H264 Decoder buffering 8 frames even in low latency mode

Hi,
I am trying to do hardware video decode in my application. It does H264 decoding using IMFTransform::ProcessInput and IMFTransform::ProcessOutput API. I set CODECAPI_AVLowLatencyMode to avoid any kind of buffering. The app works, however, I see that
ProcessOutput keeps returning MF_E_TRANSFORM_NEED_MORE_INPUT for the first 8 frames. Only after 8 frames are sent, I get a success and I get decoded output. My decode latency is still high since I am receiving an output 8 frames late.
Any ideas on how to solve this issue?

Ping..
Any help in this regards is highly appreciated.

Similar Messages

Low latency mode

Hi,
I'm doing vocal recording now,facing latency problem.Finally i'm trying to apply the LOW LATENCY MODE ,it really solve the latency when press the Rec Enable button...but howevery i try to route with Aux or Bus(Vocalist required some reverbs while singing)...it doesn't work..and level meter didn't show any input signal on the Aux or Bus!!!
and when i apply the Low Latency from the Vocal track Send button...level meter become active on Aux or Bus but the LATENCY appear again even the Low Latency Mode i still ON it.
What's my problem here...
Urgent!!!!Thanks for helps...

the better setting in recording (with Motu device) is:
Don't use "low latency mode" (it is a logic 8 feuatures but it is not so useful)
Sample buffer set to 256k
PDC set to Software and Instruments (avoid to use Aux or Multi instruments mode)
Reverbs and delay can be used on the Buss because latency on reverbs gives you a good predelay effect... (when PDC is set to All, usually, I use sample delay in reverber Buss in order to be able to have the same great reverbs sonds )
... PDC set to ALL gives you a lot of latency...
if you use Multi Instruments.. the better way is: (I never use Multi out instrument... i prefer to export as audio the parts!
Bounce the instrumental parts...
open a new project, import the instrumental parts... set PCD to "Track and Instrument)
When Mixing .. mastering PDC must be set to ALL...
in every other situation (expecially in recording phase) "track and instruments" PDC setting is the better choice!
G

Low latency mode didn't bypass latent plugins, or turn them orange

in our case it was waves' bass rider plugin. we had low latency mode enabled, the plugin was bypassed even, and still it caused the entire output to be latent.
i was under the impression that low latency mode was supposed to bypass these things and turn them orange right?
our latency limit was set to zero as well.
ideas?

There are so many factors that affect latencey and if you're making your living or it''s really important, I'd roll up my sleeves and tackle this puppy head on. Do you have a really high track count? Are you running a lot of plug ins? What is the disk buffer size set to? What do you have the processing threads set to? How much RAM do you have? Are you using an external drive that's fighting to keep up? What sample rate are you running at........
I think you see where I'm going with this! It's a slippery slope my friend but nothing you can't figure out. It really all depends on understanding what you have and how best to optimize it for what you want to do. That's why a lot of the help forums can be so confusing because there isn't a simple answer that works for everyone.
In the meantime, if you really wanna just get to the business of making music try monitoring through your desk when you're tracking and only use the plugins for playback (disable software monitoring, but only if you have the hardware that will enable you to do so).
Good luck!

Low latency mode doesn t work?

If I enable "Low latency Mode" nothing happens, Ioaded tons of Plugins and set the limit to 1 ms....Shouldnt it bypass some plugins? Someone who knows the problem

does it work for you?

Can't send to bus when in low latency mode?

My colleague and I have almost the same setup, except he is running OS X 10.8.5, and I'm running OS X 10.9.4, Both are running Logic X 10.0.7. He has Apogee hardware (Symphony 64 card to DA16X/AD16X), and I have a UA Apollo.
For some reason, this works on my setup, but not on his:
Set up a bus send on any audio or VI channel, and assign the output of that channel to outputs 3-4 (outputs 1-2 are both of our main outs). The purpose of this is to send to a cue/headphone box.
Set latency compensation to Instruments and Tracks (not ALL, setting it to ALL defeats this)
Turn on low-latency mode.
Turn on Software Monitoring (buffer is set to 64)
Record enable that track.
On my setup, I get signal on channels 3-4, on his, no signal as soon as record is enabled
Auto-input monitoring is off in both cases.
On my setup, the bus popup turns orange, but I still hear the signal, on his, it turns orange, but no signal.
All plugins removed on the track and the bus.
Weird.

Bus Send is set the same on both systems, Post Pan, Post Fader, Pre Fader ?
Good point, I'll check that
Bus is set to output on 3-4 ?
Yes, on both systems
Bus level is the same ?
Yes, similar anyway
Sure you friend doesn't have any plugins on the master output bus ?
I didn't consider that - somehow didn't think that plugins on the master bus (which is outputting on 1-2) could effect a different bus. Will check that tomorrow.
Do you by any chance have direct monitoring enabled on the Apollo, any input signal will be heard.
I do use direct monitoring on the Apollo, but that is not a factor - it's prerecorded material we are trying to monitor, not input signal.
Curious as to why are you running in Low Latency mode, it is for a specific purpose not something to be used in general recording/work.
He is running on an older Mac tower and uses a lot of plugins and prefers software monitoring, and finds that he has problems unless he enables low latency mode. I usually leave it off, I just wanted to see if the issue was present on my system as well, and it's not.
Thanks, will report back.

Low latency mode: what's the sense?

I've read it bypasses plug ins which causes latency. But if it is so what'is the sense of this function? If I understand well in this mode I don't know exactly which plug in will work and which no....
I cannot accept a situation like this.
Apart from anything else I did use low latency mode (because I listened some latency after having inserted a plug) and I checked all the plugs during playback and they all seem to work. So?
I don't understand...

Usually I use Low Latency mode in when mixing-mastering.
I use UAD DSP plugins on the master... this is cause of many issue... the MAIN problem is that Logic is able to compensate Plugin Audio Latency (PDC)... but not Graphics .. (GPDC is not available... instead Logic 7... Logic 7 have GPDC!!! ... in Logic 8 graphical plugin delay compensation is a missing features)
the resulting behavior is that All meters show anticipate by the sound...
but this is not fixable.. or better.. Apple developers must included the Graphic delay compensation in order to get metering in sync with Audio... (I hope in the 8.1 version)
Anyway.. i love UAD sound and I use them in any case..
return back to your question, I use Low Latency mode when i need to record some missing details .. without get DSP latency!
It is a helpful features (not so basical but helpfull).
Logic Pro 7 is also a great DAW!
G

When I press 'Low Latency' button I only have audio coming through left channel ?!

When I press 'Low Latency' button I only have audio coming through left channel !?
When you take it off 'Low Latency' mode the audio is fine and comes out Right and Left speaker.
Can anyone please advise why this happening and how I can fix it ?

I think Pancenter is trying to get you to read and understand the "Working in Low Latency" section of the manual as the results or problems with working in Low Latency mode are variable.
This section, for instance:
Note: The sound may change in Low Latency mode. Depending on the plug-ins in use, the changes can be anything from subtle to dramatic. If plug-ins being used do not exceed the total latency limit, there will be no audible difference.
If it was me posting about this issue I would now come back with what exactly is inserted/bussed on the tracks because it could be an accumulative problem based on what plug-ins you have active and if they are on audio tracks (stereo or mono), instrument tracks etc. Lots of variables. The section of the manual quoted above won't give specific instances as any user could be using any combination of any amount of plug-ins.
So more information is needed before anyone can help you any further.

Noise in LR 1.3.1 even at low ISOs

It seems I am getting noise in my RAW files even at low ISOs like 400, 200 and 100. I have noticed this after "upgrading" to 1.3.1. I am using an Apple I-Mac with Leopard. I am primarily shooting with a Canon Rebel XTi but noticed it with an older Canon Rebel and Canon P&S.
Is anyone else having noise problems? Its starting to drive me nuts.
Is this operator error and maybe just underexposed shots gaining noise when RAW and getting processed through LR?

The default noise reduction and sharpening have been made less aggressive with 1.3 I believe. That could be what you're seeing. Then again, we don;t know what you're talking about without images.
>I guess if the issue is underexposing then are all images underexposed going to have noise in the blacks even without making exposure adjustments.
Generally, if you underexpose one stop, it is as if you up your ISO by one stop. So you go from 100 to 200, 200 to 400, etc. You can use this as a trick to get higher ISOs out of your camera (at the cost of significantly increased noise) than it allows you to select. A camera has many different sources of noise though and it is often not that simple and the increase in noise can be much larger if you underexpose than the equivalent change in ISO.

Multithreaded Video Decoding and Seek frame accuracy for seamless m4a loop

Hi everybody,
We read that flash 11.2 and Air 3.2 will improve the seek functionality.
Right now being able to select a part of a track and precisely loop it is very difficult and complex for an mp3 but impossible with an m4a due to NetStream.
So if you could let us know if the Multithreaded Video Decoding and Seek frame accuracy improvement in 11.2 and 3.2 will bring us a solution would be great!
Basically as m4a are local files does 11.2 and 3.2 will allow to seek not only to keyframes but anyframe? This is key functionality to all music applications. If not do you think it will come soon?
Thank You

Seamless looping using seek will not work very well. It will have to be a separate feature.
You could submit it as a feature request.

Xperia Z3 always awake sometimes, even with Stamina, Ultra Stamina, Low Battery modes

Hi to all,
Has anyone every noticed his/her Z3 always awake (and thus bad battery life), even from within Ultra Stamina mode with no running apps?
This happened to me twice now with just over two weeks of using this phone. I haven't been able to take specific steps to replicate this yet, but 2 days in about 16 days means so far over 12% of the days.
Usually, the phone lasts over 24 hours with moderate use. I normally have stamina mode "on", and set to activate at 50% battery level, and then Low Battery mode to activate at 20%. Note also that all the apps I am using have been installed from the first day I've been using the phone. That is, I have not installed any new apps before I noticed this happening, and when I have "normal" battery life it is using the same set of apps.
Here is an outline of what happens:
I notice a much faster than usual power drain, despite being a typical day of usage (i.e. nothing out of the ordinary).
I go to battery usage, and see the history details, and see that "Awake" has been on constantly, including during long screen off times, and all the while with Stamina mode enabled and active.
I then restart the phone. Problem still there, still constantly awake despite screen off and stamina.
Battery dips below 20% and Low Battery mode activates.
Phone keeps on being constantly "awake", despite no usage and now both low battery AND stamina modes on.
Restart the phone. Same problem.
I turn on Ultra Stamina mode. (I did this last time at 17%)
Phone continues to be always awake. Battery quickly drains down to 5% within less than an hour and almost no use, even though Ultra Stamina is supposed to stretch the life for an order of magnitude longer than that.
I shut off the phone to avoid complete discharge. When I return home, I plug it in and wait for a decent charge before turning the phone back on.
After full charge, I take the phone off the charger.
Phone now works normally again, without always being "awake".
I can't figure out what causes this sequence of events to kick off in the first place. I'm trying to play around with things to get it to start doing this, but so far I haven't been successful in deliberately replicating the problem.
Anyone else experience anything like this before?
My android version is 5.0.2, v. 23.1.A.0.726.

Perforn a clean update and report back in 3-5 days
PC Companion (PCC)
http://support.sonymobile.com/us/tools/pc-companion/
Bridge (for Mac)
http://support.sonymobile.com/global-en/tools/bridge-for-mac/
Alternatives on How to backup Xperias
http://talk.sonymobile.com/thread/36355
"I'd rather be hated for who I am, than loved for who I am not." Kurt Cobain (1967-1994)

CPU Fan running really fast even at low temperature.

I bought a HP Envy 700-160eo desktop computer 5 or 6 months ago and for the last 2 months the CPU fan has been running really fast even at low temperature. It's like locked to about 2400RPM as soon as I boot up my PC. Even if I get 30 degrees celcius it runs at 2600RPM. I've tried resetting my CMOS and tried updating my BIOS, but the BIOS update that you are providing me on the HP Envy 700-160eo driver & update section isn't compatible with my PC. The update is for 700-0xx. I've been using the program called Argus Monitor for controlling my fan speed.

I've got the opposite problem, my AMD CPU is running around 55C and the fan hardly ever exceeds 1200 rpm most of the time its running around 950rpm. I think the 48C to 65C is running on the edge of to hot in my opinion, I like to stay around 35C-40C. Motherboard, video card and hard drive all stay in the 30C range. I tried to speed the fan up with SpeedFan program but it can not interface.
I hope someone can give us a answer, I checked the UEFI and could not find any setting allowing controlling fan speeds on my ENVY 700-200z/AMD A10 cpu.

Why do I have to check low latency every time I want to record 1 track

I-Mac, 3.06 GHz Intel core 2 duo. 4GB RAM. Logic 9.1.1 Mac 10.6.4, Apogee Duet.. Theres no need for this in my main studio. Mac Pro, Motu 896. Even with 8 tracks recording.

OK here ya go. 2 different versions of logic and 2 different versions of SL. Every time you close and reopen a project. You have to check low latency again. I guess I just never noticed it in the main studio because I never needed it, Using a Mac Pro, Motu 896, Buffers set @ 512 Medium With multiple fire wire devices, Multiple midi devices, Recording 8 audio tracks at the same time. Never a problem. On the I-Mac With Apogeee duet. 1 Track with no plugins. Buffers @ 128 I need it. Go figure. Thanks for your input everyone. Bill
Message was edited by: cowbell bill

Low-latency audio on Windows Phone 8 (hint: forget it for now...)

My business specializes in audio and music apps for the Windows ecosystem. For this new project that I’m considering (a virtual instrument of sorts), I need to achieve the lowest possible audio latency from capture to render.
A measured latency below 3 ms would be ideal, 20 ms would be ok, and anything above 30 ms would be a deal breaker. And just to be precise about my definition of latency, I’m strictly talking about “mic-to-speaker” latency (includes the processing in-between).
I’m not talking about “glass-to-speaker” latency, although this topic is also of interest to me.
I figured that I should be using WASAPI, as this is apparently the API that sits at the bottommost of the audio stack in WinRT / WinPRT. So I spent a couple of days re-familiarizing with C++ (happy to meet an old friend)
and I coded a prototype capturing, processing and rendering audio using WASAPI having in mind the goal of achieving the lowest possible latency. After trying and comparing various methods, workflows, threading mechanisms, buffer values and mix formats, the
best I could achieve using WASAPI on Windows Phone was ~140 ms of latency. This is using 3 ms buffers. And although the audio is glitch-free, I could have been using 30 ms buffers and it wouldn’t have made the slightest difference! The latency would still
be around 140 ms. As I found out, the API/driver is extremely defensive in protecting against audio starvation, so it adds quite a bit of buffering at both ends. This is very unfortunate, because it basically disqualifies real-time audio/musical
applications.
I’d love to be able to provide quality audio/musical apps for the platform (both Windows Phone and Windows 8), but right now, this latency issue is kind of a deal breaker.
I've been pointing out the importance of low latency audio to Microsoft for quite a while, and I know I'm not the only one, and I know a lot of people at Microsoft realize this is important. But in its execution,
it seems Microsoft constantly fails to deliver a truly low-latency audio stack. In the pre-XP days, I've had talks with the sysaudio devs about this, and I was told, "yeah, we're working on a new architecture that will come out after XP and
it will solve the latency problem for all audio and musical applications." Fast forward to mid-2010 (pre-Windows Phone 7), and I was still there pointing out the horrible latency figures one would get from the APIs that were about to ship. And
now that WASAPI is available on WP8 (our best hope yet for low-latency audio), I discover the overly defensive and buffer-happy architecture of WASAPI (even though one of its promises was precisely low-latency audio).
So, the question is…
Is Microsoft aware of this issue? If so, is Microsoft giving up and simply conceding the pro-audio territory to iOS? If not, I’d be glad to discuss this issue with an engineer at Microsoft. I’m serious about bringing audio apps
to the platform. I just need some assurance that Microsoft is taking action on its end, so that I can sync my development with the next product cycle.
Thanks in advance for any help, advice, or insights!
/Antoine

Any update on this issue? Is there a roadmap for a fix?
I ported a realtime sound analyzing application (requiring both audio input and output) to WinRT and was quite surprised at the high latency from WASAPI. I then searched around and posted some questions regarding this on various MS forums and got to the
following conclusions based on feedback from people including MVPs and MS employees:
A lot of those people are under the impression that WASAPI qualifies as a low-latency API.
Apparently WinRT apps are not supposed to have high CPU usage, and this is purposefully baked into the framework (minimal thread priority control, async/await everywhere which results in thread priority inversion, and other issues).
Why would anyone ever need high-priority threads? They must use a lot of CPU, right?
A lot of people think low-latency audio means high CPU usage. You can see where this is going when you look at the previous point.
Async/Await is being forced down to all levels, even though it should only be used at the UI level. What some people are actually now calling "old-school" multithreading is now being pushed out (lock, etc). Async/Await has horrible overhead and
results in thread priority inversion amongst other issues. For example, witness that there is no StorageFile.Read, just StorageFile.ReadAsync. Do some IO benchmarks with some of these async methods and you will see some horrible performance compared to desktop
file IO.
To get an understanding of what low latency audio means and why it is important, see this video. It compares the latency of Android to iOS using the exact same music app. Ever wondered
why there are no quality music apps for Android? Well now you know. And then realize that WinRT has
twice as much latency as Android.
And if anyone thinks this is a niche use case, consider that Apple created adds showcasing "musicians" playing their iPad "instruments". A use case that is essentially unavailable for WinRT apps. Why would Apple create
adds for a niche use case?
This should have been one of the high priority issues solved from the start in WinRT. MS solved this issue in Windows (desktop) long ago, with the ability to get insanely low latency there (0.3 ms in some stress tests,
see here), even beating out OSX. It is as if there is a new generation of architects at MS that know nothing about this previous work and are doomed to make the same mistakes over again. I really don't
understand why these pre-existing APIs can't be exposed in WinRT. No need to re-invent the wheel. But I guess it just isn't important enough.
I find this situation really sad since otherwise it could be a great and powerful platform.

LOW LATENCY PROBLEM via Delivery Network. Help .

I would like to have a low latency streaming option even if it does eventually mean running only a single stream.
I am running 2 Mbps H.264 stream via a content delivery network. During the first 30 seconds or so the the latency is about 1 second between source and viewer it then increases after a short while and settles down to about 3 or 4 seconds. I have changed the : Delivery Network (CDN), the bit rate, the codec from H.264 to VP6, and taken off the auto adjust option but I can't seem to get a permanent low latency stream. My colleague has dropped the buffering at his end (or at the server ) and although this makes an initial affect , we still end up with a three second delay.
I do not have any formal training on streaming (not sure who does) so I have to ask : should I by-pass my CDN and point the single stream at the viewer ?
I anticipate working in different locations so I expect a heap of problems - blocked by routers , can't find the IT guys etc etc .
Also, I have a funny feeling that RTMP/Flash will automatically resort to a three/four second delay.
Any ideas for a stable low latency stream ?

When the number of bytes read is less than the array
lengthThat doesn't necessarily mean you're at the end of the data.
Any other wahy, maybe a timer (if is not receiving
any byte for 20 seconds then cut the conexion?)Again that doesn't necessarily mean you're at the end of the data.
There is no way to know you're at the end of the data without either an application protocol containing the length, or a socket close at the sender.

ISE Endpoint losing IP after transition to Low-Impact-Mode

I've recently moved an ISE implementation into the low-impact authentication phase, and the client's security cameras are having a rough go of it. In monitor mode, they were able to stay connected as they should but in low-impact mode they are losing their IP addresses as evidenced in the auth session output below:
SWITCH-1#sh auth sess int g4/0/6            Interface: GigabitEthernet4/0/6          MAC Address: 0040.8cc7.4822           IP Address: 10.92.6.3            User-Name: 00-40-8C-C7-48-22               Status: Authz Success               Domain: DATA       Oper host mode: multi-domain     Oper control dir: both        Authorized By: Authentication Server          Vlan Policy: N/A              ACS ACL: xACSACLx-IP-PERMIT_ALL_TRAFFIC-5165e13c      Session timeout: 3600s (local), Remaining: 338s       Timeout action: Reauthenticate         Idle timeout: N/A    Common Session ID: 0AFF320A000661C965742D42      Acct Session ID: 0x00067E9F               Handle: 0x72000982Runnable methods list:       Method   State       dot1x    Failed over       mab      Authc SuccessSWITCH-1#sh auth sess int g4/0/6            Interface: GigabitEthernet4/0/6          MAC Address: 0040.8cc7.4822           IP Address: 169.254.45.196            User-Name: 00-40-8C-C7-48-22               Status: Authz Success               Domain: DATA       Oper host mode: multi-domain     Oper control dir: both        Authorized By: Authentication Server          Vlan Policy: N/A              ACS ACL: xACSACLx-IP-PERMIT_ALL_TRAFFIC-5165e13c      Session timeout: 3600s (local), Remaining: 338s       Timeout action: Reauthenticate         Idle timeout: N/A    Common Session ID: 0AFF320A000661C965742D42      Acct Session ID: 0x00067E9F               Handle: 0x72000982Runnable methods list:       Method   State       dot1x    Failed over       mab      Authc Success
This is happening approx. every 10 seconds which curiously is the timer value of my dot1x tx-period. As well, the host never has its reauthentication timer restarted but I can see the following in ISE approx. every 10-15 seconds:
Why is it going through Dynamic Authorization? Why am I losing my legitimate IP address every 10 seconds and getting an APIPA address in its place? The port configuration is as follows:
interface GigabitEthernet4/0/6 description Security switchport access vlan 292 switchport mode access ip access-group ACL-DEFAULT in power inline auto max 15400 authentication event fail action next-method authentication host-mode multi-domain authentication open authentication order dot1x mab authentication priority dot1x mab authentication port-control auto authentication periodic authentication violation restrict mab dot1x pae authenticator dot1x timeout tx-period 10 storm-control broadcast level 2.00 storm-control action shutdown spanning-tree portfast spanning-tree bpduguard enableend
And my ACL-DEFAULT is...
Extended IP access list ACL-DEFAULT    10 permit udp any eq bootpc any eq bootps    20 permit udp any any eq domain    30 permit icmp any any    40 permit udp any any eq tftp    50 deny ip any any log
Upon switch log review, I'd noticed that the ACL-DEFAULT is blocking the cameras from certain igmp and tcp/554 (RTSP) communications. To see if it would help, even though I shouldn't have to, I placed ACE's into my ACL-DEFAULT to permit this traffic and would still drop my IP address every 10 seconds. I shouldn't have to do this because the "xACSACLx-IP-PERMIT_ALL_TRAFFIC-5165e13c" is a simple "permit ip any any" ACL which should allow all of the traffic to flow.
Ideas?
Kind Regards,
Kevin

As well, the dACL is properly replacing the first "any" with the endpoint's IP:
SWITCH-1#show ip access-lists interface g4/0/6
     permit ip host 169.254.45.196 any
SWITCH-1#show ip access-lists interface g4/0/6
     permit ip host 10.92.6.3 any
Kind Regards,
Kevin

H264 Decoder buffering 8 frames even in low latency mode

Similar Messages

Maybe you are looking for