|
|||||||||
![]() |
|
Thread Tools | Search this Thread |
![]() |
#1 |
Trustee
Join Date: Mar 2004
Location: Milwaukee, WI
Posts: 1,719
|
Canon H1 analog or SDI uncompressed?
I have been trying to find out an answer to this question for a long time now.
Lets say you have a Decklink Multibridge extreme that has both component and SDI inputs. We all know that SDI foolows a strict digital format of digital 4:2:2 samples at 1920x1080. Analog however cannot really be measured by the same 4:2:2 because it isn't digital. While the analog signal may be close to 4:2:2 in terms of raw detail and signal strength it isn't exactly the same. Based on this does anybody see any advantage to using component outputs on the H1 to capture 4:4:4 RGB on the Decklink Extreme? Could an analog HD signal offer just a little bit more detail compared to a digital 4:2:2? I understand all of the drawbacks to analog such as signal noise and generation loss. I also understand that in no way would the analog equal a true 4:4:4 digital image. What I want to know is if the translation from an analog signal to a 4:4:4 digital signal could give a little bit more detail compared to just a 4:2:2 signal. I am mainly looking for opinions on the subject and to see if anybody that has a Decklink Multibridge Extreme can compare the two formats. The same could also apply to any HDV camera and choosing to either capture YUV or RGB and if RGB could gain anything even if small. |
![]() |
![]() |
![]() |
#2 |
Major Player
Join Date: Mar 2005
Location: UT
Posts: 945
|
I only have a Kona LH, but theoretically, capturing live camera head analog output into PhotoJPEG 1920x1080 at 100% should be an HD 4.4.4. image. Am I mistaken? I have serious doubts that it will look any different than the 4.2.2 HD formats, but it would be interesting to see. Might actually be better for greenscreen.
|
![]() |
![]() |
![]() |
#3 |
Trustee
Join Date: Mar 2004
Location: Milwaukee, WI
Posts: 1,719
|
Well I figured that since an analog signal is... well analog and a wave of information and not binary that the details are smooth gradations instead of blocks of pixels. This might translate into a analog to digital conversion at RGB 4:4:4 having a smoother transition between chroma pixels and not just a copy of every other horizontal pixel. It may or may not have more detail but smoother edges for keying work.
When the 12 bit analog to digital converter converts to RGB would it not look at the change in the analog wave strength to figure out the full 1920x1080 chroma spectrum. Based on the curves of the wavelengths couldn't the in between chroma pixels actually contain more accurate information than just chucking them out and duplicating the pixel right before it? how exactly does the chroma reduction work in analog anyways since it doesn't reduce pixels? I guess this would all depend on if the DSP in the H1 even sends out a full analog signal to component or if it has been already converted to an SDI signal in the DSP and then sent to the component. Hopefully somebody with a Multibridge can test this out. |
![]() |
![]() |
![]() |
| ||||||
|
|