ip输出编码比特流的流向该写回memory吗？
-
@karthik_tdk 不好意思，关于编码好的比特流是否应该写会mem我并没有明确的答案，编码好的比特流好像并不包含解码所需的sps头等，如果写回到mem我们后序应该怎样处理呢？第二个问题我的理解是load的参考帧应该是上一次store的，但仿真中似乎并不是这样，而是直接给出了参考帧的yuv文件，对于这个文件的来源我也感到奇怪，如果你有好的想法可以告诉我，谢谢
-
@damian
你好呀
基本上，我从规范和白皮书中了解到，我研究了输入应该采用 .yuv 格式，并且第一帧始终是 I 帧
，在编码器中重建 I 帧后，它用作 P 或 B 帧的参考帧，但在这个模拟中，他们最初采用 P 帧（但不知道为什么），就像从 fetch_p_cur.yuv（版本 1 代码）中获取一样,如果理解 H265 的代码，如果对该代码有疑问，请指导我，这对我来说非常重要。
顺便问一下，你看得懂英语吗？Hi there
Basically what, I understood from specifications and white papers I studied the input should be in .yuv format and in that the 1st frame is always a I Frame
, after reconstruction of that I Frame in encoder, it used as reference frame for P Or B frame , but in this simulation they are taking P frame initially (but don't know why) like fetching from fetch_p_cur.yuv(version 1 code) ,And if understand the code of H265 Please guide me if have some doubts in that code, it very important to me.
By the way do you understand English? -
@karthik_tdk I am a beginner in video encoding technology, and I have just come across this open-source project. I haven't studied it in depth, so I can't give guidance to others; besides, I know a little English.
-
@karthik_tdk In the v2.0 code, it seems that the current frame can be determined as an I-frame or a P-frame by configuring the IP input signal, sys_type_i. When it's 0, the current frame is an I-frame, and when it's 1, the current frame is a P-frame. There is a PPT that briefly introduces IP. You can check it out on this website.
-
@damian
It's okay I understood,and I am also in same situation, so is it ok with u if we help each other? What are knowledge we have!,
And in which discussion the ppt present -
This post is deleted! -
想知道V1.0中也是会出现这种情况吗
-
@wupang 我没有关注v1.0，那边的处理流程怎么样呢
-
@damian 刚准备down一下看看，可能差不多，有类似情况我回来回复。by the way，请问你有2.0的相关文档说明吗？只看代码有点太多。
-
@damian it's same but they solved some bugs and added new block , I think.