Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: lossyWAV 1.4.2 Development (was 1.5.0) (Read 107943 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

lossyWAV 1.5.0 Development

Reply #25
TAK is an alternative,  yes.
But a lossy codec is used for portable devices usually, and for this FLAC is supposed to be the codec of choice for most users.

Well... correct me if I'm wrong, but FLAC is supposed to be the lossless codec of choice for many users. If user needs lossy codec it's much easier (and, probably, more efficient) to use "traditional" MP3, OGG or AAC.

lossyWAV 1.5.0 Development

Reply #26
TAK is an alternative,  yes.
But a lossy codec is used for portable devices usually, and for this FLAC is supposed to be the codec of choice for most users.

Well... correct me if I'm wrong, but FLAC is supposed to be the lossless codec of choice for many users. If user needs lossy codec it's much easier (and, probably, more efficient) to use "traditional" MP3, OGG or AAC.

Well, he is just comparing lossyFLAC and lossyTAK, and saying that lossyTAK might not be chosen because it's basically available only on PC, and on PC lossless is preferable.

That being said, indeed MP3 (or AAC/Vorbis) is far more popular, widely supported, efficient on lower bitrate than lossyFLAC.
However, lossyFLAC also has it's own good points, reasons to choose it.
- FLAC decoder consumes a lot less CPU power than MP3 (good for battery life).
- FLAC is always gapless. No delays, no extra padding. Even by players that doesn't support gapless playback of MP3 or AAC through LAME/iTunes specific tag hacks. Even when remuxed into another container such as Matroska.

lossyWAV 1.5.0 Development

Reply #27
TAK is an alternative,  yes.
But a lossy codec is used for portable devices usually, and for this FLAC is supposed to be the codec of choice for most users.

Well... correct me if I'm wrong, but FLAC is supposed to be the lossless codec of choice for many users. If user needs lossy codec it's much easier (and, probably, more efficient) to use "traditional" MP3, OGG or AAC.

Well, he is just comparing lossyFLAC and lossyTAK, and saying that lossyTAK might not be chosen because it's basically available only on PC, and on PC lossless is preferable.

That being said, indeed MP3 (or AAC/Vorbis) is far more popular, widely supported, efficient on lower bitrate than lossyFLAC.
However, lossyFLAC also has it's own good points, reasons to choose it.
- FLAC decoder consumes a lot less CPU power than MP3 (good for battery life).
- FLAC is always gapless. No delays, no extra padding. Even by players that doesn't support gapless playback of MP3 or AAC through LAME/iTunes specific tag hacks. Even when remuxed into another container such as Matroska.


And also LossyFLAC is (to my ears) absolutely 100% transparent at standard settings and the bitrate is half the bitrate of the original. MP3 and AAC are never transparent. Off course this leaves room for discussion, but I prefer LossyFLAC above any lossy codec. Still I have my entire collection on lossless flac, because 1: Space is not an issue on my NAS servers.
2: LossyFLAC development is still going on so I don't want to turn the switch yet.

For portable use I do use LossyFLAC.

lossyWAV 1.5.0 Development

Reply #28
Current LossyWAV algorithm (with noise shaping etc) is quite slow, so the encoding time doesn't matter much. Besides, LossyWAV-specific codec was removed from the release version of TAK.

lossyWAV 1.5.0 Development

Reply #29
Not thread specific but, once again, thank you for creating this codec. I now it for archival because I have limited disk space and I haven't heard a difference.

lossyWAV 1.5.0 Development

Reply #30
My collection is entirely LossyFLAC, except some hard-to-compress tracks such as piano solo.

lossyWAV 1.5.0 Development

Reply #31
Any updates in development? Maybe plans?

Thanks

lossyWAV 1.5.0 Development

Reply #32
Any updates in development? Maybe plans?

Thanks


There's nothing in particular being worked on at the moment - so unless I have a "great idea" that turns into a development theme shortly then I'll be releasing the amended version as 1.4.1 next month.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

lossyWAV 1.5.0 Development

Reply #33
lossyWAV beta 1.4.1g attached to post #1 in this thread.

I've compiled the 32-bit and 64-bit versions with MinGW 5.2.0 (and added a few more compiler directives) - the 64-bit version in particular sees a fairly large speed boost on my development machine - the internal FFT routines seem to be about as fast as when FFTW is used.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

lossyWAV 1.5.0 Development

Reply #34
I'm very newbie sorry, I tried to convert the same song from FLAC to lossyFlac with lossyWAV v1.4.0 and then with lossyWAV v1.4.1g and the result is the same file. I mean exactly the same, tested with MD5 check.

Using Foobar "std": lossywav - --quality standard --silent --stdout|flac - -b 512 -5 -f -o%d --ignore-chunk-sizes

Is this normal? thanks

lossyWAV 1.5.0 Development

Reply #35
I'm very newbie sorry, I tried to convert the same song from FLAC to lossyFlac with lossyWAV v1.4.0 and then with lossyWAV v1.4.1g and the result is the same file. I mean exactly the same, tested with MD5 check.

Using Foobar "std": lossywav - --quality standard --silent --stdout|flac - -b 512 -5 -f -o%d --ignore-chunk-sizes

Is this normal? thanks


Entirely normal (good to hear, in fact) - there has been no change to the algorithm used in the DSP.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

lossyWAV 1.5.0 Development

Reply #36
Ha, just about to knock your door hard for the old drop dead date and here is a new one.

Thanks, Nick.

lossyWAV 1.5.0 Development

Reply #37
I can report an extremely good time gain, in my test I've converted in lossyFlac std the same 9 files, original Flac 950+ kbps total size 273mb:
lossyWAV v1.4.0 time = 3:15
lossyWAV v1.4.1g time = 2:05

total size came down to 122mb, same md5 for both lossyWav (as usual now I can say!)

Thank you very much for this update, Nick!
@Musique-Rabbit thank you also for answering to my pm time ago

lossyWAV 1.5.0 Development

Reply #38
Glad to hear that the performance improvement exists on more than one machine....

To reduce file-size further, I use "-P=4096" to reduce the FLAC file padding from a default 65536 bytes and use "-8" as the FLAC compression preset. This reduces filesize with a small increase in compression time.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

lossyWAV 1.5.0 Development

Reply #39
Glad to hear that the performance improvement exists on more than one machine....

To reduce file-size further, I use "-P=4096" to reduce the FLAC file padding from a default 65536 bytes and use "-8" as the FLAC compression preset. This reduces filesize with a small increase in compression time.


Hi Nick,

Do you know what happens when you exceed the padding size after editing? Normally the padding range is overwritten when adding metadata, but when there is more than 4096 bytes to be written, will the file be written over with a larger padding block of will changes fail from that moment on...

Thanks

lossyWAV 1.5.0 Development

Reply #40
I tend to do my conversion / tagging in foobar2000 - it seems to checks for padding overflows in some way and rewrites the file with more padding if required.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

lossyWAV 1.5.0 Development

Reply #41
lossyWAV beta 1.4.1h attached to post #1 in this thread.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

lossyWAV 1.5.0 Development

Reply #42
There's nothing in particular being worked on at the moment - so unless I have a "great idea" that turns into a development theme shortly then I'll be releasing the amended version as 1.4.1 next month.

That is good because I'd like the fix of checking function to be available in stable branch. Also I did some ABX tests on extraportable setting. From 7 sample only one fails, other are transparent. I can't say something wrong with my equipment or ears because I am able to ABX Lame or Vorbis 128 bitrate on many my samples.

Awesome tool I found. Thank you Nick.C and guys for your work!

Re: lossyWAV 1.5.0 Development

Reply #43
lossyWAV beta 1.4.1i attached to post #1 in this thread.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

Re: lossyWAV 1.5.0 Development

Reply #44
lossyWAV beta 1.4.1j attached to post #1 in this thread.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

Re: lossyWAV 1.5.0 Development

Reply #45
Any comments regarding the new hybrid shaping would be very much appreciated.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

Re: lossyWAV 1.5.0 Development

Reply #46
Trying to get lossyWAV working in foobar2000. I based my settings on the one's in Nick's signature. I keep getting an error when converting.

Code: [Select]
An error occurred while writing to file (The encoder has terminated prematurely with code 0 (0x00000000); please re-check parameters) 

Additional information:
  Encoder stream format: 96000Hz / 2ch / 24bps
  Command line: "C:\Windows\SysWOW64\cmd.exe" /d /c C:\"Program Files"\lossyWAV.exe -q C -a 4 -s h -A --feedback 1 --stdout|C:\"Program Files"\flac.exe -8 -e -p -b 512 -P=4096 -S -o%d

 

Re: lossyWAV 1.5.0 Development

Reply #47
Might be what I have reported before and that is to change
Code: [Select]
-o%d
to
Code: [Select]
-o %d
. Please let me know if that worked.

Re: lossyWAV 1.5.0 Development

Reply #48
"Hybrid" adaptive noise shaping (and the other changes in 1.4.1j):
Is there some (more detailled) information on what they do to read other than the change log?

Re: lossyWAV 1.5.0 Development

Reply #49
And, a minor cosmetic "glitch":
There's a "\n" missing in the long help text (-L) at the shaping section (before "n, nowarp"):
Code: [Select]
-s, --shaping        modify settings for noise shaping used in bit-removal:
       a, altfilter  enable alternative adaptive shaping filter method.
       c, cubic      enable cubic interpolation when defining filter shape
       e, extra      additional white noise to add during creation of filter
       f, fixed      disable adaptive noise shaping (use fixed shaping)
       h, hybrid     enable hybrid alternative to default adaptive noise shaping
                     method. Uses all available calculated analyses to create
                     the desired noise filter shape rather than only those for
                     1.5ms and 20ms FFT analyses.       n, nowarp     disable warped noise shaping (use linear frequency shaping)