foobar2000 v1.6.6 final has been released

[digitalfrost]

Download: https://www.foobar2000.org/download

Release notes: https://wiki.hydrogenaud.io/index.php?title=Foobar2000:Version_1.6_Release_Notes

Version 1.6 overview

This is a major update introducing new features and functional changes.

Compatibility

Windows 7 or newer is required, with a processor that supports SSE2 opcodes (anything made during the last 15 years should do).

Windows XP and Windows Vista are no longer supported.

1.5 series supporting these operating systems will be kept available for download indefinitely. Also, important bug fix updates for 1.5 series will be made for as long as people care to report bugs.

Notable new features

• New default WASAPI output.
• Output-independent implementation of clickless fading (same as in current foobar2000 mobile and foobar2000 for mac).
• Out-of-the-box 7zip archive reading (Archive Reader component included with setup).
• WebP cover art support.
• New Unicode-based filename character substitution in Converter & File Operations.

Removed features

• DirectSound output (replaced by WASAPI)
• Removed obsolete edit-replaygain-info menu command since Properties dialog does the same now.
• Windows volume mixer integration (buggy, rarely used).

 

[daftcombo]

1.6.7 beta is already out though.

1.6.7

• Built-in WASAPI exclusive output.
• Added 'smart stop' option in Advanced Preferences; press stop key once to stop-after-current, twice to stop now.
• Fixed WavPack lossy not working regression from 1.6.6.

1.6.6

• Made Advanced Preferences searchable.
• Added support for attaching pictures to Matroska files; fixed Matroska tag editing bugs.
• More accurate reporting of Matroska bitrate (actual content is probed if needed).
• Added itunesadvisory tag support for M4A.
• Updated embedded cuesheet handling code.
• Replaced PPHS resampler with much faster libretro resampler.
• Added support for internet radio album art.
• Converter: Encoded FLAC files are now updated with proper FLAC seektables, if FLAC.exe doesn't write them by itself.
• Converter: Workarounds for clashes when writing to a Media Library indexed folder.
• Converter: Added built-in profile for exhale AAC encoder.
• Converter: Fixed some warning prompts not being possible to permanently disable.
• Fixed slightly inaccurate bitrate reporting for Musepack and WavPack. Tags are now excluded from file size used to calculate the bitrate.
• Now compiled with Visual Studio 2019 (was 2017 before).
• Album art related regressions fixed. (beta 2)
• Worked around 7-zip reading crash. (beta 2)
• Fixed missing no-artwork pic in some scenarios. (beta 3)
• Fixed slower playlist & library search regression from earlier betas. (beta 4)
• Fixed failure to read some rare MP4/M4A files. (beta 5)
• Converter: Added a toggle for FLAC seektable fix. (beta 5)
• Fixed Default UI playlist rendering glitches with millions of items. (beta 6)
• Fixed decoding of 7-channel FLAC files. (beta 7)
• Made FLAC channel mask not cleared with 'remove tags'. (beta 7)
• Improved compatibility with badly configured HLS radio servers. (beta 7)
• Fixed odd samplerate autoresampling regression from earlier betas. (beta 8)

(I was with 1.6.5 for a while.)

 

[daftcombo]

The major change is indeed that WASAPI (shared) has become the default ouput (instead of Direct Sound).
It is a different implementation than the WASAPI component.

 

[roskodan]

Just dropping by to remind that foobar2000 has a convolver plugin, Impulse Response Convolver plugin, once installed, to EQ your headphone to Harman, just download the respective .wav file for your headphone from AutoEQ results and load it into the plugin (WARNING: just turn off "Auto level adjust", in plugin settings, as it seems to not work properly and induces digital clipping).

 

[roskodan]

If DS was dropped, then I guess when upgrading to v1.6+, any DS OS integrated DSP (like EQ APO) will be bypassed and unusable with foobar2000?

 

[Frgirard]

If DS was dropped, then I guess when upgrading to v1.6+, any DS OS integrated DSP (like EQ APO) will be bypassed and unusable with foobar2000?

I use the 1.6.6 and directsound is usable and write "default" output.

 

[digitalfrost]

If DS was dropped, then I guess when upgrading to v1.6+, any DS OS integrated DSP (like EQ APO) will be bypassed and unusable with foobar2000?

It's using WASAPI shared mode, only the exclusive mode disables any APOs. So EqAPO works.

 

[daftcombo]

No real point in using EqAPO in Foobar when you can load a correction file in convolver though.

 

[digitalfrost]

No real point in using EqAPO in Foobar when you can load a correction file in convolver though.

System-wide correction (also for movies, games etc...)

 

[daftcombo]

System-wide correction (also for movies, games etc...)

I mean: when you are watching a movie or playing a game, EqAPO will do the correction. But when you launch Foobar2000, Convolver can take over.
For instance, you can use more taps in the file used in Convolver for musing listening, because you don't need low latency.

 

[Matias]

I mean: when you are watching a movie or playing a game, EqAPO will do the correction. But when you launch Foobar2000, Convolver can take over.
For instance, you can use more taps in the file used in Convolver for musing listening, because you don't need low latency.

You can set up Convolution in Equalizer APO as well. Or if you don't want the lag, PEQ in Equalizer APO and foobar running Convolution while using ASIO to bypass Equalizer APO.

 

[daftcombo]

You can set up Convolution in Equalizer APO as well. Or if you don't want the lag, PEQ in Equalizer APO and foobar running Convolution while using ASIO to bypass Equalizer APO.

Yes, that's what I meant and what I do, actually.

 

[MrOtto]

Trying out this Convolver stuff. What is the difference between the 44,1 and 48 kHz minimum phase wav autoeq file?

 

[roskodan]

You need to feed the convoler the right sampling rate, depending on the .wav FIR file rate, for it to EQ correctly. Basically the .wav FIR file works correctly for one specific sampling rate only.

AutoEQ provides .wav FIR files for either 44.1kHz or 48kHz sample rate playback (you can use AutoEQ to generate .wav FIR files for other rates too).

I got the 44.1kHz .wav FIR file, so I set the resampler plugin in foobar2000 to convert everything that is played to 44.1kHz, so that if playing other than 44.1kHz files, it gets resampled to 44.1kHz and gets EQ-ed correctly.

 

[MrOtto]

Ok, thanks.

 

[daftcombo]

Yes, you can put the Resampler (SoX) plug-in just before Convolver.

 

[daftcombo]

According to @edechamps :

in modern versions of Windows, DirectSound and MME use WASAPI Shared behind the scenes

Source: https://github.com/dechamps/FlexASIO/blob/master/CONFIGURATION.md

Also:

Note that WASAPI Shared might still do additional processing besides sample rate and channel count conversion (e.g. sample format conversions, mixing, APOs).

 

[daftcombo]

I'll be trying the default ouput from now on (WASAPI shared) and will report.
The reason for it is I'm worried because Amir stated that Asio4all truncated to 16-bit and, although I mostly listen to redbook quality, I'm worried Foobar2000 volume control + 16-bit truncation might lead to <16-bit depth.

 

[Propheticus]

That only happens for some devices. Asio4all is generic and only needed when you need an ASIO API connection but your device does not come with a native ASIO driver (or you can't be arsed to download it). Other than low latency music production, hardly any use at all really imho.

I agree with Foobars move to Wasapi. Mainly for the automatic switching of samplerate to match.

 

[Pluto]

I agree with Foobars move to Wasapi

A couple of things to note with regard to Foobar's new defaults

• The default appears to be WASAPI shared (not exclusive, even in 1.6.7β)
• The default appears to be WASAPI push. It is generally recommended to use the event option (pull) when possible and to achieve this you have to select preferences/advanced/playback/exclusive mode/use event