TAK 2.2.0 - Alpha release Options

[Destroid]

Well, add about half a second to 02:04:16 and your files exceed the 4 GB limit of the wave file format. There is little i can do besides adding support for something like the Wave64-Format.

Ok, that makes sense. Those large WAV files were a DTS conversion of a concert DVD where all the channels were in one single WAV.

An intermediate step to the "RF64" option might be two different options: a) for the user, divide the concert into chapters; b) the program maybe could accommodate the 6-WAV approach (multiple WAV files of the the different channels) since it's likely each channel is less than 4GB by itself...

[lvqcl]

Good idea! Currently i have no time to look at the source to check if this works for files larger than 4 GB. TAK's stream functions definitely support more than 4 GB, but i am not sure, if i have somewhere coded some restriction. Anyway, you will not be able to decode such a file while TAK only supports the standard wave format as output format...

What can I say...

a) 12.7GB .W64 file was converted to 12.7GB .WAV
b) This .WAV file was converted to 5.56 GB .TAK file (takc.exe -e -ihs -p2 - out.tak < in.wav).
c) Then .TAK file was decoded back to .W64 with foobar2000.

Resulting .w64 file is identical to the source.

[TBeck]

What can I say...

a) 12.7GB .W64 file was converted to 12.7GB .WAV
b) This .WAV file was converted to 5.56 GB .TAK file (takc.exe -e -ihs -p2 - out.tak < in.wav).
c) Then .TAK file was decoded back to .W64 with foobar2000.

Resulting .w64 file is identical to the source.

Thank you! I am surprised. Really nice.

[Destroid]

Didn't have doubt about data integrity of TAK, I just can't benchmark with TAK converting via FB2K piped and get CPU% numbers to account for HDD bottlenecks. Also not really sure why FB2K writes a WAV that it can't read to-the-end correctly when converting from DTS, but the AC3filter utils' "valdec.exe" managed to do so.

This multichannel TAK implementation seems highly successful and since I will be looking to bench with chapter-split files I can run tests to compare the performance vs. ALAC-derived and have results before xmas

Great work, Thomas!

[Destroid]

And here's some test results

Not sure exactly what settings to use with MP4ALS I ended up supplying using no arguments at all. Comparing that with -p0 TAK came up a little less compressed in every file. I re-ran with -p1 and TAK edged out every MP3ALS file, so I'm guessing the closet TAK setting to default MP4ALS for compression ratio is either -p0e or -p0m.

CODE

6ch 48KHz 16bit TAK 2.2.0alpha -p1 MP4alsRM23
WAV filesize ratio enc dec ratio enc dec
----- --------- ------ ------ ------ ------ ------ ------
D1-02 106.15 MB 44.95% 42.80x 81.36x 45.68% 13.00x 34.16x
D1-03 228.73 MB 41.33% 44.94x 87.38x 42.10% 13.15x 35.49x
D1-05 407.76 MB 33.35% 48.04x 97.76x 34.04% 13.66x 36.94x
D1-08 506.63 MB 42.42% 44.79x 86.55x 43.47% 13.00x 36.78x
D1-10 140.89 MB 45.28% 44.01x 87.80x 46.21% 13.12x 36.81x
D2-05 153.60 MB 43.40% 44.08x 81.35x 43.98% 12.95x 34.61x
D2-09 201.40 MB 36.35% 47.12x 90.26x 37.13% 13.45x 37.67x
D2-10 378.39 MB 42.43% 44.40x 87.83x 43.04% 12.88x 37.27x
D2-11 296.93 MB 45.78% 42.19x 84.29x 46.29% 12.79x 36.19x
D2-18 340.04 MB 45.58% 43.54x 85.02x 46.55% 12.84x 36.79x
D3-02 87.80 MB 49.94% 42.81x 76.37x 51.55% 12.87x 37.89x
D3-05 184.21 MB 48.95% 43.27x 81.93x 50.45% 12.91x 36.44x
D3-09 211.97 MB 45.15% 44.26x 86.97x 46.62% 12.86x 35.85x
D3-11 501.20 MB 44.19% 44.04x 84.76x 45.54% 12.90x 36.61x
D3-12 191.09 MB 46.88% 44.00x 83.70x 48.29% 12.91x 36.56x

I did try the settings from the TAK 2.2.0 development thread but MP4ALS encoder was so incredibly slow I abandoned the test early on, but here's what I got:

CODE

6ch 48KHz 16bit TAK 2.2.0alpha -p2m MP4alsRM23 -7 -o160 -t6
WAV filesize ratio enc dec ratio enc dec
----- --------- ------ ------ ------ ------ ------ ------
D1-02 106.15 MB 43.72% 4.184x 76.83x 43.12% 0.206x 7.12x
D1-03 228.73 MB 40.38% 4.231x 74.86x 39.58% 0.209x 7.55x
D1-05 407.76 MB 32.70% 4.357x 84.39x 32.40% 0.220x 8.41x
D1-08 506.63 MB 41.07% 4.087x 75.10x 40.66% 0.211x 9.07x
D1-10 140.89 MB 44.11% 4.134x 79.31x 43.30% 0.209x 7.09x

[TBeck]

Didn't have doubt about data integrity of TAK, I just can't benchmark with TAK converting via FB2K piped and get CPU% numbers to account for HDD bottlenecks.

Doubts wouldn't be totally unjustified. There must be a reason, why i call it an alpha release...

Not sure exactly what settings to use with MP4ALS I ended up supplying using no arguments at all. Comparing that with -p0 TAK came up a little less compressed in every file. I re-ran with -p1 and TAK edged out every MP3ALS file, so I'm guessing the closet TAK setting to default MP4ALS for compression ratio is either -p0e or -p0m.

Usually a -p1x setting is comparable to the default mode of Mpeg4Als. Then both are using 12 predictors (like FLAC -8). The significantly better performance of -p1 provides an indication, that your files provide good opportunities for TAK's channel decorrelation algorithms. And those seem to be more efficient than Mpeg4Als' ones. This advantage may compensate for the lower predictor count (4) of -p0x.

I am really glad to see an advantage of TAK's channel decorrelation algorithms on your files, because most of the work has gone into them. Since my multi channel test file corpus is still small, i really couldn't be sure if the advantage would also manifest in other comparisons.

I did try the settings from the TAK 2.2.0 development thread but MP4ALS encoder was so incredibly slow I abandoned the test early on, but here's what I got:

More nice results.

Could you please test those 5 files with -p4m?

You really made my day. Thank you!

This post has been edited by TBeck: Jun 8 2011, 17:14

[Destroid]

The -p4m results:

CODE

6ch 48KHz 16bit TAK 2.2.0alpha -p2m TAK 2.2.0alpha -p4m MP4alsRM23 -7 -o160 -t6
WAV filesize ratio enc dec ratio enc dec ratio enc dec
----- --------- ------ ------ ------ ------ ------ ------ ------ ------ ------
D1-02 106.15 MB 44.00% 4.184x 76.83x 43.85% 3.079x 62.78x 43.12% 0.206x 7.12x
D1-03 228.73 MB 40.38% 4.231x 74.86x 40.18% 3.121x 60.16x 39.58% 0.209x 7.55x
D1-05 407.76 MB 32.70% 4.357x 84.39x 32.49% 3.312x 61.38x 32.40% 0.220x 8.41x
D1-08 506.63 MB 41.07% 4.087x 75.10x 40.94% 3.018x 68.00x 40.66% 0.211x 9.07x
D1-10 140.89 MB 44.11% 4.134x 79.31x 43.91% 3.159x 62.90% 43.30% 0.209x 7.09x

Note: I initially noticed for the file D1-02 the ratio of -p2m was better than -p4m and found it was a error in calculation or a typo, so it's fixed. I double-checked the other files' ratios for -p2m and there were no further errors/typos.

Speaking of 'no errors,' all my tests with -p0, -p1, p2m and -p4m were decompressed and bit-compared, still no errors found

This post has been edited by Destroid: Jun 8 2011, 18:35

[TBeck]

The -p4m results:
...
Speaking of 'no errors,' all my tests with -p0, -p1, p2m and -p4m were decompressed and bit-compared, still no errors found

This was fast!

Ok, nevertheless Mpeg4Als here compresses better, although at the price of about 15 times slower encoding and about 9 times slower decoding. Not to mention, that it's using bigger frames which also are not independent, so that you for instance would loose much more audio data in case of a single bit error.

That's not meant as an excuse. It simply illustrates, why i don't want to modify TAK in a similar way to improve it's compression. I really like high speeds and small, independent frames of audio data.

And thank you again!

[Destroid]

Ok, nevertheless Mpeg4Als here compresses better, although at the price of about 15 times slower encoding and about 9 times slower decoding. Not to mention, that it's using bigger frames which also are not independent, so that you for instance would loose much more audio data in case of a single bit error.

That price of roughtly 0.5% for the speed isn't worth it, IMO. Although this ALS is unoptimized I have yet to seen any optimizations from this lossless format.

This was fast!

You can thank TAK for that, and the fact I didn't re-run the ALS tests. Actually, once I get my initial test script written it's really easy to change the settings. It was the first run of testing where that error of the -p2m compression ratio came from (sorry about that) and I got the script errors fixed after.

This post has been edited by Destroid: Jun 8 2011, 20:12

[Destroid]

What can I say...

a) 12.7GB .W64 file was converted to 12.7GB .WAV
b) This .WAV file was converted to 5.56 GB .TAK file (takc.exe -e -ihs -p2 - out.tak < in.wav).
c) Then .TAK file was decoded back to .W64 with foobar2000.

Resulting .w64 file is identical to the source.

Thank you! I am surprised. Really nice.

I somehow overlooked lvqcl using pipe input off redirect instead of using FB2K. Now I should start saving up for glasses, apologies.

I also wanted to inquire: would there be interest for other testing 24-bit DTS->WAV files? I can see one point of interest (to measure TAK's ability on that bit-depth) as well as a point of non-interest (persons don't much convert DTS->WAV anyway, much less at 24-bit since there nearly zero audible advantages). At any rate, the DTS decoder gave me the option of 16 or 24 bits. I suppose it's more interesting to test on 24/48 or even 24/96 multichannel material but I have no such material.

[TBeck]

I also wanted to inquire: would there be interest for other testing 24-bit DTS->WAV files? I can see one point of interest (to measure TAK's ability on that bit-depth) as well as a point of non-interest (persons don't much convert DTS->WAV anyway, much less at 24-bit since there nearly zero audible advantages). At any rate, the DTS decoder gave me the option of 16 or 24 bits. I suppose it's more interesting to test on 24/48 or even 24/96 multichannel material but I have no such material.

I agree. Testing of true 24-bit material would be nice, but testing of an 24-bit version of your DTS files would not really be informative. The results may be similar to the 16-bit results. If not, the difference may be attributed to the output stage of the decoder, whose properties are of little interest for us.

Maybe you could test your files with FLAC and WavPack to obtain a more complete picture?

Well, somehow this release is a bit problematic because it adds a feature only few users will be using and this means even less testers. I am not sure, if i will immediately release a final version or insert a beta release. The latter may be tried by more users who possibly don't want to touch alpha releases. In fact this only means a different label, because only the wave reader has been slightly modified.

[bryant]

Maybe you could test your files with FLAC and WavPack to obtain a more complete picture?

I would be happy if someone included WavPack in the comparison, but I believe that the configuration used in the other testing thread (-hhx) might be suboptimal for multichannel and high sampling rates. The first issue is that multichannel WavPack files will use shorter blocks and those end up being less efficient with the longer headers of the very high mode, so it might even be that the regular high mode (-h) is better (and certainly faster to encode and decode). Also, the -x1 to -x3 options are optimized for regular 44.1 KHz audio, so they are not ideal for the LFE channel of a multichannel stream or 96 KHz sampling rates. The -x4 to -x6 options generate custom decorrelation filters, so they work better (sometimes much better) than the lower ones in these cases.

Maybe a good comparison configuration (that would not be too slow at encoding) would be -hx4 (or even -x4 if that was still too slow). These might improve WavPack's compression a little compared to the other test (especially at 96 or 192 KHz), but I'm sure it still won't get close to the values for this new TAK (which is a great piece of work, Thomas, congratulations!)

David

[RastaMan]

Tested TAK 2.2.0 Alpha on an HDCD album "Sara Evans - Stronger (2011)" [Peak extension disabled]

Original album size [compressed using Monkey's Audio High (-c3000)]

256 MB

Size of album after decompressing to 24-bit WAV, then compressed with TAK 2.2.0 Alpha (-p4m)

255 MB

Nice to know that you could listen to this type of HDCD album in foobar2000 without having to use the HDCD plugin, save 1 MB an album in the process, and also use ReplayGain with the TAK files as well.

MD5 - CRC Comparison:

CODE

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\01. Desperately.tak"
MD5: EA41B8E89DCF097D78D6F7F7A52F9CC4
CRC32: 99BB3CD7
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\01. Desperately.wav"
MD5: EA41B8E89DCF097D78D6F7F7A52F9CC4
CRC32: 99BB3CD7
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\02. A Little Bit Stronger.tak"
MD5: F323EB8CC18FA1467A08BD966948A702
CRC32: 5D6CB218
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\02. A Little Bit Stronger.wav"
MD5: F323EB8CC18FA1467A08BD966948A702
CRC32: 5D6CB218
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\03. My Heart Can't Tell You No.tak"
MD5: EFBDB2BA5D2CF4090CBBC52CB859BE60
CRC32: FF639028
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\03. My Heart Can't Tell You No.wav"
MD5: EFBDB2BA5D2CF4090CBBC52CB859BE60
CRC32: FF639028
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\04. Anywhere.tak"
MD5: 486998CCBB67C144CE6786019371E036
CRC32: CD82EF1D
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\04. Anywhere.wav"
MD5: 486998CCBB67C144CE6786019371E036
CRC32: CD82EF1D
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\05. Alone.tak"
MD5: 243DE8364D81FFC8FE33107273E65863
CRC32: B7F53113
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\05. Alone.wav"
MD5: 243DE8364D81FFC8FE33107273E65863
CRC32: B7F53113
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\06. Ticket To Ride.tak"
MD5: C45F19A945CB72CCA8B81150851A932A
CRC32: 76B7E44F
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\06. Ticket To Ride.wav"
MD5: C45F19A945CB72CCA8B81150851A932A
CRC32: 76B7E44F
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\07. Life Without Losing.tak"
MD5: B0526DE645D9B10EC1D2B63D1F6DBB3B
CRC32: 3A2723C1
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\07. Life Without Losing.wav"
MD5: B0526DE645D9B10EC1D2B63D1F6DBB3B
CRC32: 3A2723C1
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\08. What That Drink Cost Me.tak"
MD5: B9E747A5210E29796CAF720035DF2809
CRC32: A5D464CE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\08. What That Drink Cost Me.wav"
MD5: B9E747A5210E29796CAF720035DF2809
CRC32: A5D464CE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\09. Wildfire.tak"
MD5: 045DAAC28BA8FBF5B6F06D84D2C42C5A
CRC32: 5739ECFE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\09. Wildfire.wav"
MD5: 045DAAC28BA8FBF5B6F06D84D2C42C5A
CRC32: 5739ECFE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\10. Born To Fly (Bluegrass Version).tak"
MD5: 1563D1F2FB08BAE898CDC1B3B3EDCE47
CRC32: 75AFC5AC
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\10. Born To Fly (Bluegrass Version).wav"
MD5: 1563D1F2FB08BAE898CDC1B3B3EDCE47
CRC32: 75AFC5AC
No problems found.


All items decoded successfully.

Bit Comparison:

CODE

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\01. Desperately.tak"
MD5: EA41B8E89DCF097D78D6F7F7A52F9CC4
CRC32: 99BB3CD7
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\01. Desperately.wav"
MD5: EA41B8E89DCF097D78D6F7F7A52F9CC4
CRC32: 99BB3CD7
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\02. A Little Bit Stronger.tak"
MD5: F323EB8CC18FA1467A08BD966948A702
CRC32: 5D6CB218
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\02. A Little Bit Stronger.wav"
MD5: F323EB8CC18FA1467A08BD966948A702
CRC32: 5D6CB218
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\03. My Heart Can't Tell You No.tak"
MD5: EFBDB2BA5D2CF4090CBBC52CB859BE60
CRC32: FF639028
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\03. My Heart Can't Tell You No.wav"
MD5: EFBDB2BA5D2CF4090CBBC52CB859BE60
CRC32: FF639028
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\04. Anywhere.tak"
MD5: 486998CCBB67C144CE6786019371E036
CRC32: CD82EF1D
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\04. Anywhere.wav"
MD5: 486998CCBB67C144CE6786019371E036
CRC32: CD82EF1D
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\05. Alone.tak"
MD5: 243DE8364D81FFC8FE33107273E65863
CRC32: B7F53113
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\05. Alone.wav"
MD5: 243DE8364D81FFC8FE33107273E65863
CRC32: B7F53113
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\06. Ticket To Ride.tak"
MD5: C45F19A945CB72CCA8B81150851A932A
CRC32: 76B7E44F
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\06. Ticket To Ride.wav"
MD5: C45F19A945CB72CCA8B81150851A932A
CRC32: 76B7E44F
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\07. Life Without Losing.tak"
MD5: B0526DE645D9B10EC1D2B63D1F6DBB3B
CRC32: 3A2723C1
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\07. Life Without Losing.wav"
MD5: B0526DE645D9B10EC1D2B63D1F6DBB3B
CRC32: 3A2723C1
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\08. What That Drink Cost Me.tak"
MD5: B9E747A5210E29796CAF720035DF2809
CRC32: A5D464CE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\08. What That Drink Cost Me.wav"
MD5: B9E747A5210E29796CAF720035DF2809
CRC32: A5D464CE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\09. Wildfire.tak"
MD5: 045DAAC28BA8FBF5B6F06D84D2C42C5A
CRC32: 5739ECFE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\09. Wildfire.wav"
MD5: 045DAAC28BA8FBF5B6F06D84D2C42C5A
CRC32: 5739ECFE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\10. Born To Fly (Bluegrass Version).tak"
MD5: 1563D1F2FB08BAE898CDC1B3B3EDCE47
CRC32: 75AFC5AC
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\10. Born To Fly (Bluegrass Version).wav"
MD5: 1563D1F2FB08BAE898CDC1B3B3EDCE47
CRC32: 75AFC5AC
No problems found.


All items decoded successfully.

Ran another test on another HDCD album "Carly Simon - Boys In The Trees (1978)' [Peak extension enabled] This time, though, I decompressed the original album into 24-bit WavPack files (Extra High) and compared it to TAK 2.2.0 Alpha to see how it handled block size

WavPack (Extra High)

254 MB

TAK 2.2.0 Alpha (-p4m)

231 MB

That's a 23 MB difference in file size (in TAK 2.2.0 Alpha's favor)

MD5 - CRC Comparison:

CODE

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\01. Desperately.tak"
MD5: EA41B8E89DCF097D78D6F7F7A52F9CC4
CRC32: 99BB3CD7
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\01. Desperately.wav"
MD5: EA41B8E89DCF097D78D6F7F7A52F9CC4
CRC32: 99BB3CD7
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\02. A Little Bit Stronger.tak"
MD5: F323EB8CC18FA1467A08BD966948A702
CRC32: 5D6CB218
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\02. A Little Bit Stronger.wav"
MD5: F323EB8CC18FA1467A08BD966948A702
CRC32: 5D6CB218
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\03. My Heart Can't Tell You No.tak"
MD5: EFBDB2BA5D2CF4090CBBC52CB859BE60
CRC32: FF639028
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\03. My Heart Can't Tell You No.wav"
MD5: EFBDB2BA5D2CF4090CBBC52CB859BE60
CRC32: FF639028
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\04. Anywhere.tak"
MD5: 486998CCBB67C144CE6786019371E036
CRC32: CD82EF1D
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\04. Anywhere.wav"
MD5: 486998CCBB67C144CE6786019371E036
CRC32: CD82EF1D
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\05. Alone.tak"
MD5: 243DE8364D81FFC8FE33107273E65863
CRC32: B7F53113
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\05. Alone.wav"
MD5: 243DE8364D81FFC8FE33107273E65863
CRC32: B7F53113
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\06. Ticket To Ride.tak"
MD5: C45F19A945CB72CCA8B81150851A932A
CRC32: 76B7E44F
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\06. Ticket To Ride.wav"
MD5: C45F19A945CB72CCA8B81150851A932A
CRC32: 76B7E44F
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\07. Life Without Losing.tak"
MD5: B0526DE645D9B10EC1D2B63D1F6DBB3B
CRC32: 3A2723C1
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\07. Life Without Losing.wav"
MD5: B0526DE645D9B10EC1D2B63D1F6DBB3B
CRC32: 3A2723C1
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\08. What That Drink Cost Me.tak"
MD5: B9E747A5210E29796CAF720035DF2809
CRC32: A5D464CE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\08. What That Drink Cost Me.wav"
MD5: B9E747A5210E29796CAF720035DF2809
CRC32: A5D464CE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\09. Wildfire.tak"
MD5: 045DAAC28BA8FBF5B6F06D84D2C42C5A
CRC32: 5739ECFE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\09. Wildfire.wav"
MD5: 045DAAC28BA8FBF5B6F06D84D2C42C5A
CRC32: 5739ECFE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\10. Born To Fly (Bluegrass Version).tak"
MD5: 1563D1F2FB08BAE898CDC1B3B3EDCE47
CRC32: 75AFC5AC
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\10. Born To Fly (Bluegrass Version).wav"
MD5: 1563D1F2FB08BAE898CDC1B3B3EDCE47
CRC32: 75AFC5AC
No problems found.


All items decoded successfully.

Bit Comparison:

CODE

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\01. Desperately.tak"
MD5: EA41B8E89DCF097D78D6F7F7A52F9CC4
CRC32: 99BB3CD7
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\01. Desperately.wav"
MD5: EA41B8E89DCF097D78D6F7F7A52F9CC4
CRC32: 99BB3CD7
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\02. A Little Bit Stronger.tak"
MD5: F323EB8CC18FA1467A08BD966948A702
CRC32: 5D6CB218
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\02. A Little Bit Stronger.wav"
MD5: F323EB8CC18FA1467A08BD966948A702
CRC32: 5D6CB218
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\03. My Heart Can't Tell You No.tak"
MD5: EFBDB2BA5D2CF4090CBBC52CB859BE60
CRC32: FF639028
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\03. My Heart Can't Tell You No.wav"
MD5: EFBDB2BA5D2CF4090CBBC52CB859BE60
CRC32: FF639028
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\04. Anywhere.tak"
MD5: 486998CCBB67C144CE6786019371E036
CRC32: CD82EF1D
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\04. Anywhere.wav"
MD5: 486998CCBB67C144CE6786019371E036
CRC32: CD82EF1D
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\05. Alone.tak"
MD5: 243DE8364D81FFC8FE33107273E65863
CRC32: B7F53113
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\05. Alone.wav"
MD5: 243DE8364D81FFC8FE33107273E65863
CRC32: B7F53113
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\06. Ticket To Ride.tak"
MD5: C45F19A945CB72CCA8B81150851A932A
CRC32: 76B7E44F
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\06. Ticket To Ride.wav"
MD5: C45F19A945CB72CCA8B81150851A932A
CRC32: 76B7E44F
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\07. Life Without Losing.tak"
MD5: B0526DE645D9B10EC1D2B63D1F6DBB3B
CRC32: 3A2723C1
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\07. Life Without Losing.wav"
MD5: B0526DE645D9B10EC1D2B63D1F6DBB3B
CRC32: 3A2723C1
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\08. What That Drink Cost Me.tak"
MD5: B9E747A5210E29796CAF720035DF2809
CRC32: A5D464CE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\08. What That Drink Cost Me.wav"
MD5: B9E747A5210E29796CAF720035DF2809
CRC32: A5D464CE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\09. Wildfire.tak"
MD5: 045DAAC28BA8FBF5B6F06D84D2C42C5A
CRC32: 5739ECFE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\09. Wildfire.wav"
MD5: 045DAAC28BA8FBF5B6F06D84D2C42C5A
CRC32: 5739ECFE
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\10. Born To Fly (Bluegrass Version).tak"
MD5: 1563D1F2FB08BAE898CDC1B3B3EDCE47
CRC32: 75AFC5AC
No problems found.

Item: "C:\Documents and Settings\Default User\My Documents\My Music\Sara Evans\2011 - Stronger (HDCD)\10. Born To Fly (Bluegrass Version).wav"
MD5: 1563D1F2FB08BAE898CDC1B3B3EDCE47
CRC32: 75AFC5AC
No problems found.


All items decoded successfully.

[Destroid]

Maybe you could test your files with FLAC and WavPack to obtain a more complete picture?

I would be happy if someone included WavPack in the comparison, but I believe that the configuration used in the other testing thread (-hhx) might be suboptimal for multichannel and high sampling rates. The first issue is that multichannel WavPack files will use shorter blocks and those end up being less efficient with the longer headers of the very high mode, so it might even be that the regular high mode (-h) is better (and certainly faster to encode and decode). Also, the -x1 to -x3 options are optimized for regular 44.1 KHz audio, so they are not ideal for the LFE channel of a multichannel stream or 96 KHz sampling rates. The -x4 to -x6 options generate custom decorrelation filters, so they work better (sometimes much better) than the lower ones in these cases.

Maybe a good comparison configuration (that would not be too slow at encoding) would be -hx4 (or even -x4 if that was still too slow). These might improve WavPack's compression a little compared to the other test (especially at 96 or 192 KHz), but I'm sure it still won't get close to the values for this new TAK (which is a great piece of work, Thomas, congratulations!)

David

I am embarrassed to admit that I did not remember WavPack handled multichannel (nor FLAC). My apologies. I have a good idea why I might have missed knowing so.

At any rate, I am most happy to run some more tests. Here's what I have got to work with: DTS extracted from DVD's, hard rock music at live venues, 5.1 channels, 48KHz. I suppose (unless advised otherwise) WavPack -h and FLAC -8.

I will also use a slightly different test corpus than previous tests. I noticed D1-05 was noticeably different from other files and plan to focus on those (FYI, I recall D1-05 was a solo acoustic guitar performance w/no vocals).

Thanks to all (especially Thomas) for their feedback

[TBeck]

Maybe a good comparison configuration (that would not be too slow at encoding) would be -hx4 (or even -x4 if that was still too slow). These might improve WavPack's compression a little compared to the other test (especially at 96 or 192 KHz), but I'm sure it still won't get close to the values for this new TAK (which is a great piece of work, Thomas, congratulations!)

Thank you, this means a lot to me!

Tested TAK 2.2.0 Alpha on an HDCD album "Sara Evans - Stronger (2011)" [Peak extension disabled]
...
All items decoded successfully.
...
Ran another test on another HDCD album "Carly Simon - Boys In The Trees (1978)' [Peak extension enabled] This time, though, I decompressed the original album into 24-bit WavPack files (Extra High) and compared it to TAK 2.2.0 Alpha to see how it handled block size
...
All items decoded successfully.

Great! Thank you.

At any rate, I am most happy to run some more tests. Here's what I have got to work with: DTS extracted from DVD's, hard rock music at live venues, 5.1 channels, 48KHz. I suppose (unless advised otherwise) WavPack -h and FLAC -8.

I will also use a slightly different test corpus than previous tests. I noticed D1-05 was noticeably different from other files and plan to focus on those (FYI, I recall D1-05 was a solo acoustic guitar performance w/no vocals).

And... time for another 'Thank you'.

[bryant]

Ran another test on another HDCD album "Carly Simon - Boys In The Trees (1978)' [Peak extension enabled] This time, though, I decompressed the original album into 24-bit WavPack files (Extra High) and compared it to TAK 2.2.0 Alpha to see how it handled block size

WavPack (Extra High)

254 MB

TAK 2.2.0 Alpha (-p4m)

231 MB

That's a 23 MB difference in file size (in TAK 2.2.0 Alpha's favor)

Thanks for including WavPack in your test!

I'm not sure what command-line options you used, but adding "--blocksize=256 --merge-blocks" can make a huge improvement for decoded HDCD files. WavPack blocks are pretty big by default, and so special options are required to get good results from these files (or lossyWAV files, which have similar kind of redundancy).

[RastaMan]

The original tests were run using the highest compression levels of both codecs, without modification. Didn't think to add the relative switches for block size control. So I redid the test again with them in the mix.

Carly Simon - Boys In The Trees (1978) [HDCD] (Peak extension enabled)

TAK -p4m -fsl512

220 MB


WavPack -h -x2 --blocksize=256 --merge-blocks

219 MB


This time around, WavPack has a 1 MB edge. Extremely close.

[Ljubo44]

Any chance to play tak format with windows media player ?

[lvqcl]

Ljubo44: http://liviocavallo.altervista.org/

[Ljubo44]

Tired before, installed from C:/Temp from bat as admin without errors on w7 64. But..

Windows Media Player cannot play the file. The Player might not support the file type or might not support the codec that was used to compress the file.

Tired to replace newer "tak_deco_lib.dll" in windows/system32. But same message from wmp

Windows Media Player cannot play the file. The Player might not support the file type or might not support the codec that was used to compress the file.

anyway thanks for help.

[TBeck]

I want to prepare a final release.

Therefore, if anyone has encountered any problems with this alpha release, now is a good time to report them.

Currently the only difference between this alpha and the final release consists in a modified wave file reader, which supports meta data located at the end of the wave file. Not enough to initiate an additional beta release.

[Manlord]

I dont have any multichannel file, but if any test with standard files is needed I will be happy to help.

[CoRoNe]

*bump*

Hello Thomas,

On the Hydrogenaudio KB I can see cue sheets and cover art are already on your to-do list, but how much of a priority are they?
Currently I'm using WavPack (-hhx -w "Cuesheet=@*.cue" --write-binary-tag "Cover Art (Front)=@*.png"). I've already been experimenting a little with TAK and of course I can add cue sheets and cover art afterwards with Mp3tag because of the APEtag, but once TAK supports it natively I'm really thinking of switching over.
Keep up the good work!

[Destroid]

Therefore, if anyone has encountered any problems with this alpha release, now is a good time to report them.

No problems here, all the TAK files decoded with no differences. Obviously I didn't get my last comparison done, there was desktop computer disaster (an electricity outage in my area made my harddrives disappear from the BIOS, turned out half the power regulating capacitors on the mainboard blew). I'll slate that comparison after final release of 2.2.0.

[TBeck]

I dont have any multichannel file, but if any test with standard files is needed I will be happy to help.

Since i have performed a lot of code optimizations which affect standard files as well, i would be happy if you tested the final release.

*bump*

Hello Thomas,

On the Hydrogenaudio KB I can see cue sheets and cover art are already on your to-do list, but how much of a priority are they?
Currently I'm using WavPack (-hhx -w "Cuesheet=@*.cue" --write-binary-tag "Cover Art (Front)=@*.png"). I've already been experimenting a little with TAK and of course I can add cue sheets and cover art afterwards with Mp3tag because of the APEtag, but once TAK supports it natively I'm really thinking of switching over.
Keep up the good work!

The inclusion of the cuesheet should already work, although not with the '*' placeholder:

CODE

-tt #      Add textual tag item #, where # is a key/value pair: "key=value",
           for instance "TITLE=A nice song". "key=@file" will read the value
           from the text(!) file "file" in the source directory.

The support for binary items like pictures is now on my todo list for V2.2.1. If nothing intervenes.

No problems here, all the TAK files decoded with no differences.

Fine!

Obviously I didn't get my last comparison done, there was desktop computer disaster (an electricity outage in my area made my harddrives disappear from the BIOS, turned out half the power regulating capacitors on the mainboard blew). I'll slate that comparison after final release of 2.2.0.

Good idea!

Thomas