eac3to - audio conversion tool

[Wilbert]

@xkodi and others,

I'm using this version:
3) CinemasterAudio.dll v4.3.0.169, SonicHDDemuxer.dll v4.3.0.89

When using a 5.1_96kHz_24bit MLP stream (i'm sure it's 96kHz by looking at the MLP header), the stream is "converted" to (passed through as ?) 88.2kHz:

http://www.geocities.com/wilbertdijkhof/sonic_hddemuxer.jpg

MLP header info:

http://www.geocities.com/wilbertdijkhof/mlp_header.jpg

Do you also have this problem? (When playing the graph the audio seems fine, so it's not mangled up ...) When loading the graph in AviSynth with directshowsource.dll the samplerate is also reported as 88.2kHz.

[ACrowley]

i don't have experience with DD+, so someone else should tell.

SonicCinemasterAudioDecoder applies DRC on EAC3 Decoding.
NeroAudioDecoder2 not
So its recommend to use only Nero Decoder to decode EAC3

[honai]

Which is the default decoder on the eac3to command-line for DD+, Sonic or Nero?

[madshi]

When using a 5.1_96kHz_24bit MLP stream (i'm sure it's 96kHz by looking at the MLP header), the stream is "converted" to (passed through as ?) 88.2kHz
That's a bit strange, but it's not completely without sense, as 48khz and 44.1khz are the usual frequences and 88.2khz happens to be exacty 2 * 44.1khz. Personally, I don't have any MLP files with 96khz, so I can't reproduce the problem. My only MLP sources are movie tracks and they seem to be 48khz only at this point in time.

[madshi]

Which is the default decoder on the eac3to command-line for DD+, Sonic or Nero?
Sonic is currently the default decoder. You can force nero with a switch (see first post of this thread).

[xkodi]

That's a bit strange, but it's not completely without sense, as 48khz and 44.1khz are the usual frequences and 88.2khz happens to be exacty 2 * 44.1khz. Personally, I don't have any MLP files with 96khz, so I can't reproduce the problem. My only MLP sources are movie tracks and they seem to be 48khz only at this point in time.

i will test that in a few hours and post back the results ...

[hristoff2]

Nero filters seem to apply DRC though...need to compare Sonic Decoder and Nero output first but I'm kind of disappointed after comparing DVD and Nero output. (dvd center wav vs. nero center wav)

[arty]

thank you xkodi, ACrowley, hristoff2!

this means we should NOT convert/playback DD+ or TrueHD with any codec because of the DRC :( bad news ... really bad.

even if we'll have a hdmi audio output we can not enjoy the HD discs with our PC :( the whole industy s*cks :( we are forced to use expensive set-top-hddvd/bd/players 'coz maybe intervideo/cybelink decoders also produce bad output we just don't know ... :(

An old dvd will sound better with its low bitrate ac3 track than a DRC'ed TrueHD track :(

[hristoff2]

Actually my version current version of Nero (and it's filters) could be corrupt. (7.9.6.0)

Trying 7.8.5.0 now...latest test I did (30mins before) I encountered that Sonic @ "quiet environment" was even more dynamic than the output of my Nero 7.9.6.0 and it's filters. :mad:
Nero's DRC can be turned off in Nero Showtime though...but until now I thought this would be directed to TrueHD decoding only...I wonder if anyone actually made tests like this:

Sonic vs. Nero
Nero vs. DVD
Sonic vs. DVD

http://nwgat.net/woot/files/51/comparison_thumb.png (http://nwgat.net/woot/files/51/comparison.png)

I could provide all this stuff (the ac3 files) and the eac3 file with the correct timecodes if someone can explain how to cut an eac3 file.

[honai]

I encountered that Sonic @ "quiet environment" was even more dynamic

Do you mean it was "more dynamic" or "more compressed"? (compression as in dynamic range compression) EDIT: Never mind, just took a look at the screenshot and saw what you mean.

@arty, hristoff2

Thanks for your investigation. I've made a post in the other forum summarizing the situation.

[madshi]

Wow, that looks really bad. It's about time we get a useful ffmpeg E-AC3 decoder. Hopefully Google Summer of Code will help!

[ACrowley]

Mh.... i thought too Nero will not apply DRC on DD+ ?
But Sonic does surely

Sonic Decoder Output sounds always louder...on AC3 too compared to AC3 Filter

Mhh i dont think we can 100% compare a DVD AC3 with EAC3 .Shouldnt EAC3 sound mor dynamic so the Waveform is different...
But Sonic shows no differenc between quite and normal..it was reported in few Threads

But no doubt....the comaprisons Screens show Nero is more Flat

However, encoded AC3 from DD+ sounds good imho. Also i usually mux the untouched EAC3 into my reencodes, so....

This is a sample from EAC3 ( Serenity HDDVD 1536kbps)
Both Center waves are loaded into Audacity
1st in Nero 2nd Sonic


http://nwgat.net/woot/files/57/Comparison_thumb.png
(http://nwgat.net/woot/files/57/Comparison.png)

[hristoff2]

Wow, that looks really bad. It's about time we get a useful ffmpeg E-AC3 decoder. Hopefully Google Summer of Code will help!

It gets worse o_O
...
Just saw it and thought it's relatively comparable...take a look at this huge shit:

http://nwgat.net/woot/files/51/comparison_2_thumb.png (http://nwgat.net/woot/files/51/comparison_2.PNG)

...so we've got 2x shit, but nero is even shittier because it's 1/2 loud. (compare overall volume from Nero to DVD and Sonic to DVD...Sonic has at least the correct "overall" volume)

[honai]

As long as the range remains the same as in the original it's no big deal, you just turn up the volume knob. From the screenshot it looks as if Nero's output is really just quieter but not less dynamic than Sonic's.

[hristoff2]

As long as the range remains the same as in the original it's no big deal, you just turn up the volume knob. From the screenshot it looks as if Nero's output is really just quieter but not less dynamic than Sonic's.

Sure, but that means it's 1:0 for Sonic because it has one issue less than Nero. :D

[ACrowley]

Sure, but that means it's 1:0 for Sonic because it has one issue less than Nero. :D

yeah..but what is the Reference ? As i say i dont know i f we can compare EAC3 and AC3 so easy..dont think so

However..most reencodes from EAC sound better as DVD AC3

[honai]

most reencodes from EAC sound better as DVD AC3

... in your own humble opinion. ;)

Maybe it's time for another double-blind listening test: DRCed EAC3->AC3 vs AC3 from DVD. :)

[ACrowley]

... in your own humble opinion. ;)

Maybe it's time for another double-blind listening test: DRCed EAC3->AC3 vs AC3 from DVD. :)


Hey Kollega..."humble opinion" ? Cool down!
Sometimes a reencode from a 1536Kbps EAC3 sounds better as a 384Kbps 5.1 from DVD

[honai]

Sorry if it came across as offensive, it wasn't meant to be. I was just trying to point out that the general claim "most reencodes from EAC sound better as DVD AC3" shouldn't be made without more empirical data.

[ACrowley]

Sorry if it came across as offensive, it wasn't meant to be. I was just trying to point out that the general claim "most reencodes from EAC sound better as DVD AC3" shouldn't be made without more empirical data.

No Problem....
Ofcourse you can say it Generally..
But for example a reencode from 1.5Mbps EAC3 to 640Kbps AC3 or to 768kbps 24Bit WMA 5.1 (when i encode WVC1) sounds better ,compared with a standard 384kbps DVD AC3..not always, but often.
Thats what i mean

And about the AC3-EAC3 Comparison .. How was the Wave decoded from AC3 ?
I mean there can be a huge difference between the Methods how the wave was decoded from AC3
Example :
Source is AC3 (untouched) 448kbps.
Decoded with :ffdshow raw dump(no DRC)/ SteinbergNuendo(no DRC)/ Tranzcode 0.4(no DRC)/ Besweet(no DRC in Command)

Example :

http://nwgat.net/woot/files/57/other/Audacity%20AC3_thumb.png
(http://nwgat.net/woot/files/57/other/Audacity%20AC3.png)

You can see there the Waveform is different. All without DRC.
I would say Steinberg Nuendos AC3 Decoder is Refernce because its a off. Pro. Dolby Decoder/Encoder ( same Result with Digigram,Softencode Decoder)

[arty]

interesting ... I made my comparsions with azid too ! (besweet uses azid)

maybe it decodes wrong? I haven't applied normalisation either but the waveform looked similar to your example...

[ACrowley]

interesting ... I made my comparsions with azid too ! (besweet uses azid)

maybe it decodes wrong? I haven't applied normalisation either but the waveform looked similar to your example...
Maybe it has something to do with the Downmix Overflows or DRC is active even its not present in the commandline

A made a RMS Calculation on all Center Wave samples.
Source is AC3 448Kbps 5.1 has -27db DialogNormalization

- Nuendo AC3 Decoder,Soft Encode Decoder, ffdshow Dump : -30,4 db
- Besweet : -20db
- Cyberlink AudioDecoder :-24db
- SonicCinemasterAudioDecoder 4.2 : - 22,08db
- NeroAudioDecoder2 : -30,4db

Heres the Wavestructure in Audacity from Reference Nuendo compared with Nero and Besweet and Sonic4.3


http://nwgat.net/woot/files/57/ok_thumb.png
(http://nwgat.net/woot/files/57/ok.png)

I cant see anything wrong with NeroAudioDecoder2, its has the same Values and Structure as the Dolby Decoder from Nuendo and SoftEncode



PS, Sry my English is terrible

[ACrowley]

Now the Tests with AC3 and EAC3 Samples

1. AC3 with Nuendo (its the same Result with ffdshow dump and Softencode Dolby Decoder)
2. EAC3 with Nero
3. EAC3 with Sonic


http://nwgat.net/woot/files/57/other/eac3_thumb.png
(http://nwgat.net/woot/files/57/other/eac3.png)

The AC3 was decoded with SteinbergNuendo 3.2
EAC3 with Nero and Sonic

RMS Values :
- AC3 with Nuendo : -30,9db ( same with ffdshow dump and Softencode decoder)
- EAC3 with Nero : -30,5db
- EAC3 with Sonic : -22,4db

[arty]

thanks for the explanation, seems really professional :)

don't you have the same track in TrueHD ? So you could add a decoded trueHD wav into the comparsion ... (1 x ac3 + 2 x eac3 + 1 x truehd)

[ACrowley]

thanks for the explanation, seems really professional :)

don't you have the same track in TrueHD ? So you could add a decoded trueHD wav into the comparsion ... (1 x ac3 + 2 x eac3 + 1 x truehd)


i will see...I cant say anthing sure for TrueHD at the Moment

[tebasuna51]

interesting ... I made my comparsions with azid too ! (besweet uses azid)

maybe it decodes wrong? I haven't applied normalisation either but the waveform looked similar to your example...
To clarify the behavior of ac3 decoders I make a test with:

- NicAudio.dll (AviSynth plugin)
- Azid (standalone or with BeSweet)
- Cyberlink PowerDvd Audio decoder (Dolby Digital certified ?)
- ffdshow

Decoder test over a 5.1 regular ac3 with Dialog Norm. = -27 dB
All decoders without Dynamic Range Compression.

(dB values) NicAudio Azid() Azid(-n1) Cyberlink ffdshow
-------------- -------- ------- --------- --------- -------
RMS power ch0: -29.31 -29.31 -33.30 -33.31 -33.29
RMS power ch1: -27.90 -27.90 -31.91 -31.92 -31.91
RMS power ch2: -20.21 -20.21 -24.20 -24.22 -24.20
RMS power ch3: -47.67 -47.68 -51.68 -51.68 -51.67
RMS power ch4: -31.46 -31.46 -35.47 -35.48 -35.46
RMS power ch5: -33.61 -33.62 -37.62 -37.63 -37.62

Like you can see NicAudio and Azid without parameters are 4 dB loud than Cyberlink, ffdshow and Azid(-n1).

When we send the -n1 parameter to Azid is: apply the Dialog Normalization (31 - 27) = 4 dB attenuated.

Then Cyberlink and ffdshow apply this attenuation by default (correct because are players).

But if we want recode the signal, we don't need this attenuation, then the adequate decoders are Azid() or NicAudio (or apply +4 dB to others decoders)

Edit: SoftEncode Decoder have four decode modes with results between NicAudio and more attenuated (drc applied).

[ACrowley]

My RMs Values was taken from Adobe Audition..
On the Center Channel only..

With certified i mean the Decoder/Encode from Nuendo is a prof. Dolby Software, not a Dshow Decoder like Cyberlink with a Playback license from Dolby
I think Nuendo is a Reference Decoder.
The Values from Nuendo (or Digigran,Softencode) are always around -30db, Azid always around -22-25db

Softencode Decoder with DRC "Line Out" should be (?) "no DRC" it has ~ -30db in my Tests
Line Out means for me it goes out as it comes in, without the touch from DRC...i think so

[tebasuna51]

With certified i mean the Decoder/Encode from Nuendo is a prof. Dolby Software

If you are a owner of a certified encoder/decoder you can make a definitive test:

Use a 5.1 wav and encode to ac3, after decode to wav, the rms values from source and decoded wav must be the same.

[ACrowley]

If you are a owner of a certified encoder/decoder you can make a definitive test:

Use a 5.1 wav and encode to ac3, after decode to wav, the rms values from source and decoded wav must be the same.

Yep ,its the same..

I decode AC3 to wave in Steinberg Nuendo with the Dolby Decoder.
Then the reencode with Steinberg Nuendos Dolby encoder.

Always with -31db DialogNormalization and no DRC.
So i get the same Volume Level compared to the wave source

And as i say, Nero is always right with ~- 30db

[tebasuna51]

I decode AC3 to wave in Steinberg Nuendo with the Dolby Decoder.
Then the reencode with Steinberg Nuendos Dolby encoder.

Always with -31db DialogNormalization and no DRC.
So i get the same Volume Level compared to the wave source

And as i say, Nero is always right with ~- 30db
Calling a0.ac3 the ac3 source, decoded to a0.wav.

Encode to a1.ac3 with -27db (or other value distinct of -31) DialogNormalization and no DRC.

Decode to a1.wav with Steinberg Nuendo Dolby decoder, to a2.wav with Azid and to a3.wav with Azid -n1.

Please put:
Rms a0_center (source)
Rms a1_center (Nuendo decoded)
Rms a2_center (Azid decoded)
Rms a3_center (Azid -n1 decoded)

[ACrowley]

Your requested Test on AC3 from DVD 448kbpos Center Ch, (90% without) Dialog


http://nwgat.net/woot/files/57/other/test_thumb.png
(http://nwgat.net/woot/files/57/other/test.png)

RMS a0_Center : -25.23db
RMS a1_Center : -25.24db
RMS a2_Center : -25.24db
RMS a3_Center : -25.25db

No difference in Waveform. No difference in loud Environmet Sounds.

##################################

Other Tests on the untouched AC3 Source (this new Sample works better as my Test above)
RMS Values :

a0_Center Nuendo : -25.24db
a0_Center azid no DRC : - 12.38db
a0_Center azid with DRC : -21.22db
a0_Center Tranzcode, no Option : -21.23db


The disabled DRC from azid works well, cause some Parts in the Wave with loud Environmet have Full Dynamics, no Compression
With DRC the whole wave is small.
Nuendo and Tranzcode wave looks flat all time long...I dont know if its only a Volume difference, because i cant imagine Nuendo applies DRC ?


http://nwgat.net/woot/files/57/other/nuendo_thumb.png
(http://nwgat.net/woot/files/57/other/nuendo.png)

When you compare it now with the Results above...why are all louder Parts from the Source wave are not present as like as from Besweet/ffdshow ?
I cant Image Nuendo applies DRC..but it looks like DRC, no dynamic Parts anymore. When i compare it with Nero its the same

[tebasuna51]

With your last test seems:

- Nuendo apply DRC and Dialog Norm.
- Azid-DRC only DRC
- Azid nothing (full dynamic)
- ffdshow only Dialog Norm.
- Tranzcode only DRC

[madshi]

You guys are confusing me. Now I'm totally unsure which E-AC3 and AC3 decoder should be used...

:confused:

Anyway, just keep on researching. Hopefully we'll find out exactly which decoder is the way to go. Personally, I've just done some TrueHD checks. Here are my results:

(1) In Nero ShowTime the DRC setting shows an effect. Basically with DRC turned on the peaks are not as loud as without DRC turned off, but the rest of the audio track is more or less unchanged.

(2) With eac3to or GraphEdit, the Nero output is identical to Nero ShowTime with DRC turned *OFF*. It doesn't matter if DRC is turned on/off in ShowTime. In eac3to and GraphEdit DRC seems to be always turned off.

(3) PowerDVD knows 3 different settings: Quiet, Normal and Noisy. In my tests Quiet and Normal are identical and the WAV looks similar to Nero's output with DRC turned off. When using PowerDVD with "Noisy" the output is even more compressed than with Nero DRC turned on.

(4) Sonic is totally useless. It's much louder than anything else. And there's noise in every channel. Strangely, if you compare the Sonic WAV to Nero/PowerDVD, the center information seems to talk over into Left/Right channels in the Sonic decoder. Something fishy going on there.

(5) Unfortunately Nero's output for 16bit TrueHD tracks contains full 24bit information. That means, Nero must be doing some processing on the 16bit decoder output. It doesn't seem to be DRC to me. Must be something else. This is bad for reencoding, obviously.

(6) Unfortunately PowerDVD always only output 16bit, even if the TrueHD track is 24bit!! As a result I'm not totally sure if PowerDVD leaves the 16bit decoder output untouched. Maybe it does, maybe not.

It seems that there is no really good TrueHD available yet. Nero's and PowerDVD's WAVs look quite similar in Audacity, but the bits are different. PowerDVD is always 16bit, Nero is always 24bit.

All I wrote above is strictly only about TrueHD decoding.

[madshi]

Nero DRC off / Nero DRC on
http://madshi.net/TrueHD/Nero.noDRC_.png (http://madshi.net/TrueHD/Nero.noDRC.png)http://madshi.net/TrueHD/Nero.DRC_.png (http://madshi.net/TrueHD/Nero.DRC.png)

PowerDVD DRC off / PowerDVD DRC "noisy"
http://madshi.net/TrueHD/PowerDVD.noDRC_.png (http://madshi.net/TrueHD/PowerDVD.noDRC.png)http://madshi.net/TrueHD/PowerDVD.Noisy_.png (http://madshi.net/TrueHD/PowerDVD.Noisy.png)

Sonic
http://madshi.net/TrueHD/Sonic_.png (http://madshi.net/TrueHD/Sonic.png)

[d0ORk]

I tried to convert the audio stream that EVODemux made.
But I'm getting an error:
Muxing evo's first audio track to raw. Please wait...
The demuxer didn't accept the source file.

Is it an issue with EVODemux?

[madshi]

Is it an issue with EVODemux?
That's very hard to say. It could be an issue with EvoDemux. But more likely the Sonic filters refuse to work on your PC. Try playing the Evo with PowerDVD (if you have that installed). With the latest update patch PowerDVD player EVO files with only a TrueHD track just fine for me. If PowerDVD plays your EVO fine, then the EVO is probably ok.

[d0ORk]

mhh. the video works, but not the audio file that EVODemux made. At least, PowerDVD breaks down if I want to play it.
Any other idea how to demux the audio? Good is, that EVODemux merges the 2 evo files into one (video)

[madshi]

@d0ORk, which audio format is that?

[d0ORk]

it's DD+ on the Test HDDVD

[arty]

I think I completly lost :D

... of course thank you for your efforts!

[ACrowley]

With your last test seems:

- Nuendo apply DRC and Dialog Norm.
- Azid-DRC only DRC
- Azid nothing (full dynamic)
- ffdshow only Dialog Norm.
- Tranzcode only DRC

yep, youre right...theres obviously no other possibility. So it makes my 1st Test useless!!

Nuendo certified Dolby Decoder seems to apply both, DRC and Dialog Normalization.
Can be true ,because you can see in the Options from Soft Encode Dolby Decoder and from Digigram, (both older certified Dolby Dec-Enc Versions) that youve few Options to change DRC and Dialog Norm for Ac3 to pcm decoding
But You cant change Decoder Options in Nuendo.
So Nuendos/Cubase default seems to be DRC/Dialog Norm Output, not nice..all my last AC3 reencodes from last the Years are all from Nuendo 3.2 :)

Anybody know which Decoder Option is 100% "no" DRC AND Dialog Normalization in SoftEncode ? There are following DRC Decoder Options :
-Custom Analog /Dialog Norm
-Custom Mode / Digital Dialog Norm
-Line Out
-RF Remod Mode

The Softencode wave and RMS Values looks as like as Besweet (no DRC enabled) when RF Remod Mode is active. All other Modes are as like as Nuendo or Tranzcode.

I found this :
"Normal is similiar to 'Line out mode' and Heavy is similiar to 'RF-remod mode' of Sonic Foundary Soft Encode. It seems the WinDVD directshow filter (iviaudio.ax) invokes Heavy DRC also. However, Heavy is NOT recommended
because it usually results in obvious volume transitions, especially for music titles. "-Hiroko Hori (author of DVD2AVI) "

So ,why is the Wave with RF Mod from SoftEncode Output (DRC according to the Posting) the same compared with Besweet and "no" DRC ? Its all reversed ?!
Soft Encode Setting at "Line Out" is as like as Nuendo

All i want is to find a reference to decode AC3 to Wave "without DRC and Dialog Normalization.
Then i can compare TrueHD/EAC3 with AC3

[madshi]

it's DD+ on the Test HDDVD
Well, demuxing + decoding via eac3to should work fine. There's no need to keep DD+ inside the EVO.

[d0ORk]

It cant decode it

Q:\test>"D:\_eac3to\eac3to.exe" "Q:\test\PEVOB_1_MERGED.rebuilt.
EVO" "Q:\test\PEVOB_1_MERGED.DD.6ch.s03.German 5.1.ac3" -448 -24
Muxing evo's first audio track to raw. Please wait...
Converting the raw file to wav. Please wait...
Converting the wav file to ac3. Please wait...

Aften: A/52 audio encoder
Version 0.07
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

invalid or empty chunk in wav header
invalid wav file: -b
"aften" reported error code "1". A valid *.wav file was created sucessfully, tho
ugh.

Q:\test>

[madshi]

Try another Aften build. Maybe revision 449 or 433.

[d0ORk]

I just watched the .raw file while demuxing. It's just the name like 'example.raw' but it's always 0byte :(

[tebasuna51]

All i want is to find a reference to decode AC3 to Wave "without DRC and Dialog Normalization.
Then i can compare TrueHD/EAC3 with AC3

Like I say you, use Azid-BeSweet, NicAudio-AviSynth or Foobar also.

[ACrowley]

Like I say you, use Azid-BeSweet, NicAudio-AviSynth or Foobar also.


Yeah..sure.
To bad that Nuendo decodes with DRC/DN without Options to change it ( 1000$ Software!)

But still 2 Questions :

1. Which Softencode DRC Mode is without DRC/DialogN. ?
2. Besweet Downmix Overflows
Its no Problem so far i know..only a warning and is relevant when downmixing to 2Ch right ?

greets

[d0ORk]

Somehow I got it to work. :)
But theres a problem left...

Q:\test>"D:\_eac3to\eac3to.exe" "Q:\test\PEVOB_1_MERGED.DD.6ch.s
03.German 5.1.ddp" "Q:\corpse_demux\PEVOB_1_MERGED.DD.6ch.s03.German 5.1.ac3"
-448
E-AC3, 5.1 channels, 1:17:12, 640kbit/s, 48khz
Muxing eac3 file to raw. Please wait...
The NCBENUM return code is: 0x0
The NCBENUM return code is: 0x0
Audio Decoder output was 24 bit this time.
Unfortunately the Audio Decoder has output only 2 channels.
Converting the raw file to wav. Please wait...
Converting the wav file to ac3. Please wait...

Aften: A/52 audio encoder
Version SVN
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

input format: Signed 24-bit 48000 Hz 5.1-channel
output format: 48000 Hz 3/2 + LFE

progress: 100% | q: 233.2 | bw: 34.0 | bitrate: 448.0 kbps

Done.

Q:\test>

How can I get it to be 5.1?

Thanks for your help

[arty]

Yeah..sure.
To bad that Nuendo decodes with DRC/DN without Options to change it ( 1000% Software!)

But still 2 Questions :

1. Which Softencode DRC Mode is without DRC/DialogN. ?
2. Besweet Downmix Overflows
Its no Problem so far i know..only a warning and is relevant when downmixing to 2Ch right ?

greets

I got overflows even when set -6ch in besweet ... (decode to 6 mono waves). So the resulted wavs' waveform seems a bit "flat" where they reached the "maximum" ... :(

[madshi]

Unfortunately the Audio Decoder has output only 2 channels.
Try changing the Sonic filter settings (see Control Panel -> CinePlayer).