PerfectTUNES discussion

[Porcus]

Re: PerfectTUNES discussion

Screw my last message, I'm upgrading again


Q: Was this version supposed to ditch my database, having to listen to the entire library again? (Would it matter whether I uninstalled before reinstalling?)

[dreamslides]

Re: PerfectTUNES discussion

Can I get some more screenshots of PerfectTunes? Will it replace dbpoweramp?

Should I use it to RIP my collection instead?! HELP!

[dbfan]

Re: PerfectTUNES discussion

Perfecttunes does not rip cds, or convert music, so will not replace dbpoweramp.

This update to perfecttunes will rescan the whole collection again.

[Porcus]

Re: PerfectTUNES discussion

This update to perfecttunes will rescan the whole collection again.

Ah well. Resorting to a subset for testing then ...


Wishlist for options for the final version:

- Those of us who use EAC or dBpoweramp, which submit to the database, would probably like to set a threshold at 2, so that PerfectTunes does not verify as Accurate based on our files matching our own submissions only. Indeed, we might want to reduce positive AR scores by 1 on some (search-criteria based) part of our collection.

- In particular, those of us who have a licenced dBpoweramp, might even sacrifice our privacy if it still is possible to identify that our machine submitted this particular track -- maybe by linking PerfectTunes to our dMC installation?

- And those of us who use dBpoweramp and write AccurateRip results to our tags, might want to either update with new tags, or another tag for retro-checked status (in which case I suggest using the same tag name as CUETools does). Also, we might want an option to skip tracks that originally (or at last check) verified as Accurate.

[Spoon]

Re: PerfectTUNES discussion

The AccurateRip checker is very efficient with caching, so it is as good as skipping already tested tracks.

Good idea about the conf of 2.

[Porcus]

Re: PerfectTUNES discussion

First test round with the AccurateRip checker. Some debugging seems to be in order:


- It claims many of my files to be corrupt, where every other tester I've tried, verify as OK. (Those are flac -t, the AudioTester.exe tool, and foobar2000's integrity verifier -- the latter probably only uses flac -t).
Indeed, this includes cases where foobar2000's AccurateRip verifier confirm the album as Accurate (and, since I have submitted myself: sometimes with status 02 or higher).

- It is inconsistent about 24-bit files. Yeah, I regret converting with hdcd.exe in the chain.
But some of these files are reported uncheckable as 'NOt Lossless CD Quality', while in at least one case, the album is checked (and of course one track reported inaccurate).

- It misses tracks. Stating 'Incomplete album'. Here too, this includes cases where foobar2000's AR verifier confirms the album as Accurate (and no data tracks nor HTOA).
(Of course, it *also* includes cases where a track is unreadable or 24-bit.)

- At least one album is reported twice. (With same status though, so they show up next to each other -- this is merely a nuissance.)


All the issues seem to occur both on various artists (all files in same folder) and ordinary releases, except I didn't search for more than one case where a 24-bit-file was actually checked for AR status.

[dbfan]

Re: PerfectTUNES discussion

If an album is reported twice, either it is in two separate folders, or it is stored in two different file types.

Please send on Hdcd file which is incorrectly looked up, and one flac which is reported as corrupted.

[Porcus]

Re: PerfectTUNES discussion

pm for *********

[Arnold_Layne]

Re: PerfectTUNES discussion

Just a suggestion. Is it possible for you to include a context menu entry for the perfect tunes accurate rip checker similar to dbpoweramp's ID tag and music converter?

A_L

[bort]

Re: PerfectTUNES discussion

Which version of AccurateRip is used by PerfectTunes? I've got tons of albums that are reported as bad or missing rips by Beta 5, but the albums verify against AR v1 and/or v2 successfully.

[Spoon]

Re: PerfectTUNES discussion

It uses the normal AccurateRip, but it has to crossreference a CD TOC to get the AccurateRip lookup ID, this TOC database is filling as we speak.

[bort]

Re: PerfectTUNES discussion

That is very cool. How long will the TOC database take to fill? Will it be possible to save a .toc file in each album directory after it is found?

[TJOz]

Re: PerfectTUNES discussion

Spoon

Any chance of:
* adding the fingerprint id as a tag to a track and/or
* listing duplicates in DeDup where the artist and title are not an exact match (this will assist in identifying poorly tagged tracks)

Tony

[Spoon]

Re: PerfectTUNES discussion

It is difficult to say how long the toc filling process will take (it is all done from accuraterip outside of perfecttunes).

---
The fingerprint id is unique to your system for dedup, so would not really mean anything.

Other request noted.

[TedSmith]

Re: PerfectTUNES discussion

After a couple of days of use here are my notes:

De-Dup:
De-Dup at times gives some completely bogus false positives: in one case it thought that 30 seconds of spoken text sounded like a 5 minute symphonic piece, and also that they sounded like another 30 second or so snippet of a jazz trio. Another group had essentially all of the tracks of a CD along with some other tracks containing rain, white noise, and some other test tones.

De-Dup found 5 or 6 albums which were obviously ripped twice, thanks.

AccurateRip:
A nit: The "Not Lossless CD Quality" wording is confusing: I have many more higher than CD resolution tracks than lower than CD resolution tracks.

In the Cannot Check view, the Albums Not Present... view and the Total Albums view you can't tell which album they are talking about, e.g. I have multiple albums with essentially the same metadata (separate releases mastered slightly differently) and also 16/44.1, 24/96, 24/88.1, etc., and DSD versions of some albums. An open folder button like the De-Dup would be nice (or simply clicking on the album cover.)

Tho rescanning my library from scratch takes quite a while it doesn't take too long to do an incremental scan. HOWEVER the "Checking with AccurateRip" pass takes hours every time: if this delay is unavoidable a progress bar or at least a time estimate would be nice.

It also would be nice to get a log or to at least be able to save the corrupted files list so I can fix them over a few days without running the AccurateRip over if I have to reboot.

I have a number tracks which are reported as corrupted: I don't know how: they play fine, the dbPoweramp Test Conversion doesn't report any problems, "flac --test --warnings-as-errors xxx.flac" doesn't find any problems, and they open in Adobe Audition and look fine... I rescanned from scratch with PerfectTUNES and now I get a different set of "corrupted" files.

Album Art:
It's too easy to click on the wrong cover while I'm trying to use the mouse to scroll (The scroll bar is pretty narrow on my monitor).

Sometimes I'd like to compare a few possible covers at full resolution before I pick one.

It doesn't take too much time to find possible new covers for each album, but after trying to fix every missing and low resolution cover in my collection I didn't find a single one that was better than the ones I already had. I guess much of my collection would be considered obscure.