capable of 4:2:2 10 bit via its SDI-HD.

Venture elsewhere on the net and you will find a thread by Jim Arthurs comparing the EX1 with the RED ONE. The test was a comparison on how well they key. The EX1 looked great. In fact the 1080P comparison to the RED was very close. The RED ONE 4K stuff is in another league beyond the EX1 and way beyond the HVX200.

Noah Yuan-Vogel · May 16th, 2008, 05:32 AM

im going to agree with leonard and buzz, if i had to shoot a green screen i would choose the EX1 in a second over the HVX200. consider this:

color sensor resolution, color output resolution of EX1: 1920:1080, 960x540
color sensor resolution, color output resolution of of HVX: 960x540, 640x1080

so basically, HVX would appear to have horizontal limiting resolution of 640, and vertical limiting resolution of 540 or less, and while the horizontal res is supersampled at the sensor, its scaled by 66% (less ideal than 50%). whereas EX1 gives you 960x540 limiting resolution (960x1080 over SDI) which is supersampled at the sensor by 200% in each direction. supersampling is important from what i understand because you get images that have better contrast/resolution at the limiting resolution than in systems where there is no supersampling. moral of the story is that it is simply by virtue of the fact that the EX1 DELIVERS as much or more detail than the HVX's SENSORS can hope to resolve, EX1 will most likely give you the best results.

also, i have seen some odd color artifacting in the video from HVX200. im not sure if it is from the dvcprohd codec or because the image is being scaled in odd ways, but it scares me a bit.

that said, i have used the hvx for greenscreen work and not the ex1. in my opinion, the HVX creates very soft images where the actual image resolution is far less than the delivered resolution, and that along with the noisiness of the sensor make it a pretty poor choice for greenscreen work. maybe im expecting too much from prosumer camcorders, but i dont think so.

4:2:0, 4:2:2 etc mean nothing out of context. you could be shooting in 4:4:4, but if your camera is SD (some refer to the HVX as an SD camera because of its sensor resolution) you could be in for some soft images depending on your final delivery specs. DVCPROHD is especially misleading since 4:2:2 sounds great but with its downsampled luma, it really should be called 2.6:1.3:1.3

Alister Chapman · May 16th, 2008, 05:42 AM

All I can add to this is that I have done some blue screen work with the EX1 and it keys very well. I used 1080P as I find the P colour space keys much better than interlace. I've been doing green/blue screen with XDCAM HD for a couple of years and it has always produce clean keys when lit properly.

Sebastien Thomas · May 16th, 2008, 06:47 AM

Have been doing key also with footage from the EX1 without any problem. I would go with the EX1 over the HVX.

When keying, NEVER shoot interlace !! Use progressive. Shoot at native resolution (1080p for the EX1). If in a studio, you can do live recording in 4:2:2 color space using the HD-SDI out of the EX1 plugged to an aquisition card (kona, blackmagic...) and get even better result.

David Heath · May 16th, 2008, 06:53 AM

Quote:

Originally Posted by Noah Yuan-Vogel

4:2:0, 4:2:2 etc mean nothing out of context. ........... DVCPROHD is especially misleading since 4:2:2 sounds great but with its downsampled luma, it really should be called 2.6:1.3:1.3

There are many misconceptions about colour space, and it may be worth looking at a bit of history. The "4" notation derives from the very early days of digital video, and when video meant composite. In which case sampling frequency was locked to colour subcarrier frequency. For NTSC, the rate chosen was 14.32MHz - 4x3.58MHz (see where we're going!?) and for PAL was 13.29MHz - 3x4.43MHz.

As component systems were introduced, it was considered an advantage to standardise the sampling frequency to 13.5MHz for luminance, where the numbers worked out well for both systems, and to 6.75MHz for chrominance, but 13.5MHz was still considered a nominal 4x subcarrier. Hence 4:4:4, 4:2:2 etc.

Roll on to HD, and everything changes. How to describe colour space now? One possibility would be to say 4=720 samples, so a 1080 system should be based on 10.66. The EX would be 10.66:5.33:0 colour space. Hmmm. In practice, "4" was redefined to be 1920, and such as HDCAM become nice figures like 3:1:1.

Next come 720 systems such as Varicam, and here "4" gets redefined once again to be 960. DVCProHD extends to the 1080 system and let's redefine "4"
again as 1280! And if Sony really wanted consistency, shouldn't HDV and XDCAM-HD be referred to 3:1.5:0?

Noah - you are spot on when you say "4:2:0, 4:2:2 etc mean nothing out of context". Buzz - you need to inform your client he is not comparing like with like. To do so, he really needs to say "4:2:0 v 2.6:1.3:1.3" or "6:3:0 v 4:2:2"!!

Also worth noting is that 4:2:2 originally came about with an interlace system. Move to progressive, and it's advantages over 4:2:0 are lessened.

Marc Faletti · May 16th, 2008, 09:20 AM

I keyed 22 episodes of the digital series The Five Minute Drill (http://fiveminutedrill.tv) with the HVX200, and I was always battling noise, particularly in the blacks and in dark shirts. I often had to jump through some absurd hoops to get anything remotely clean. It was a top-notch green screen studio with great lighting; the shortcomings were entirely a product of the HVX.

Now, I shoot with the EX1 in a smaller studio with less professional lighting and it comes out 100x better with minimal effort. You can see a sample test at http://www.vimeo.com/810225 if you're curious. The Drill will be much easier this fall thanks to the bigger, badder chips on the EX1.

Gabe Strong · May 16th, 2008, 04:26 PM

Anyone can claim anything they want. I'd caution anyone that if at all possible they should look at a test of this with their own eyes rather than taking the word of 'internet experts'. I know this isn't always possible, but if you can, you owe it to yourself to do it. I know that because I saw greenscreen images and images in general from XDCam HD (I think it was from a F 350 not an EX 1 though) and the same images from a HVX200 on a calibrated HD monitor, you will see that those numbers (4:2:2 vs. 4:2:0) don't tell the whole story. There's a LOT of other variables of course, chip size, pixel count and such. Look at it for yourself! People do have different preferences.

May 15th, 2008, 09:43 PM	#1
Buzz Remde New Boot Join Date: Apr 2008 Location: LA and Phoenix Posts: 21	EX1 Vs. HVX200 with Green Screen I have been talking with a client about shooting a series of "Green Screen's" for a TV series. He is getting info about the HVX200 outperforming the EX1 in this area. 4:2:2 vs. 4:2:0 Any tips? You help is always appreciated.

May 16th, 2008, 05:42 AM	#5
Alister Chapman Inner Circle Join Date: Dec 2005 Location: Bracknell, Berkshire, UK Posts: 4,957	All I can add to this is that I have done some blue screen work with the EX1 and it keys very well. I used 1080P as I find the P colour space keys much better than interlace. I've been doing green/blue screen with XDCAM HD for a couple of years and it has always produce clean keys when lit properly. __________________ Alister Chapman, Film-Maker/Stormchaser http://www.xdcam-user.com/alisters-blog/ My XDCAM site and blog. http://www.hurricane-rig.com

May 16th, 2008, 06:47 AM	#6
Sebastien Thomas Major Player Join Date: Nov 2007 Location: Paris, France Posts: 210	Have been doing key also with footage from the EX1 without any problem. I would go with the EX1 over the HVX. When keying, NEVER shoot interlace !! Use progressive. Shoot at native resolution (1080p for the EX1). If in a studio, you can do live recording in 4:2:2 color space using the HD-SDI out of the EX1 plugged to an aquisition card (kona, blackmagic...) and get even better result. __________________ DoP, Video/Camera assistant http://www.lecentre.net/blog

May 15th, 2008, 10:07 PM	#2
Leonard Levy Trustee Join Date: Feb 2006 Location: USA Posts: 1,684	I can only tell you what I've read on the web and conversations I've had with people . The consensus is that the Ex-1 vastly outperforms the HVX200 for green screen. The HVX was never that good as it had comparatively lower resolution and was very noisy. The 4:2:0 thing seems to be over valued. Its a ratio of chroma sampling to luminance and the the resolution of the green channel is so much higher on the Ex at 1080 and the number of pixels so much greater that it at least equals the HVX in chroma info . Also you can shoot 4:2:2 if you can deal with the SDI output and in a few months with a Matrox II that should be easy. A number of people have said that it equals or actually outperforms an HDCAM on green screen. Adam Wilt told me that just in terms of specs it should. That's all heresay to but it convinced me. Go elsewhere on the web and search for Jim Arthurs whose tested this stuff extensively. Lenny Levy

May 15th, 2008, 10:12 PM	#3
Steven Thomas Trustee Join Date: Sep 2005 Location: Gilbert, AZ Posts: 1,896	The tip is - don't believe everything you hear. The EX1 is capable of 4:2:2 10 bit via its SDI-HD. Venture elsewhere on the net and you will find a thread by Jim Arthurs comparing the EX1 with the RED ONE. The test was a comparison on how well they key. The EX1 looked great. In fact the 1080P comparison to the RED was very close. The RED ONE 4K stuff is in another league beyond the EX1 and way beyond the HVX200.

May 16th, 2008, 05:32 AM	#4
Noah Yuan-Vogel Major Player Join Date: Dec 2004 Location: New York City Posts: 613	im going to agree with leonard and buzz, if i had to shoot a green screen i would choose the EX1 in a second over the HVX200. consider this: color sensor resolution, color output resolution of EX1: 1920:1080, 960x540 color sensor resolution, color output resolution of of HVX: 960x540, 640x1080 so basically, HVX would appear to have horizontal limiting resolution of 640, and vertical limiting resolution of 540 or less, and while the horizontal res is supersampled at the sensor, its scaled by 66% (less ideal than 50%). whereas EX1 gives you 960x540 limiting resolution (960x1080 over SDI) which is supersampled at the sensor by 200% in each direction. supersampling is important from what i understand because you get images that have better contrast/resolution at the limiting resolution than in systems where there is no supersampling. moral of the story is that it is simply by virtue of the fact that the EX1 DELIVERS as much or more detail than the HVX's SENSORS can hope to resolve, EX1 will most likely give you the best results. also, i have seen some odd color artifacting in the video from HVX200. im not sure if it is from the dvcprohd codec or because the image is being scaled in odd ways, but it scares me a bit. that said, i have used the hvx for greenscreen work and not the ex1. in my opinion, the HVX creates very soft images where the actual image resolution is far less than the delivered resolution, and that along with the noisiness of the sensor make it a pretty poor choice for greenscreen work. maybe im expecting too much from prosumer camcorders, but i dont think so. 4:2:0, 4:2:2 etc mean nothing out of context. you could be shooting in 4:4:4, but if your camera is SD (some refer to the HVX as an SD camera because of its sensor resolution) you could be in for some soft images depending on your final delivery specs. DVCPROHD is especially misleading since 4:2:2 sounds great but with its downsampled luma, it really should be called 2.6:1.3:1.3

May 16th, 2008, 09:20 AM	#8
Marc Faletti Regular Crew Join Date: Oct 2004 Location: Brooklyn, NY Posts: 50	I keyed 22 episodes of the digital series The Five Minute Drill (http://fiveminutedrill.tv) with the HVX200, and I was always battling noise, particularly in the blacks and in dark shirts. I often had to jump through some absurd hoops to get anything remotely clean. It was a top-notch green screen studio with great lighting; the shortcomings were entirely a product of the HVX. Now, I shoot with the EX1 in a smaller studio with less professional lighting and it comes out 100x better with minimal effort. You can see a sample test at http://www.vimeo.com/810225 if you're curious. The Drill will be much easier this fall thanks to the bigger, badder chips on the EX1.

May 16th, 2008, 04:26 PM	#9
Gabe Strong Major Player Join Date: Jul 2002 Location: Juneau, AK Posts: 814	Anyone can claim anything they want. I'd caution anyone that if at all possible they should look at a test of this with their own eyes rather than taking the word of 'internet experts'. I know this isn't always possible, but if you can, you owe it to yourself to do it. I know that because I saw greenscreen images and images in general from XDCam HD (I think it was from a F 350 not an EX 1 though) and the same images from a HVX200 on a calibrated HD monitor, you will see that those numbers (4:2:2 vs. 4:2:0) don't tell the whole story. There's a LOT of other variables of course, chip size, pixel count and such. Look at it for yourself! People do have different preferences.