Thetis v2.8.11 - UPDATES and current issues

tok2010
Posts: 36
Joined: Sun Apr 09, 2017 9:18 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby tok2010 » Mon Oct 25, 2021 12:51 am

Hi Richie, All,

From yesterday I started to use the update version here.

I would like to report the comparison between Thetis original 2.8.11 and the updates on Pure signal and TX/RX sound.

First of all, PS Amp View is strange between them. Please look at the screenshot. In 2.8.11, Mag. Corr. curve (red line) is displayed always, however in 21j this red curve is not consistent when I speak. Most of time, it disappear. 2-Tone test is quite OK. But speaking not. The amplifier setting, and others are the same. Since this test is done within 10 minutes.

As for TX sound quality, my local ham reported Thetis 2.8.11 is much better than 21j this morning.

Concerning RX sound, the sound is the same between 21j and 2.8.11, however RX sound changes between21k and 2.8.11. I prefer the RX sound of 2.8.11.

My RX sound is from Orion MKII, not through VAC. But I use VAC for TX.

My PC and the firmware of Orion are shown in the attached pictures. The setting of Thetis for 21j/21k remains unchanged except enabling TUNE STEP by mode, which is not in Thetis 2.8.11.

73,
Tokio
JA1CCA
Attachments
Firmware.PNG
Firmware.PNG (4.25 KiB) Viewed 12433 times
cpu.PNG
cpu.PNG (3.18 KiB) Viewed 12433 times
PS amp view.PNG
PS amp view.PNG (76.51 KiB) Viewed 12433 times
nubbyless
Posts: 26
Joined: Sun Jul 12, 2020 9:42 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby nubbyless » Mon Oct 25, 2021 7:06 pm

For the the amp view issue I would suggest unchecking low res bottom right in your screenshot and see if that helps.
User avatar
W4ATL
Posts: 15
Joined: Tue Mar 09, 2021 9:14 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby W4ATL » Mon Oct 25, 2021 7:21 pm

It seems not many people using Thetis run CW but I hope a few do. One nagging issue I have running CW with Thetis on my new Anan 7000DLEMKII is when in QSK there are 2 tones coming through. The Sidetone (when enabled) and what I believe to be the transmitted signal. I can get rid of one of the tones by disabling the sidetone so that only one tone remains. But that tone is delayed so that using a paddle with it is almost impossible when operating over 20wpm. Is there a workaround, besides just switching to semi break-in? It would be nice to enable the sidetone and disable the transmitted tone.
User avatar
w-u-2-o
Posts: 5525
Joined: Fri Mar 10, 2017 1:47 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w-u-2-o » Mon Oct 25, 2021 8:12 pm

When running QSK you need to have sidetone disabled. Search on "QSK" in this forum.

Reminder: this topic is for bug reporting on Richie's developmental versions, not new feature requests (there's a separate topic for that), and not for help with Thetis (search for existing topics, or create new ones if necessary).
w2gs
Posts: 36
Joined: Sun Apr 09, 2017 5:18 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w2gs » Tue Oct 26, 2021 12:31 pm

Hi Ritchie,
Using ver 21k7. I am having problems with transmit profiles. When I change modes via the band buttons and band stack the SSB mode Transmit Profile doesn't always change to the parameters that were saved in the SSB profile. A simple example would be if a setup a profile "SSB" with MIC, COMP, TX EQ, and TX FL turned on and save it under Setup>Transmit. When I change from 40M CW to 20M SSB the profile SSB is shown in the Transmit Profile box, but COMP, TX EQ and TX FL are not turned on. I have done many test with different saved profiles and some work and some don't. If I change modes and bands as in the above example and the saved parameters don't open properly even if I open the profile in Setup>Transmit the expected parameters don't open. Not sure what is causing this inconsistent operation. Any thoughts?

73, George
W2GS
User avatar
w-u-2-o
Posts: 5525
Joined: Fri Mar 10, 2017 1:47 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w-u-2-o » Tue Oct 26, 2021 2:19 pm

George -- in Setup > Transmit, do you have either of the "Auto Save TX Profile" options turned on? If so, try turning them off and resaving your profiles.
tok2010
Posts: 36
Joined: Sun Apr 09, 2017 9:18 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby tok2010 » Tue Oct 26, 2021 9:13 pm

Hi nubbyless,

Thank you for your comment on my posting.
Low Res on PS Amp View panel is for resolution. It doesn't affect the issue I posted.

I suppose the problem would be related to the baseband (audio) processing modification in Richie's development.
In addition, in 21j PS improvement is lower than Thetis original's HERE.

73,
Tokio
JA1CCA
w2gs
Posts: 36
Joined: Sun Apr 09, 2017 5:18 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w2gs » Wed Oct 27, 2021 12:24 pm

Hi Scott,
Yes both auto save TX profile boxes are checked.
73, George
W2GS
User avatar
w-u-2-o
Posts: 5525
Joined: Fri Mar 10, 2017 1:47 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w-u-2-o » Wed Oct 27, 2021 1:15 pm

w2gs wrote:Hi Scott,
Yes both auto save TX profile boxes are checked.
73, George
W2GS

Did you try turning them off as I suggested and manually resaving the various profiles you want?
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Thetis v2.8.11 - UPDATES and current issues

Postby W1AEX » Thu Oct 28, 2021 2:01 am

On 160 meters tonight three of us running Thetis 2.8.11 v21k7 compared notes on the CFC Visualizer and we found it was not working correctly for any of us. Two of us were running VAC and we thought that might be the issue but the third user was not running VAC and his description sounded like what I see in my screenshot with everything squeezed into the left side of the display. Have we mis-configured something or do others see the same thing we see?

Thanks and 73, Rob W1AEX
Attachments
CFC_visualizer.jpg
CFC_visualizer.jpg (68.27 KiB) Viewed 12086 times
"One thing I am certain of is that there is too much certainty in the world."
dynamicfusion
Posts: 57
Joined: Thu Oct 14, 2021 5:23 am

Re: Thetis v2.8.11 - UPDATES and current issues

Postby dynamicfusion » Thu Oct 28, 2021 2:24 am

I discovered a bug with the "attenuation during tuning". I set mine to be 30db attenuation. Save profile and all that. Close thetis. Open it back up, and it arbitrarily gets set back to default, or random numbers. In some cases it gets changed to 0 ! This caused me to get a ADC overload message which was my only clue something changed. What the h is going on with the tuning/transmit attenuation values? 21k7 version

UPDATE 1:

It appears the tune power variable is tied to what band is selected. OK, so I went through each band and set it to 30db atten. Saved profile. Exported. Closed thetis, reopened, and it 20 meters changed back to 17db atten, the heck? So random. I believe this variable is either being modified from something else, or it has a goal to always reset to default value.
Attachments
tuning power changes.jpg
tuning power changes.jpg (104.96 KiB) Viewed 12080 times
Last edited by dynamicfusion on Thu Oct 28, 2021 5:28 am, edited 1 time in total.
User avatar
W9BHI
Posts: 127
Joined: Wed Apr 17, 2019 3:30 pm
Location: Serena Il.

Re: Thetis v2.8.11 - UPDATES and current issues

Postby W9BHI » Thu Oct 28, 2021 3:45 am

I don't use the CFC function but I checked it out and I am seeing the same thing as W1AEX.
W9BHI
User avatar
ramdor
Posts: 1468
Joined: Wed Jul 03, 2019 3:07 pm
Contact:

Re: Thetis v2.8.11 - UPDATES and current issues

Postby ramdor » Thu Oct 28, 2021 9:37 am

dynamicfusion wrote:I discovered a bug with the "attenuation during tuning". I set mine to be 30db attenuation. Save profile and all that. Close thetis. Open it back up, and it arbitrarily gets set back to default, or random numbers. In some cases it gets changed to 0 ! This caused me to get a ADC overload message which was my only clue something changed. What the h is going on with the tuning/transmit attenuation values? 21k7 version

UPDATE 1:

It appears the tune power variable is tied to what band is selected. OK, so I went through each band and set it to 30db atten. Saved profile. Exported. Closed thetis, reopened, and it 20 meters changed back to 17db atten, the heck? So random. I believe this variable is either being modified from something else, or it has a goal to always reset to default value.


If you use PS it will be adjusted. There are also times when it is reset to 31.

Richie.
NC3Z
Posts: 463
Joined: Sun Oct 29, 2017 8:57 pm
Location: Merritt, NC

Re: Thetis v2.8.11 - UPDATES and current issues

Postby NC3Z » Thu Oct 28, 2021 11:13 am

W1AEX wrote: Have we mis-configured something or do others see the same thing we see?


Rob, it does not work for me either, but not sure what it would even show me that would help improve the audio over what it is now.
Gary NC3Z
User avatar
kc2rgw
Posts: 165
Joined: Mon Jun 22, 2020 5:44 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby kc2rgw » Thu Oct 28, 2021 12:35 pm

NC3Z wrote:
W1AEX wrote: Have we mis-configured something or do others see the same thing we see?


Rob, it does not work for me either, but not sure what it would even show me that would help improve the audio over what it is now.


I was one who requested it. The idea is to show how much impact or reduction each band of the CFC is engaging. It’s useful to understand the spectral distribution where without it you are just going by ear but may not realize just how hot a hot spot is or that your overall gain staging is not in balance. I.e. lets say you are just hammering 63 Hz into the CFC, it would be better, cleaner, to attenuate 63 Hz before it even gets to that stage of processing rather than pounding into a limiter.

It’s too fast to see on my system, didn’t seem to register in VAC mode but I went over to analog wired mode and it works but it’s so fast it’s hard to make any sense of currently.
User avatar
ramdor
Posts: 1468
Joined: Wed Jul 03, 2019 3:07 pm
Contact:

Re: Thetis v2.8.11 - UPDATES and current issues

Postby ramdor » Thu Oct 28, 2021 1:40 pm

kc2rgw wrote:
NC3Z wrote:
W1AEX wrote: Have we mis-configured something or do others see the same thing we see?


Rob, it does not work for me either, but not sure what it would even show me that would help improve the audio over what it is now.


I was one who requested it. The idea is to show how much impact or reduction each band of the CFC is engaging. It’s useful to understand the spectral distribution where without it you are just going by ear but may not realize just how hot a hot spot is or that your overall gain staging is not in balance. I.e. lets say you are just hammering 63 Hz into the CFC, it would be better, cleaner, to attenuate 63 Hz before it even gets to that stage of processing rather than pounding into a limiter.

It’s too fast to see on my system, didn’t seem to register in VAC mode but I went over to analog wired mode and it works but it’s so fast it’s hard to make any sense of currently.


It is a snapshot of the compression, and does not have any history/hang time. Also, it is only queried around 20 times a second so it is no doubt missing lots of information. It needs more work, and will require some adjustments in the cfc dsp code. We may have to keep peak values in the snapshot buffer, and then reset them after the the query is made. I will talk to Warren to see if we can improve it.

A quick look at the issue : https://youtu.be/9OlfLs-h5pk

R.
User avatar
w-u-2-o
Posts: 5525
Joined: Fri Mar 10, 2017 1:47 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w-u-2-o » Thu Oct 28, 2021 5:40 pm

All,

IMHO, the CFC compression display is probably not worth pursuing, and it might be best to simply remove it from the codebase.

Using professional audio industry UI trends as a guide, it is easy to see that audio engineers have not found real-time display of compression gain vs. frequency to be a very useful tool, and it seems that many here are drawing the same conclusions.

The most advanced multi-band compressor UIs concentrate on displaying the audio input spectrum with a method to draw the desired compressor response vs. frequency as an overlay. This way the engineer can visualize where the energy is and where they want to add or subtract compression. The rest is done primarily by ear, not by eye, hence the lack of sophisticated real-time displays of compression gain.

This is well illustrated in this video review of several high-end, multi-band compressors:



However, the "knee" type display in the BlueCat MB-5 compressor is very compelling. I have used that exact compressor in some of my other work and that specific type of display is very useful and intuitive. But it would potentially be a lot of work for Richie and I'm not sure the juice is worth the squeeze.

73,

Scott
User avatar
kc2rgw
Posts: 165
Joined: Mon Jun 22, 2020 5:44 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby kc2rgw » Thu Oct 28, 2021 8:48 pm

I disagree having worked professionally with compressors. If you don’t have an indicator of the gain reduction it can be hard to understand the threshold, level/depth of reduction, attack and release times etc. When you don’t understand how the compressor is applied to and affecting the signal, you can spend time poking at other things as symptoms where the compressor itself is the cause.

Does the average Anan user care? Probably not, but to say that visual feedback is not useful with compression, that’s not true.

One of the primary reasons I hang onto my aging Symetrix is to see the indicators for the gate and compressor to aid in adjusting them. One of the worst pieces of gear I own is an Aphex Channel that has no threshold adjustment and a display that doesn’t match its impact with the compressor it has in it. The compressor in that thing engages long before you see any indication. By the time the meter really indicates, you are just hammering into it and the results are not good.

The CFC is kind of a black box right now as is. When you raise a slider are you reducing the threshold for that frequency band or increasing it? Are the sliders thresholds or are they attack time? Yes, you can wing it by ear but it’s a lot more useful if you know what it is doing and visual feedback is useful.

The sound quality of the CFC vs using a clipper (as in EESDR) is superior, but I dislike black boxes.
User avatar
w-u-2-o
Posts: 5525
Joined: Fri Mar 10, 2017 1:47 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w-u-2-o » Thu Oct 28, 2021 10:32 pm

kc2rgw wrote:I disagree having worked professionally with compressors.

We'll just have to agree to disagree. Certainly the industry tends to side towards having no such displays by a large margin.

The CFC is kind of a black box right now as is. When you raise a slider are you reducing the threshold for that frequency band or increasing it? Are the sliders thresholds or are they attack time?
Neither. Raising the slider increases the compression ratio. There is no control over threshold, attack, hold or release time (I was able to get those controls into the downward expander/gate/VOX section, though).
User avatar
kc2rgw
Posts: 165
Joined: Mon Jun 22, 2020 5:44 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby kc2rgw » Thu Oct 28, 2021 11:27 pm

Neither. Raising the slider increases the compression ratio. There is no control over threshold, attack, hold or release time (I was able to get those controls into the downward expander/gate/VOX section, though).


Interesting, so what happens if you drop the slider?

I may have missed this in the w0dsp docs or something, do you have a handy link to where this is, if it exists in docs?
dynamicfusion
Posts: 57
Joined: Thu Oct 14, 2021 5:23 am

Re: Thetis v2.8.11 - UPDATES and current issues

Postby dynamicfusion » Thu Oct 28, 2021 11:38 pm

ramdor wrote:
dynamicfusion wrote:
If you use PS it will be adjusted. There are also times when it is reset to 31.

Richie.


Thanks for the reply. I am not quite understanding why in some cases it goes all the way up to zero attenuation? Won't this blow out some receiver module? Shouldn't it be hardlocked to at least 15db attenuation? Furthermore why is Pure Signal related to tuning tone output in the first place? I reckon we dont need to be correcting a single tone being sent out for tuning purposes, but that's only me. Thanks.
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Thetis v2.8.11 - UPDATES and current issues

Postby W1AEX » Thu Oct 28, 2021 11:39 pm

Richie,

Thank you for putting together the CFC video to show what's happening. I was able to duplicate your setup and observed perfect sweeps across the CFC visual display and the "Auto-Scale" set up the vertical range perfectly. I was mostly curious if I had missed a setting somewhere when I tried to observe my transmitted audio in the CFC display and only saw the deflection squeezed into the left part of the horizontal range. Normally, after setting all my levels in the software audio chain from front to back I usually close my eyes and listen to the monitor while running a recording of my voice played through the Thetis voice processing software. Then I make my geeky compression adjustments to add punch and clarity and run with that. I guess the visual representation is not a necessity for me but understand that it might be helpful to someone else. Thanks for all the progress you have made with advancing Thetis!

73, Rob W1AEX
"One thing I am certain of is that there is too much certainty in the world."
NN3RP
Posts: 65
Joined: Sun Feb 25, 2018 1:53 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby NN3RP » Tue Nov 02, 2021 8:40 pm

dynamicfusion wrote:
ramdor wrote:
dynamicfusion wrote:
If you use PS it will be adjusted. There are also times when it is reset to 31.

Richie.


Thanks for the reply. I am not quite understanding why in some cases it goes all the way up to zero attenuation? Won't this blow out some receiver module? Shouldn't it be hardlocked to at least 15db attenuation? Furthermore why is Pure Signal related to tuning tone output in the first place? I reckon we dont need to be correcting a single tone being sent out for tuning purposes, but that's only me. Thanks.



I recently updated to 21k7 and I am having the same issue...the ATT on Tx I set it to 31 as I always have...then when TX it immediately reverts to 0 (zero). I does not matter if saving the TX profile with ATT set to 31.....
PS-A Enabled.jpg
PS-A is enabled but does not allow 31 ATT
PS-A Enabled.jpg (100.34 KiB) Viewed 11654 times


A curious / Interesting finding: If I disable PS-A, the S-ATT shows 31!!! Even if the ATT on TX is set to Zero, disabling the PS-A changes it back to 31!!! Pure signal is not showing as working, of course, under Liniearity / Amp View....
PS-A Disabled.jpg
PS-A is disabled and does allows 31 ATT
PS-A Disabled.jpg (152.89 KiB) Viewed 11654 times


I do not think that, when PS-A is enabled, as setting of zero is helping with the signal.

This not use to be like this.
Rafael / NN3RP
Nation's Capital USA
User avatar
w-u-2-o
Posts: 5525
Joined: Fri Mar 10, 2017 1:47 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w-u-2-o » Tue Nov 02, 2021 9:28 pm

These attenuator issues do not seem like a bug at this time. People are not reporting a lot of problems in this area, and these issues can't be duplicated here.

When PureSignal is active (PS-A button illuminated), the PureSignal Auto Attenuate feature is also active by default. Auto Attenuate will set the step attenuator in accordance with the feedback level it sees either from the internal or external coupler. Coupler selection is manual via rear panel jumper on 8000's, or via software selection on other models that support PureSignal. Improper coupler input selection will result in bad attenuator settings. Improper external coupler connections will also result in bad attenuator settings. Broken or intermittent coupler or cable connections will also result in bad attenuator settings.

Auto Attenuate will only kick-off a reading if ALC is at 0dB. If audio or other adjustments are not driving to 0dB ALC then the step attenuator will not be set properly (and PureSignal corrections will also not be updated). When using speech as the source of audio, this can sometimes be problematic. For testing this can be overcome by using the two tone output feature with the "Use Drive Power" option enabled.

Finally, when using an external coupler, if proper feedback levels are not obtained as described here, this can also cause the step attenuator to be driven to either 0dB or 31dB depending on if the feedback levels are too high or too low.

Internal feedback levels are unambiguously safe at all step attenuator values. External feedback levels should never exceed +13dBm in order to be unambiguously safe at all step attenuator values.

If for some reason PureSignal operation is desired but without Auto Attenuate functionality, this can be disabled in Linearity > Advanced and the TX atten value set manually.

If you feel that there is some problem with your PureSignal setup or operation, please post in the PureSignal section of the forum.
NN3RP
Posts: 65
Joined: Sun Feb 25, 2018 1:53 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby NN3RP » Tue Nov 02, 2021 10:54 pm

Thanks, Scott.

After reading your post, and knowing that I don't have any coupler / cable / ALC settings in the wrong direction, I added a new default TX profile to my arsenal. While using this added one everything is working as expected.

So I conclude that my TX profile had some corruption somewhere that is not easy to fix....

Thanks to all you amigos who make this software better every day....now back to my spinning wheel.

73s
Rafael / NN3RP
Nation's Capital USA
User avatar
ramdor
Posts: 1468
Joined: Wed Jul 03, 2019 3:07 pm
Contact:

Re: Thetis v2.8.11 - UPDATES and current issues

Postby ramdor » Thu Nov 04, 2021 3:44 pm

Just an update. Slow going at the moment, as free time is minimal. I am currently looking at some CTUN issues when using split. It seems like much of that has been broken or not working as it should for years.

Hopefully get something out next week with that stuff resolved and a number of other fixes.

Richie.
Aivars
Posts: 77
Joined: Sun Apr 09, 2017 11:12 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby Aivars » Mon Nov 08, 2021 8:27 pm

One my friend ask ;) He complain, split frequency doesn't store in transverter mode. I hope I understand him in right way.
Attachments
photo_2021-11-08_22-18-38.jpg
photo_2021-11-08_22-18-38.jpg (100.46 KiB) Viewed 11253 times
YL2GVC 73!
Joe
Posts: 190
Joined: Tue Aug 22, 2017 5:58 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby Joe » Sat Nov 13, 2021 2:18 pm

Ok, chomping at the bit here! Any news on Thetis updates? I know Richie has been very busy but sure do miss his work with the others that put work in also and it is really appreciated. Also, Is there any update on when there will be a complete update with full source code and install from GitHub, I enjoy working and learning the source when I can. Thanks and 73’s.

Thanks,
Joe
WD5Y
Joe
Posts: 190
Joined: Tue Aug 22, 2017 5:58 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby Joe » Wed Nov 24, 2021 6:14 pm

Just checking, any news on updates in the works and source code released? It’s been some time since information on this.

Thanks,
Joe
WD5Y
User avatar
w-u-2-o
Posts: 5525
Joined: Fri Mar 10, 2017 1:47 pm

Re: Thetis v2.8.11 - UPDATES and current issues

Postby w-u-2-o » Wed Nov 24, 2021 9:57 pm

Joe,

All dev's, and Richie in particular, are currently on a coding hiatus. With the holiday season upon us, it would not be unreasonable to expect this to continue until after the new year.

If there is any activity you will certainly see posts about it, so inquiries of this nature are not necessary.

73,

Scott

Return to “Thetis”