foo_dsp_continuator, crossfading plugin Options

[kelesh]

what bug did this fix? did it fix the cracking/popping/static bug? i will test, of course. thanks Cpt. Footure!

[rectangle]

Finding (really) stupid bug...
Banging head against wall... 
Fixing bug...

Thanks Cpt. I've been running it for 8 hours now without a snap, crackle or pop!

[rectangle]

O-Oh! I've just got this failure.txt & i noticed I'm getting this in the console:

INFO (CORE) : location: "file://E:\Lambchop\How I Quit Smoking\06. Suzieju.mpc" (0)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 16647 samples (377 ms) removed (from beginning)

Can this be correct?

[kamien001]

Cool plugin ...

This post has been edited by kamien001: Feb 25 2004, 11:36

[Cpt. Footure]

Cool plugin  ...

thnx.

O-Oh! I've just got this failure.txt & i noticed I'm getting this in the console:

INFO (CORE) : location: "file://E:\Lambchop\How I Quit Smoking\06. Suzieju.mpc" (0)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 277536 samples (6293 ms) overlapped (from end)
INFO (foo_dsp_continuator) : 16647 samples (377 ms) removed (from beginning)

Can this be correct?

No, i fixed the bug and disabled all console writing in addition. Will anybody miss it?

[Melomane]

Cpt. Footure, thank you for your Continuator!

Question, what is the function of "empty buffer up to overlap point of track change" ?

[Cpt. Footure]

Question, what is the function of "empty buffer up to overlap point of track change" ?

if checked:
- reaching end of track
- search point/time where new track will start according to threshold setting
- flush/output internal buffer contents up to this point (this may need some megs of memory, depending on buffer size)
- start new track while fading out old

if unchecked:
- reaching end of track
- search point/time where new track will start according to threshold setting
- no buffer flushing
- start new track while fading out old

The resulting track changes will be where they belong if "empty buffer..." is checked, in contrast to sqrsoft's ACF. It is especially useful if used with the diskwriter to create an Audio CD with already faded tracks.

bye

[Lyx]

hmm, i still get a short (0,3 - 0,7 sec.) "playback freeze", at the moment continuator kicks in at the end of a track..... hmm, lets say "max overlap length" is set to 8000, and the track is 4:00mins long.... then the skip would often happen at 3:52.

It seems that this only happens on my system, if the tracks are already gapless(not sure).

Settings:
max. overlap lengt: 8000
empty buffer: on/off (happens in both cases)
threshold: 13.0
force fade-out: on
fade-rate: 5.5
disable track overlapping if seamless...: on
force track length...: off

This post has been edited by Lyx: Mar 1 2004, 22:24

[Cpt. Footure]

i will try to lower the cpu consumption on track changes, in the meantime try to enlarge the buffer length of your preferred output plugin.

[Lyx]

you are probably right about cpu consumption, since at the point when the skips happen, cpu-usage spikes from 1% up to about 55% for a short moment.

Running foobar all the day, i can say that at least on with the albums i played today, it -only- happens, if the tracks are already seamless - if they are not and overlapping is used instead, then theres no problem.

Thanks for your work

- Lyx

[Cpt. Footure]

new version (0.1.7) uploaded with reduced cpu consumption on track changes

[Lyx]

Illegal operation:
Code: C0000005h, flags: 00000000h, address: 008623CAh
Access violation, operation: read, address: 00CD8BE0h
Call path:
playback_core_i::g_thread_func=>playback_core_i::thread_func=>dsp::run
This is the first crash logged by this instance.
Code bytes (008623CAh):
0086238Ah:  C8 C1 F9 03 8B 55 14 D1 EA 2B CF 3B CA 89 55 10
0086239Ah:  73 1D 8B 76 08 8B 4B 04 8D 3C F8 3B FE 0F 84 1E
008623AAh:  01 00 00 DD 07 DD 19 83 C1 08 83 C7 08 EB EC 8D
008623BAh:  04 F8 D9 EE 89 45 08 8D 1C D0 3B C3 74 0B D9 C0
008623CAh:  DC 00 83 C0 08 DD D9 EB F1 83 65 FC 00 DD 55 F0
008623DAh:  89 55 F8 DF 6D F8 8B 4D 0C 8D 45 E8 50 DC 3D 80
008623EAh:  41 86 00 DD 55 F8 D9 C1 D8 C9 DD 5D E8 DD D8 DD
008623FAh:  D8 E8 90 03 00 00 8B 45 14 03 F8 C1 E7 03 8B 46
Stack (04F0FD24h):
04F0FD04h:  000020D6 00000000 00000000 BAE46C48
04F0FD14h:  BFE9E33B 817D708C 00000000 00000000
04F0FD24h:  00C17590 00C17540 00000000 04F0FD54
04F0FD34h:  00000000 00C17540 00000000 00C17590
04F0FD44h:  04F0FD94 04F0FD94 00861FC5 00CD8BE0
04F0FD54h:  00C175A0 00000005 0000000A 00000000
04F0FD64h:  00C17540 004384A0 00000004 47000000
04F0FD74h:  00000000 C0590000 42556EE8 3DDAADD3
04F0FD84h:  0001BA0E 00000000 C0590000 0043E6C8
04F0FD94h:  04F0FDBC 00862E30 021705B0 00000000
04F0FDA4h:  0040AAA7 04F0FEF0 00C4A35C 00000001
04F0FDB4h:  00000000 021705B0 04F0FE4C 0040A988
04F0FDC4h:  0000003C 04F0FEF0 00C4A35C 00000001
04F0FDD4h:  04F0FE00 00C2F5C8 00C0D980 00C6AFA4
04F0FDE4h:  00000003 00C6AFA4 021705B0 00000030
04F0FDF4h:  00443B28 00000000 00000728 00000000
04F0FE04h:  00000000 0109001F 00000010 7F6BAF16
04F0FE14h:  40A41B65 D69331BB 6F1CB1DE 00000010
04F0FE24h:  10001DE2 00C61020 10001E17 00000000
04F0FE34h:  04F0FE54 004365E9 004365F4 00000000
Registers:
EAX: 00CD8BE0, EBX: 00CD8C08, ECX: 000000D6, EDX: 00000005
ESI: 00C17590, EDI: FFFFFF7B, EBP: 04F0FD48, ESP: 04F0FD24
Crash location: "foo_dsp_continuator", loaded at 00860000h - 00868000h

Loaded modules:
foobar2000                    loaded at 00400000h - 00451000h
ntdll                            loaded at 77880000h - 77901000h
KERNEL32                      loaded at 77E70000h - 77F33000h
USER32                        loaded at 77E00000h - 77E65000h
GDI32                            loaded at 77F40000h - 77F7C000h
ole32                            loaded at 77A40000h - 77B2C000h
RPCRT4                        loaded at 77D20000h - 77D8E000h
ADVAPI32                      loaded at 77DA0000h - 77DFD000h
COMCTL32                      loaded at 77B40000h - 77BC9000h
utf8api                          loaded at 10000000h - 1000D000h
comdlg32                      loaded at 76B00000h - 76B3E000h
SHLWAPI                          loaded at 77C60000h - 77CAA000h
SHELL32                          loaded at 77580000h - 777CC000h
MSVCRT                        loaded at 78000000h - 78046000h
foo_ape                          loaded at 00800000h - 0081F000h
foo_cdda                      loaded at 00820000h - 0082E000h
foo_clienc                    loaded at 00830000h - 0083B000h
foo_diskwriter                loaded at 00840000h - 00852000h
WINMM                            loaded at 77540000h - 77571000h
foo_dsp_continuator              loaded at 00860000h - 00868000h
foo_dsp_extra                    loaded at 00870000h - 00885000h
foo_dsp_wider                    loaded at 00890000h - 00897000h
foo_flac                      loaded at 008A0000h - 008B6000h
foo_id3v2                        loaded at 008C0000h - 008E6000h
MSVCP60                          loaded at 780C0000h - 78121000h
foo_infobox                      loaded at 008F0000h - 00903000h
foo_input_std                    loaded at 00910000h - 009A9000h
MSACM32                          loaded at 773E0000h - 773F3000h
foo_masstag                      loaded at 009B0000h - 009C0000h
foo_mod                          loaded at 009C0000h - 009C9000h
BASS                          loaded at 009D0000h - 00A2A000h
foo_osd                          loaded at 00A30000h - 00A3B000h
foo_out_dsound_ex                loaded at 00A40000h - 00A49000h
DSOUND                        loaded at 51080000h - 510DA000h
VERSION                          loaded at 77810000h - 77817000h
LZ32                          loaded at 75940000h - 75946000h
foo_playlistgen                  loaded at 00A60000h - 00A6B000h
foo_read_http                    loaded at 00A70000h - 00A79000h
WS2_32                        loaded at 74FA0000h - 74FB3000h
WS2HELP                          loaded at 74F90000h - 74F98000h
foo_rgscan                    loaded at 00A80000h - 00A8B000h
foo_ui_std                    loaded at 00A90000h - 00AA7000h
foo_vis_manager                  loaded at 00AB0000h - 00AB8000h
foo_vorbisenc                    loaded at 00AC0000h - 00BBA000h
foo_web                          loaded at 00BC0000h - 00BD8000h
wdmaud                        loaded at 77530000h - 77539000h
KsUser                        loaded at 5EF80000h - 5EF84000h
imagehlp                      loaded at 77910000h - 77933000h
DBGHELP                          loaded at 72970000h - 7299D000h

Stack dump analysis:
Address: 00861FC5h, location: "foo_dsp_continuator", loaded at 00860000h - 00868000h
Address: 004384A0h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 0043E6C8h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00862E30h, location: "foo_dsp_continuator", loaded at 00860000h - 00868000h
Address: 0040AAA7h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 0040A988h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00443B28h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 10001DE2h, location: "utf8api", loaded at 10000000h - 1000D000h
Symbol: "uCallStackTracker::uCallStackTracker" (+00000000h)
Address: 10001E17h, location: "utf8api", loaded at 10000000h - 1000D000h
Symbol: "uCallStackTracker::uCallStackTracker" (+00000035h)
Address: 004365E9h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004365F4h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 0041E3A1h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 0041F171h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00438978h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004389ECh, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004389E4h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004389D0h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004389E4h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004389D0h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439EF4h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439EE0h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439F70h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439EB0h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439ED8h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439EC4h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439ED8h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439EC4h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004392E4h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004384A0h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 004384A0h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439EF4h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00439EE0h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 00444A00h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 0041F501h, location: "foobar2000", loaded at 00400000h - 00451000h
Address: 77E787DDh, location: "KERNEL32", loaded at 77E70000h - 77F33000h
Address: 77E91856h, location: "KERNEL32", loaded at 77E70000h - 77F33000h
Address: 77E7AE88h, location: "KERNEL32", loaded at 77E70000h - 77F33000h
Address: 0041F4D2h, location: "foobar2000", loaded at 00400000h - 00451000h

Version info:
foobar2000 v0.8
UNICODE

crash happened while playing albumgained MP3 VBR files
DSP-Chain: continuator->convert to 4ch->foo_wider
Win2k, DirectSound v2 Output

[Cpt. Footure]

crash happened while playing albumgained MP3 VBR files

same procedure as in last reply:
new version (0.1.8) uploaded

[Lyx]

works great now - thank you

Feature request:
Would it be possible to choose between overlapping-mode(as it is now) and old fashioned crossfading(difference to case's variant being that if seamless transition is detected, no crossfade happens - and of course the built-in gap-killer)? Maybe like a radio-knob switch in the pref-page.

- Lyx

[Cpt. Footure]

Feature request:
Would it be possible to choose between overlapping-mode(as it is now) and old fashioned crossfading(difference to case's variant being that if seamless transition is detected, no crossfade happens - and of course the built-in gap-killer)?

Something like this?
Note: New checkbox in settings (Disable theshold...).

[Melomane]

WONDERFUL!

[Canar]

Okay... I've been trying to write a crossfaded audio CD using foobar's audio CD writer and the continuator DSP.

I'm presently using the Continuator DSP and Advanced Limiter as DSPs, and I have the audio CD writer options set to use DSP, don't reset between tracks, use ReplayGain, and dither.

To test it all, I've been writing image files and mounting them in Daemon tools, then playing them through foobar.

I'm getting fade-out at the end of the tracks, but no crossfade.

[tom_vienna_at]

Okay... I've been trying to write a crossfaded audio CD using foobar's audio CD writer and the continuator DSP.

I'm presently using the Continuator DSP and Advanced Limiter as DSPs, and I have the audio CD writer options set to use DSP, don't reset between tracks, use ReplayGain, and dither.

To test it all, I've been writing image files and mounting them in Daemon tools, then playing them through foobar.

I'm getting fade-out at the end of the tracks, but no crossfade.

I've been trying this too, but actually have been converting mpc-files to a single-file output, using only continuator as a dsp. Works very well, except for the very last track of the mix, which is cut off at the end. Found a work-around: I add a silent one-minute-mpc-track at the end of the playlist that I want to be a crossfaded mix, convert single-file-output, perfect mix as a result, last tracks does fade out now. I burn that to CD (with Feurio) and have a nice mix on CD (with one long track on it - must look for a way to have it split in songs again).
Tom.

[edit] grammar

This post has been edited by tom_vienna_at: Mar 19 2004, 08:32

[Enig123]

Footure, I have tested your great plugin and tried to seamlessly playback with MP3, but failed. I know MP3 is not a naturally seamless format. If there's a way to make MP3s playback seamlessly, your plugin will be near perfect.


Regards,

[Cpt. Footure]

Works very well, except for the very last track of the mix, which is cut off at the end.

This is the failure of the diskwriter. As i just discovered it doesn't send dsp::FLUSH on end-of-playback (ie. last track) if "don't reset dsp between tracks" is checked. As i rely on that flag, the end of the last track is cut off.

[Cpt. Footure]

(with one long track on it - must look for a way to have it split in songs again)

Why don't you just use Convert->Run conversion? Make sure you check "Force track length..." in Continuator and burn the wavefiles without pauses between tracks.

[tom_vienna_at]

(with one long track on it - must look for a way to have it split in songs again)

Why don't you just use Convert->Run conversion? Make sure you check "Force track length..." in Continuator and burn the wavefiles without pauses between tracks.

Thanks Cpt. Footure... seems the easy, natural solution... too blind to see that.
I'll use the simple "run conversion" next time I do a mix.
Fot the cut-off last track in the mix: the one-minute silent mpc-track as the last track of the mix does the trick.
Thanx for the Continuator, love it... no music without it here; using it on my radioshow for pre-produced mixes.
Tom.

[Melomane]

Okay... I've been trying to write a crossfaded audio CD using foobar's audio CD writer and the continuator DSP.
.......
I'm getting fade-out at the end of the tracks, but no crossfade.

I had the same problem with foo_burminate.dll from foobar installer.
i have re-download foo_burminate.dll from
http://foobar2000.org/components.html#foo_burninate

now work fine

and the end of the last track is NOT cut off.

[Lyx]

i just ripped a seamless ambient album and encoded it with lame 3.96beta1 and 3.90.3. With both lame versions, at some trackchanges audible clicks happen. I guess thats a limitation of how mp3-encoding works.

Maybe a simple way for continuator to fix this could be to even if seamless transition is detected, a small amount of frames gets overlapped? Just an idea, i dont know if it would work.

- Lyx

[Cpt. Footure]

Works very well, except for the very last track of the mix, which is cut off at the end.

This is the failure of the diskwriter. As i just discovered it doesn't send dsp::FLUSH on end-of-playback (ie. last track) if "don't reset dsp between tracks" is checked. As i rely on that flag, the end of the last track is cut off.

I found the bug in diskwriter.
@ zZzZzZz
line 1167 should read

CODE

if (!(b_dsp_noreset && file_ptr < files.get_count()-1))  

@ all
Download the fixed diskwriter from here.