This page is part of document INS-DAS-29,
Operations manual for UltraDAS
between UltraDAS and ancestral systems
Differences between s9.1 for IDS and s7.1 for IDS
S7.1 was released on the INT for IDS in February
1999. It was the culmination of the Unix-based observing systems including
the "Data-Cell DAS" (i.e. the DAS based around the image-capture interface
supplied by Data-Cell limited) and was the last system to be commissioned
before the advent of UltraDAS.
S9.1 replaced s7.1 for IDS in July 2000. The intermediate
systems, s8.1..s8.4 were tried out with IDS but were never released for
These are differences between s9 and and s7 for
IDS that are apparent to observers.
These are changes that are only visible to operations
In s7, you logged in to lpss15 using a VT220 terminal
i.e. a physical, character-only terminal. In s9, you log in to the DAS
computer of UltraDAS using a terminal window on the system computer.
Because a different DAS computer is used, a different
set of data discs is available.
Readout is faster. For example, readout of an EEV
CCD typically takes one minute less in s9 than it did in s7.
In s7, you had a choice of five readout speeds called
slow, standard, quick, turbo and nonastro. In s9, you have a choice of
two speeds called slow and fast. See the rspeed command for details.
In s7, you could choose between speeds at which the
CCD cleared itself. S9 allows only one speed.
Windowing is controlled differently:
The window command
of s9 takes an IRAF image-section as an argument, e.g. window 1 "[500:1500,1:4200]",
instead of a set of four separate numbers.
The enable_win and disable_win commands
of s7 are not present in s9. Instead, you can enable and disable individual
windows, e.g. window 1 disable.
The trick of setting a window to [0:0,0:0] to remove
it doesn't work in s9. Use window <n> disable instead.
The mimic for the CCD is changed in appearence; but
the information displayed is largely the same.
The FITS files have changed:
Each image in the file appears as a FITS image extension.
This is true even when there is only one image in the file.
The keywords describing the location of illuminated
pixels have changed: look for TRIMSEC, BIASSEC and WINSECn, n
The integration-time keywords are the same, but they
are in the extension headers, not in the main header.
See document INS-DAS-26
details of the new format.
The handling of the data discs is somewhat different.
See the obsdata command for details.
SDSU CCD-controllers replace "Dutch" controllers.
Fibre runs for the CCD-controllers go to the CLIP
centre, not to the control room.
Controllers are self-configuring if exchanged between
CCDs; there is no equivalent of setting up a RAM-disc in the controller.
The set-up tools for the Data-Cell DAS in s7 (dasprofile,
daspackets, and the CCD profiles) are disused. UltraDAS in s8 replaces
these with configuration files in /int/etc.
Differences between s8.2 for INT WFC and s7.1 for
S7.1 was released on the INT for WFC in February 1999, at hte same time
release for IDS. S8.2 for WFC was released in August 1999 and replaced
s7.1 for INT
WFC only (s7.1 remained in use for IDS). S8.2 was the first system
to include UltraDAS.
During 1999 and 2000, various upgrades were added to S8.2 as patches,
some extra features from s8.3 at the WHT.
These are differences between s7 and s8 that are apparent to observers:
In s7, you logged in to lpss15 using a VT220 terminal i.e. a physical character-only
terminal. In s8, you log in to the DAS computer of UltraDAS using a terminal
window on the system computer.
Because a different DAS computer is used, a different set of data discs
Readout is faster by a factor of roughly two.
In s7, you had a choice of two readout speeds called slow and standard.
S8.2 gives you no choice at all.
In s7, you could choose between speeds at which the CCD cleared itself.
S8 allows only one speed.
In s7, you could bin the CCDs. S8 does not support binning.
The FITS files have changed:
Each image in the file - i.e. the image from each CCD - appears as
a FITS image-extension. This is true even when there is only one image
in the file.
The handling of the data discs is somewhat different. The system choose
the disc for you and there is no obsdata command.
The readout noise is slightly increased.
The reliability is much better.
These are changes that are only visible to operations staff:
An SDSU CCD-controller replaces the "Dutch" controller.
Fibre runs for the CCD-controller go to the CLIP centre, not to the control
The set-up tools for the Data-Cell DAS in s7 (dasprofile, daspackets, and
the CCD profiles) are disused. UltraDAS in s8 replaces these with configuration
files in /int/etc.
Differences between s8.3 for UES and PFIP and the
The VAX-ADAM DAS, which employs the DMS, was first
released on the WHT in the telescope's first year of operation. That system
has been gradually adapted and improved over the years; the part visible
to the observer dates from around 1993.
S8.3, including UltraDAS, was introduced in September
1999 to drive the new WHT WFC, a two-chip mosaic camera which produces
too much data to handle in the ADAM DAS. The new system replaces the ADAM
DAS (but not yet the ADAM ICS) when the mosaic camera is in use at UES
These are some differences of s8.3 from the ADAM DAS.
Observations are faster. Readouts are much faster
(by a factor of six or more) and data-saving operations are a little faster.
Multi-channel cameras are possible: this allows the
WHT WFC to go into service.
The content of the FITS headers has changed.
The command-line interface for UltraDAS is on a separate
terminal form ICL.
The semantics of DAS commands are very different.
The ADAM commands return the ICL prompt immediately and are then queued
until the system can execute them. The UltraDAS commands block until complete
(but can be put into the background, using hte normal mechanism of Unix'
The DMS and its SPARCstation aren't used with UltraDAS.
The data-reduction computer can see the raw observation
discs directly and IRAF can read the observation files. There is no need
to transfer files via a scratch disc.
The data discs are handled more transparently; the obsdata command
is no longer needed.
The mimicry is very different.
UltraDAS starts up about 20 time faster than the ADAM DAS.