delorie.com/archives/browse.cgi   search  
Mail Archives: opendos/2001/02/14/16:27:11.1

X-Apparently-From: <pmoran22 AT yahoo DOT com>
Message-ID: <006a01c096cb$ebc33420$73822a40@dbcooper>
From: "Patrick Moran" <pmoran22 AT yahoo DOT com>
To: <opendos AT delorie DOT com>
References: <3A8A8327 DOT BF3B2805 AT compuserve DOT de>
Subject: Re: prob audio CD on 2nd CD-ROM with NWCDEX
Date: Wed, 14 Feb 2001 08:41:27 -0700
MIME-Version: 1.0
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.3018.1300
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.3018.1300
Reply-To: opendos AT delorie DOT com

----- Original Message -----
From: "Utz Zarwell" <UtzZarwell AT compuserve DOT de>
To: <opendos AT delorie DOT com>
Sent: Wednesday, February 14, 2001 6:07 AM
Subject: Re: prob audio CD on 2nd CD-ROM with NWCDEX


> >Personally I don't see why you would need to put it
> >with the fast hard drive. They are slow to begin with.
> Because it reads with 24x max.

I am not certain why that matters. I have an old 4X Toshiba SCSI CDROM.
My brother has a Creative Labs 24X. I really have not been able to tell
any difference between his and mine. In fact faster CDROMs have more of
a tendency to skip on audio CDs that slower ones. I have seen this many,
many times. An audio CD will play fine in a 2X or 4X drive and skip like
crazy in a 12X, 16X, or higher speed drives.

Even when loading software like NT which is huge, I have not notice my
drive taking longer than his drive takes and we both have comparable
systems as far as speed it concerned. However, when I bought mine I got
one with very low access time, so that may in itself make up the
difference in speed for loading software. Mine is something like 80ms
access time.

Maybe if you play those huge stupid shoot 'em up games and play directly
from the CDROM, there might be a big difference, but I always do a full
install of a game to my HD which is many times faster than any CD drive.
I just never could see wasting money to buy something faster that would
never be needed. I am perfectly happy with this old 4X CDROM and will
only change it out when ever I get one of those DVD RAM drives. Those do
not have blazing CDROM read speeds either.

For audio CDs you only need a 1X or 2X drive (I don't recall which the
audio CD speed is.)  You are not going to get any better fidelity on
audio CDs with a higher speed drive as they all play at that slow 1X or
2X speed.

As for loading software, if a high speed drive does auctually load the
software faster, it's no big deal to me. Even when installing Linux on
my old 386DX40, I never noticed a difference it the time it took to
install it from a CDROM or my much faster Syquest cartridge drive. Even
if it did, who cares if I save 15 minutes. I just start the install and
go do something else while it's installing. Same with stupid WINDOZE.
With WINDOZE, I copy all the files to a temporary directory (like
\WIN95) and install it from the HD because the CDROMs are so slow
installing it. But who really cares, it still takes a long time to
install it. When I worked for Gateway Customer Service, I had to reload
WINDOZE many, many, many times. I would after remormatting the drive, I
would have the customer start the install and tell them it's going to
take a while and that they can take a break. So I would put them on hold
and go outside and smoke a couple of cigarettes, go nack in and get back
on the phone and it still was not done loading the first phase of the
WINDOZE install. This was after copying the files to the hard drive and
installing from the hard drive. On my old DX40, I could start an install
of WINDOZE 95 or Linux, then go do my grocery shopping and get back home
and it would still be installing the OS.<grin> It is about like loading
an AUTOCAD drawing on a 4.77MHz XT without a math co-processor!!!!!!

> >I believe they only have 1mbs maximum burn rate anyway.
> >That is the same as a 2.88 floppy drive.
> >I know that you can use them through a USB port and not
> >lose any speed and that has a maximum of 1mbs or 1.5mbs
> >for burning. For playback you could lose some speed,
> >depending on the playback speed of the drive. Since you
> >got an IDE drive, I am not aware of any IDE adapters for
> >USB like there is for SCSI.
>
> SCSI, USB, floppy? Sorry I cannot follow.


What I am talking about is data transfer rates. Whatever devices you
have in your system each have a maximum transfer rate. If you have a
drive that can transfer at say 5mbs, but you controller can only
transfer at 1mbs, the max data rate will only be 1mbs. If you have a
controller that can transfer at 160mbs and a drive that can only
transfer at 5mbs, the max transfer rate you can get is 5mbs. Since most
CDROMs write trandfer rate is very slow, you can connect a SCSI CD
burner to a SCSI port coverter connected to a USB port and not lose any
speed. If the CDROM read rate is higher than 1mbs or 1.5mbs, then you
would have a slow down of reads if you used the USB port. There are
inexpensive SCSI ports that you can connect to the USB port and daisy
chain these devices. This would be good for slow devices like CDROMs,
Printers, Scanners, Plotters, Tape drives and other SCSI devices and not
have to connect them to a good high speed expensive SCSI controller. You
would use that controller for hard drives.

> >You may wind up just having to remove the other CDROM drive.
>
> This means you advise to remove a CDROM drive ??

Yes, If you cannot get them both to work, you may have no choice but to
remove one. However, usually you can find some configuration that will
work. But you may not find one. That is only one of the main reasons I
chose SCSI over IDE.


> >It may be purely a hardware problem.
>
> Less likely. Despite of the IDE issue there is more evidence
> beeing a API or _protocoll_ problem. At which level remains open...

It most likely is a combination of both. I just had a thought. What if
you were to disable one of the drives in the CMOS setup and enable the
one you want to use each time you want to change drives? I don't know if
this will work. I don't have a lot of experience playing with IDE stuff.
If the master and slave settings are purely a hardware thing, then maybe
you can do it that way. But if you had a slave without a master and it's
not just a hardware thing, then maybe you cannot do this in the CMOS
setup. I only have one IDE device in my system. I do have an old 40MB
ST-157 drive laying around and could play with that and install it, but
that may not help with things like CDROMs, ZIP drives, Jazz drives,
LS-120 drives that use IDE interface. I have avoided IDE as much as
possible because of all the weird problems I have seen reported. When I
first got my Syquest drive, they sent me an IDE version. I immediately
sent it back and exchanged it for a SCSI because that is what I had
ordered. Who knows what kind of problems I may have had with it. It was
probably an EIDE drive and it is as fast as most hard drives were when I
got it, only access time was slower and that is around 19ms which was
decent back then (over 5 years ago.)

>
>
> >Someone mentioned something about using the secondary controller
> >for IDE drives and the Primary for EIDE drives.
>
> I mentioned to put the fast HDD and the CDRW to the onboard EIDE
> controller which is _first_ in system and to cable the slow HDD
> and the CDROM to the on board IDE channel.

Aren't both channels EIDE??? It would appear to be the case on both my
AMD 5x86-133 and my brother's Cyrix 5x86-133 systems. They both have LBA
and block modes. I did put my brother's old 120MB none EIDE drive on the
same channel as the CDROM drive as I have heard that if you combine a
slow drive with a fast drive on the same channel, they both will be
slow. i.e. you cannot use block transfers if a non EIDE drive is on the
same controller as an EIDE drive. Even then the maximum bloack transfer
size is limited to whatever drive has the least capability. This may
also be true when setting the mode a well.

> I do not see alternatives.
> I am aware that usage of HDD and CDRW at the same cable is often
> said to be a risk. (Not for SCSI.) But bandwith together with
> hardware caches IMHO should give lower risk.

I have also heard this. Hardware caches may indeed help this, because
the transfer rate will be from the cache and not the controller/drive
until the cache is emptied. But once you have to do direct reads from
the drive, you will be back to point zero and have the problem with a
combination of hast and slow drives.

Now that IDE has finally got busmastering and DMA capabilites, this may
not be as big a problems as it once was. SCSI has always had this
capability. (Although some of the cheap early SCSI-I stuff did not have
busmastering capability, but we are talking many years ago, when
considering this.)

Pat



_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019