LAV Filters - DirectShow Media Splitter and Decoders

[huhn]

doesn't all proper player support audio delay/time shift? why is that not good enough or what am i missing?

[STaRGaZeR]

Am I reading this right? :)

http://git.videolan.org/?p=ffmpeg.git;a=commit;h=217e4ff4d1f845b76e44634e29371cd09313d1c2

[nevcairiel]

Its quite buggy though, and contrary to the commit message not lossless.
However, there is another alternative, and the next LAV version will include full lossless DTS-HD MA decoding, regardless of that commit there. ;)

[STaRGaZeR]

Good news!

I noticed after posting that they even say that the output is not lossless in the changelog. Still, it's progress!

[06_taro]

Great news, DTS-HD is the only reason to use 32-bit players in LAVFilters' side. And the only pity in the chain would be madVR. I believe madshi has a timeline in his mind, to let us throw away all 32-bit stuffs. :)

[madshi]

However, there is another alternative, and the next LAV version will include full lossless DTS-HD MA decoding, regardless of that commit there. ;)
Ready for prime time? And something that could be used for eac3to, too?

[nevcairiel]

Ready for prime time? And something that could be used for eac3to, too?

https://github.com/foo86/dcadec

This project has really clean code, and he was very responsive to an issue I reported. It may not be 100% perfect yet, but its in a much better state than the libav decoder in any case, and the developer is still actively working.

[madshi]

Cool, having a fully featured "free" DTS decoder available would be awesome!

[kalston]

Omg, that's awesome news!

[stax76]

Is it correct that the Haali Splitter setup silently disables LAV Filters features? Isn't this very unfortunate for LAV Filters?

I'm asking because of software that rely on Haali splitter like eac3to, gdsmux/dsmux, TS-Doctor, ffms2 etc.

[huhn]

not sure if you need something special from pot player but with MPC-HC/BE you could use lav audio for decoding, ffdshow for processing and still use the build in time shift from the audio switcher/audio renderer.

and a new audio renderer for MPC-HC is coming "soon" too.
http://forum.doom9.org/showpost.php?p=1709247&postcount=1722

[nevcairiel]

Is it correct that the Haali Splitter setup silently disables LAV Filters features? Isn't this unfortunate for LAV Filters.

I'm asking because of software that rely on Haali splitter like eac3to, gdsmux/dsmux, TS-Doctor, ffms2 etc.

It doesn't "disable" them as such, it just puts itself into their place.
If you install LAV after Haali, it'll do the same. Both compete for similar functionality afterall, and only one source filter can be registered with DirectShow for each format.

[Sparktank]

https://github.com/foo86/dcadec

This project has really clean code, and he was very responsive to an issue I reported. It may not be 100% perfect yet, but its in a much better state than the libav decoder in any case, and the developer is still actively working.

This is wonderful news!

I was waiting for something like this to show up in the open-source world.

Really can't wait to see all the new updates from everyone. :goodpost:

[leeperry]

not sure if you need something special from pot player but with MPC-HC/BE you could use lav audio for decoding, ffdshow for processing and still use the build in time shift from the audio switcher/audio renderer.
Good point, anyway I've run more tests and it seems obvious that LAV's "Auto A/V Sync Correction" helps quite a lot IME when seeking while having Ozone5's GUI running on top of mVR, it doesn't completely avoid random desync when O5's GUI appears but it does seem to kick in and resync by itself rather quickly so all is well, all roads lead to Rome and sorry for the e-drama.

I'm more than ever impressed by what you can achieve with PotPlayer+mVR+LAV+ffdshow. Turns out Ozone5's linear phase EQ latency together with my TV input lag call for -262ms delay in LAV audio for perfect lipsync @24Hz, too good :)

:thanks:

[Thunderbolt8]

so I guess the only thing now missing is a free Dolby Atmos decoder?

[LigH]

That will quite certainly stay missing. Atmos relies on measuring the equipment to achieve optimal utilization.

[xooyoozoo]

That will quite certainly stay missing. Atmos relies on measuring the equipment to achieve optimal utilization.

Rendering to an "idealized" speaker system setup would probably also work. After all, that describes essentially every multichannel mix we've had to date. :)

I think the important part is getting a renderer that can map arbitrary objects at arbitrary points in a 3D space into (semi-customizable) loudspeaker positions. Unfortunately, that part is beyond what's usually expected of simple software decoders.

[huhn]

i guess we need new hardware to support the 32 channel HDMI 2.0 setup.

without that we can't even send a decoded DD atmos stream to a over 7.1 setup.
i guess this will happen some time in the future this is kinda like room correction software nothing totally new.

and headphone DD atmos is something that makes totally sense for a PC too and a full decoded stream is needed for this.

so i hope it will be added in the feature. but i guess bitstreaming will be the best way to deal with it.

[Liandri]

I'd like more options for sound mixing.

Right now, I can only choose between mono, stereo, 4.0, 5.1, 6.1 and 7.1.

I have a 5.1 headphones and some videos come with 5.1 channels, but in most of them I got a bad impression due to voices being too silent in 5.1 mode. So I prefer using stereo output.

I'm using KMPlayer and it has an internal "KMP Audio Codec" which has a lot of output options, including "2/0+sub; 2.1 stereo" option. In this mode, the output is stereo but also I really can feel the subwoofer (and I love it).

But with LAV Audio (which I'd really like to use as a main option) I can only set sound mixing to Stereo, with no subwoofer output. Setting to 5.1 only gives me an old result which I don't like.

tl;dr: I want "2.0+sub" output mode for sound mixing.

[Pat357]

I'd like more options for sound mixing.

Right now, I can only choose between mono, stereo, 4.0, 5.1, 6.1 and 7.1.

I have a 5.1 headphones and some videos come with 5.1 channels, but in most of them I got a bad impression due to voices being too silent in 5.1 mode. So I prefer using stereo output.

I'm using KMPlayer and it has an internal "KMP Audio Codec" which has a lot of output options, including "2/0+sub; 2.1 stereo" option. In this mode, the output is stereo but also I really can feel the subwoofer (and I love it).

But with LAV Audio (which I'd really like to use as a main option) I can only set sound mixing to Stereo, with no subwoofer output. Setting to 5.1 only gives me an old result which I don't like.

tl;dr: I want "2.0+sub" output mode for sound mixing.

If you really want this, move the slider for LFE (Low Frequency ...) to the right in the LAV-audio mixing tab. You need full range speakers for this. Also disable the "normalize matrix" button, just leave the "clipping protection" enabled.
If the surround is to loud (or voices to low), try the to move the slider for surround to the left (less surround) and the slider for the center to the right (more center) in the down-mixing.
You'll need to re-adjust the volume on your AMP/headphones after these to compensate.

[Jtacdf]

Earlier this month, Intel released a new driver that supposedly add video playback of HEVC video format at 4K.
https://downloadcenter.intel.com/download/24785/Intel-Iris-and-HD-Graphics-Driver-for-Windows-7-8-8-1-64-bit

I'm not sure if it is software, hybrid or QS? Anyway, I'm unable to test it as the HEVC is greyed out with QS as an option.

[Liandri]

If you really want this, move the slider for LFE (Low Frequency ...) to the right in the LAV-audio mixing tab. You need full range speakers for this. Also disable the "normalize matrix" button, just leave the "clipping protection" enabled.
If the surround is to loud (or voices to low), try the to move the slider for surround to the left (less surround) and the slider for the center to the right (more center) in the down-mixing.
You'll need to re-adjust the volume on your AMP/headphones after these to compensate.
I tried and it seems LFE slider does nothing when mixing mode is Stereo (don't feel anything, also can see in Status that there is no LFE bar). I suppose your recommendation is for case when 5.1 mixing is used. I tried what you said with 5.1 mode and I really didn't like the results.

I'd like to use Stereo mixing mode with subwoofer/LFE working.

[mastan]

Earlier this month, Intel released a new driver that supposedly add video playback of HEVC video format at 4K.
https://downloadcenter.intel.com/download/24785/Intel-Iris-and-HD-Graphics-Driver-for-Windows-7-8-8-1-64-bit

I'm not sure if it is software, hybrid or QS? Anyway, I'm unable to test it as the HEVC is greyed out with QS as an option.

On Intel HD 4400 LAV uses dxva2cp direct with Sintel_4k_27qp_24fps_1aud_9subs.mkv 4k-video.
QS does not support this(yet?)

DXVAChecker x64 benchmark on the same file shows 40 fps while using 26% of each of CPU kernels on Core i5-4200U.

[jmone]

https://github.com/foo86/dcadec

This project has really clean code, and he was very responsive to an issue I reported. It may not be 100% perfect yet, but its in a much better state than the libav decoder in any case, and the developer is still actively working.

This is great news and the feature list already looks pretty spot on! Do you know if it will be multi thread safe? eg in MC if you used the convert Video -> Audio feature with the dtsdecoderdll.dll you had to use only one conversion at a time or else odd things happened.

Thanks
Nathan

[Arm3nian]

I'd like more options for sound mixing.

Right now, I can only choose between mono, stereo, 4.0, 5.1, 6.1 and 7.1.

I have a 5.1 headphones and some videos come with 5.1 channels, but in most of them I got a bad impression due to voices being too silent in 5.1 mode. So I prefer using stereo output.

I'm using KMPlayer and it has an internal "KMP Audio Codec" which has a lot of output options, including "2/0+sub; 2.1 stereo" option. In this mode, the output is stereo but also I really can feel the subwoofer (and I love it).

But with LAV Audio (which I'd really like to use as a main option) I can only set sound mixing to Stereo, with no subwoofer output. Setting to 5.1 only gives me an old result which I don't like.

tl;dr: I want "2.0+sub" output mode for sound mixing.

I don't know what model of headphones you have, but the "5.1" is most certainly marketing BS. Does your headphone actually have 6 drivers? Doubt it, it most likely has 2. The reason 2.1 is not an option in lav, or pretty much anything is because there is no 2.1 content. There are however 2.1 systems, which include a left and right speaker, the low frequency content is automatically fed to the subwoofer based on the default or specified crossover frequency. The reason the volume is lower when you select 5.1 is because the extra channels are lost.

The '.1' is the dedicated channel for the subwoofer, but you don't have a subwoofer, so what you're asking for doesn't really make sense. As mentioned, moving the LFE slider up is what you're looking for. It mixes the low frequency content from the subwoofer channel instead of throwing it away. I can hear a little difference, using HD650s and an o2odac, I'm not sure how your setup compares to my headphones, the HD650 is known to have good bass reproduction. Another reason why you might not be hearing a difference is because the content you are watching might not actually contain a lot of LFE content. It really is only noticeable on very low frequencies, even on my surround sound system. The bars in the status tab only show what is being played, so if you have stereo selected only the left and right bars are going to show activity. If you leave it on 5.1/7.1 you will see the LFE bar move, and will also see how infrequent and low amplitude it is.Careful what you set the slider to however, headphones can reproduce the low frequency sounds, but are really bad at doing anything else at the same time. You need a dedicated subwoofer for that. I have my settings set to down mix to stereo and the LFE slider set to 1.58. Experiment as settings vary with different equipment.

[huhn]

I don't know what model of headphones you have, but the "5.1" is most certainly marketing BS. Does your headphone actually have 6 drivers? Doubt it, it most likely has 2. The reason 2.1 is not an option in lav, or pretty much anything is because there is no 2.1 content. There are however 2.1 systems, which include a left and right speaker, the low frequency content is automatically fed to the subwoofer based on the default or specified crossover frequency. The reason the volume is lower when you select 5.1 is because the extra channels are lost.

The '.1' is the dedicated channel for the subwoofer, but you don't have a subwoofer, so what you're asking for doesn't really make sense. As mentioned, moving the LFE slider up is what you're looking for. It mixes the low frequency content from the subwoofer channel instead of throwing it away. I can hear a little difference, using HD650s and an o2odac, I'm not sure how your setup compares to my headphones, the HD650 is known to have good bass reproduction. Another reason why you might not be hearing a difference is because the content you are watching might not actually contain a lot of LFE content. It really is only noticeable on very low frequencies, even on my surround sound system. The bars in the status tab only show what is being played, so if you have stereo selected only the left and right bars are going to show activity. If you leave it on 5.1/7.1 you will see the LFE bar move, and will also see how infrequent and low amplitude it is.Careful what you set the slider to however, headphones can reproduce the low frequency sounds, but are really bad at doing anything else at the same time. You need a dedicated subwoofer for that. I have my settings set to down mix to stereo and the LFE slider set to 1.58. Experiment as settings vary with different equipment.

a LFE crossover is defiantly not a bad idea. i recommend to add this with the panned addition of up mixing. https://code.google.com/p/lavfilters/issues/detail?id=11&colspec=ID%20Type%20Status%20Priority%20Filter%20Summary%20Modified

and there are a lot of headphones with multiply driver and even a driver for LFE. here an example picture: http://www.tacticalsites.com/~squinty/images/medusaheadphone/openearphone.jpg 4 is the LFE.
these system have a huge problem with bass reproduction when a source is played that's isn't using a LFE channel so a LFE crossover for stereo is a must have for these.

a lot of smaller 5.1 or 7.1 system work the same way and get into the same problem without a LFE channel when they are used analog.

[nevcairiel]

Its far more likely that I will decide not to add any kind of upmixing, rather than also consider adding a cross over. You should find someone that builds you a new audio processing filter, LAV isn't the place where this is really going to happen.

[Liandri]

I don't know what model of headphones you have, but the "5.1" is most certainly marketing BS. Does your headphone actually have 6 drivers? Doubt it, it most likely has 2. The reason 2.1 is not an option in lav, or pretty much anything is because there is no 2.1 content. There are however 2.1 systems, which include a left and right speaker, the low frequency content is automatically fed to the subwoofer based on the default or specified crossover frequency. The reason the volume is lower when you select 5.1 is because the extra channels are lost.

The '.1' is the dedicated channel for the subwoofer, but you don't have a subwoofer, so what you're asking for doesn't really make sense. As mentioned, moving the LFE slider up is what you're looking for. It mixes the low frequency content from the subwoofer channel instead of throwing it away. I can hear a little difference, using HD650s and an o2odac, I'm not sure how your setup compares to my headphones, the HD650 is known to have good bass reproduction. Another reason why you might not be hearing a difference is because the content you are watching might not actually contain a lot of LFE content. It really is only noticeable on very low frequencies, even on my surround sound system. The bars in the status tab only show what is being played, so if you have stereo selected only the left and right bars are going to show activity. If you leave it on 5.1/7.1 you will see the LFE bar move, and will also see how infrequent and low amplitude it is.Careful what you set the slider to however, headphones can reproduce the low frequency sounds, but are really bad at doing anything else at the same time. You need a dedicated subwoofer for that. I have my settings set to down mix to stereo and the LFE slider set to 1.58. Experiment as settings vary with different equipment.
I have Somic E-95 v2010. Marketing or not, it has more than 2 speakers. I can easily feel 5.1 sound if the source has it. And it has a subwoofer, I've checked it and it works - disabled all channels except LFE and I could hear what I usually hear from my sub from Defender Blaze 60 (2.1) system, plus vibration. So my headphones have .1 channel, and if they do not, I still have a working 2.1 system right here. Extra channels are not lost, otherwise I'd not hear something in 5.1 content, but I can hear everything in 2.0, 2.1 and 5.1 modes. I just don't feel like using 5.1. (and please don't suggest that it's broken)

I can clearly hear the difference. I've checked with the same content. Even for stereo-only content, when KMP Audio is set to 2.1 - LFE channel works (as for how it works, I'm not sure - it probably gets added by mixing both channels) and I can feel bass with vibration, with 2.0 - not. With LAV Audio I can only feel that in 5.1 and above.

As I already mentioned, moving LFE slider did not do anything when Stereo mode was active. If you are suggesting to use 5.1 mode, I can only repeat that I prefer using 2.0/2.1 over 5.1 right now.

LFE in both my 2.1 system and 5.1 headphones can work with both Stereo and 5.1 content. Even if you say there is no 2.1 content, I don't see why there shouldn't be such an option.

Its far more likely that I will decide not to add any kind of upmixing, rather than also consider adding a cross over. You should find someone that builds you a new audio processing filter, LAV isn't the place where this is really going to happen.
That's too bad. Thanks for the honest answer.

[Virtual_ManPL]

@ nevcairiel - CUDA 7 was released some days ago.
Will it have any reasons like performance, stability or others ones to update it in LAV Video Decoder for NVIDIA CUVID hardware acceleration decoder mode?
I remember that CUDA 4.0 was used in LAV CUVID Decoder which was next merged into LAV Video Decoder.

On the other hand will The Boost C++ Libraries or similar ones help in performance in LAV Filters, MPC-HC or even madVR?

[nevcairiel]

LAV doesn't use CUDA, it only uses the video decoder, which is always the same performance.

[Virtual_ManPL]

Hmm, but doesn't LAV Video Decoder in NVIDIA CUVID hardware acceleration decoder mode use NVIDIA CUDA Video Decoder API (https://docs.nvidia.com/cuda/video-decoder/index.html), which on other hand utilizes either CUDA hardware or Video Processor? :p
In LAV Filters source code (hhttps://github.com/Nevcairiel/LAVFilters/tree/master/decoder/LAVVideo/decoders) I'm seeing that CUDA is in version 6.5. ;)

What about second question?

[Telion]

That's too bad.
That's perfectly fine, LAV Audio is more of a decoder than a full-blown sound processor. You should just install AC3Filter (http://www.ac3filter.net/wiki/AC3Filter), it has a plethora of channel processing options including upmix to 2.1. In its "System/Use AC3Filter for" disable everything except PCM, and it'll work as a sound processor only while LAV will continue to work as a decoder.

[LigH]

ffdshow has some audio postprocessing features too. Just as well, you may disable all formats except "raw" to use it as PP filter only.

[leeperry]

I have a 5.1 headphones and some videos come with 5.1 channels, but in most of them I got a bad impression due to voices being too silent in 5.1 mode.
Sounds like what you want is increase the coeff of the center channel, ffdshow audio can easily post-process LAV but it might just be that the mixing matrix of your 5.1'ish phone is wonky to begin with.

And going back to lipsync, a 1ms difference is a lot with headphones IME......I mean ±1 ms and it's easy to see that sync is broken, I'd love a decimal figure tbh.

[huhn]

Sounds like what you want is increase the coeff of the center channel, ffdshow audio can easily post-process LAV but it might just be that the mixing matrix of your 5.1'ish phone is wonky to begin with.

And going back to lipsync, a 1ms difference is a lot with headphones IME......I mean ±1 ms and it's easy to see that sync is broken, I'd love a decimal figure tbh.

how will you notice 1 ms of a difference there is usually only a frame every ~42 ms. even reclock with default settings has a max latency of 100 ms (20% of audio buffer which is 500 ms at default). i mean how do you even know it is 1 ms in the first place.

i wouldn't be shocked that it is totally impossible with current hardware to get a perfect sync better than +/-1 ms.
i mean soundcard/DAC delay, audio renderer delay, leo bodner tester "- Provides accurate measurement within 1 millisecond accuracy." and even that is questionable. and all the clocks in a PC that fight each other...

[leeperry]

i wouldn't be shocked that it is totally impossible with current hardware to get a perfect sync better than +/-1 ms.
I run Invariant TSC in W7SP1 with a 0.5ms granularity together with an async USB DAC(running off two low-jitter clocks for 44.1/48kHz multiples), using a 1m long DVI cable to my TV and a 1.8m long cable to my headphones. I also use a hub that reclocks USB data for better timings and Slysoft Reclock in 88.2kHz Exclusive mode WASAPI prevails(it runs its audio thread in realtime priority BTW).

Fact is that ±1ms delay in LAV kills lipsync in mVR, I'm quite happy with my current rig but I wouldn't mind a decimal figure to play around with that's all :devil:

Of course 5.1 loudspeakers in a room would be a completely different story and only a proper AVR calibration could compute the perfect delay required for each speaker to reach the hot spot simultaneously.

[nevcairiel]

A 0ms perfect sync is practically impossible with DirectShow, just due to the way it syncs playback and how video presentation works. Anyone that claims 1ms makes a difference is delusional.

[leeperry]

Right, 42 ms @24Hz ought to be enough for anybody LOL gotta love humorous forums ^^

[nevcairiel]

Its not about whats enough or whatnot, its about technical fact, and the fact is that DirectShow is not capable to give you 100% perfect 0ms sync. In fact, I would be surprised if any PC player can really do that, simply because of how VSYNC works.
Knowing that your sync is never going to be precise to the millisecond, its also clear that a 1ms difference is not going to change anything, since its within the margin of error we're going to have anyway.

[Sulik]

Here is a 1ms sync adjustment you can do yourself: just move one step back or one step forward (sound waves travel at about one foot (~36cm) per second) :)

[James Freeman]

..sound waves travel at about one foot (~36cm) per second) :)

man thats sllloooowwww.... ;)

[Liandri]

I'd like to correct myself about 2.1 bass feature. Choosing 2.1 in KMP Audio is not the only thing that is required to make it work. It has a checkbox for "LFE Redirection", which makes it work for stereo source. Same with PotPlayer.

[jmonier]

Here is a 1ms sync adjustment you can do yourself: just move one step back or one step forward (sound waves travel at about one foot (~36cm) per second) :)

I think you mean about 1 ft/ms.

[nevcairiel]

He probably just meant to say milliseconds, since thats what we talked about. :p

[huhn]

I'd like to correct myself about 2.1 bass feature. Choosing 2.1 in KMP Audio is not the only thing that is required to make it work. It has a checkbox for "LFE Redirection", which makes it work for stereo source. Same with PotPlayer.
this may work in your sound driver too. you may have a LFE redirection or you just have to set your speaker to small/not full range or something like that. my creative/ESI soundcards have options like that.

[e-t172]

Fact is that ±1ms delay in LAV kills lipsync in mVR.
Right, 42 ms @24Hz ought to be enough for anybody LOL gotta love humorous forums ^^

Your claim is absolutely ridiculous (not quite as ridiculous as your previous claims about high sample rate audio, but close). 1 ms is equivalent to sound traveling 35 cm. Are you seriously claiming that you can perceive the difference in lipsync just by moving 35 centimeters away from the speakers? For reference, my arms are longer than that. That must cause you real trouble in the real world, because that means you would notice lipsync issues just by looking at people talking to you from the other end of a room…

Also, the international standard for sync requirements in professional video production, ITU-R BT.1359 (http://www.itu.int/rec/R-REC-BT.1359-1-199811-I/), states that detection thresholds are as high as 45 ms, while the recommendation allows for delays as high as 90 ms (end to end). This means that your personal sync requirements of <1 ms are two orders of magnitude more stringent than the equipment the material is produced with in the first place.

[infiniter]

Hello! I searched this thread, but couldn't find the proper answer, so I ask here.

I got LAV Splitter installer 0.64 downloaded and only installed the splitter (don't need the rest). I got Windows 7 Ultimate x64. Neither the x86, nor the x64 version of lavsplitter.ax would install with regsvr32 lavsplitter.ax. RegSvr32 always said "cannot load module".
I used of course admin shell and there is no former version of LAV splitter on my system. Is this because the installer put all the files in the x86 path of C:\program files? Any idea? Thanks in advance.

Update: I found a ZIP (https://1f0.de/downloads/) of the splitter, unpacked to C:\program files\lavcodecs, ran the install_splitter.bat and it worked. Though the BAT actually does the same as I tried before. It can only mean, that when using the installer of LAVSplitter and only selecting "Splitter (x86)" and "Splitter (x64)" it does not install all files necessary. Doh!

[Liandri]

this may work in your sound driver too. you may have a LFE redirection or you just have to set your speaker to small/not full range or something like that. my creative/ESI soundcards have options like that.

The device is from C-Media, and as far as I can see it doesn't have such options.

[Arm3nian]

a LFE crossover is defiantly not a bad idea. i recommend to add this with the panned addition of up mixing. https://code.google.com/p/lavfilters/issues/detail?id=11&colspec=ID%20Type%20Status%20Priority%20Filter%20Summary%20Modified

and there are a lot of headphones with multiply driver and even a driver for LFE. here an example picture: http://www.tacticalsites.com/~squinty/images/medusaheadphone/openearphone.jpg 4 is the LFE.
these system have a huge problem with bass reproduction when a source is played that's isn't using a LFE channel so a LFE crossover for stereo is a must have for these.

a lot of smaller 5.1 or 7.1 system work the same way and get into the same problem without a LFE channel when they are used analog.

If a system has problems with bass when a LFE channel is not present then that system is a piece of crap...

Almost all content, besides the audio on bluray do not have an LFE channel. So when listening to music or watching videos on youtube there would no bass... but of course there is. The bass is either reproduced by the left and right speakers or fed into to the subwoofer at the crossover frequency. The purpose of an LFE channel is to isolate specific sound effects from the other sounds, so the subwoofer reproduces them instead of the other speakers (since it does a better job). LFE can contain frequencies up to 120hz. An LFE channel can allow you have a higher crossover frequency if your main speakers cannot reproduce low frequency sounds. Headphones do not need this feature as they can reproduce the audible range of tones, but some speakers might. Quality systems focus on a specific frequency response for each speaker.

Basically, any quality headphone has 2 drivers, you should know this. What hifi gear have you seen with 8 drivers that costs $80. That makes no sense when a proper headphone with 2 drivers costs more. Any decent headphone can reproduce bass without a dedicated subwoofer driver/cone. Notice how standalone subwoofers have huge cones? You need them to produce the low frequencies at audible volumes. A half inch subwoofer inside a headphone isn't going to provide much benefit compared to the main drivers.

The surround sound aspect does not make much sense either. Notice how a proper surround system has the surround sound speakers on the opposite sides of the other speakers? The auditory system is good in positioning, but it doesn't make up for the fact that the surround speakers are right next to the other ones in a headphone. Any proper headphone will have a good enough sound stage and imaging capabilities to mimic what multiple drivers do in such a confined space.

tldr: Why should software support badly designed hardware that sells purely on marketing?

I have Somic E-95 v2010. Marketing or not, it has more than 2 speakers. I can easily feel 5.1 sound if the source has it. And it has a subwoofer, I've checked it and it works - disabled all channels except LFE and I could hear what I usually hear from my sub from Defender Blaze 60 (2.1) system, plus vibration. So my headphones have .1 channel, and if they do not, I still have a working 2.1 system right here. Extra channels are not lost, otherwise I'd not hear something in 5.1 content, but I can hear everything in 2.0, 2.1 and 5.1 modes. I just don't feel like using 5.1. (and please don't suggest that it's broken)

I can clearly hear the difference. I've checked with the same content. Even for stereo-only content, when KMP Audio is set to 2.1 - LFE channel works (as for how it works, I'm not sure - it probably gets added by mixing both channels) and I can feel bass with vibration, with 2.0 - not. With LAV Audio I can only feel that in 5.1 and above.

As I already mentioned, moving LFE slider did not do anything when Stereo mode was active. If you are suggesting to use 5.1 mode, I can only repeat that I prefer using 2.0/2.1 over 5.1 right now.

LFE in both my 2.1 system and 5.1 headphones can work with both Stereo and 5.1 content. Even if you say there is no 2.1 content, I don't see why there shouldn't be such an option.


That's too bad. Thanks for the honest answer.

I was not suggesting to use 5.1 but I can see why you are not using it. Your headset has a hardware mixer attached. If you have left the settings on your computer as default, and the voices are too quiet, then the mixer is configured wrong... Blurays are professionally mastered (at least I hope most are), so it should sound correct with no downmixing. I would suggest lowering the volume of all the other channels and leaving the center higher, but from what I can tell the mixer level controls are also the attenuator, meaning you can't amplify after mixing, another limitation of your setup.

Once again, if you are certain the clip you are testing has significant LFE content, and the slider does nothing, then your left and right drivers can't reproduce the sounds. If the "subwoofer" in your headset is not doing anything at all when you have stereo selected in LAV then your headset doesn't even have a passive filter to send to low frequency content to the sub. So if 5.1 doesn't work correctly, and the sub doesn't work unless there is a dedicated feed, then what good is your headset?

Which leads to 2.1. Seems your options for decent bass are to use a directshow filter or to get another pair of headphones :)
And to be clear, I'm going completely off what you're posting. I don't know how your headset performs, but it seems to me it's causing more trouble than it's worth. You're losing the benefit of the extra speakers for movies, and the bass is weak in content without a dedicated LFE stream. The configuration in your player is basically doing what ffdshow/ac3filter would do, which is either downmixing to 2.1 or upmixing 2.0 to 2.1 for the headset to use the subwoofer/manage the bass properly.

[kalston]

Your claim is absolutely ridiculous (not quite as ridiculous as your previous claims about high sample rate audio, but close). 1 ms is equivalent to sound traveling 35 cm. Are you seriously claiming that you can perceive the difference in lipsync just by moving 35 centimeters away from the speakers? For reference, my arms are longer than that. That must cause you real trouble in the real world, because that means you would notice lipsync issues just by looking at people talking to you from the other end of a room…

Also, the international standard for sync requirements in professional video production, ITU-R BT.1359 (http://www.itu.int/rec/R-REC-BT.1359-1-199811-I/), states that detection thresholds are as high as 45 ms, while the recommendation allows for delays as high as 90 ms (end to end). This means that your personal sync requirements of <1 ms are two orders of magnitude more stringent than the equipment the material is produced with in the first place.

Haha, lipsync issues in real life, I hadn't thought of that but that's a fun way of illustrating how ridiculous his claim was.