![]() ![]() At the right of the screen, you see the same image but in a YUV breakdown. On the left side of the screen you see the RGG version of an image, you see the breakdown of the red, green and blue. I was reading several pages (doing some research on this matter) and I think I have to mention that YUV is not equal to YCbCr, the naming just stuck as the official name like… not go there). If you want to map the YUV back to the old days, YPbPr colour model was used in analogue component the video digital version, we used, later on, is YCbCr. ![]() The Y is the brightness of the picture and the UV add the colour to the picture. YUV can be split into the Y which is the luminance and UV which are the Chromian components. Together with the Chroma Subsampling, there is also the YUV format. Copyright Wikipedia YUV Copyright Wikipedia Copyright Wikipedia The differences between samplings are explained, one of the examples they use is seen here. If you want more information on 4:4:4 and the other samplings look at Wikipedia for one or Adobe blog and you find a ton of information. It was one of the reasons many customers still stayed on PCoIP even though VMware is putting all their effort in Blast. VMware realized that also and the first versions for Blast were using the not so optimal 4:2:2 YUV sampling. For text, however, you need good quality. For a lot of scenario’s lesser sampling will do fine. If you go further back all the way to 3:1:1 you would end up with about 40% of the file size but you would lose quality fast.Īs you see the sampling rate defines the size and the quality, I guess you’re not surprised so far. So if your image is 100MB in size delivered with 4:4:4, with 4:2:2 it would be 70MB. 4:2:2 (which VMware used to use with H.264 and Blast is around 70% the size of what a 4:4:4 sampling will deliver. The cost of this result is that the file size is larger. If you look at pixel density 4:2:2 is far better up to +30% from 3:1:1.Ĥ:4:4 is an uncompressed option of Chroma Subsampling and will deliver an even better result. 3:1:1 is offering more colour information than e.g 4:2:0 or 4:2:2 (not with 4:4:4). You could do a math on the differences and you would see that e.g. PAL, HDTV and Internet video are using 4:2:0 Subsampling. If we look back a bit we see that you had 3:1:1 with the old HDCAM. There are more Chroma Subsampling options, 4:4:2 and 4:4:4 are not the only two. Putting this all together will get you a good picture, sharp with the right colour and without bandwidth issues. In the YUV format luminance is ever further reduced and only uses one-third of the original size. By reducing the amount of colour you can get the picture transported and displayed and you make sure the size is reduced by 50%. With Chroma Subsampling you make sure luminance is sent and you drop some of the colour. Colour does not make a picture sharper, just nicer perhaps. So you want to make sure the luminance of the signal is not harmed as it makes the picture “sharp”. Luminance is the contrast of the picture.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |