Hello!

I read that using "usefua" with Plextor drives can enhance the ripping speed. If I start EAC with that option, should I also tick the box that the drive is caching audio data or not? Also, does "usefua" have any impact if I rip tracks using another device (I also have a NEC and a BenQ unit installed)?

Sebastian

The commandline option -usefua will (in some cases) make EAC use the "Force Unit Access" instead of "normal" MMC read commands. According to André Wiethoff, this works with Plextor drives only.

FUA will make the drive read directly from the disc instead of reading out of its physical buffer which (if implied properly which it isn't at the moment) renders the EAC cache flushing trick obsolete since using the FUA command makes a caching drive non-caching ... you can deactivate the checkbox that says "Drive Caches Audio Data" in that case and get a ~ 50% improvement in ripping speed.

If you enable -usefua, you should run the drive features detection again ... if you are lucky, EAC will report your formerly caching Plextor drive as "non-caching".

From my experience, the NEC DVDRW drives don't cache and the BenQ 1640 DAE cache size (according to Feurio!) is some 150 kB which is smaller than the EAC read block size of 2MB (IIRC) ... it should be possible to deactivate the cache checkbox with a BenQ as well ... I recently did some testing and received some positive results.

There seems to be some evidence that the CD you use makes a difference. EAC should find that your drive does not cache; if it says it does, try with another CD.

This may be mumbo-jumbo, but if you look at the other thread there are a few of us who believe this to be true.

-Just wanted to clarify for you that you should UNCHECK the "This drive chaches audio" box
-The command only works with plextors but I dont know if it would have any effects good or bad on other drives.

As for it not being properly implemented in EAC is that true? As far as I can tell from what I have read its just not tested but I have not read anything about it not being properly or fully implemented.

Well, should I create two shortcuts for EAC then and use the "-usefua" version when I want to rip with Plextor and then when I want to switch to NEC for example start EAC using the "regular" shortcut or what?

More precisely, EAC will issue an additional cache flushing command
using the FUA bit before reading again. I can confirm that this command
is correct and is the same command used by Plextools that I mentioned
2+ years ago on this board (and that Wiethoff refuted at the time).
So no need to test forever, it works (of course, don't forget to also
uncheck "Drive Caches Audio Data").

Using this FUA command on a drive which does not support it will not harm
your ripping other than the speed hit ; however, unless a miracle occur it
will not do anything, because the semantics of the command are non-standard
and therefore require specific firmware code to handle them.

Any idea what happens if I use "usefua" on a non-Plextor device that does not support the FUA bit and the "Drive caches audio" box is ticked?

I dont think anyone can answer that question at this point since its still not even clear exactly what and how well its working with plextors. My guess would be that other drives dont recognize the command and could either ignore it or cause some sort of error. The easiest solution is just to have 2 shortcuts, one with the command for plextor drives and one without. Some testing should be done on non-plex drives.

In this case EAC mixes the useless FUA commands with overlapping cache
busting reads, so you're wasting time for nothing. Since the semantics of the
Plextools FUA command are very specific it is very unlikely that any other
brand will ever support this command (which does not mean that the FUA
bit would not work on other drives), so you should use two shortcuts.

Wrong, all drives will recognize the command and none of them will report an error.
You might want to first find out what command we're talking about before discussing
what it does.

I think I pretty clearly statted that it was "my guess", a google search doesnt turn up much on the usefua command. You seem to speak with more authority on the subject then anyone else I have heard so far spath. Do you have any other references on the command?

@Eli

The FUA bit is documented in the SCSI-3/MMC drafts at t10.org. It can be set on some commands like Read (10) and Read (12) and defines wether the drive shall acces the media or the cache when performing the command.

I have searched the EAC forum for info on Andre's implementation of the -usefua switch, and i found this comment from Andre :

"But I just used a sniffer to have a look how PlexTools is clearing the cache... Nice idea how they do it... It is possible that it would also work with other drives that caches.
I am not sure whether I have implemented it correctly, so it would be a good idea to test a modified version with several drives that cache (but not only "Test Drive Features", but really a comparison of how many errors go unnoticed in comparison to the old version where the "Cache" flag is used. (And also compare speeds). Because of several aspects, the "Test Drive Features" could tell that a drive has now no cache anymore, but use it nevertheless for extraction."

Source :
http://www.digital-inn.de/77551-post16.html

One of the reasons is that instead of blindly trusting Wiethoff or relying on google
I prefer to look at what is really going on. In this case, I studied Plextools and EAC
behaviour and I wrote my own tool to detect all kind of possible FUA support
(because some drives also support FUA, but not in the same way as Plextor).

As a side note, I never understood how "audiophiles" could accept to have
so little information about the tools they use ; while people interested in
burning quality are modifying firmwares´write strategies, people interested
in DAE are still wondering how many sectors their beloved tool read at a time.

I can sadly confirm that the -usefua commandline doesn't work at all in my system environment ... so André is right about not recommending his rather "quick and dirty" or untested implementation.

I recently started a thread over at the EAC board to invite Plextor RW drive owners to do some testing in various system environments (and maybe help André to find a better implementation).

I hope some people will join this test ...

Just posted my results - basically, it doesn't work on my system.

Hmm, very interesting stuff. I just got through doing a clean install of XP Pro on my Asus A7N8X Dlx 2.0 and have no apps on it. I'll give it a try and report back. Oh, Premium and 712A.

I've been using -usefua for a couple months now on my Plextor PX-W2410A. It gives PlexTools-like speeds up to 40x, but some errors seem to be ignored as opposed to ripping without -usefua. However, I have yet to find an audible glitch on these tracks where, usually, an difficult error would be found and the drive would grind away for hours on a single track (and produce seemingly the same result.) I do use/recommend Cool Down Drive. I have not done extensive CRC comparisons or anything like that, I'd rather be listening to my bit-perfect flac's. =P

Interesting side effect: with -usefua, EAC will always report a Sync Error before it reports a Read Error on my system. Still, no glitches can be heard at this "Sync Error." This occurs on slightly damaged discs (i.e. no label side damage but the scratches are enough to normally invoke a Read Error.)

I just like using it in T&C mode. Nearly the same as burst speeds with a at least some error correction and the security of T&C. If AR results match their is an added bonus. I just wish I knew about the switch months ago.

To those of you who have found -usefua doesn't work correctly with EAC, do you have DAEMON Tools installed on your system?

I have discovered that the installation of SPTD (SCSI Pass Through Direct) on my system has prohibited -usefua from working.

After removing SPTD, -usefua is working once again.

You can remove it by your system by downloading the installer and typing "sptdinst_x86.exe remove" at the command prompt. Typing "sptdinst_x86.exe add" will re-install it.

sptdinst_x86.exe (direct link to the executable) can be downloaded from this page:
http://www.daemon-tools.cc/dtcc/download.p...ory&catid=5

Also, remember that using EAC's Detect Read Features is not the proper way to determine if -usefua works correctly.

I wonder whether Plextools DAE (which basically uses FUA) is affected by SPTD as well ...

>I have discovered that the installation of SPTD (SCSI Pass Through Direct) on my system has prohibited -usefua from working.

Would that have also prohibited the FUA option in dBpowerAMP R12 from not working (you mentioned yor results on the Secure thread)?

On this note, R12 detected FUA as working with my Lite-On drive...I don't know if this is a fluke or not.

I tried the drive with EAC and the -usefua flag and it just crashes when I hit an unrecoverable read error

greynol, how did you discovered that? what symptoms? I have Alcohol 120% on my system and it also installed SPTD but apparently I have no problem with -usefua.

Cheers!

Sergio

>On this note, R12 detected FUA as working with my Lite-On drive...I don't know if this is a fluke or not.

What timing values were shown?

I'm not intentionally ignoring the other questions at the moment, but I thought I could comment on this. I don't have a lot of time today.

I have been running Alchol 120% on my system for a while now and have had no problems with -usefua.

After stripping my system of all sorts of software (both through uninstalling and manually removing system files), I started adding programs one by one and testing. There were a few pieces of software that I installed since I noticed the problem: iTunes7, spoon's r12a5, DAEMON Tools and a set of updates for WinXP. I intentionally left DAEMON Tools until last since it is what I suspected as being the problem.

DAEMON Tools prompted me about installing SPTD, "This program will install SCSI Pass Through Direct...not compatible with kernel mode debuggers..." I clicked ok, it installed SPTD and said it needed a reboot. Upon rebooting, the DAEMON Tools installation resumed and I cancelled it. DAEMON Tools was not installed on my system this time around, but the damage was done, -usefua would no longer work correctly with EAC. spath's cache explorer gave 4/5 for the FUA test. I did not try spoon's software to see how it responded, nor did I try PlexTools.

Hopefully I can do some more testing in the next couple of days.

I'll have to get back to you on that, this computer is so screwed up

I can't get it to detect a consistent cache size on either my LiteOn LTR-24102B or my Artec DHM-G48R

It's not your software, I tried cache explorer and got the same problem, no matter which cd I try. I think it might be due to me running Windows 2003?

Both drives came back with a 5/5 FUA support rating though in cache explorer

I just tested it and the answer is yes.

Test time without: 578 ms
Test time with: 1078 ms

greynol,

thanks for pointing to spath's cache explorer: very interesting!

I made some tests (cachex -r 0xD8 -p <driveletter>) both with and without SPTD installed (from Alcohol 120% 1.9.5.4327) and with the three Plextors I have: Premium, PX-W2410A and PX-712.

My results are that the Premium always give a 5/5 while the others two give inconsistent results ranging from 3/5 to 5/5 independently of the fact that SPTD is installed or not.

At this point I suspect the culprit is the SPTD that comes with Daemon Tools.

Is there something more I can do to test the reliability of Alcohol 120% SPTD?

Best regards.

Sergio

I checked the version of Alcohol 120% that I am using. It does not install SPTD though it does install these files into the system32\drivers directory:
- a347scsi.sys
- a347bus.sys

Whether they serve a similar purpose, I'm not sure.

EDIT: Formatting

Ok, I just finished trying this out. It's probably not a definitive test on the matter, but it's good enough for me. What I did was removed that spdt thing with that file from that place, then tried ripping a cd with usefua on, and with it off. I made the test conditions the same for both trials, and came up with similar results each time I did the test (I did it twice). The cd was System of a Down, Toxicity, and here's what I came up with:

- no usefua: total time to rip cd to wav: 3:22

-with usefua on: total time to rip cd to wav: 3:42

So with no usefua, it was faster. I'd like to hear your guys's interpretations, but for me, the myth of usefua is busted.

edit: and I bit-compared the .wav files with foobar and they were identical.

Hmmmmmm.... if I were I gambler i'd bet 10000 Euro that you had "Drive caches audio data" checked also when you were testing usefua. Try without.

Cheers!

Sergio

Edit: Of course assuming you are using a "genuine" Plextor...

Edit 2: If I were a gambler I'd be a poor man! Try instead "Drives caches audio data" unchecked AND "Drive is capable of retrieving C2 error information" checked! Then take a look to this post.

Yeap! Apparently Alcohol Software switched to SPTD in their latest build. This is not the first time they change the underling technology to their "Virtual Drive" feature.

Maybe you should start gambling because you'd be right in this case. I'll do it again. This time with caching turned off, but I'm going to leave c3 error info also turned off for both tries.

Just out of curiosity, give a try to caching off/c2 on, as well... ;-)

Cartman_Sr,

What is the make and model of your drive? If it's a Plextor PX-230A, you should be aware that this drive is not made by Plextor and likely won't support the FUA command.

Use either spoon's new alpha ripper or spath's cache explorer to determine if your drive is compatible.

I mentioned earlier that EAC's detection of audio caching is not the proper way to test for FUA. Ripping speed isn't the proper way either.

Ok, here's what I get. All the rips were identical according to foobar. The drive is a 230a. All other EAC settings stayed the same.

usefua | c2 | caching | ripping time
on | on | off | 2:15
on | on | on | 3:42
on | off | off | 2:45 ** this is the setting I'll use from now on
off | on | on | 3:22
off | off | on | 4:58

Obviously fua is doing something. More tests would probably be needed by others to confirm my results. Most of my cd's are already extracted, but from now on I'm going to use fua. Thanks SMZ.

You're welcome!

If I were you I'll make some more testing:

A) Do what greynol suggested and test your drive with Spath's Cache Explorer.

B) Find one or more problematic CD (with many-many C1 errors and many C2 errors, as can bee seen with a PlexTools Pro Q-Check) but that can be ripped without errors in secure mode and check if using the settings you choosed the rip correspond to the "secure" one.

BTW, you have by far less difference in ripping speed with and without C2 used with your PX-230 compared to the one I had at the time of testing with my PX-712A (I had 25.6 against 6.5, in terms of speed). I never made exhaustive testing with the Premium (just used caching-no C2-yes and usefua), but I'll see if I can do something in the next few days.

Take care! (of your rippings, too )

Sergio

Did you try:
off | off | off |
?

Like I said earlier, speed is not the way to determine if usefua is actually working properly.

In fact, during re-reads, -usefua (if working properly) makes the rip take longer than not using -usefua with caching unchecked in both cases; as it's supposed to.

The PX-230A is not a real Plextor, and I would be surprised if it did anything with the FUA command except cause your rips to be insecure.

EDIT: This isn't to say that Plextors are the only drives compatible with the FUA command. It's certainly possible that the PX-230A is compatible, and I hope that it is, but I don't think it is a good idea to use this experimental command without first testing to see if it works.

230A is not compatible with FUA.

I'm having the strangest issues with 2 of my drives

both my LiteOn and Artec drives have a 5/5 fua support rating, yet when I try to detect cache size, or enable -usefua I get semi lock-ups

It'll rip for a few seconds then freeze, but then I'll push the eject button and it'll kick back in. I've tried unhooking one of the drives since they're on the same IDE channel, but it's still behaving the same way. I tried reverting from Windows 2003 to XP to no avail also.

Like I said, the freeze-ups aren't from FUA specific, it does the same thng when R12 or cache explorer tries to detect the cache size

Great news!

The new 1.49 public beta of SPTD appears to not interfere with EAC's FUA command.

Thanks so much to HisInfernalMajesty for bringing the issue to the SPTD developers!

This is very exciting! Now I can actually install and use Daemon Tools.

I just tried it and it still shows caching for me on my Plextor PX-716A. Without Daemon Tools or any SPTD loaded, the FUA command works fine (shows no caching).

Robert

Just installed SPTD 1.49 and in cache explorer I get:



For my Lite-On SHW-160P6S PS0A

But running EAC with -usefua still detects caching in drive features (tried two different CDs in case it matters).

As I've already mentioned, EAC's test for caching is not a reliable way to test for the command. You need to see the speed at which EAC performs re-reads and whether EAC can report read errors in addition to sync errors. It is true that with some test CDs, EAC says my PX-716A no longer caches, but with other test CDs it says the drive still does.

...my 2 cents.

Regarding Cache Explorer, the developers of SPTD aren't concerned if their program affects its results as they don't feel that it uses Microsoft SPTI correctly:
http://forum.duplexsecure.com/showthread.php?p=703#post703

I don't know if what they say is true or not, I'm simply passing along the information.

Does this affect the newer model Plextor PX-240A?

I've been reading about adding the command line - USEFUA to the exe line, but how exactly is this done? I can't find the answer anywhere. Then again, this new drive of mine may not be effected by this such as everyone has found the PX-230A has (no effect).

There may still be an effect: even slower ripping speed when the cache setting is selected. The PX-230A isn't a real Plextor and it appears that the PX-240A isn't either, though I have a non-Plextor drive that accepts the command.

"C:\<path to EAC>\EAC.exe" -usefua

Thanks-

Where might one type in- "C:\<path to EAC>\EAC.exe" -usefua- exactly?

I just tack -usefua to the end of target line in my shortcut to EAC.