Hi Ravi, Do you have any update regarding this approach?
Thanks Jagan On Mon, Feb 16, 2009 at 6:20 PM, rktb <[email protected]> wrote: > > You might not be able to set one uncompressed format and one > compressed format on the same input component. However, let me get > back to you on this. This could probably be addressed. > > -Ravi > > On Feb 16, 2:53 pm, Jag Dev <[email protected]> wrote: > > Audio Encoder: > > > > File: AndroidAudioInput::getParametersSync() > > Can I also advertise more than One MIME support? I can return PVMF_PCM16 > and > > also PVMF_AMRNB? > > Use AndroidAudioInput::serParametersSync() to set PVMF_AMRNB as the > > configuration? > > > > Once I get the Encoded data back, I can use the fileoutputnode to write > the > > encoded data (no encode node needed) to the file? Does this suffice? > > > > If I will make the corresponding changes to the Audioflinger and other > > modules, which will directly read AMRNB encoded data from the driver? > > > > Thanks > > Jagan > > > > On Mon, Feb 16, 2009 at 11:40 AM, rktb <[email protected]> wrote: > > > > > On Feb 16, 11:03 am, Jag Dev <[email protected]> wrote: > > > > Hi Ravi, > > > > > > Thanks for your response. This is what I have in mind: > > > > > > Audio Encoder: > > > > ---------------------- > > > > > > If the data read from the hardware is compressed (already in the > required > > > > encoded format), then the MIO can directly use this data to be > written to > > > a > > > > file. > > > > > > MIO (Compressed input) -> Composer Node -> File. In this case we dont > > > need a > > > > Encoder node to be instantiated. > > > > > > Can you please let me know how we can specify to the MIO / Composer > that > > > the > > > > Author is receiving a compressed data, that needs no encoding? > > > > > The MIO that you are going to use as part of AddDataSource() should > > > advertize the format it can support. E.g., take a look at > > > android_audio_input.cpp. The call to > > > AndroidAudioInput::getParametersSync() for the key > > > OUTPUT_FORMATS_CAP_QUERY is returning the mime type "PVMF_MIME_PCM16". > > > Similarly, the MIO that you are going to add should advertise the > > > compressed format that it can support, e.g., PVMF_MIME_AMR. > > > > > > Audio Decoder: > > > > > > Right now, as you said: > > > > > > PVPlayerEngine -> DataPath -> PVOMXDecNode : Decoder Node (Decoded > data) > > > -> > > > > MIO component -> Hardware. > > > > > > Pre-requisite: > > > > --------------------- > > > > > > If I support a tunnel mode OMX component: > > > > > > PVPlayerEngine -> DataPath -> PVOMXDecNode : Decoder Node -> OMX > > > Component > > > > -> Hardware. > > > > > > WIll this work? Without using the output MIO and no decoded data back > to > > > the > > > > MIO output is it possible to acheive the app working? The only issue > here > > > is > > > > that the Clock is not getting updated to the application. This > results in > > > > the application not updating the UI (We dont see any progress bar > > > updation > > > > or not even the Music playback screen). > > > > > > Can you please suggest if my understanding is right? and also some > > > pointers > > > > if we can achieve this? > > > > > I don't think this is possible. You need to pass data out through the > > > mediaoutputnode, by making your MIO accept compressed data. > > > > > > Thanks > > > > Jagan > > > > > > On Mon, Feb 16, 2009 at 6:46 AM, rktb <[email protected]> wrote: > > > > > > > Hi, > > > > > > > I don't think we have a direct support for tunneling in OpenCORE > 2.0. > > > > > There may be ways to achieve it with some mods, but nothing that I > > > > > know which is straightforward. > > > > > > > Typically we would prefer the codecs to lie under our OMX node. > This > > > > > is the regular mode of operation ... SourceNode --> OMXDecoderNode > --> > > > > > MediaOutputNode. However, there could be cases where it is > preferable > > > > > to remove the OMXDecoderNode and send in compressed data directly > to a > > > > > MIO that can decode + render at the same time. For this, one would > > > > > probably need to write a MIO that would interface with the > > > > > MediaOutputNode using the MIOControl and MediaTransfer interfaces > on > > > > > one side and on with the OMX components on the other side. > > > > > > > Please watch out for upcoming posts that will discuss in detail the > PV > > > > > plans to support tunneling mode. > > > > > > > -Ravi > > > > > > > On Feb 15, 8:19 pm, Jag Dev <[email protected]> wrote: > > > > > > Hi Ravi, > > > > > > > > Can you please comment on this? If we can achieve tunneled mode > > > encoders > > > > > and > > > > > > decoders using the following approach: > > > > > > > > Can you please suggest me if I implement my own MIOs can I > acheive > > > > > tunneled > > > > > > decdoding and encoding for audio? > > > > > > > > > 1. My understanding was MIOs are used for writing to hardware > > > (decoded > > > > > > > frames) and Reading from hardware (frames to be encoded)? > > > > > > > > > 2. Decoding and Encoding should still happen using PVMF nodes > > > (PVData > > > > > paths > > > > > > > and using PV player engine and Pv Nodes)? > > > > > > > > > Please correct me if I am wrong? > > > > > > > > > Thanks for your time. > > > > > > > Jagan > > > > > > > > > On Sat, Feb 14, 2009 at 10:02 AM, Dave Sparks < > > > > > [email protected]>wrote: > > > > > > > > >> My understanding is that PV is working on improving support > for > > > > > > >> tunneled mode. I'm not sure when you'll see those changes in > the > > > open > > > > > > >> source tree, perhaps a PV representative can comment. > > > > > > > > >> On Feb 13, 9:03 pm, Jag Dev <[email protected]> wrote: > > > > > > >> > Hi, > > > > > > > > >> > I am planning to implement tunneled mode for only audio > decoders > > > and > > > > > > >> > encoders. I see that the PVOMXAudioDecNode always expects > > > > > > >> FillBufferDones. > > > > > > > > >> > The behavior is that the decoding is going through fine > (using > > > the > > > > > OMX > > > > > > >> > decoders), but the UI is not getting updated? > > > > > > > > >> > Thanks > > > > > > >> > Jagan > > > > > > > > >> > On Thu, Feb 12, 2009 at 7:29 PM, Dave Sparks < > > > > > [email protected] > > > > > > >> >wrote: > > > > > > > > >> > > Are the frames arriving in the MIO > > > (android_surface_output.cpp)? > > > > > If > > > > > > >> > > so, do you have appropriate code in there to deliver the > > > frames to > > > > > > >> > > your video device? > > > > > > > > >> > > On Feb 12, 10:34 am, Jag Dev <[email protected]> wrote: > > > > > > >> > > > Hi, > > > > > > > > >> > > > 1. Do we have support for MIO (Modular I/O library) in > > > android? > > > > > > > > >> > > > 2. For now we have support only for non-tunneled mode > > > encoders > > > > > and > > > > > > >> > > decoders > > > > > > >> > > > support in PV opencore. Is there anyway we can enable > > > tunneled > > > > > mode > > > > > > >> > > > operation? I tried that, but the application is not > behaving > > > > > > >> properly > > > > > > >> > > (the > > > > > > >> > > > UI updates are not happening) > > > > > > > > >> > > > Thanks > > > > > > >> > > > Jag > > > > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "android-framework" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/android-framework?hl=en -~----------~----~----~----~------~----~------~--~---
