[x265] P-frames referencing B-frames, Generalised B-frames in HEVC

[Roshantha Mendis]

Okay thank you for clearing that. Someone in the doom9 forum said its
legal, but not efficient, so had to check. Thank you !

On 26 October 2015 at 08:47, Deepthi Nandakumar <
deepthi at multicorewareinc.com> wrote:

> Mario - Yes, and P slices use only I-slices or other P-slices as reference.
>
> On Mon, Oct 26, 2015 at 2:04 PM, Mario *LigH* Rohkrämer <contact at ligh.de>
> wrote:
>
>> The question was about P slices.
>>
>> I would be surprised if P slices were allowed to have B slices as
>> reference, which are below in hierarchy. From what I know (which may be
>> partially limited to AVC), usually only more dependent types can reference
>> less dependent types, to avoid error accumulation (quality reduction and
>> even possible data errors with decoding artefacts in slices used as a
>> reference will remain in all dependent slices):
>>
>> I(ntra) slices are independent, have no references (prediction generation
>> 0).
>>
>> P(redicted) slices can have I slices as reference (prediction generation
>> 1), possibly also previous P slices (not sure about this).
>>
>> B(idirectionally predicted) slices can have I and P slices as reference
>> (prediction generation 2), and in a pyramid hierarchy even B slices which
>> have no other B slices as possible reference.
>>
>> The more dependent a slice is, the more loss of quality may accumulate
>> with each further prediction generation, therefore it should only be
>> visible for a short time and less likely be used as reference for even more
>> dependent slices and even more accumulation of quality loss. Because P
>> slices could potentially be a reference of many dependent slices, it
>> wouldn't make sense to allow them to have references in B slices which are
>> already predicted in second or even third generation, compared to an
>> original I(DR) slice of the current GOP.
>>
>> Disclaimer: I may not yet know about exceptions in HEVC. The statement
>> above is based on knowledge from MPEG-1/2 to AVC.
>>
>>
>>
>> Am 26.10.2015, 04:31 Uhr, schrieb Aarthi Priya Thirumalai <
>> aarthi at multicorewareinc.com>:
>>
>> Hi,
>>>
>>> When b-pyramid is on, then b slices can refer to both B , P and I slices.
>>> We are creating a hierarchy among B frames, that is some B frames can be
>>> used as reference for other b frames.
>>>
>>> Regards
>>> Aarthi
>>> On Oct 26, 2015 3:23 AM, "Roshantha Mendis" <hrm506 at york.ac.uk> wrote:
>>>
>>> Hi,
>>>>
>>>> When --b-pyramid  is set to true, can P-SLICES refer to B-SLICES ?
>>>>
>>>> I haven't come across a video sequence that this happens, but just want
>>>> to
>>>> know if this can happen.
>>>>
>>>> Also if this situation can occur, then is it valid according to the HEVC
>>>> specs ?
>>>>
>>>> Many thanks
>>>>
>>>> --
>>>> Rosh Mendis
>>>> Research Student - EngD in LSCITS
>>>> Dept. of Computer Science
>>>> University of York
>>>>
>>>> Disclaimer: http://www.york.ac.uk/docs/disclaimer/email.htm
>>>>
>>>> _______________________________________________
>>>> x265-devel mailing list
>>>> x265-devel at videolan.org
>>>> https://mailman.videolan.org/listinfo/x265-devel
>>>>
>>>>
>>>>
>>
>> --
>>
>> Fun and success!
>> Mario *LigH* Rohkrämer
>> mailto:contact at ligh.de
>>
>>
>> _______________________________________________
>> x265-devel mailing list
>> x265-devel at videolan.org
>> https://mailman.videolan.org/listinfo/x265-devel
>>
>
>
>
> --
> Deepthi Nandakumar
> Engineering Manager, x265
> Multicoreware, Inc
>
> _______________________________________________
> x265-devel mailing list
> x265-devel at videolan.org
> https://mailman.videolan.org/listinfo/x265-devel
>
>


-- 
Rosh Mendis
Research Student - EngD in LSCITS
Dept. of Computer Science
University of York

Disclaimer: http://www.york.ac.uk/docs/disclaimer/email.htm
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20151026/7a549ba3/attachment.html>