CD Spec vs Ripper Configuration

**EliC** · September 04, 2009, 04:08 PM

Re: CD Spec vs Ripper Configuration

Originally posted by BigN

RIPPING Ripping goes ahead OK with 100% AccurateRip on Pass 1 (Confidence 19) for the first disc... but CD Ripper seems to have done this without using either C2 Error Pointers or Cache support!

Is the disc you ripped the one with the damage? Try ripping the discs with damage that dont report any C2 errors. My experience with many of the latest TSST drives is that they "see through" a lot of damage and are able to read the discs as though there was no damage. If thats the case, they shouldn't report C2 errors. Try a disc that the drive can't rip and see if C2 is reported.

**Spoon** · September 04, 2009, 08:34 PM

Re: CD Spec vs Ripper Configuration

Also be aware that some drives report that they support c2 on the technical page, when infact they do not (they feature lie). Make sure there is real damage on the disc, like in the setup guide (1 cm wide black triangle, which a drive cannot correct for)

>Ripping goes ahead OK with 100% AccurateRip on Pass 1 (Confidence 19) for the first disc... but CD Ripper seems to have done this without using either C2 Error Pointers or Cache support!

If there were no errors then c2 or cache invalidation would not have been used.

**BigN** · September 07, 2009, 01:24 PM

Re: CD Spec vs Ripper Configuration

Thanks for quick replies from a grateful dBpoweramp noob! Elic, the first disc I ripped was NOT the simulated damaged audio CD I used to [Detect C2 Support].

It was a perfectly good choral work disc , Allegri Miserere by the Tallis Scholars. I assume your comment that TSST drives can "see through" a lot of damage augurs well for future error-free rips on this drive.

I have not been idle. I had another post ready very late on Friday night, but a careless key press deleted the lot and I couldn't face typing it again so late! Went to bed, and have since tried new things and learnt lots about this excellent application.

DISC 1 (as per my previous post)
I suspect Spoon is probably correct and that although it says it can, I do not now believe the TS-H352C actually does support C2 Error Pointers. I can assure him that the simulated damaged audio disc certainly has the 1Cm wide black triangle as per the Setup Guide, plus some extra very deep gouges added! I noticed also that when [Detect C2 Support] test is active, this drive's LED does not blink at all, as if the disc is not actually reading.
Q1) Is this as it should be?
Q2) Neither of you answered my second point... are Buffer Size and Cache the same thing when applied to a CD drive spec?

DISC 2
I next ripped Bruckner's Mass No 3, and AccurateRip Confidence ranged from 2 to 4 for the six tracks. In doing so, I came across the useful Text Document Digital Audio Extraction Log file for this rip and it clearly shows:-
"Ripping with Drive 'E:[TSSTcorp - DVD-ROM], Drive Offset: 6**, Over-read Lead-in/out: No, AccurateRip: Active, Using C2: No, Cache: None, FUA Cache Invalidate: No". ** +6 offest is now confirmed as correct for this drive model by reference to "AccurateRip CD Drives Offset List".

DISC 3
Palestrina: Missae Papae Marcelli has every track disagreeing with at least two other people, but nevertheless tracks are listed as Secure. As the PerfectMeta album art is different from my CD jewel case, it is reasonable to suppose this is simply a different pressing. Indeed the Extraction Log last line says "12 Tracks Ripped Securely (AccurateRip: Different Pressing?". So AR agrees. This software reallly is quite impressive so far.

DISC 4
Berlioz Te Deum... a disc that I know may have some problems on some tracks from earlier experimentation!
First, I ripped the tracks in Burst mode to see what would happen:-
Track 1 Inaccurate (4)
Track 2 Accurate (4)
Track 3 Inaccurate (3)
Track 4 Inaccurate (4)
Track 5 Inaccurate (4)
Track 6 Inaccurate (3)
Repeated in Ultra Secure mode:-
Track 1 Accurate (4)
Track 2 Accurate (4)
Track 3 Accurate (3)
Track 4 AR (4) i Secure
Track 5 AR Error
Track 6 AR Error
Extraction Log says "6 Tracks Ripped: 3 Accurate, 1 Secure (Warning), 2 Could Not Be Ripped".
Significantly there are a few scratches on the outer edge of this disc, although I would not have thought bad enough to cause catastrophic read fails! But if my understanding is correct, the CD Red Book spec places Track 1 on the data spiral inner, so problems with Tracks 5 & 6 might make sense for this particular CD. The Log Drive & Settings first three lines are as previously with the other discs, notably that no C2 or Cache has been used, despite there being plenty of errors around.

RESULT I am now convinced that this TSSTcorp drive uses neither C2 Error Pointers nor Cache.

SECOND DRIVE
At this stage I was struggling a bit as, despite repeated attempts, Tracks 5 & 6 could not be ripped at all on the TSSTcorp drive.
Q3) Maybe there is something else I should have tried?
I then realised that perhaps it might be worth using my second drive, a Philips DVD+/-RW DVD8701 model.
First Offset is set to +618, again conforms with the AccurateRip List alluded to earlier. So no problem.

dBpoweramp CD Ripper: Options pane, and clicking on CD ROM [Technical] shows:-
Manufacturer: Philips
CD Drive: DVD+ROM DVD8701
Firmware: 5D24
Serial: shown
Maximum Speed: 8467KB/sec (x47)
Current Speed: 8467KB/sec (x47)
Spin-down After: 4 minutes
Buffer Size: 2MB
Accurate Stream: Yes
C2 Error Pointers: No
Reads ISRC: Yes
Read UPC: Yes

So this drive is a) potentially faster b) definitely DOES NOT use C2 Error Pointers. But what about Cache? Running cachex -i -c -n 50 F: lists 50 passes with results varying from 4KB/2 Sectors to 18KB/8 Sectors, but with most at 13KB/6 Sectors. So despite the Technical Spec showing a Buffer Size of 2MB, I reckon it is actually very small, and therefore leave the program's Drive Read Cache 1024KB "no-brainer" default setting.

DISC 4 RE-RIP ON DVD8701, DRIVE READ CACHE = 1024KB, C2 Error Pointers for Error Detection = Not checked
Ultra Secure Mode:-
Track 1 Accurate (4)
Track 2 Accurate (4)
Track 3 Accurate (3)
Track 4 Accurate (4)
Track 5 AR (4) i Secure
Track 6 AR Error {after 2 Passes, 4 Ultra Passes and ~1600 re-ripped frames}.
While it took ages (overnight), it still failed to rip all tracks satisfactorily.

DISC 4 RE-RIP ON DVD8701, DRIVE READ CACHE = None, C2 Error Pointers for Error Detection = Not checked
Tried again with the Drive Read Cache set to None, although technically according to Cache Explorer it should not be:-
Ultra Secure Mode:-
Track 1 Accurate (4)
Track 2 Accurate (4)
Track 3 Accurate (3)
Track 4 Accurate (4)
Track 5 AR (4) i Secure {after 2 Passes, 4 Ultra Passes and ~21 re-ripped frames}.
Track 6 AR (3) i Secure {after 2 Passes, 4 Ultra Passes and ~1510 re-ripped frames}.
So finally all tracks have been ripped successfully. This latter rip was repeated several times, with entirely consistent results for all Tracks, and unlike the previous 1024KB Cache setting, it did not take too long to complete the task.
Q4) Would someone care to explain this apparent anomoly?
Q5) Can you please confirm that my Configuration settings are now optimized?

I just came across another thread "Ultra Secure Settings" where trtlrock seems to be asking very similar questions re C2 for the same model TSSTcorp drive. Will now try to read and understand this in light of my rapidly improving knowledge. I can't help but feel that a bit more detailed explanation in the Setup Guide for C2, Caching etc may be useful to complete ripping newbies such as me!

FINALLY Must commend the comprehensive features of dBpoweramp, most of which I have only glanced at so far, also the attractive GUIs, but most importantly the quick response and detailed replies from the various enthusiasts and gurus happy to share their knowledge. It all helps reassure me that my current project to move from my Imerge SoundServer system to dBpoweramp (for ripping), MediaMonkey (for music and playlist management) and Sonos for audio streaming is worth pursuing, as good forum support will be the key to success. All these products seem well-supported in this respect.

Keep up the good work!

**bhoar** · September 07, 2009, 06:41 PM

Re: CD Spec vs Ripper Configuration

I will note two things:

1. The drive's advertised cache is for data usage (cd-roms, dvds, etc.). The audio cache is usually much smaller than the advertised data cache, but you must test it. If you get repeatable results on a "clean" system, you should use that value or round up to the next power of two if you want to be conservative. Note that the LARGER the cache is, the LONGER secure reads will take and each factor of two size increase can lead to substantially longer read times.

2. If you set the cache to a lower-than-actual value, dbpoweramp may report "secure" when it wasn't really a "secure" result (as defined by spoon), since matching reads were not rereads of the disc but rereads from the cache (which always returns the same value). So, beware setting the cache value too LOW. I've seen people convince themselves that a lower than real setting is good because it rips faster, or because they tested on a system that had a Filter Driver installed that interfered with the cache testing results.

**BigN** · September 07, 2009, 07:16 PM

Re: CD Spec vs Ripper Configuration

OK, many thanks bhoar. I thought that my last rip reading all six tracks successfully might be too good to be true! So given my cachex.exe result in my earlier post, with an 18KB max, presumably 2 to power 5 = 32KB should be correct?

But I just decided to run cachex again, with 100 passes using a good disc. As before 13KB/6 sectors was most common value returned, but range was from 4KB/2 Sectors to 66/29. Repeated test again... 4/2 to 20/9 this time. So perhaps 2 to power 7 = 128KB would be a better value to try? Take your point about increased re-read times!

Q1) What exactly do you mean by a "clean" system for this test?
Q2) Why is the reported cache size not a constant value, like the Drive Offset value determined during AccurateRip configuration?

**BigN** · September 07, 2009, 08:07 PM

Re: CD Spec vs Ripper Configuration

Using the Philips DVD8701, and with Drive Read Cache entry in dBpoweramp set to 128KB, I have now re-ripped my Berlioz Te Deum disc, with this result...
Ultra Secure Mode:-
Track 1 Accurate (4)
Track 2 Accurate (4)
Track 3 Accurate (3)
Track 4 Accurate (4)
Track 5 AR (4) i Secure {after 2 Passes, 4 Ultra Passes and 12 re-ripped frames}.
Track 6 AR (3) i Secure {after 2 Passes, 4 Ultra Passes and 1490 re-ripped frames}.

It did not take too long, except for the last track with the errors but even this track was much faster than with the 1024KB Drive Read Cache value tried previously. bhoar has already explained why this is!

Q1) Can anyone now convince me that this outcome is as good as I can hope to get with this album on this particular CD Drive using CD Ripper?
Q2) Or is there yet something else to try?

Thanks in anticipation.

**bhoar** · September 07, 2009, 09:25 PM

Re: CD Spec vs Ripper Configuration

Originally posted by BigN

Q1) What exactly do you mean by a "clean" system for this test?
Q2) Why is the reported cache size not a constant value, like the Drive Offset value determined during AccurateRip configuration?

A1: Clean for me depends on two factors: filter drivers and timing. Some software installs Filter Drivers that can interfere with optical drive IO. Some other software that runs on your machine can steal CPU or otherwise interfere with IO, which leads to unreliable testing of cache size, for example.

A2: The audio cache size is never reported by the drive. Therefore one has to probe the drive with repeated attempts to create discontinuities in response that *might* give away the audio cache size. So, the reported cache size is a guess based on testing the amount of time it takes for repeated actions to return. The read results, even on a clean system are "noisy" in the sense that other parts of the system can introduce delays and you cannot tell for sure if the delay is due to drive behavior or external-to-drive factors. Finally, it may be the case that the effective size of the audio cache used may depend on positioning of the data, which is another level of uncertainty, which is why I like to round up to the closest power of 2.

Short version: if you want to properly test cache size, it's probably best to do so on a newly installed windows system before installing any third party software.

**bhoar** · September 07, 2009, 09:27 PM

Re: CD Spec vs Ripper Configuration

Originally posted by BigN

Q1) Can anyone now convince me that this outcome is as good as I can hope to get with this album on this particular CD Drive using CD Ripper?
Q2) Or is there yet something else to try?

A1: With that disc and that drive, probably.
A2: You might try cleaning or scratch-repairing the disc, if there seems to be damage on the read side.

-brendan

**Spoon** · September 07, 2009, 10:44 PM

Re: CD Spec vs Ripper Configuration

It is likely that Track 5 and Track 6 have errors, a consistent error which would have been detected with a drive with c2 pointers (AccurateRip should match those tracks).

**BigN** · September 08, 2009, 06:59 PM

Re: CD Spec vs Ripper Configuration

Thanks to bhoar and spoon for sharing their vast dBpoweramp experience. I do hope other readers find this thread of interest!

OK re Drive Read Cache size testing. I understand what you are saying. Eventually I expect to re-build this computer to be optimized for my audio work, but first it is important that I understand the applications (dBpoweramp & MediaMonkey) plus the Sonos system that I intend to buy. Meantime it occurred to me that the nearest thing to a clean OS install may be to run the Drive Read Cache test in safe Mode, with minimal 3rd party stuff active to colour results.

DVD8701 (which I now think may be a re-badged BenQ 1640) -- when started up from cold, the drive's Drive Read cache varied from 6KB/3 Sectors to 147KB/64 Sectors. After just 34 passes of the cachex Explorer utility, the reported result became absolutely stable at 147KB/64 Sectors, with no deviation from this value as further tests were carried out. Like the drive had to warm up and settle down! So will now try Ultra Secure Drive Read cache at next power of 2, which is 256KB.

TS-H352C -- Cachex Explorer always reports "no cache detected". End of Safe Mode working.

Incidentally, the Philips DVD8701, now with Drive Read Cache entry in dBpoweramp set to 256KB, re-ripped my Berlioz Te Deum disc, with this result...
Ultra Secure Mode:-
Track 1 Accurate (4)
Track 2 Accurate (4)
Track 3 Accurate (3)
Track 4 Accurate (4)
Track 5 AR (4) i Secure {after 2 Passes, 4 Ultra Passes and 19 re-ripped frames}.
Track 6 Error {after 2 Passes, 4 Ultra Passes and 1499 re-ripped frames}.

So [I]spoon[I] is probably correct in that Tracks 5 & 6 are still having error problems that were being masked by earlier incorrect Drive Read Cache settings.

I find compromise an ugly word! Having spent loads of cash on my Imerge and audio system, network kit, Network Attached Storage and so on, I am reluctant to have an inexpensive CD (Read) Drive be the weakest link in my workflow chain. Cost of replacement vs investment to date or pending is trivial.

Everywhere on the forums I see reference to Plextor being "best" and that Accurate Stream seems almost UNIVERSAL; Caching is BAD; C2 Error Detection is GOOD; HTOA is GOOD; and Overread Both is also GOOD.
Q1) My drives are IDE. Do you think it would make a difference (worth the outlay) to replace the TS-H352C with a Plextor unit eg PX-860A?
Q2) Alternate recommendation?

FINALLY
And perhaps not strictly for this forum... but I have noticed that when I play tracks from this Berlioz rip using MediaMonkey, the dying notes of each movement seem to get cut off prematurely, as if the track is clipped. Kind of like the break between tracks is slightly ahead of where it should be. Consecutive tracks seem to play closer together than they should.
Q3) A dBpoweramp or MediaMonkey issue -- any thoughts?

**Spoon** · September 08, 2009, 09:28 PM

Re: CD Spec vs Ripper Configuration

Which audio format are you ripping to?

**BigN** · September 09, 2009, 09:52 AM

Re: CD Spec vs Ripper Configuration

Hi spoon
FLAC is my chosen format.
As for Q3, I have now found a setting in MediaMonkey that may be the answer... Tools | Options | Output Plug-ins, Configure button and uncheck "Remove silence at beginning and end of track". Looks promising fix, but still to run some thorough tests. I obviously want the ripped version of the disc to run from my NAS HDD FLAC file as if I was playing the CD in a conventional player. Any guidance welcome!

Norrie

**BigN** · September 11, 2009, 09:57 AM

Re: CD Spec vs Ripper Configuration

Q1) Any comment re PX-860A as replacement for my TS-H352C, or any alternative recommendation? I have researched this Plextor model and it looks good, and is readily available. But not sure if it 100% meets all dBPA requirements as no mention of C2 Error Pointers for example is made on the manufacturer's spec.
Any guidance appreciated from the audio experts.

**Spoon** · September 11, 2009, 12:23 PM

Re: CD Spec vs Ripper Configuration

If I was choosing a drive, I would make sure it supported C2 pointers.

CD Spec vs Ripper Configuration

CD Spec vs Ripper Configuration

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment