Value of basic ripper use in comparison ripper?

Discuss the current and future development of Max.
Post Reply
khead
Posts: 66
Joined: Mon Jan 23, 2006 3:40 pm

Value of basic ripper use in comparison ripper?

Post by khead » Sat Sep 09, 2006 1:58 pm

I’ve put in an insane amount of time trying to objectively come up an answer to what is the best ripping option with Max.
That is I have six drives available (one non catching), and have ripped two different problem CDs (one quite scratched and one with damage to the foil layer) with every about every possible method - over different 20 rips of each CD - and imported files into Audacity to invert and mix waveforms to objectively compare differences.

One thing I can safely say at this point is that the comparison ripper can pass through timing errors right now with it’s standard settings.
Might it not be better to use paranoia, or overlap checking as a basis for the comparison ripper to ensure you‘re not passing through repeating bad read errors - skipped or dropped frames?

There are also some pretty bad problems with the basic, and comparison ripper with external drives -- Hard systen freeze on my G3 and G4 systems requiring a power button restart. And failure to work with a Plexor on my G5.

User avatar
sbooth
Site Admin
Posts: 2445
Joined: Fri Dec 23, 2005 7:45 am
Location: USA
Contact:

Post by sbooth » Sat Sep 09, 2006 4:25 pm

Was this with Max 0.6.1 or with one of the svn versions? I added C2 error detection to the comparison ripper after the release of 0.6.1 which might help locate some of the problematic reads.

I also have been spending a lot of time thinking about the best way to improve ripping in Max- but I haven't really come up with a good answer yet.

khead
Posts: 66
Joined: Mon Jan 23, 2006 3:40 pm

Post by khead » Sat Sep 09, 2006 6:12 pm

That was with C2 checking on.
I’ve been doing this over the last week or so, and have been keeping up with the latest builds.

I really haven’t been able to figure out exactly what a C2 error is .... other than an error ....

I do know that different drives are considered to be less reliable than others in how accurately they report them.

Still I kinda don’t think they would affect skipping, or the sort of time shifting drop or drift corrections that cdparanoia performs.
Thats why I wonder if it might be best to pretty much do what rubyripper does - Do the comparisons off the various cdparanoia rip options.
Its just a cleaner rip to begin with at a basic level, and the option to perform multiple comparisons on top of that would only seem to be a good thing.

Another thing that seems to be emerging from my comparisons so far is that the catching issue seems to be overblown - at least thats the way it looks to me.

Have you tried the subtracting waveform trick?
It can really reveal a lot of whats going on.
Let me know if you need any info on that.

khead
Posts: 66
Joined: Mon Jan 23, 2006 3:40 pm

Post by khead » Wed Sep 20, 2006 8:46 pm

Just wanted to follow up on the G3-G4 freezing issue.

It seems to be gone with the 801 build.
I’d been avoiding trying to freeze those systems, and hadn’t checked on the recent builds.

khead
Posts: 66
Joined: Mon Jan 23, 2006 3:40 pm

Post by khead » Fri Sep 22, 2006 12:06 am

Yikes!

I spoke too soon on the freezing issue.
I froze my G4 eMac again with the basic ripper.

User avatar
sbooth
Site Admin
Posts: 2445
Joined: Fri Dec 23, 2005 7:45 am
Location: USA
Contact:

Post by sbooth » Fri Sep 22, 2006 12:44 am

khead wrote:Yikes!

I spoke too soon on the freezing issue.
I froze my G4 eMac again with the basic ripper.
I'll look into this once I get the code base stabilized...

khead
Posts: 66
Joined: Mon Jan 23, 2006 3:40 pm

Post by khead » Sat Sep 30, 2006 3:29 pm

I've taken a screenshot of one example of how the different rippers in MAX deal with a problem CD that is prone to skipping.

http://homepage.mac.com/davesprou/PhotoAlbum15.html

What this shows is the results of ripping the same track ripped three different times on the same drive with:

1- Comparison Ripper C2 off
2- Comparison Ripper C2 on
3- Paranoia

After doing these rips, I repaired the CD to where I get perfect rips (matching results across several drives).
Then, in the audio editor, I inverted the perfect rip and mixed it with each of the original rips. This cancels out everything that is the same between the files, and leaves only the differences. The differences in this case are the ripping defects.

The two Comparison Ripper columns show where skipping occurs - the wave forms are jumbles out of phase, and don’t cancel each other out. These are plainly audible problems.

The Paranoia column shows a couple of slight clicks in the same area. These sort of defects can often be difficult to impossible to hear, and this is where the invert and mix method can help objectively compare relative rip qualities.

So what this shows here is that in with this problem track, on these particular rips, with this particular drive (LiteOn SHOR 5239), the Comparison ripper skips, and Paranoia does not.
It also looks like C2 off skips a bit less than C2 on.

Also, as far as I know, the LiteOn is a catching drive. It seems to work fine with Paranoia.
I did this exact same test with several different drives, and I’m not seeing any clear correlation between catching and how Paranoia performs as far as skipping. In fact the only drive that I have that pretty certainly caches zero audio skips quite badly on this track - my Pioneer which does cache doesn’t skip....

User avatar
Fuga
Posts: 391
Joined: Mon Jun 05, 2006 8:30 pm
Location: Texas

Post by Fuga » Sat Sep 30, 2006 7:14 pm

Would be curious to see this rip from EAC.

User avatar
sbooth
Site Admin
Posts: 2445
Joined: Fri Dec 23, 2005 7:45 am
Location: USA
Contact:

Post by sbooth » Sun Oct 01, 2006 12:13 am

That screenshot hurts my feelings- I thought the comparison ripper was quite good :).

RonaldPR
Posts: 433
Joined: Tue May 30, 2006 8:27 am
Location: Amsterdam, Netherlands

Post by RonaldPR » Sun Oct 01, 2006 9:27 am

I use Max on my PowerBook G4, with MATSHITADVD-R UJ-816 drive, to rip my CDs and encode them to FLAC. I never did any kind of tests like "khead" did, I just listened to the results. So far, the results with Paranoia ripping were perfect each time, with no audible problems at all. The results with Comparison ripping were disappointing with several CDs, sometimes with serious audible defects or artifacts. All those CDs are without visible scratches or stains and used to play fine in a 10 year old Denon CD player. I stopped trying Comparison ripping and now use Paranoia ripping only.

khead
Posts: 66
Joined: Mon Jan 23, 2006 3:40 pm

Post by khead » Sun Oct 01, 2006 5:26 pm

The example I posted above is the result from a circular scratch that comes from CD drive mechanisms. These kind of scratches are the worst kind because they run directly along the data track on the CD, and are the most likely cause of skipping problems.
They also aren't always all that easy to see just by taking a quick look at the CD - this one wasn't very visible.

I've been running through these kind of tests for a few weeks now and have been pretty much just torture testing with problem disks. I'd have to say that I haven't seen a case yet where the Comparison ripper has done better than Paranoia --- at least with the drives I've been using.
But after posting this yesterday I started trying something different - using a new scratch free CD. Guess what....
Using Paranoia has been giving inconsistent rips across a couple of drives while using the Comparison ripper has resulted in perfect matches over several trials.
The problems I've seen so far with Paranoia in this case have not been anything that could be reasonably considered audible, but still this is a surprise result.
I think I might start using the Comparison ripper on really clean disks and Paranoia on the more beat-up ones --- at least on some drives... Need to look into this more.

From all I can tell, developing a secure ripper is a pretty tough project. EAC has been getting seriously developed for years, and still not entirely perfect.
I've been sort of following the goings on with the dBpowerAMP ripper --- lots of work with direct testing with many drives.
It seems the huge variabilities between how different drives work is the biggest issue.

I think for us MAX users it be a have to be a bit of a trial and error process of finding out what drives work best with which rippers for what type of CD problems.
I found the same thing as Roland with my Matshita Superdrive - It's a bad drive with iTunes, made really (and I mean, REALLY) bad by the Comparison ripper, made quite good by Paranoia.

And finally - If the comparison component could be grafted onto Paranoia, I think that would be a great thing to do.
It would be nice to have a Hyperparanoia option for real problem disks.

Again - Thanks for all your work on this program.

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests