🚀 go-pugleaf

>>The results from my drives are:
>I slightly doubt your test.
>
>>Ultraplex
>>Average DAE Speed: 22.1x
>>DAE quality: 10 (perfect DAE)
>Known good drive.
>
>>AOpen CD924E
>>Average DAE Speed: 6.9x
>>DAE quality: 10 (perfect DAE)
>I believe all versions of this drive are not good at DAE by data
>mismatch.
>

I must admit that this drive's behaviour is a bit weird for DAE.
I have been able to do perfect DAE at 13.5x with it (extracted track twice
with CD Copy and compared them with fc/b).
However when reading raw sectors without any delay (as in the benchmark)
then it has a very hard time finding those sectors which explains the
strange graph.
The only thing that is wrong with the DAE quality is the offset.
The offset is about 2000 bytes too early. This means that the 2000 first
bytes from each track are actually the 2000 last bytes from the previous
track.
Problems like this cannot be detected by CD Speed.

Anyway, I never have to use jitter correction with this drive.
So unless the cd is badly damaged I always get perfect copies.

>>Ricoh MP6200A IDE
>>Average DAE Speed: 6.0x
>>DAE quality: 10 (perfect DAE)
>This is what tells me something is wrong. No ROM version to date can
>produce reliable DAE on this drive.

Tested with v2.20.

The DAE quality test is very simple. CD Speed reads the first, middle and
last 800 sectors from a disk and writes them to HD (audio.raw).
Then it reads those sectors again and compares them with the file.
If no differences are found then it gives a 10 for DAE quality.
This is the method that's mostly used for DAE testing (similar to the fc /b
method).

IMO the most reliable DAE quality test is when you extract tracks from a
known source but for obvious reasons this is impractical.

Anyway, if you have any suggestions for a more reliable DAE test, let me
know :)

Erik

btw. you can listen to the extracted sectors from the last test by opening
the audio.raw with a wav editor like Cool Edit (44100Hz / stereo / 16 bit /
PCM-Intel).

----------------------------------------------------------------
email: erikd@glo.be
CD-ROM benchmark: http://user.glo.be/~erikd/cdspeed.htm
----------------------------------------------------------------