BobDiaz
The Professor
There's a great deal of interest in the video community regarding the new Sony PMW EX-1 HD Camera. The big questions from many is, should I get this camera rather than the Panasonic HVX-200? Forums like this one are an ideal place to share information. While I'm no expert on this camera, I'll tell you what I do know and address areas that I'm unsure about.
My 3 basic guidelines are:
(1) The perfect video camera does NOT exist. Every camera is a tradeoff between different things.
(2) Paper specification can only tell you so much; the real test is in a side by side comparison.
(3) Everyone has different views as to that is critical to them. Thus, what may be seen as minor to one person, may be major to another.
DATA RATES:
The Panasonic HVX-200 records at up to 100Mb/second in 1080,60i/30p and in 720,60p. For other modes, like 720,24P, the data rate is lower. I'm unaware of anyone reporting problems with compression noise with the HVX-200 data rates.
It is possible that someone might have a gripe that the HDPRO100 format does sub-sample the HD image. That is, 1920h x 1080v is recorded as 1280h x 1080v and 1280h x 720v is recorded as 960h x 720v. The key advantage to the HDPRO100 format is that every frame of video sits on its own and is not compressed with the other frames in an MPEG format.
The Sony EX-1 has lower data rates of 25Mb/s and 35 Mb/s. It does record the full HD resolutions of 1920h x 1080v and 1280h x 720v at the higher data rate of 35Mb/s. However, in order to do so at 35Mb/s, it must use MPEG-2, Long GOP compression.
This turns out to be both a blessing and a curse at the same time. Users of the Sony Z1 discovered that a very fast pan could overload the CODEC with too much data and generate compression noise blocks. The Z1 recorded a sub-sampled HD image (1440h x 1080v) to tape at 25Mb/s and while MPEG-2 does very well when the number of changes in the image aren't too great, it does fall apart when the number of changes are excessive.
Not all CODECs are equal, so some do a better job of compressing the data than others. The higher the recorded data rate, the less likely that compression noise could appear; even if you do a rapid pan and try to overload the CODEC.
You might assume that with a HQ mode of 35Mb/s, this is an increase of 1.4x the number of bits per second compared to 25Mb/s found on the Sony Z1. However, the horizontal recorded resolution also increases between the Z1 (and any other 1080 HDV camera) from 1440h at 25Mb/s to 1920h at 35Mb/s. This means that the horizontal data increases by a factor of 1.3333...x.
When we take into account the increased horizontal resolution of the HQ, 35Mb/s mode, the data rate only increases by 1.05x. While we do increase the recorded horizontal resolution in the HQ mode on the EX-1, we are not lowering the compression ratio very much compared to the 25Mb/s SP mode.
Does this matter? It all depends on what you are shooting and how good is the CODEC inside the EX-1. Unfortunately, I can't say if this is a big issue or not. We'll have to wait for feedback from users to know if the EX-1 has a problem with excessive motion or not.
The good news is that if your shoot at 1080, 24P, the amount of data you generate is about 80% that of 1080, 60i/30p. Also, if you shoot at 720, 60p, the amount of data generated is 89% of the 1080, 60i/30p mode. The higher progressive frame rate would be better for motion anyway. If you drop the frame rate to 720, 24P, the data rate would be only 36% of the highest mode.
The lower resolutions and frame rates are less likely to encounter compression noise due to overloading the CODEC with too much data. One interesting factor is that at 35Mb/s, the camera records in VBR (Variable Bit Rate) mode. Thus, if it needs the higher data rate, to avoid compression noise, the camera will take it, BUT if the higher data compression rate isn't needed, the camera will record at a lower data rate. Thus, with some shots, you may find you time is longer than 50 minutes per 16GB in HQ mode.
IMAGE QUALITY:
To me the images from the Sony Ex-1 looked very good. Of interest is the following WEB page:
http://www.genkosha.co.jp/vs/sp/200712/
I took parts from each image and placed them in an 800x600 image for you to see the differences. In the first image below, we see the Panasonic HVX-200 on top and the Sony EX-1 below. Both look good, but the Sony seems to be sharper, just look at the print on the bottle on the left of the image. However, when I look at the colors, the HVX-200's colors just looks richer.
EDIT ADDED; IMPORTANT: When I wrote this I failed to point out that the images for the HVX looked a bit off. Because I pulled someone else's work, I really don't know how much effort they took, if any, to make the HVX look proper. In retrospect, to me it looks like whoever generated the original images, did a poor job. So take the following images with a grain of salt
In the 720, 30p mode, the sharpness difference is not as much as the 1080p mode. However, the colors on the HVX are still richer than the EX-1.
LOW LIGHT PERFORMANCE:
The specification sheet for the EX-1 says, 0.14 lux. However most people miss the fact that the long shutter must be used or an exposure time of 1.07 seconds. OK, for a still shot, but not OK for anything with movement.
If we take into account an exposure like 1/24 of a second, the minimum illumination would be, 3.6 lux. This is slightly above the HVX's specification of 3 lux. However, this is where a paper specification can be misleading.
To me the EX-1's image noise seems less at +18dB of gain than the HXV appeared at +12dB of gain. I could be wrong about this, because I was not able to do a side by side comparison. Until someone does a side by side test, my best guess is that the EX-1 may do slightly better in very low lighting conditions.
In trying to point the camera to a dark area in the booth, the image still looked OK at +18dB of gain. It had a lot of noise, but was still useable; still some may limit the gain to +12 or even +6 dB gain.
IN CLOSING:
There's so much more I could say about the camera, but it's getting late and I've written almost 3 pares of comments on my word processor.
Let me just close by saying, that the EX-1 camera appears to be very interesting in terms of what it has to offer. Should someone get it instead of the Panasonic HVX-200? That's not so easy to answer, we already know the weak and strong point of the HVX, but the EX-1 is yet to be fully tested in the field. So, for now, the jury is out on that issue...
Bob Diaz
My 3 basic guidelines are:
(1) The perfect video camera does NOT exist. Every camera is a tradeoff between different things.
(2) Paper specification can only tell you so much; the real test is in a side by side comparison.
(3) Everyone has different views as to that is critical to them. Thus, what may be seen as minor to one person, may be major to another.
DATA RATES:
The Panasonic HVX-200 records at up to 100Mb/second in 1080,60i/30p and in 720,60p. For other modes, like 720,24P, the data rate is lower. I'm unaware of anyone reporting problems with compression noise with the HVX-200 data rates.
It is possible that someone might have a gripe that the HDPRO100 format does sub-sample the HD image. That is, 1920h x 1080v is recorded as 1280h x 1080v and 1280h x 720v is recorded as 960h x 720v. The key advantage to the HDPRO100 format is that every frame of video sits on its own and is not compressed with the other frames in an MPEG format.
The Sony EX-1 has lower data rates of 25Mb/s and 35 Mb/s. It does record the full HD resolutions of 1920h x 1080v and 1280h x 720v at the higher data rate of 35Mb/s. However, in order to do so at 35Mb/s, it must use MPEG-2, Long GOP compression.
This turns out to be both a blessing and a curse at the same time. Users of the Sony Z1 discovered that a very fast pan could overload the CODEC with too much data and generate compression noise blocks. The Z1 recorded a sub-sampled HD image (1440h x 1080v) to tape at 25Mb/s and while MPEG-2 does very well when the number of changes in the image aren't too great, it does fall apart when the number of changes are excessive.
Not all CODECs are equal, so some do a better job of compressing the data than others. The higher the recorded data rate, the less likely that compression noise could appear; even if you do a rapid pan and try to overload the CODEC.
You might assume that with a HQ mode of 35Mb/s, this is an increase of 1.4x the number of bits per second compared to 25Mb/s found on the Sony Z1. However, the horizontal recorded resolution also increases between the Z1 (and any other 1080 HDV camera) from 1440h at 25Mb/s to 1920h at 35Mb/s. This means that the horizontal data increases by a factor of 1.3333...x.
When we take into account the increased horizontal resolution of the HQ, 35Mb/s mode, the data rate only increases by 1.05x. While we do increase the recorded horizontal resolution in the HQ mode on the EX-1, we are not lowering the compression ratio very much compared to the 25Mb/s SP mode.
Does this matter? It all depends on what you are shooting and how good is the CODEC inside the EX-1. Unfortunately, I can't say if this is a big issue or not. We'll have to wait for feedback from users to know if the EX-1 has a problem with excessive motion or not.
The good news is that if your shoot at 1080, 24P, the amount of data you generate is about 80% that of 1080, 60i/30p. Also, if you shoot at 720, 60p, the amount of data generated is 89% of the 1080, 60i/30p mode. The higher progressive frame rate would be better for motion anyway. If you drop the frame rate to 720, 24P, the data rate would be only 36% of the highest mode.
The lower resolutions and frame rates are less likely to encounter compression noise due to overloading the CODEC with too much data. One interesting factor is that at 35Mb/s, the camera records in VBR (Variable Bit Rate) mode. Thus, if it needs the higher data rate, to avoid compression noise, the camera will take it, BUT if the higher data compression rate isn't needed, the camera will record at a lower data rate. Thus, with some shots, you may find you time is longer than 50 minutes per 16GB in HQ mode.
IMAGE QUALITY:
To me the images from the Sony Ex-1 looked very good. Of interest is the following WEB page:
http://www.genkosha.co.jp/vs/sp/200712/
I took parts from each image and placed them in an 800x600 image for you to see the differences. In the first image below, we see the Panasonic HVX-200 on top and the Sony EX-1 below. Both look good, but the Sony seems to be sharper, just look at the print on the bottle on the left of the image. However, when I look at the colors, the HVX-200's colors just looks richer.
EDIT ADDED; IMPORTANT: When I wrote this I failed to point out that the images for the HVX looked a bit off. Because I pulled someone else's work, I really don't know how much effort they took, if any, to make the HVX look proper. In retrospect, to me it looks like whoever generated the original images, did a poor job. So take the following images with a grain of salt
In the 720, 30p mode, the sharpness difference is not as much as the 1080p mode. However, the colors on the HVX are still richer than the EX-1.
LOW LIGHT PERFORMANCE:
The specification sheet for the EX-1 says, 0.14 lux. However most people miss the fact that the long shutter must be used or an exposure time of 1.07 seconds. OK, for a still shot, but not OK for anything with movement.
If we take into account an exposure like 1/24 of a second, the minimum illumination would be, 3.6 lux. This is slightly above the HVX's specification of 3 lux. However, this is where a paper specification can be misleading.
To me the EX-1's image noise seems less at +18dB of gain than the HXV appeared at +12dB of gain. I could be wrong about this, because I was not able to do a side by side comparison. Until someone does a side by side test, my best guess is that the EX-1 may do slightly better in very low lighting conditions.
In trying to point the camera to a dark area in the booth, the image still looked OK at +18dB of gain. It had a lot of noise, but was still useable; still some may limit the gain to +12 or even +6 dB gain.
IN CLOSING:
There's so much more I could say about the camera, but it's getting late and I've written almost 3 pares of comments on my word processor.
Let me just close by saying, that the EX-1 camera appears to be very interesting in terms of what it has to offer. Should someone get it instead of the Panasonic HVX-200? That's not so easy to answer, we already know the weak and strong point of the HVX, but the EX-1 is yet to be fully tested in the field. So, for now, the jury is out on that issue...
Bob Diaz
Last edited: