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: CUETools versions 1.9.5 through 2.1.6 (Read 1910171 times) previous topic - next topic
0 Members and 3 Guests are viewing this topic.

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2725 –
HDCDs. CUETools and hdcd.exe seem not to output bit-identical files. Known issue? If so, with which one of them?
I have noticed that both halve the volume, so peaks are the same.

(foo_hdcd agrees with neither,  it seems.)


Quote from: Porcus on
(Decoding HDCD upon ripping to lossless is bad, though.)


Yeah ... I found a few CDs which I had, ignorantly, ripped with HDCD decoding (dBpoweramp, uses hdcd.exe), and re-ripped them. Then I started to play with decoders ...

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2726 –
Quote from: Porcus on
HDCDs. CUETools and hdcd.exe seem not to output bit-identical files. Known issue? If so, with which one of them?
I have noticed that both halve the volume, so peaks are the same.

(foo_hdcd agrees with neither,  it seems.)


Quote from: Porcus on
(Decoding HDCD upon ripping to lossless is bad, though.)


Yeah ... I found a few CDs which I had, ignorantly, ripped with HDCD decoding (dBpoweramp, uses hdcd.exe), and re-ripped them. Then I started to play with decoders ...

Just tried Knopflers Golden Heart that has a good amount of hdcd peak extension and the music content is 100% identical with using CUEtools against hdcd.exe at cmd.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2727 –
Quote from: Wombat on
Quote from: Porcus on
HDCDs. CUETools and hdcd.exe seem not to output bit-identical files. Known issue? If so, with which one of them?
I have noticed that both halve the volume, so peaks are the same. [...]

Just tried Knopflers Golden Heart that has a good amount of hdcd peak extension and the music content is 100% identical with using CUEtools against hdcd.exe at cmd.


No peak extension on this one, but low-level gain: King Crimson: Larks' Tongue (30th Anniversary editions from y2k)


For laziness reporting only track 2 (which is short)


1) CUETools converted (to 24 bits, stop looking after 750) vs HDCD.exe converted:

Compared 7755720 samples.
Differences found: 4752 values, starting at 0:00.000000, peak: 0.0002804 at 0:00.023583, 1ch
Detected offset as 0 samples.


2) As 1), but not stopped looking: same result.


3) foobar2000 converted using foo_hdcd postprocessing, halve output volume: "Always" (because the other two do) - compared with and w/o "Don't reset DSP" :
Bit-identical.


4) fb2k-converted against hdcd.exe-decoded file:

Compared 7755720 samples.
Differences found: 805810 values, starting at 0:00.033651, peak: 0.0000024 at 2:47.694104, 2ch
Detected offset as 0 samples.


4) fb2k-converted file against CUETools-converted file:

Compared 7755720 samples.
Differences found: 808751 values, starting at 0:00.000000, peak: 0.0002804 at 0:00.023583, 1ch
Detected offset as 0 samples.




CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2729 –
I don't have this cd but tried Oldfield, Killing Fields and Camel, A Nod And A Wink. Complety image files and no single bit different. Is this Krimson cd accurately ripped? Only an idea but maybe a ripping error creates undefned values?

my hdcd.exe 143.360 Bytes, md5 e17a38e6723ae2a0fadd6c927c878767
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2730 –
Quote from: Abelenki link=msg=0 date=
CUETools don't work under Windows 10

hello,

first of all, thanks for great application.

i use CUETools occasionally for un-merging some image+cue FLAC's into track FLAC's. but recently i switched to Windows 10, and no versions of CUETools work there, even beta (unstable).

prerequisites are met (.NET 2.0 and 3.5 are installed via Windows Features; VC 2005-2013 are installed, including 2008, of course), but CUETools still don't work.

can you please look into Windows 10 compatibility? i miss CUETools badly. 

it just crashes with "CUE Tools has stopped working" message.

If CUETools cannot access one of the plugins or other support files when it starts, that can make it crash. Check the file/folder permissions where you are trying to run CUETools from or try unpacking the full contents of the CUETools_2.1.5.zip to a different location such as C:\ (you'll end up with C:\CUETools_2.1.5\) and running CUETools from there.
CUETools is working fine so far on my Windows 10 Pro system.
korth

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2731 –
I've uploaded a file with some megabytes off this CD, and can share the link and password with those who want to test. Wombat, you got pm in a minute.


Quote from: Wombat on
Is this Krimson cd accurately ripped? Only an idea but maybe a ripping error creates undefned values?

Could have been a point, but it was verified as Accurate upon ripping.

Funny CUETools log though - but even if there were issues with the last two tracks, it was track #2 I did the more extensive testing on.

Code: [Select]
[CUETools log; Date: 12.08.2015 00:07:09; Version: 2.1.6]
HDCD: peak extend: none, transient filter: none, gain: -4,0dB..0,0dB
[CTDB TOCID: OcB1KSGANlnyXItjuYwf6tst64k-] found.
Track | CTDB Status
  1   | (175/203) Accurately ripped
  2   | (179/203) Accurately ripped
  3   | (179/203) Accurately ripped
  4   | (179/203) Accurately ripped
  5   | (174/203) Accurately ripped
  6   | (  5/203) Accurately ripped, or (169/203) differs in 6711 samples @00:00:02-00:00:08,06:37:21,06:37:74,06:38:16,06:38:34,06:40:42,06:41:02,06:41:20
[AccurateRip ID: 000bd8ff-003da00e-560aed06] found.
Track   [  CRC   |   V2   ] Status
 01     [9343727c|a1b9c3e0] (008+005/486) Accurately ripped
 02     [6d99968e|8642cbfb] (008+005/492) Accurately ripped
 03     [9080d78e|2465d650] (008+005/489) Accurately ripped
 04     [0606b398|95d6cd24] (008+005/493) Accurately ripped
 05     [d15f2c09|23e6644d] (008+005/491) Accurately ripped
 06     [13a1a829|8ad124a0] (008+005/480) Accurately ripped
Offsetted by 744:
 01     [8de29f83] (003/486) Accurately ripped
 02     [db698deb] (003/492) Accurately ripped
 03     [640108d0] (003/489) Accurately ripped
 04     [87b0947b] (003/493) Accurately ripped
 05     [54739ad2] (003/491) Accurately ripped
 06     [0d0672ba] (003/480) Accurately ripped
Offsetted by 1472:
 01     [a19fdd43] (000/486) No match (V2 was not tested)
 02     [6cf35d72] (000/492) No match (V2 was not tested)
 03     [58723100] (000/489) No match (V2 was not tested)
 04     [f53201e2] (000/493) No match (V2 was not tested)
 05     [09c5921c] (000/491) No match (V2 was not tested)
 06     [faa4398d] (000/480) No match (V2 was not tested)
Offsetted by 1488:
 01     [820c7cdf] (003/486) Accurately ripped
 02     [50298290] (003/492) Accurately ripped
 03     [bcfedf17] (003/489) Accurately ripped
 04     [4a581808] (003/493) Accurately ripped
 05     [3ddacb42] (000/491) No match (V2 was not tested)
 06     [c5dd3f33] (000/480) No match (V2 was not tested)
Offsetted by 2071:
 01     [1d1a1196] (046/486) Accurately ripped
 02     [2f111ed0] (046/492) Accurately ripped
 03     [518aa9c1] (047/489) Accurately ripped
 04     [0562979e] (047/493) Accurately ripped
 05     [e4d30a91] (000/491) No match (V2 was not tested)
 06     [61298651] (000/480) No match (V2 was not tested)
Offsetted by 2085:
 01     [0df27c7c] (016/486) Accurately ripped
 02     [d3688d0b] (016/492) Accurately ripped
 03     [d3ccc1ba] (015/489) Accurately ripped
 04     [57931379] (016/493) Accurately ripped
 05     [bc755f47] (000/491) No match (V2 was not tested)
 06     [ecd11f85] (000/480) No match (V2 was not tested)
Offsetted by 2139:
 01     [661e87d9] (016/486) Accurately ripped
 02     [4811206d] (016/492) Accurately ripped
 03     [22386e59] (016/489) Accurately ripped
 04     [b40de01a] (016/493) Accurately ripped
 05     [b7adb2bb] (000/491) No match (V2 was not tested)
 06     [3d0dfae8] (000/480) No match (V2 was not tested)
Offsetted by 2157:
 01     [793c751b] (122/486) Accurately ripped
 02     [fe30e051] (127/492) Accurately ripped
 03     [5523646a] (125/489) Accurately ripped
 04     [4da32041] (127/493) Accurately ripped
 05     [5d6d82f6] (000/491) No match (V2 was not tested)
 06     [25fe02fa] (000/480) No match (V2 was not tested)
Offsetted by 2163:
 01     [140ca029] (002/486) Accurately ripped
 02     [02bccdbd] (002/492) Accurately ripped
 03     [4a98a984] (002/489) Accurately ripped
 04     [b67d6082] (002/493) Accurately ripped
 05     [4f5f9ec8] (000/491) No match (V2 was not tested)
 06     [436d6caf] (000/480) No match (V2 was not tested)

Track Peak [ CRC32  ] [W/O NULL] 
 --   90,1 [76EB613F] [3077AA68]           
 01   90,1 [CA8F86BC] [0FB89500]           
 02   54,6 [A7EBAB43] [69475155]           
 03   65,8 [47D5973F] [5C3E5A4F]           
 04   90,1 [45C56739] [E076D49D]           
 05   90,1 [B713769F] [1511EC8F]           
 06   90,1 [504B09B6] [1B6F261A]         

Curious as I am, I tried a repair. The pressings are patently different, as I lose Accurate-ness with offsets 0 and 744, but achieve an AccurateRip match at six other offsets.

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2732 –
Ok, lets make it simple. The files decoded with CUEtools and my hdcd.exe are identical again.
I found the cjk folder online and tried an older version.
The hdcd i use is the newest from 14.11.2008 simply called hdcd.zip
Using an old verions from 2007 indeed creates a different file.
Near the end of the song we have some seconds constant noise down ~100db difference.
Using the hdcd-0.2-x64 from the same place has the same content as the older versions and so is different to my used non 0.2 with the same newer date.
Since this very silent noise hardly can be called creating different files we have to ask the developer how this comes.
Sine Grigory most likely had some more or less official code for his dll i guess this one is right.
No idea atm how to bring this further and how to contact the dev.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2733 –
Quote from: Wombat on
Ok, lets make it simple. The files decoded with CUEtools and my hdcd.exe are identical again.
I found the cjk folder online and tried an older version.
The hdcd i use is the newest from 14.11.2008 simply called hdcd.zip
Using an old verions from 2007 indeed creates a different file.


The one you use to get bit-identical results, is the one in the zip file, that when unpacked has an md5sum of e17a38e6723ae2a0fadd6c927c878767 ?
That is precisely the one I used to get different results.

So ... WTF? I use CUETools 2.1.6. Everything on 64-bit Windows 7. (Intel Core2 CPU.)

But that November 2008 date is interesting, because my original rip log says October 2008. And the file generated back then, using whatever version followed dBpoweramp version 13, is bit-identical to the one I create today using the November 2008 version.






Quote from: Wombat on
Near the end of the song we have some seconds constant noise down ~100db difference.


Of course - it has to be the low-level gain implementation that makes the difference, because there is no peak extension in the file.

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2734 –
Quote from: Porcus on
The one you use to get bit-identical results, is the one in the zip file, that when unpacked has an md5sum of e17a38e6723ae2a0fadd6c927c878767 ?

Yup, w10pro x64 here. Did you notice the included CUEtools hdcd.dll is from 13.08.2008? At the cjk download location are dlls in hdcd-0.2-dev.zip that are one day newer. I am a bit to lazy to test these with CUEtools atm but could be interesting.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2735 –
Quote from: Wombat on
Did you notice the included CUEtools hdcd.dll is from 13.08.2008? At the cjk download location are dlls in hdcd-0.2-dev.zip that are one day newer. I am a bit to lazy to test these with CUEtools atm but could be interesting.


I did not know, but if I replace CUETools' hdcd.dll file with the cjk file from hdcd-0.2-dev.zip or hdcd-dev-0.2.zip, I still get different results.

How do I write a .bat file that calls a dll?

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2736 –
Quote from: Porcus on
How do I write a .bat file that calls a dll?

No idea, you may use another GUI like eac3to?
I think all we do is fishing in the dark. I don't know who to ask for these differences or how to contact the dev.
One thing for you to be certain may be to capture the output of the Krimson track decoded with WMP to know what way of decoding on your system is the right one.
There must be a howto around for that.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2738 –
Quote from: Porcus on
OK, now I have chosen to bother kode54 for foo_hdcd as well.

Thats the spirit!
Nonetheless these decoders to my understanding must create identical files to WMP output.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2739 –
OK, now I have differences between FLAC and WavPack output ... or maybe I have to reboot between replacing .dll files?!? I need to reset results and try over, but not today.

Quote from: Wombat on
Quote from: Porcus on
OK, now I have chosen to bother kode54 for foo_hdcd as well.

Thats the spirit!


Yeah, I cannot repair, but I am darn good at criticizing those who actually do something



Quote from: Wombat on
Nonetheless these decoders to my understanding must create identical files to WMP output.


But that does not convert, does it?

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2740 –
Quote from: Porcus on
But that does not convert, does it?

All hdcd decoders offered are simply based on mimic the WMP output.
Here someone checks it. http://lukeskaff.com/?p=17
Do the same and you will know what solution of yours is right.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2741 –
Quote from: Wombat on
Quote from: Porcus on
But that does not convert, does it?

All hdcd decoders offered are simply based on mimic the WMP output.
Here someone checks it. http://lukeskaff.com/?p=17
Do the same and you will know what solution of yours is right.


Right. Googling a bit, I found the discussion between the two of you back then.

Well that WMP plugin works only on XP. Which I have, a fragile thing I do not want WMP on. It will have to wait until I get a virtual machine up and going.

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2742 –
Sonata in d'oh minor coming up. The differences are due to applying HDCD track-based or image-based.
I store as tracks. If I use fb2k (w/o HDCD processing) to generate wav+cue, and process the .wav using hdcd.exe, I get the very same results as if I use CUETools.
Presumably, the CUETools output is even more correct than the per-track based HDCD.exe, because a DAC that is fed an S/PDIF stream does not see track boundaries.
(Last time I got a major difference in one track, I cannot explain that; maybe it was after fiddling around with .dlls, maybe I dumped the files into fb2k before completely done causing a read/write conflict ... I think I will write that off without worrying.)


foo_hdcd is not bit-exactly the same as hdcd.exe. Official build:
Differences found: 21246657 values, starting at 0:00.104626, peak: 0.0000091 at 46:21.342222, 1ch
(A new version kode54 built for the discussion, brings it down to ...73.)

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2743 –
Just curious: when CUETools reports "CTDB: verified OK, confidence 10, or verified OK, confidence 3" - what is then the difference between those in the 10-group and those in the 3-group? Offset, like AccurateRip?
The log only says 13/15 (for the lowest track).

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2744 –
Different TOC due to Data track or disc pregap (HTOA) but stored under the same TOCID (the TOCID is calculated using audio Tracks only).

A pressing with a Data track is stored separately under a different CTDBID than one without (even when the Tracks have the same CRCs). Same for a pressing with a disc pregap (HTOA) or if the length of a Data track or disc pregap (HTOA) is different.

Edit: You can see the CTDBID in the accurip log by enabling the Detailed log.
Edit2: Changed phrasing because bad rips also have a different CTDBID but matching tracks are grouped together in the log.
korth

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2745 –
Quote from: korth on
Edit: You can see the CTDBID in the accurip log by enabling the Detailed log.

Yay! Now I realize that I do not get the most verbose log by merely ticking "Verbose" under the AccurateRip tab ...


Quote from: korth on
Edit2: Changed phrasing because bad rips also have a different CTDBID.

OK, so for some CD the detailed log shows three different CTDBIDs;
- one is a different rip with presumably an error, so it gets a different checksum and thus a different CTDBID;
- one verifies mine, and has a data track like mine has;
- and then one with a higher confidence, but no data track.


That gets me curious again; which of the following would/could be the case?
(1) There is a different pressing which has the same audio bits and hence the same audio checksum, but no data session, hence a different TOC and hence a different CTDBID - but CUETools knows from either (a) the audio checksum or (b) the tracklengths that it should be compared to mine?
(2) The other submissions could very well be from the same pressing, but the information concerning the data track is not preserved in a way accessible to CUETools, so it treats it as if it were case (1)?
(3) None of the above?
I verify by ACCURATERIPID extracted from dBpoweramp tags, so even without cuesheets I assume CUETools knows quite a lot from me. 


I have only one rip with both HTOA AND data track, and then all CDTB entries (and all AR entries) perfectly agree.  Anyway, still a bit annoyed that CUETools cannot guess that a track 00 in a folder is a HTOA ...

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2746 –
The TOCID is a calculated TOC (Table of Contents) of the audio tracks only with the 1st track starting at zero (Data track and disc pregap (HTOA) info are excluded in this calculation but are stored separately). Each related CTDBID is stored under a TOCID.

The TOCID would be calculated from your rip and searched, then if found the CRC32 of your rip (except for some leadin and leadout samples) would be checked against any existing CTDBID (using an offset-finding checksum that will detect any offset difference between the rip in the database and your rip, even if your rip contains some errors).
korth

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2747 –
Is there any way to get CUERipper to run a user-definable external process after it finishes a rip?  I've automated a bunch of post-rip workflow, formerly with EAC+REACT, now with my own scripting configured as an EAC "external compressor". Now I'd like to move everything to CUERipper so I can get rid of certain steps I have to do myself if I use EAC, and because it writes UTF-8 cuesheets.  But I can't see any way to have it actually trigger my external processing once it's done with the rip.

The console ripper doesn't really work for me; I want to be able to tune all the metadata before I kick off the rip, not clean it up afterwards.

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2748 –
Not that I know of.

Note: CUERipper only passes the audio to the encoder. Tags are added after all the encoded files are written using taglib-sharp (for most formats).
korth

CUETools versions 1.9.5 through 2.1.5 (current)

Reply #2749 –
What are the possible error messages CUETools can return upon trying to verify corrupt files? I know two: "unsupported subframe coding" and "frame crc mismatch".  (Also of course "Audio format is not Red Book PCM", but I have not seen that arise from corruption.)

I just put a few thousand through it (89 uploads, that's my meager contribution of the week) and I cannot manually read through every line.