Michael Galvan
April 9th, 2006, 12:43 AM
So I've done my testing today on some HDV 24F to various SD codecs for higher chroma resolution in SD. My results ... downrezzing HDV 4:2:0 to a SD codec does retain the full color information the codec is capable of ...
I took the same clip and downconverted to DV, DVCPro50, and PhotoJPEG (at 100%). THen I brought it into Final Cut Pro (with each clip in their own respective timelines with the same properties as the clips themselves) and looked at the individual Cr and Cb channels of each clip in a side by side comparison. And sure enough, the PhotoJPEG had the same chroma res as its luma res. The DvcPro50 file was 4:2:2 and the DV 4:1:1. I did apply the same various filter sets to each of the clips and rendered ... the PhotoJPEG and DVCPro50 held up much cleaner than the DV clip.
And to finally confirm my tests, I took the DV compressed clip and bumped it up to DVCPro50 and dropped it side by side with my "native" DVCPro50 clip. After looking at the individual Cr and Cb channels again, the DV-to-DVCPro50 clip, showed the same chroma resolution of the DV clip, even though it was encoded up to DVCPro50, proving that your footage isn't any better than the highest compression scheme it has gone through.
This is great, as I never expected this kind of SD workflow from my H1. You really can get 4:2:2 or even 4:4:4 (or at least close to it) by downconverting HDV to the right SD codec. THis is great for any work I put out on DVD. Thanks for everyone's insight on this.
I took the same clip and downconverted to DV, DVCPro50, and PhotoJPEG (at 100%). THen I brought it into Final Cut Pro (with each clip in their own respective timelines with the same properties as the clips themselves) and looked at the individual Cr and Cb channels of each clip in a side by side comparison. And sure enough, the PhotoJPEG had the same chroma res as its luma res. The DvcPro50 file was 4:2:2 and the DV 4:1:1. I did apply the same various filter sets to each of the clips and rendered ... the PhotoJPEG and DVCPro50 held up much cleaner than the DV clip.
And to finally confirm my tests, I took the DV compressed clip and bumped it up to DVCPro50 and dropped it side by side with my "native" DVCPro50 clip. After looking at the individual Cr and Cb channels again, the DV-to-DVCPro50 clip, showed the same chroma resolution of the DV clip, even though it was encoded up to DVCPro50, proving that your footage isn't any better than the highest compression scheme it has gone through.
This is great, as I never expected this kind of SD workflow from my H1. You really can get 4:2:2 or even 4:4:4 (or at least close to it) by downconverting HDV to the right SD codec. THis is great for any work I put out on DVD. Thanks for everyone's insight on this.