Quisk Version 3.7.6  September 2015
===================================
The PulseAudio code was completely rewritten by Eric Thornton, KM4DSJ.  It now uses
the asynchronous interface. This will:
1)Improve stability/performance. We can now manage latency events how we want to and prevent
ever increasing latency due to CPU load or network delays. latency_millisecs in the quick_conf.py
file will established the target buffer size for playback. Realtime latency for pulse audio
streams is available on the config/status tab.

2)Allow connection to two different PulseAudio servers. The default connection is to the local
machine for normal usage. A separate machine can be specified to pass IQ audio via the network.
This is useful to utilize a Softrock or Peaberry via a raspberry pi or other remote computer.
Specify IQ_Server_IP = "your.ip.here" in your quisk_conf.py to utilize this option. Be sure to
set latency_millisecs to something reasonable (300 works for me over wifi). NOTE: To keep network
congestion down, all modes except CW will "cork" the idle up/down IQ stream.
3)Quisk will attempt to utilize native 16 bit LE format if a PulseAudio sink is configured this way.
Default fallback is Float 32. This reduces network overhead for remote IQ audio streaming while not
losing resolution for any local streams that are configured as floating point. 

Additional changes:
-Remote control of Softrock receivers via usbsoftrock. Add usbsr_ip_address = "your.ip.here",
usbsr_port=port#, and import hardware_net.py in quick_conf.py to communicate with a usbsoftrock
daemon. Start usbsoftrock on the remote machine with the -D option.
-Added a macports target to the makefile to utilize dependency libraries from macports.
This option now allows use of PulseAudio on OSX.

Eric
KM4DSJ


Quisk Version 3.7.5  September 2015
===================================
I changed the code for full-screen mode.  You can now enter window sizes directly.  See
window_width in the file quisk_conf_defaults.py.

I added two more bits to tx_control for the FPGA protocol. I updated the PyUSB
package that ships with Quisk to version 1.0.0b2.

The received characters from FreeDV now appear in the upper left of the graph.  I added
the new FreeDV 700 and 700B modes.  Right-click the FDV button to select the mode.

To make it clear when buttons have added functions, I added a miniature slider to buttons
that can be right-clicked, and a circular arrow to buttons that can be clicked repeatedly.
This can be turned off in your config file.

Quisk Version 3.7.4  August 2015
================================
The Spot button transmits a variable level carrier in all modes, not just SSB.  The new config
file option "spot_button_keys_tx = True" will cause the Spot button to key the transmitter.  The
default is False.  The up button for the Favorites screen used to set the frequency in Hertz,
but the frequencies must be in Megahertz.  The ctcss was also broken.  This has been fixed.

I added a new Hermes-Lite model configuration file hermes/quisk_conf2.py.  This is for
use by people who can program in Python and want to add features to Quisk.  It is set up
to send the Spot button status on the J16 connector.  Quisk now operates in CW mode using
the new Hermes-Lite key input.  For other modes, use the PTT button.  On the Hermes,
use the PTT button with Spot, even for CW.  The CW key always transmits full power CW even
if the mode is not CW.

There is now an option to receive microphone samples from a UDP device.  This is used for
SDR hardware that has a built-in codec.  See quisk_conf_defaults.py.

The FreeDV code was completely rewritten.  It is now simpler, and it is easier to change to a
newer libcodec2.  See the file freedvpkg/README.txt. If anyone is running on Apple OSX, and
has problems, email me, and look at freedv.c.

The config file option graph_width=0.8 sets the width of the graph.  If this is set to exactly 1.0,
Quisk will run in full-screen mode.  This is meant for built-in screens, tablets etc. that lack
window management.

Quisk Version 3.7.2  July 2015
==============================
I made some changes for Hermes-Lite.  You must add rx_udp_clock to your config file.  You can adjust
rx_udp_clock slightly to correct the frequency display.  See hermes/quisk_conf.py.
I added Hermes_BandDict to the config file.  It controls the bits on the J16 connector of the Hermes-Lite.

Quisk will now keep the same filter for a given mode when changing bands.  I added RepeaterOffset() to
the softrock hardware file so that repeater offsets work.  I added mode DGT-FM for digital FM modes.


Quisk Version 3.7.1  July 2015
==============================
This is a bug fix release to correct an incompatibility between my FPGA firmware and that of
Stephen, DL2STG.  He was using the byte in the control record that I used to send the sidetone volume.
I now send the sidetone volume as byte 17.

I added the config file parameter use_sidetone to add/remove the Sto sidetone volume control.  It
used to be necessary to use the hardware file for this.  The default is no sidetone.  To add the
control, add "use_sidetone=1" to your config file.

I fixed a bug in the Audio Plackback button and Hermes-Lite.

Quisk Version 3.7.0  July 2015
==============================
Quisk now supports a new radio, the HermesLite.  If you have a HermesLite, copy hermes/quisk_conf.py
to your config file and modify the copy.  You should only need to change the soundcard names.
If you use Quisk on multiple radios, remember that you can have multiple config files.
Use "python quisk.py -c my_config_file.py" or its equivalent to choose among config files.  This
code is currently under test and may have problems.

The favorites screen now has two more columns for repeater offset and CTCSS tone.  See the file
quisk_conf_defaults.py.  You must set do_repeater_offset=True, and your hardware must be capable
of performing the shift.

The FDV feature is now in "final" form.  Please see freedvpkg/README.txt.  Note that FDV is
a moving target.

There is now an option to send radio sound to a UDP device.  This is used to send radio sound to
SDR hardware that has a built-in codec.  See quisk_conf_defaults.py.

The SoftRock hardware file now supports the key line as a PTT button for voice modes.  Use this if you
have a hardware mic button that you can connect to the key line.

I added the sidetone volume to the HiQSDR control data to accommodate new hardware.  I softened the
volume controls Vol and Sto for a smoother response.  I added the new config file option cw_delay.

Quisk Version 3.6.22 April 2015
===============================
FreeDV is the combination of the codec2 codec and the fdmdv modem.  It provides digital voice
in 1200 Hz bandwidth suitable for HF transmission.  You can use FreeDV by downloading the FreeDv
program from freedv.org, and connecting it to Quisk using the usual digital mode DGT-U.  But now
there is a simpler way.  Quisk has a new mode button FDV.  Just push the FDV mode and talk.  See
http://www.rowetel.com/blog/?page_id=452.

Quisk will add the FDV mode button unless your config file contains the line
    add_freedv_button = 0
If there is a problem with the freedv module, the button will be grayed out.  See README.txt in the
freedvpkg subdirectory for more information.  The "Split" button currently fails with FDV.

With the mouse in the frequency display, roll the mouse wheel to change the digit.

The new config option hamlib_ip specifies the Hamlib IP address; default "localhost".

The IMD button now has a right-click level control like the Spot button.  This is more
convenient than having a few fixed levels.

The Spot button can now transmit a carrier at zero amplitude.  This is useful for testing
the output noise.  If you wrote your own OnSpot() method, level -1 is now Spot button off,
and the level is now 0 to 1000.

Quisk Version 3.6.21 March 2015
===============================
Quisk can now transmit a message from a WAV file.  Record your message at
a high level (near clipping) at 48 ksps, 16-bit, one channel (monophonic).
Then enter the file name on the Config/Config screen.  Press the "File play"
button to transmit.  Quisk will press the PTT button for you, and release it
during pauses.  To interrupt playback, press PTT or release FilePlay so you can answer.

The "Split" button has been replaced with a "Splt" button and a "Rev" button.  The "Splt"
button splits Rx and Tx; and if you click it with the right mouse button instead of the
left, it also locks the Tx frequency so tuning changes the Rx frequency.  The "Rev"
button reverses the Tx and Rx frequencies.  These features were suggested by Mario, DH5YM.

I added a new parameter mic_agc_level to the config file to control the mic AGC.  Input levels
below mic_agc_level are ignored.  The default is 0.1.  Increase this (up to 1.0) to reduce the
AGC range, and reduce it to increase the AGC mic gain boost.

Philip Lee contributed a patch to the PulseAudio code to set the play buffer size.

Quisk Version 3.6.20 December 2014
==================================
Thanks to Graeme, ZL2APV, quisk now changes both channels of a control when using mixer_settings[].
He reports that setting boolean values does not work, but it works on my machine.  More testing
is needed.

The new config file dictionary bandTransverterOffset[] gives the offset in Hertz for bands that are
used with a transverter; for example 144000000 - 28000000 for a two meter transverter.  This
currently works for HiQSDR, SoftRock and SdrIQ.  I also changed the graph X axis code so gigahertz
frequency labels are not too wide.

Quisk Version 3.6.19 October 2014
=================================
Mario, DL3LSM, contributed changes for MacOS support.  Thanks!

I added device names to PulseAudio.  The PulseAudio name "pulse" still refers to the default
device.  Otherwise, enter a PulseAudio name such as "pulse:alsa_input.pci-0000_00_1b.0.analog-stereo".
See quisk_conf_defaults.py or docs.html.  PulseAudio support enables you to connect to recent
versions of wsjt-x.  To turn this off, set show_pulse_audio_devices = False in your config file.

Quisk Version 3.6.18 June 2014
==============================
The Windows installer now works with the new WxPython 3.0.  Previously it required 2.x.
I fixed an annoying sound loop that happened on Windows when quitting Quisk.

The Rx Filter screen now displays the bandwidth of the filter at the 3 and 6 dB points.
I improved the calculation of FFT size for fft_size_multiplier==0. I added more FFT buffers
to improve performance at high data rates.

When you right-click the S-meter, there are new options to change the S-meter time
constant, and to measure the audio voltage.

Quisk Version 3.6.17 June 2014
==============================
Philip G. Lee contributed code to provide native PulseAudio support for Quisk.  You will need
to install the package "libpulse-dev" to compile.  Thanks Philip!

Stephen, K6BSD, provided patches to support FreeBSD.  Thanks Stephen!

The Spot button now works in CW mode with softrock USB hardware. There is a new
slider control for SoftRock transmit level on the Config/Config screen.

The new config file options file_name_audio and file_name_samples are the initial names
for saving audio and samples on the Config/Config screen.
I added more sample rates to support different hardware.

Quisk Version 3.6.16 March 2014
===============================
I re-wrote the CW transmit logic for SoftRock transmitters.  It now implements semi break-in
CW operation.  The config file options are key_poll_msec and key_hang_time.

I added the config file option sample_playback_name to play the raw samples to a loopback or
VAC device.  You can access the raw samples by reading them back with another program.
See quisk_conf_defaults.py.  The sample rate is the same as the hardware sample rate.

Quisk Version 3.6.14 November 2013
==================================
I restored audio to the "Tx Audio" screen.  I increased the digital (DGT-U etc.) transmit
bandwidth to 5 kHz.  The Ys and Yz settings that control the waterfall colors are now
saved for each band.  Just adjust Ys and Yz on a band, and these settings are restored when
you return to the band.

I improved the receive AGC to reduce distortion.  The AGC button now has separate slider values
for On and Off.  Right click the AGC button to set the sliders.  Push the "Help" button to read
the description of the AGC.

I corrected some bugs related to dark color schemes, and added color_bg_txt to control the
slider text on the main screen.

Quisk Version 3.6.13 October 2013
=================================
I added VOX, voice operated relay.  I improved the transmit audio speech processing.  I
added controls for the audio clip and preemphasis levels to the Tx Audio tab of the Config
screen.  Press the "Help" button to read the documentation on how to set your audio levels.

Quisk Version 3.6.12 August 2013
================================
I fixed a bug that produced slight distortion in the transmitted signal.  I added the Enum
type to the mixer control "mixer_settings[]".  Be sure to use a Python float if you want a
decimal fraction 0.0 to 1.0; that is, use "1.0" for 100%, not "1".

The config file now defines hot keys that will press the PTT button when the keys are pressed.

I added a new tab to the Config Screen for testing transmit audio.  You can record and play
back the processed microphone output to hear what it sounds like.  I plan to add audio
controls to this screen in a future release.  I think the record/playback is more useful
than a simple loopback because I find it difficult to talk and listen at the same time.

Quisk Version 3.6.11 July 2013
==============================
I increased the size of the SoftRock status line.

You can now change frequency by left clicking on the digits in the frequency display.  Click on
the upper part of a digit to increase it, and the lower part to decrease it.

I added some GREAT features from Christof, DJ4CM:
  Symbols for buttons.  But you can change back to text with use_unicode_symbols = False.
  Two new buttons for direct entry and recall of items on the favorites screen.
  A new window to display saved stations, favorites and dx cluster data below the frequency axis.
  A feature to query a dx cluster using telnet, and display the stations.
The symbols use Unicode, and were tested on Linux and Windows, and on computers in Germany and the USA.
But if your Windows does not support Unicode, you will need to add use_unicode_symbols = False.
Read quisk_conf_defaults.py (as usual), and look for quisk_typeface, btn_text_, use_unicode_symbols,
sym_stat_, station_display_lines, and dxClHost to see how to use these features by Christof.

Quisk Version 3.6.10 May 2013
=============================
I made some changes to my n2adr hardware and widget files.  I increased the timeout for select()
in the UDP code.  I added a patch from David Turnbull, AE9RB, to make Quisk USB control work with
the Peaberry.  The problem was discovered and solved by ON7VQ.

Windows changes the Documents folder name for different languages, and has no standard way to
find it.  I added "Mine Dokumenter" to the list.

The AGC release time and mic gain control are now parameters in the config file.

I added some features from Christof, DJ4CM: additional color control, double click tune on favorites
screen, and format changes.

Quisk Version 3.6.9 April 2013
==============================
The new config file parameter digital_output_level controls the sound level to Fldigi, etc.
I fixed some small bugs in the Windows DirectX sound system.

When using the Split feature, Quisk can now receive on both the Rx and the Tx frequencies, and play them
in stereo.  When working DX, you can now hear the DX in one ear and the pileup in the other.  I like
the lower frequency signal to be left, and the higher one right.  This is controlled by the new config
file option split_rxtx that can be 1, 2, 3 or 4.  See the file quisk_conf_defaults.py (as usual).

This version of Quisk supports the new features in my HiQSDR FPGA firmware version 1.4.  If you are
using Quisk to set your hardware IP to rx_udp_ip, please be sure that rx_udp_ip_netmask is correct.
The default is 255.255.255.0.

Quisk Version 3.6.8 March 2013
==============================
The new config file parameter button_font_size can be changed to reduce the size of the button font.
This is useful for a netbook when the lower screen resolution results in crowded buttons.

There is a new button "Notch" for an automatic notch filter to get rid of carriers in SSB signals.  It
works in CW too, but is less useful because the sharp filters can get rid of unwanted signals.  The "Notch" 
is an advanced feature based on the FFT and not the LMS algorithm.

Quisk Version 3.6.7 February 2013
=================================
I added a check for the correct wx version.  Thanks to Mario, DH5YM.  Selecting "Config/Favorites/TuneTo"
now changes to the configured default_screen instead of the waterfall screen.  Thanks to Detlef, DL7IY.
The mute button and volume now controls only the radio sound, and not the digital output.  Thanks to Mario, DH5YM.

Previously, Fldigi XML-RPC control only worked if a digital mode was selected and a digital audio device
was specified.  Now it is always active unless you turn it off in the config file.

The transmit power level and the digital transmit power level can now be adjusted on the config screen.  The
levels are a percentage of tx_level as set in the config file.  The meaning of digital_tx_level was changed.
It is now the maximum percentage value for the slider.  Thanks to Hubert, DG7MGY.

There is a new feature to save frequencies and return to them.  When you have tuned in a signal of interest,
press the "Save" button to save the frequency, band and mode.  Repeat for more signals.  Now press "Next",
to switch to the next saved signal, and press "Next" repeatedly to cycle through the list.  To delete a
saved signal, first tune to it with "Next" and then press "Delete".  If you save a large number of signals,
right click the "Next" button, and you will get a popup menu so you can jump directly to a station.  Thanks
to Detlef, DL7IY.

I had to renumber the columns for the Quisk buttons. That won't matter to you unless you add your own widgets
to the bottom of the screen by importing your own quisk_widgets.  Then you will need to renumber your columns.
See n2adr/quisk_widgets.py for an example.

Quisk Version 3.6.6 January 2013
================================
I corrected the frequency measurement feature so it works with RIT.  This feature was successfully
verified in the November 2012 ARRL Frequency Measurement Test.  I fixed a bug in Fldigi frequency
control thanks to Hubert, DG7MGY.

The config file has a new parameter mouse_wheelmod to control the mouse wheel
sensitivity.  Thanks to DG7MGY for the patch.

I changed the digital filters so they have a bandwidth up to 20 ksps.  This accommodates more digital modes.
There are now three digital modes: DGT-U and DGT-L decode the audio as upper or lower sideband.  The old
DGTL mode is now DGT-U.  The mode DGT-IQ does not decode.  It sends the I/Q samples directly to the
sound card device.  It is interesting to listen to the DGT-IQ signal, as it provides binaural reception.
Also, the PTT can be controlled by either Quisk or Fldigi, not just by Quisk.

There is a new tab on the Config screen where you can enter the frequencies and modes for favorite
stations.  This can be used to list and tune to repeaters or nets.  The page works like a spreadsheet.
Fill in the rows with an arbitrary name, the frequency as integer Hertz or decimal megahertz, the mode and
an arbitrary description.  Right click on the left row label for a popup menu that allows you tune to that
row.  Thanks to Brian KF7WPK for suggesting this feature.

Quisk Version 3.6.5 November 2012
=================================
I added a waterfall method ChangeRfGain() that enables you to keep the waterfall colors constant
for changes in the RF gain control.  See my n2adr directory for an example of how to use it.

I added the ability to connect Quisk to the hamlib rigctld daemon.  This enables Quisk to work
with any rig compatible with hamlib.  For an example, see quisk_conf_kx3.py.  Push the Help button
for documentation.

I added a feature to measure the frequency of a continuous RF signal.  Right click the S-meter window
to turn it on.  Push the Help button for documentation.  It is meant for precise frequency measurement
such as would be needed to characterize crystals for a filter.  Precision is 0.01 Hertz.

Quisk Version 3.6.4 September 2012
==================================
I added Hamlib control to Quisk.  Set your digital or logging program to rig2,
device localhost:4575.  See the Help and docs.html.  This is used to control Quisk
from other digital mode programs such as WSPR.

I added the Y scale to the graph above the waterfall.

Quisk can now record the speaker audio and the digital samples to a WAV file.  Set the
file names using the config screen, and then use the "FileRec" button to start recording.
Press the Help button for more information.

Quisk Version 3.6.3 July 2012
=============================
Thanks to Steve Murphy, KB8RWQ for the patch adding additional color control, and for his 
dark color design. 

I am using Quisk with my AR8600 receiver 10.7 MHz IF output as a general coverage receiver.
My config file is n2adr/quisk_conf_8600.py.  This covers the VHF and UHF bands, and so
I needed to add some FM repeater and scanner features.  I added a Squelch button for FM.
Right-click the button to adjust the squelch level.  The Squech and AGC buttons
are combined to save space.  The new configuration file items freq_spacing and
freq_base are used to round frequencies to channel spacings on VHF.  There is scanner
logic in my config file.  You should look at this if you use Quisk with a transverter for
the higher bands.  With my hardware it is able to scan known repeater frequencies jumping
across bands as it scans.  The 960 ksps rate of Quisk and HiQSDR is very useful at VHF
and higher.

I added tabs to the config screen, and cleaned it up.

I added a record and playback button.  Press Record to start a new recording of radio
sound.  The maximum recording length is set in the config file, and the default is 15
seconds.  After this limit, the most recent 15 seconds of sound is retained.  To play
the recorded sound, press the Play button.  If you are transmitting, the recorded sound
is transmitted provided the microphone and playback sample rate are both 48000 sps.  The
transmitted recorded sound is not subjected to the usual audio processing.  That means
that you can play another ham's audio back and give him/her a good idea of how it sounds.

Quisk Version 3.6.2 May 2012
============================
I added a display of the filter bandwidth to the graph screen.  This is based on code
provided by Terry Fox, WB4JFI.  Thanks Terry!  See the file quisk_conf_defaults.py.

I added detailed information on each sound device to the config screen.  The Test
button now generates AM and FM as well as CW and SSB.

The receive filtering has been re-written to improve the shape of the filters and to
reduce the CPU time.  Quisk now runs on my fan-less Shuttle Atom machine at speeds
up to 480 ksps.  The CW filters are particularly nice.

Quisk Version 3.6.1 April 2012
==============================
There is a new "DGTL" mode to send Quisk audio to an external digital mode program
such as Fldigi.  Read the file quisk_conf_defaults.py to see the new config file
options available.  Use the Help button for basic information, and see docs.html.

I changed the 60 meter operation to agree with new FCC rules (for the USA).  See
the configuration file for items to control 60 meters.

Quisk Version 3.6.0 March 2012
==============================
There are no new user features in this release, and no changes to the HiQSDR code.
This version adds a new feature for those writing C-language extension modules
that need to access C code from the _quisk extension module.  Examples are the
SDR-IQ and the Charleston extension modules.  This feature was requested by
Maitland Bottoms, AA4HS, and he also provided patches.

Previously, symbols from the _quisk module were linked to sub-modules with the
C linker.  Now _quisk exports symbols using the Python CObject or Capsule
interface.  The documentation is in import_quisk_api.c.  Only minimal changes
to extension modules are required, as most changes are in _quisk.  The linker
method still works on Linux, but the new interface is highly recommended.

Quisk Version 3.5.12 February 2012
==================================
There are no changes to Quisk, but this version includes the new Quisk VNA program
that enables you to use my original transceiver hardware and the newer HiQSDR
hardware as a vector network analyzer.  Use "python quisk_vna.py" to run it.

Quisk Version 3.5.11 December 2011
==================================
I fixed a bug that caused the microphone to freeze when sending the mic sound to
the SoftRock for transmit.

Quisk Version 3.5.10 December 2011
==================================
Lucian Langa contributed a patch to return the primary display size for dual
displays.  If you decreased graph_width for dual displays you will need to change it
back to 0.80 (or similar).

I improved the transmit audio filters to reduce spurs and decrease processing time.
The mic sample rate can now be either 48000 (as before) or 8000 samples per second.
The plan is to make Quisk run effectively on small laptops or even tablet computers.
Remember to adjust mic_clip and mic_preemphasis in your config file.

Quisk Version 3.5.9 November 2011
=================================
I fixed a bug in Windows that occurs only when using the mic for transmit.

Quisk Version 3.5.8 October 2011
================================
The Windows version is now equal to the Linux version, and transmits properly.

I added a parameter agc_off_gain to the config file.  It controls the audio gain
when AGC is off.  Reduce it if sound with AGC off is too loud.  Note that even with
AGC off, the output is limited to the clip level.  I made some other improvements to AGC.

I added FM transmit.  The modulation index can be set in your config file.

I made some improvements to demodulation, and to the SSB transmit filter.   I fixed an
array out of bounds bug in transmit.

Quisk Version 3.5.7 September 2011
==================================
This is a quick release to fix two bugs in 3.5.6, the message "expected integer" and
faint audio for FM.  I also added a new parameter agc_max_gain to the default
configuration file to control the scale of the AGC slider.

Quisk Version 3.5.6 September 2011
==================================
The Spot button now has a level adjustment instead of fixed values.  Right-click the
button to adjust.  There are now three buttons with a slider adjustment, namely AGC,
Spot and the right-most filter button.

I added a feature to measure and remove any DC component in the UDP samples.  I fixed a
problem with the waterfall display when zoomed and using band up-down.

There is a new adjustable  AGC control.   Right click it to show the slider adjustment.  The
full up position corresponds to the old AGC 1.

I removed the 1650 Hertz offset when transmitting SSB.  It was not necessary and cluttered the code.

Quisk Version 3.5.5 July 2011
=============================
These changes only affect the N2ADR 2010 transceiver and the improved version, the HiQSDR.

I moved all the files into a new package directory hiqsdr.  The old n2adr directory
has only the special files I use at my shack.  Please change your config file as follows:

from hiqsdr import quisk_hardware		# Special hardware file
use_rx_udp = 1	# Use this for the N2ADR-2010
use_rx_udp = 2	# Use this for the HiQSDR

The sample config file quisk_conf.py in hiqsdr can be used as is for the HiQSDR.
There is a new dictionary tx_level in the config file to set the transmit level.  See
quisk_conf_defaults.py for other features that can be set for the HiQSDR.

There is a new FPGA firmare version 1.1 available to support the new HiQSDR features.
Note that your firmware version is shown on the Config screen.  It is not necessary to
update your firmware unless you use the HiQSDR and you want the new HiQSDR control lines
to work.  If you do update your firmware, you must run Quisk 3.5.5 or later.

Quisk Version 3.5.4 June 2011
=============================
I added another slider labeled "Zo" to zoom (expand) the graph screen scale
so that narrow signals can be examined.  The center of the graph is changed
to the tuning frequency when zoom is turned on.  To cancel zoom, move the
slider back to the bottom position.  You can tune as usual even if zoom is on.

I put the band buttons on one line so I could add more control buttons.

Quisk Version 3.5.3 May 2011
============================
I added "Documents" as a possible config file location (for Windows 7).

These changes are specific to my 2009/2010 transceiver hardware:
I now detect and display the firmware version.  The files conf_transceiver.py
and hardware_transceiver.py are now the basic config and hardware files for my
transceiver.  The spot button now appears without a special widgets file, so no
widgets file is necessary.  The file quisk_hardware.py is still the hardware file
used in my station, but it is mostly useful as an example of what is possible,
not as a starting point for use by others.


Quisk Version 3.5.2 April 2011
==============================
I added code from Ethan Blanton, KB8OJH, to provide direct frequency control
of the Si570 chip in many SoftRocks.  I added AM transmit and improved AM
receive.  I added FM de-emphasis to receive.  I added a noise blanker.

It is now possible to delay samples (tx_channel_delay) and correct the amplitude
and phase for the sound card play device (SoftRock transmit).  Unfortunately
receive sound card corrections will need to be re-entered.

The filter bandwidths for each mode can now be set in the config file.  And
you can right-click the right-most filter button to adjust its bandwidth.

Quisk Version 3.5.1 February 2011
=================================
The phase correction control has been improved to allow multiple correction
points per band.  Unfortunately this will require re-entering corrections.

I added mic_preemphasis and mic_clip to the config file to control Tx audio processing.

Quisk Version 3.5.0 January 2011
================================
Starting with this version, a Windows version of Quisk is available (alpha code).
I changed the amplitude/phase correction control, and added config file
options (rx_max_phase_correct) to control the maximum available correction.

Quisk Version 3.4.14 January 2011
=================================
The "alsa:" names can now be used for mixer settings.  I added simplified
config and hardware files for my 2010 transceiver hardware. I moved the one
sample delay for some sound cards into the config file instead of using the
#define FIX_H101 (which remains for backward compatibility).  I added more
buttons "GraphP2" to the Graph button to activate a peak hold function.  There
are config file options graph_peak_hold_1 and _2 to control the time constant.

Quisk Version 3.4.13 December 2010
==================================
I decreased the microphone speech processing preemphasis and clipping.
I added a config parameter key_poll_msec to control the SoftRock USB
poll for key status.  I improved the config screen.  Alsa names can
now be strings like "alsa:NVidia" that match the card/device info.
Thanks to Joachim Schneider, DB6QS, I made some improvements to SoftRock
USB control.

Quisk Version 3.4.11 November 2010
===============================
Thanks to Sid Boyce, G3VBV, for sending me SoftRock hardware to work with.
The "mic_play" logic was re-written so that transmit I/Q samples can be
sent from a sound card to hardware that uses QSD up-conversion.  I added
USB access through pyusb to control recent SoftRock models.  A new package
"softrock" directly supports several SoftRock models.  

Change Spot Button to transmit at carrier frequency.

Add a Split button to enable split receive and transmit frequencies.

Fix band change data for pan adapter users.

Try to make easy_install work better.

Quisk Version 3.4.8 August 2010
===============================
A new config file option "playback_rate" can set the radio sound play rate.

I added a button to the config screen to change the decimation rate for
hardware that supports this.  See the new "VarDecim" methods in
quisk_hardware_model.py.  I added this feature to the SDR-IQ hardware
file sdriqpkg/quisk_hardware.py, and to n2adr/quisk_hardware.py.

Thanks to John Nogatch AC6SL for a bug fix.

Quisk Version 3.4.6 July 2010
=============================
I improved the mouse tuning by eliminating a tendency to tune backward.

I made the sdriq extension and my n2adr code into packages in the directories
"sdriqpkg" and "n2adr".  The new package architecture will make it easier for
authors to write Quisk extensions.  See the example config files quisk_conf_sdriq.py
and quisk_conf_n2adr.py to see how to change your imports:

  from sdriqpkg import sdriq
  from sdriqpkg import quisk_hardware
  from n2adr import quisk_hardware
  from n2adr import quisk_widgets

Thanks to Terry Fox, WB4JFI, for improvements to Quisk:
    Code to support the Charleston hardware (libusb-dev required).
    Code to add a third FFT data block.

Quisk Version 3.4.3 June 2010
=============================
The hardware open() method now returns a string for the config screen.  If
you have a custom hardware file, create a string or return the base class string.

I made the SDR-IQ code into a separate Python extension module "sdriq".
This module can serve as a model for other hardware extensions.  It is
the model for the Charleston hardware extension module.  The sdriq.so
file needs _quisk.so, so put both in the same directory.

I corrected the decimation for sample rates greater than 240 ksps, and
improved the filters for all decimations to reduce "images".

The following changes are only relevant if you use the SDR-IQ for capture:

  You need to add these lines to your config file (see quisk_hardware_defaults.py):
        import quisk_hardware_sdriq as quisk_hardware
        display_fraction = 0.85
  There is now a special hardware file for the SDR-IQ.  If you have a
  custom hardware file that uses the SDR-IQ you need to use
  quisk_hardware_sdriq as its base class (instead of quisk_hardware_model).

[See version 3.4.5 for further SDR-IQ changes]

Quisk Version 3.4.2 May 2010
============================
The config file has a new option to add an external demodulation module.  I
I added the ability to play in stereo, and corrected the sidetone logic.

The config file has a new option to add a full duplex button.

I added the ability to use PortAudio for sound card access.  PortAudio
can also be used to connect Quisk to other programs.

I added a key up delay to the is_key_down() serial port code and fixed a
sound card CW bug.

A new config file entry can make amplitude/phase corrections independent
of band.  This is needed for a panadapter.

I fixed the compressed graph labels at high sample rates.

Quisk Version 3.3.7 April 2010
=================================
If you get samples from a UDP port, you can specify the decimation rate in the
config file.

If you send samples to a sound card for transmit, CW now works (as does SSB).

Quisk Version 3.3.6 February 2010
=================================
I added BandEdge to the config file, and added code to Quisk to make the
frequency and band changes more rational.  I changed the config file
attribute freqTime to bandTime (see changes).

You can now define a class named "Hardware" in your config file, and then
you don't need a separate hardware file.  This is only recommended for
simple hardware needs.  See docs.html.

If you use the microphone and send samples with UDP, the audio is now
centered at 1650 Hertz, and you must add/subtract this offset when
setting the transmit frequency.

A number of valuable patches were submitted by Andrew Nilsson, VK6JBL,
and these were incorporated into Quisk:

  The band buttons displayed can be changed in the config file (bandLabels).

  The 6 meter band was added (change bandLabels to show it).

  Turn on add_imd_button in the config file to generate 2-tone test signals.

  The two new functions QS.capt_channels(i, q) and QS.play_channels(i, q)
  will set the capture and playback channel numbers at any time.

  If you set the key method to "", the new function QS.set_key_down(1) will
  set the key state up or down.  This enables you change the key state using
  either C or Python; for example, to add a "MOX" button.

  The microphone samples can now be output to a sound card for transmit.  See
  the additional items in the config file.

  The new config file parameter mouse_tune_method causes mouse drag tuning
  to change the VFO frequency, not the Quisk tuning frequency.

I moved microphone_name and tx_ip (for the microphone) to the config
file from the hardware file so that all the mic parameters are together.

Quisk Version 3.3.1 December 2009
=================================

For sound card input, I added controls to correct amplitude and phase
balance.  Press the new button on the config screen.  A different
correction is saved for each band.  See the help file.

I added the new band "Audio".  It sets the VFO frequency to
zero and is meant to be used with a sound card.  I changed the WWV and CHU
bands to a new Time band.  The time frequencies are named freqTime and can
be changed in your config file.

In the file quisk_conf_defaults.py I changed the default for persistent_state
to True, and added graph_width=0.8 to specify the graph width.

A period "." in the frequency entry box means megahertz.

Quisk Version 3.3.0 November 2009
=================================

I fixed a bug in the SDR-IQ decimation that produced slight audio
distortion at decimations other than 500.

If the play device is the null string "", Quisk no longer tunes and
demodulates the signal.  This saves CPU cycles when Quisk is used
as a panadapter.

I added decimation (reduction of sample rate) before the filters so that
Quisk can handle higher sample rates or slower computers.

I made the waterfall into a splitter window with a graph display at the top.
There are new attributes in the config file to control this feature.

The numeric value of Ys and Yz are now shown so that the values can be added
to the config file more easily.

There is a new config option to save the state (band, frequency, etc.) on
exit, and restore it on startup.  Only certain bits of state are saved; the
others are still taken from the config file.

The default config file sets fft_size_multiplier to zero, and this specifies
that Quisk should calculate it for you.

Quisk Version 3.2.3 September 2009
==================================

Fixed a bug that prevented tuning the SDR-IQ when using the
default hardware file.  Started adding code to capture sound
from a UDP socket.

Quisk Version 3.2.2 June 2009
=============================

The microphone access was re-written to make it work with more
sound cards.  The config file has a new parameter "mic_channel_I"
to specify which sound card channel is used for the mic.

Added Documentation.html.  Fixed lack of poll to ReturnFrequency().

Quisk Version 3.2 May 2009
==========================

Quisk now uses wxPython instead of Tkinter for its graphical
user interface.  You must install the python-wxgtk2.8 package.
Get the latest version available.  If you still want to run
the Tkinter version, it is quisk_tk.py.  The wxPython version
is much faster.

Quisk now runs in two threads; a GUI thread and a sound thread.

I moved the colors to the config file so you can change the
colors more easily.


Quisk Version 3.1  April 2009
=============================

New hardware file to control the AOR AD8600.

I added filtering to FM audio to remove CTCSS tones and provide
-6 dB / octave de-emphasis.

I removed the tkdirect C-language module and replaced it with a
pure Python equivalent.  This reduces compilation problems.

I improved the speed of the screen updates so that Quisk will run
without clicks on slower computers.


Quisk Version 3.0.0  April 2009
===============================

Thanks to Leigh L. Klotz, Jr. WA5ZNU, my special hardware control was
removed to separate files so that Quisk now has a cleaner design that
is more useful to others.

