Color Sampling: EX1 & EX3

Terry_Lasater

Terry_Lasater

All about the Caption
I'm sure this info exists within this forum, but I haven't stumbled across it yet. Can someone tell me the maximum color sampling of the EX1 and EX3 cameras?

Thanks in advance!
 
Thanks for the reply, booth.

btw, do you happen to know the color sampling for the HPX170? Is it lesser when recording to P2 than when using the SDI output?

Sorry for the noob-like questions but I'm still operating in SD Land...
 
Fortunately there is less impact of 4:2:2 vs 4:2:0 in HD compared to SD. Especially when you have full raster 4:2:0 sampling vs upscaled 4:2:2 sampling such as in the HVX series.
 
GuyB said:
Fortunately there is less impact of 4:2:2 vs 4:2:0 in HD compared to SD. Especially when you have full raster 4:2:0 sampling vs upscaled 4:2:2 sampling such as in the HVX series.
Click to expand...

Thanks for your post, Guy! That is interesting, but a bit confusing. Would you care to explain that in a bit more detail?

fwiw, I still use my DVX100A for my own projects. I do some work for a production company that owns an EX1. I also use a Sony DSR-500 and a few Beta SP cameras. I've used the EX1 several times but I'm not involved with their post, so I don't know much about the edit workflow.
 
He's saying that on the EX1, the 4:2:0 color subsampling is comparable to the HVX's 4:2:2, seeing that the EX1 records at 1920x1080 and the HVX downsamples to 1280x1080. If you do the math (about 1/4 of the EX1's 1920x1080, and 1/2 of the HVX's 1280x1080), you'll see that the number of chroma pixels recorded on the EX1 and HVX are close.
 
also, because in HD each pixel is a much smaller sample of the image compared to SD, the variations in colour between these samples, on average, will be less, so 4:2:0 fares better in HD than in SD.

An oversimplified way of looking at it is the more samples you have of the image, the less important each sample is to the image as a whole.

Also, the HVX also has a sensor with less resolution than the codec it records in (1440x1080 in 25fps DVCProHD) and the sensor is no where near that resolution). Therefore an amount of the sampling information stored in the 4:2:2 encoding is extrapolated data, not real data.

So in the end there are a lot more things to consider than just quoted colour space of the codec when comparing two competing cameras.
 
Seriously! I looked it up and your last thread outside the cafe (besides two or three in the marketplace) was in june 2007 :D This one: Bliss Lights!

Ok, enough thread jack :evil:

PS. I really DO appreciate your posts in the cafe, so more power to you :beer:
 
Another important factor that makes 4:2:0 look acceptable is Progressive Scan. The vertical subsampling is inherently better in Progressive, further lessening the differences between 4:2:0 and 4:2:2. However, if you shoot interlaced, the differences are quite dramatic!
 
You must log in or register to reply here.
Back
Top