Hydrogenaudio Forums

Full Version: DAE

Hydrogenaudio Forums > CD-R and Audio Hardware > CD Hardware/Software

kotrtim

Jul 27 2003, 05:23

I used CDex and i turned of the jitler correction.
turning of jitler correction save lots of time
Read Sectors set to 52
I donno whats the real meaning of read sectors,thought my drive is 52X,
so i set it to 52........
Spin up time to 0

I have two cdrom drive and this are the results with eaqual.exe

QUOTE

Samsung (first extract) - Samsung (second extract)

EAQUAL - Evaluation of Audio Quality
Version: 0.1.3alpha
Author:    Alexander Lerch, zplane.development
_______________________________________________________
Reference File:    samsung1.wav
Test File: samsung2.wav
Sample Rate: 44100
Number of Channels:    2

Press Escape to cancel...

Processed: 223.49 seconds of audio file
Time elapsed:   159.69

Resulting ODG:   0.00
Resulting DIX:   2.91

BandwidthRef 19338.9920
BandwidthTest   19338.9920
NMR    -133.5697
WinModDiff1    0.0000
ADB    0.0000
EHS    0.0000
AvgModDiff1    0.0000
AvgModDiff2    0.0000
NoiseLoud    0.0000
MFPD 0.0000
RDF    0.0000

ARTEC (first extract) - SAMSUNG (first extract)

EAQUAL - Evaluation of Audio Quality
Version: 0.1.3alpha
Author:    Alexander Lerch, zplane.development
_______________________________________________________
Reference File:    artec1.wav
Test File: samsung1.wav
Sample Rate: 44100
Number of Channels:    2

Press Escape to cancel...

Processed: 223.49 seconds of audio file
Time elapsed:   145.19

Resulting ODG: -3.21
Resulting DIX: -1.49

BandwidthRef 19109.2762
BandwidthTest   19054.7014
NMR    -4.8669
WinModDiff1    37.3058
ADB    2.1426
EHS    0.3945
AvgModDiff1    36.6426
AvgModDiff2    81.0012
NoiseLoud    1.5198
MFPD 1.0000
RDF    0.8136

ARTEC (first extract) - ARTEC (second extract)

EAQUAL - Evaluation of Audio Quality
Version: 0.1.3alpha
Author:    Alexander Lerch, zplane.development
_______________________________________________________
Reference File:    artec1.wav
Test File: artec2.wav
Sample Rate: 44100
Number of Channels:    2

Press Escape to cancel...

Processed: 223.49 seconds of audio file
Time elapsed:   139.86

Resulting ODG:   0.00
Resulting DIX:   2.91

BandwidthRef 19353.8576
BandwidthTest   19353.8576
NMR    -133.5781
WinModDiff1    0.0000
ADB    0.0000
EHS    0.0000
AvgModDiff1    0.0000
AvgModDiff2    0.0000
NoiseLoud    0.0000
MFPD 0.0000
RDF    0.0000

With the results above, can i disable jitler correction savely.
If u see the documentation, the example given is only Plextor
Only Plextor CDROM is good?

QUOTE CDex:
This setting controls if the jitter error correction function is enabled.
Unless you have a very good CD-ROM drive (like a Plextor drive),
I would advice you to turn on the jitter correction,
to avoid click/plops and other artifacts in the extracted WAV of MP3 file.
However, if you have a good CD-ROM drive (e.g. Plextor drives),
you can turn off the jitter correction. Doing so can improve the ripping speed considerably!

But both samsung and artec extracts audio so differently
diff. drive will extract diff. audio file, very weird?

minix

Jul 27 2003, 05:41

QUOTE(kotrtim @ Jul 27 2003, 01:23 PM)

With the results above, can i disable jitler correction savely.

If your drives are "Accurate Stream" and this feature works as it should, then you can disabled jitter correction.

To test this, I use the "Test Device" in Feurio CD-Manager (Program Parameters menu).
If both drives are really modern, they should be "Accurate Stream".

http://www.feurio.com/English/faq/faq_voca...le_jitter.shtml

QUOTE

If u see the documentation, the example given is only Plextor
Only Plextor CDROM is good?

Yes

QUOTE

But both samsung and artec extracts audio so differently
diff. drive will extract diff. audio file, very weird?

I don't know how eaqual works, but "Compare WAVs" in EAC will consider offsets, which is probably the reason why the files are different.

Pio2001

Jul 27 2003, 06:22

Eaqual is a program that tries to evaluate the audible loss when a file is encoded to MP3 for example. It won't provide any useful information about ripped wav files.

minix is right,

The tool needed to compare wav files is EAC's compare wavs.

The tool needed to know if you can disable jitter correction is Feurio's "test device".

Different drives has a different offset, thus return diffent files. See EAC's website, in the "technology" section.

AngelGR

Jul 27 2003, 07:18

QUOTE

The tool needed to compare wav files is EAC's compare wavs.

Or CDex's compare WAVs. I think it tries to align the data of the two files, so the offsets aren't a problem.

ScorLibran

Jul 27 2003, 07:25

QUOTE(kotrtim @ Jul 27 2003, 07:23 AM)

Read Sectors set to 52
I donno whats the real meaning of read sectors,thought my drive is 52X,
so i set it to 52........

"Read Sectors" may be referring to the read offset of the drive. If so, read this for more information about offsets.

minix

Jul 27 2003, 08:03

QUOTE(ScorLibran @ Jul 27 2003, 03:25 PM)

"Read Sectors" may be referring to the read offset of the drive.

No. It's the number of sectors read in every operation.
It's in the CDEx help:
"Some drive can't read more than 26 sectors at once, thus if you get silent WAV files, make sure that this value is less than 27. If your drive can read more than 27 sectors at once, you can try to increase the value to say about 400, this can speed up the ripping process considerably!"

According to Feurio's help, it's more complicated:
"Basically it would be useful to chose the biggest possible value, because both the performance will be increased and no jitter errors can occur when reading a block.

Though SCSI host adapters can only transfer a certain amount of data per command, Feurio!® will inquire the maximum amount of data from the SCSI host adapter and appropriate restrict the transfer rate.

Attention: Besides the limit by the host adapter (e. g. 4 GB by Adaptec® adapters) Windows® limits the maximum transfer size as well!

If the drive is accessed via ”DeviceIoControl” (native WinNT/2K/XP interface) a special function will return the maximum transfer size supported by Windows®, so Feurio!® will be able to address the device correctly.

However, if the drive is accessed via ”ASPI”, there is (according to the ASPI documentation and the possible status parameters) no limit. Severals tests have shown, that there still IS a limit. Partially the limit is the same as in ”DeviceIoControl”, partially it´s higher.

Usually the limit was 69632 bytes (i. e. 29 sectors).

Exceeding this limit caused errors (entries in the system protocol) like e. g. ”The driver has detected a controller error in \Device\ScsiPort1”.

Furthermore some SCSI controllers as well as most of the IDE controllers have problems if the transfer size exceeds 65535 bytes (i. e. 27 sectors).

The value should usually be (not higher than) 27 sectors!

From: Feurio! CD-Writer online help © by Fangmeier Systemprogrammierung"

kotrtim

Jul 27 2003, 09:17

thanks
i used EAC to compare wavs
these are the results

QUOTE

Samsung (first extract) - Samsung (2nd extract)
Error = 0

Artec (first extraxt) - Artec (2nd extract)
Error = 0

Artec (first extract) - Samsung (first extract)
Artec
Error = 0

Samsung
Error Type = 503 repeated samples
Position = 0:00:00.075

Nero CDSpeed
DAE quality
Both Samsung and Artec have 10 points

Artec is not so bad after, its so cheap
I just spend $40 to buy its 52X CDRW Drive
Now can i disable jitler?

kotrtim

Jul 27 2003, 09:25

CDex compare waves results

QUOTE

Samsung (first extract) - Samsung (2nd extract)

File sizes of first/second are 36093836/36093836 bytes

No differencee have been found
A total of 36093786 bytes have been compared

Artec (first extraxt) - Artec (2nd extract)

File sizes of first/second are 36093836/36093836 bytes

No differencee have been found
A total of 36093786 bytes have been compared

Artec (first extract) - Samsung (first extract)

Second file (Samsung) has an offset of 0 bytes

File sizes of first/second are 36093836/36093836 bytes

File are different, starting at location 13262

minix

Jul 27 2003, 09:28

QUOTE(kotrtim @ Jul 27 2003, 05:17 PM)

Samsung
Error Type = 503 repeated samples
Position = 0:00:00.075

That's the different offsets.
No problem there.

QUOTE

Nero CDSpeed
DAE quality
Both Samsung and Artec have 10 points

That means nothing.
Well, it means that it can read audio correctly most of the time.

Even Feurio's test can't confirm that there will be absolutely no problem. But if it fails, you know that you can't trust your drive. Feurio has a test especially for checking jitter problems, unlike CDSpeed.

Pio2001

Jul 27 2003, 14:55

The different/repeated samples need to be investigated closer.

If this is an offset problem, why does CDex reports offset=0 ? It would mean that it's offset detection routine doesn't work.
It may be 503 repeated samples inside the track, at position 13262 (in bytes, that is 0.075 in seconds).

To know it, the files should be substracted from each other. If music can be heard in the result, then they are offsetted and CDex didn't see it. If there is some data at the beginning and the rest is silent (check with statistic analysis (highest sample, lowest sample), not by listening), then there was a read error into a file.

minix

Jul 27 2003, 15:04

QUOTE(Pio2001 @ Jul 27 2003, 10:55 PM)

The different/repeated samples need to be investigated closer.

Yes, you're right.
0.075 seconds would be almost 6 sectors, which is not a normal value for offsets (usually less than 1 sector).

I have problems interpreting "repeated samples" or "missing samples" in EAC

kotrtim

Jul 28 2003, 01:05

Checked with CEP
there is nothing but only a straight horizontal line at 0.075 sec.
then it is ok now?

Pio2001

Jul 28 2003, 05:06

Zero everywhere exept a rectangle ? Then one of the files is wrong.

Zoom on the place where the rectangle is, but in the original files. It should be possible to see the problem in the wrong one as a discontinuity in the waveform.

This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.