Page 17 of 19

Re: 2.6.8

Posted: Fri Nov 01, 2019 4:52 am
by wa1oxt
Wow... that was quick !!!

Great job on d6-1.

Thank's Rich

wa1oxt / / garyradio

Re: 2.6.8

Posted: Fri Nov 01, 2019 5:04 am
by ramdor
wa1oxt wrote:Wow... that was quick !!!

Great job on d6-1.

Thank's Rich

wa1oxt / / garyradio


I don't really know if it fixed the issue as I couldn't replicate it here. However I did find something related, which may have been causing some oddities.

Richie.

Re: 2.6.8

Posted: Fri Nov 01, 2019 10:33 am
by W3MMR
Installed D6 this morning and im having an issue with the data line. #1, its not as "smooth" as the data line in D5, and it looks like the data line when I have GDI+ selected instead of DirectX. But if I have the line size below 1.0, i keep mine on 0.8 normally, it becomes spotty. This used to happen to me if I was to go from DirectX, to GDI+, and back to DirectX. Then closing an reopening the program would always fix it. But now on D6, nothing fixes it. Its all the time, no matter what I try. I tried restarting my PC, powering the radio itself down, and still the same result. Anyone else experiencing this? And yes, i put all the files in the directory this time hi hi :)

Perry

This is what im talking about
data.PNG
data.PNG (974.73 KiB) Viewed 2922 times


This is how it looks in D5, with the same settings.
good_data.PNG
good_data.PNG (983.58 KiB) Viewed 2922 times


These are my settings
appearence.PNG
appearence.PNG (28.3 KiB) Viewed 2922 times

Re: 2.6.8

Posted: Fri Nov 01, 2019 11:22 am
by ramdor
Hi Perry,

Yes, this is due to anti-aliasing being disabled. To draw a line narrower than 1 pixel nicely anti-aliasing needs to be enabled. Best you can do at the mo is set line width to 1.0 and pick colour that is dimmer/darker than you really want.

I made the change to bring it into line with the GDI+ display as I disliked the way the line seemed to fade as it became sub 1 pixel wide and gdi+ didn't really do anything below 1.0 anyway. I will add an option to enable/disable anti-aliasing.

Richie.

Re: 2.6.8

Posted: Fri Nov 01, 2019 11:34 am
by vk1hx
_

Re: 2.6.8

Posted: Fri Nov 01, 2019 12:06 pm
by NC3Z
vk1hx wrote:Don't want to be a pain. But.... Can all these updated dll's and .exe files be combined into a new FULL 2.6.8 release.


A handful of messages back about the "D" version that is what Ritchie said "All source/changes for the (d) builds have just gone off to Doug".

But for these quick fixes and testing you will have to settle for the individual files, price to pay for being a tester, and well worth the price. It also makes going back and forth between versions for troubleshooting much, much easier with the individual files.

Re: 2.6.8

Posted: Fri Nov 01, 2019 1:08 pm
by ULTIMAX
Richie,

You know everyday I look forward to your updates and awesome work, is like a nice morning coffee can't live without
and I appreciated that very much, thank you so much Brother.

BTW. love the meter choice option Gary good thought !!!

Al
73

Re: 2.6.8

Posted: Fri Nov 01, 2019 2:35 pm
by W1AEX
ramdor wrote:Hi Rob,

If you have a mo, could you try : https://www.dropbox.com/s/cg4yqonzcynke ... 1.zip?dl=0

Richie.


(EDIT: MY MISTAKE... THE D6_1 RELEASE DID NOT FIX THE CW OFFSET ISSUE WHEN CYCLING THROUGH THE BANDSTACK.)

Hi Richie,

The d6_1 release fixes the strange bandstack problem nicely here. No sign of the "offset" changing value or migrating to other frequencies saved in the bandstack.

I'll definitely echo Al's comment above and once again thank you for tracking down the little oddities that lurk in the code! It's getting better and better!

73, Rob W1AEX

Re: 2.6.8

Posted: Fri Nov 01, 2019 2:50 pm
by K2BU
Hi Richie and Rob,

I just loaded D6_1 and thought the problem was solved with the band stacking jump in frequency with both receivers on and VFO Sync activated. However, if you make one of the band stacking frequencies a CW mode frequency, the problem is still there. For example, one of my frequencies for me is W1AW on 1802.5. If I have both receivers on with VFO Sync active, multiple stacking frequencies will jump every time I go through the five band stacks. So we are getting closer, but still need to tweak the code. Thank for all the interest.

Re: 2.6.8

Posted: Fri Nov 01, 2019 3:20 pm
by W1AEX
K2BU wrote:Hi Richie and Rob,

I just loaded D6_1 and thought the problem was solved with the band stacking jump in frequency with both receivers on and VFO Sync activated. However, if you make one of the band stacking frequencies a CW mode frequency, the problem is still there. For example, one of my frequencies for me is W1AW on 1802.5. If I have both receivers on with VFO Sync active, multiple stacking frequencies will jump every time I go through the five band stacks. So we are getting closer, but still need to tweak the code. Thank for all the interest.


Uh-oh Chas. I just ran the test again (RX2 enabled and VFO Synch enabled with at least one frequency in the bandstack set to CW) and you are right, the CW mode frequency started jumping frequency again and an offset appeared on the first bandstack frequency after the one set for CW mode. Not sure why it did not do that the first time I checked. I must have missed something.

Sorry about that Richie but D6_1 did not stop the issue.

73, Rob W1AEX

Re: 2.6.8

Posted: Fri Nov 01, 2019 10:07 pm
by ramdor
UPDATE (last for D run I hope)

Hi Rob, Chas, everyone

Please try : https://www.dropbox.com/s/crioh3xc0cdpy ... 2.zip?dl=0

Please let me know if the CW stack anomaly is fixed.

Doug is going to give me a black mark when I send him the source again (for the 3rd time).... eeek :lol:

Richie.

(10/7/19) d6_2
- fix: sync vfo’s was not setting dspmode before changing frequency. Consequently cwshifts were applied when they shouldn’t have been if leaving a CW bandstack entry for example.
- add: anti alias option in Setup->Display->General. If you want sub pixel line widths and no jaggies enable this, but there will be some GPU overhead, really noticeable on GDI+, no discernible impact when in DirectX

Re: 2.6.8

Posted: Sat Nov 02, 2019 1:25 am
by W1AEX
Hi Richie,

I installed d6_2 and then checked the bandstack function with RX2 and VFO SYNCH enabled on 160 meters:

Bandstack frequencies: 1802.800 (CW), 1885 (AM), 1930 (SSB), 1962 (SSB), 1993 (SSB)

In the video below it cycles through the bandstack twice perfectly with VFO SYNCH disabled, but at around 20 seconds when VFO SYNCH is enabled it falls apart quickly and scrambles some frequencies stored in the bandstack. Somehow enabling VFO SYNCH seems to corrupt the first and second frequencies in the bandstack. Not sure if that helps but this is what seems to be happening.

Thanks for taking a look but maybe the bandstack issue is a bit too tangled up for Doug's next release. We don't want you getting ANY black marks for your efforts!

73, Rob

Note: There's no audio in the video. It's definitely easier to view if you click in the lower right of the video and watch it in full screen mode.

Re: 2.6.8

Posted: Sat Nov 02, 2019 1:40 am
by ramdor
Cheers Rob. Will keep looking. Do me a favour, try exactly the same test, but click band button to move stack very slowly, one once per 5-10 seconds.

Richie.

Re: 2.6.8

Posted: Sat Nov 02, 2019 1:48 am
by W1AEX
Roger that, will run the bandstack cycle at that rate and be back in a few minutes.

Re: 2.6.8

Posted: Sat Nov 02, 2019 2:08 am
by ramdor
I can't replicate it here.... very very odd.



Richie.

Re: 2.6.8

Posted: Sat Nov 02, 2019 2:21 am
by W1AEX
NOTE: Just saw your video above. That is perfect and it's strange what Chas and I are seeing here. I'm stumped!

Richie,

Here is what happened with the following 5 frequencies in the bandstack:

1. 1802.8 CW
2. 1885 AM
3. 1930 LSB
4. 1960 LSB
5. 1993 LSB

Starting at 1802.8 (CW) with RX2 enabled and VFO Synch enabled. Clicking the band button at a rate of once every 10 seconds it jumped to the following frequencies and modes:

1. 1803.8 AM
2. 1803.8 LSB
3. 1965 LSB
4. 1993 LSB
5. 1991.8 CW
6. 1993 LSB
7. 1930 LSB
8. 1965 LSB
9. 1993 LSB
10. 1991.8 CW
11. 1993 LSB
12. 1930 LSB
13. 1965 LSB
14. 1993 LSB
15. 1991.8 CW

At this point it keeps repeating the pattern of frequencies and modes shown in the list from click 5 to click 9 in the list above. So the first and second original bandstack frequencies are corrupted but the last three seem to be preserved.

Sorry this took so long but I ran it twice to double check it.

73, Rob

Re: 2.6.8

Posted: Sat Nov 02, 2019 2:41 am
by devnull
small issue to report.
In Andromeda (Top and bottom enabled) having CTUN enable and switching to TX the display changes and displays TX on a different frequency.
See attached screeners.

BTW LOVE THE INTERFACE !!! Slick and modern!

Cheers,
Hauke

Re: 2.6.8

Posted: Sat Nov 02, 2019 2:55 am
by W1AEX
Hi Richie,

I exported my current database and then reset Thetis to install a fresh database. Running with the fresh database I ran the same frequencies that I used before and it came up with exactly the same repeating pattern of frequencies as in my earlier test. I'm really stumped now!

Thanks for looking into this but since it works correctly with your installation I'm not sure what you can do with this one and I suspect you have many other things of greater importance to attend to! Tomorrow I'll try rolling the versions from 2.6.8 through the b, c, and d variations to see if it shows up with a certain version. That might be useful information.

73, Rob

Re: 2.6.8

Posted: Sat Nov 02, 2019 3:20 am
by ramdor
I think I have tracked the issue Rob, but it will take me a bit to resolve it. Running out of steam here now 3:20am so will pick back up tomorrow.

Ok about the DB, yes I went and checked mine here, and ran it in US region, but all looked ok.

Also all noted about Androm issue Hauke.

Doug has released a new build setup/package over on his github btw, and is waiting for these gremlins to be sorted over the next couple of days before it goes over the TAPR release.

Richie.

Re: 2.6.8

Posted: Sat Nov 02, 2019 3:22 am
by K2BU
Hi Richie and Rob,
I have just tested the latest version d6_2 and the band stacking situation is worse than before. Make sure you are testing the correct setup. The first band stack is on a CWL or CWU mode. Both RX1 and RX2 are on and VFO Sync is turned on. Using this setup, the stacking frequencies on 160 and 80 change every time you go through the five frequencies per band. The last correctly working software I have used was d1 which I am back using now. In my station, I use RX2 solely for my beverage receive antennas. So VFO Sync is on all the time, and one of my stacking frequencies on every band is on a CW frequency using CWU. Looking at the code setup for d1 may be the answer as it works fine at my station. Again, this was the last update that worked correctly for me with this issue. I hope that helps.

Re: 2.6.8

Posted: Sat Nov 02, 2019 3:28 am
by devnull
Oh and one more thing, a wish if I might be so bold .....
On the Andromeda button bar I could do with CTUN and an Antenna Selector.

Oh and PLEASE do not make the S-Meter bigger in this view.

Thanks in advance,
Hauke

Re: 2.6.8

Posted: Sat Nov 02, 2019 3:30 am
by ramdor
Cheers all for the feedback.

d2... there is a massive change that had to be done to resolve ui thread from locking from time to time

- change: vfoa/b/suba now all use property get/set with single point of text conversion, and member variables to hold frequencies instead of direct access to text boxes

This is the cause of the current band stack problem when in SYNC mode. I do believe band stacks will be ok if you run the vfo's un-synced. It is not so much a problem with the band stack, but the fact that vfoB thinks vfoA is not on the correct frequency, so changes it. Then the bandstack thinks you changed the frequency of the active stack, and boom, all in a mess.

I'll fix it tomorrow, but having an hour or so in Farming Simulator to remove the stress then zzzz... rofl.

73 for now :) Richie.

Re: 2.6.8

Posted: Sat Nov 02, 2019 5:15 am
by w2ner
One thing thats its doing now is, the Rec ant is coming on by default when I power it up. Even if I set it, save the profile it still comes up with the Rec antenna enabled on power up.

Re: 2.6.8

Posted: Sat Nov 02, 2019 9:06 am
by ramdor
w2ner wrote:One thing thats its doing now is, the Rec ant is coming on by default when I power it up. Even if I set it, save the profile it still comes up with the Rec antenna enabled on power up.


need more info Nicholas. I can click it to change to tx ant and rx ant, and closing down brings it back in the state it was last in, and state of it is not related to a profile as far as I can see in the code.

Richie.

Re: 2.6.8

Posted: Sat Nov 02, 2019 9:43 am
by ramdor
devnull wrote:small issue to report.
In Andromeda (Top and bottom enabled) having CTUN enable and switching to TX the display changes and displays TX on a different frequency.
See attached screeners.

BTW LOVE THE INTERFACE !!! Slick and modern!

Cheers,
Hauke


Can you provide steps to get this to happen?

Cheers Richie.

Re: 2.6.8

Posted: Sat Nov 02, 2019 10:50 am
by W3MMR
Thanks Richie. The data line with anti-aliasing disabled is not nearly as smooth, and this is above a 1.0 line size, as it is when enabled. The data line with it disabled, looks like the data line when I have GDI+ selected. Anyway, thanks for putting in an "on/off" switch hi hi. ;) I enjoy it much better with it enabled and DirectX selected. And I want to thank you again Richie for all the work you're putting into this. Keep up the great work!

Now its hard to really see what I mean, just from the small attachment sizes, but they're definitely different.
DirectX, 1.0 Data Line Size
smooth.PNG
smooth.PNG (1.09 MiB) Viewed 2677 times


GDI+ 1.0 Data Line Size
gdi+.PNG
gdi+.PNG (1.05 MiB) Viewed 2677 times

Re: 2.6.8

Posted: Sat Nov 02, 2019 11:14 am
by ramdor
W3MMR wrote:Thanks Richie. The data line with anti-aliasing disabled is not nearly as smooth, and this is above a 1.0 line size, as it is when enabled. The data line with it disabled, looks like the data line when I have GDI+ selected. Anyway, thanks for putting in an "on/off" switch hi hi. ;) I enjoy it much better with it enabled and DirectX selected. And I want to thank you again Richie for all the work you're putting into this. Keep up the great work!


yes with it on direct x can do sub pixel lines, personally I prefer it off, and 1.0 line thickness but the option is there now. It is amazing the impact it has on GDI+ frame rates. The following are with 1.0 line width.

DirectX AA On (MSAA=1 AM=Per-primitive)
dx_aa_on.jpg
dx_aa_on.jpg (9.36 KiB) Viewed 2668 times


DirectX AA Off (MSAA=1 AM=Per-primitive)
dx_aa_off.jpg
dx_aa_off.jpg (11.74 KiB) Viewed 2668 times

Re: 2.6.8

Posted: Sat Nov 02, 2019 12:26 pm
by Helmut
Hi all,

for your info - with the D6 version scrolling the qrg with the mousewheel won´t work in the panadapter
when I before tried to scroll in the frequency window. Have now 4.1d and all o.k.

73, Helmut

Re: 2.6.8

Posted: Sat Nov 02, 2019 12:35 pm
by ramdor
Helmut wrote:Hi all,

for your info - with the D6 version scrolling the qrg with the mousewheel won´t work in the panadapter
when I before tried to scroll in the frequency window. Have now 4.1d and all o.k.

73, Helmut


Which d6 version? and where are you using the mouse wheel?

If you are just using mouse wheel in the window somewhere, then you need the following checked...
Capture.jpg
Capture.jpg (9.5 KiB) Viewed 2641 times


However, even with that unchecked, you will be able to use mousewheel over VFO boxes and over spectral area.

After some testing, it seems the area in purple is the only spot wheel has no effect. Wheel over the 20m extra ssb and tx button works as well.

Capture.jpg
Capture.jpg (9.67 KiB) Viewed 2636 times


edit: also make sure you didnt lock the vfo.

Richie.

Re: 2.6.8

Posted: Sat Nov 02, 2019 1:05 pm
by w2ner
ramdor wrote:
w2ner wrote:One thing thats its doing now is, the Rec ant is coming on by default when I power it up. Even if I set it, save the profile it still comes up with the Rec antenna enabled on power up.


need more info Nicholas. I can click it to change to tx ant and rx ant, and closing down brings it back in the state it was last in, and state of it is not related to a profile as far as I can see in the code.

Richie.


Never mind, I see what it was. My error. Thanks Richie