K3 Noise Reduction
Locked
12
12
 
|
Is the a DB gain setting somewhere to compensate for loss of sensitivity when in the Noise Reduction (NR) mode? There was on the K2, and I can not seem to find reference to such for the K3.
TNX, Fred kt5x K3 # 0144 _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
I see it the same as Dave. However, I've about given up on using NR.
Strange thing! When I installed Firmware 1.99, it seemed to have improved! Audio level was equalized between NR on/off. And, it seemed less distorted. Then over the next few weeks, it gradually went back to (apparently) more distortion and definitely much less volume. Does newest firmware improve it? Windy KM5Q K3 #764 > Yes, I'd hope for further improvement of the NR on phone. Even if > it's not too effective I still nearly always have it on, but the radio > powers up with it off and blasts me with the increased audio. It > seems like it would be nice if it powered up with the NR in the same > state it was before. Is this a problem for anyone else? > > Dave W5DHM _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
I've about given up too, Windy. The NR must be only for the CW ops.
I'm about ready to order one of the speakers with NR and see if it's any better. If it is, I might as well be using it with my Drake instead. What we have now seems about as effective as the NR on the low end Icoms that I've had, which is minimal. That's the only other NR I've used. Are any of the other radios any better? Maybe some owners of the competing brands could comment. I think that one of the happier things about the K3 is the continuing hope of greater performance and features. Nothin' like hope. It keeps us going. Dave W5DHM > I see it the same as Dave. However, I've about given up on using NR. Strange > thing! When I installed Firmware 1.99, it seemed to have improved! Audio > level was equalized between NR on/off. And, it seemed less distorted. Then > over the next few weeks, it gradually went back to (apparently) more > distortion and definitely much less volume. > > Does newest firmware improve it? > > Windy KM5Q > K3 #764 > >> Yes, I'd hope for further improvement of the NR on phone. Even if >> it's not too effective I still nearly always have it on, but the radio >> powers up with it off and blasts me with the increased audio. It >> seems like it would be nice if it powered up with the NR in the same >> state it was before. Is this a problem for anyone else? >> >> Dave W5DHM Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by KM5Q
Windy,
The amount of audio reduction depends on how much noise reduction you have cranked into the K3. It is not fixed in the latest firmware, but the situation has been identified. The method used to correct it has not yet been settled, but it will be corrected - perhaps by allowing the user to set the amount of gain increase when NR is active. I find the K3 NR is quite effective, but the audio level change is disconcerting. This is not a problem when the NR is used on CW, it is a SSB/AM problem only. 73, Don W3FPR KM5Q wrote: > I see it the same as Dave. However, I've about given up on using NR. > Strange thing! When I installed Firmware 1.99, it seemed to have > improved! Audio level was equalized between NR on/off. And, it seemed > less distorted. Then over the next few weeks, it gradually went back > to (apparently) more distortion and definitely much less volume. > > Does newest firmware improve it? > > Windy KM5Q > K3 #764 > >> Yes, I'd hope for further improvement of the NR on phone. Even if >> it's not too effective I still nearly always have it on, but the radio >> powers up with it off and blasts me with the increased audio. It >> seems like it would be nice if it powered up with the NR in the same >> state it was before. Is this a problem for anyone else? >> >> Dave W5DHM > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Cloud Runner-3
But the strange thing is that just after 1.99 installation, it really
improved. It had been fairly good (for SSB), then it got really good, with volume equalized between on/off. Since then, I have NOT updated firmware, but it deteriorated over a period of weeks to the point where it does practically nothing now except introduce more or less distortion (normal DSP distortion) according to the 1-4 etc. setting that I use. I have not updated firmware since 1.99 but I will try that shortly. I wonder if anyone else has seen the NR performance deteriorate over time with NO change to the firmware. It's like adaptive in reverse. Windy KM5Q _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Cloud Runner-3
Continuing from my previous message -- I just loaded the current
Firmware 2.02. I think NR works better now, but I'm not really sure. In the process however, I found that I was being distracted by the more "powerful" ranges of 3-x and 4-x. For SSB, the best setting seems to be "1-3". Before, I didn't think it was so effective. The 1 range maintains good volume equalizing between on/off. Higher ranges cause too much distortion. Maybe they are OK on CW, but any advantage seems to be not worth the time to continue experimenting. For now, I'm going to use "1-3" and forget about the higher ranges of NR. A good improvement would be if the NR processing could reduce automatically with signal strength, to avoid unnecessary distortion on strong signals. I trust we'll see some further improvement eventually but meanwhile, I can enjoy using it on low signal/noise ratio conditions and switching it off on stronger signals. Windy KM5Q _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Dave Martin-3
Dave Martin wrote:
> Are any of the other radios any better? Maybe some > owners of the competing brands could comment. I've used a few NR systems (not a lot, so this is just anecdotal) including the system that Bob Heil was marketing for a while (ClearSpeech speaker). The latter probably worked better than anything else I've tried, but I can say categorically that I have never seen any NR system that actually makes an SSB signal easier to copy. In other words, if you can't copy it through the noise, you're not going to be able to copy it any better with NR applied to it. My current radio (soon to be replaced by the K3) is a Kenwood TS-2000, and the NR on it is absolutely useless on SSB. In fact, it's pretty useless on CW, too, IMHO. I never use it. NR is very far from being a mature technology! I think we'll see DSP technology take huge leaps forward over the next few years, and since the DSP in the K3 is wide open to change, I'm looking forward to continuing improvement. That said, I haven't yet experienced the K3 NR. I should have that opportunity within the next couple of days. :-) Bill W5WVO _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by KM5Q
Hi Windy,
I wonder if adjusting the AGC SLP would have an impact on NR processing? I have AGC SLP at 12 and when I need NT it is set for F1-2. Maybe a higher AGC SLP would cause less distortion with stronger signals? 73, N2TK, Tony -----Original Message----- From: [hidden email] [mailto:[hidden email]] On Behalf Of KM5Q Sent: Thursday, June 19, 2008 12:47 AM To: [hidden email] Subject: Re: [Elecraft] K3 noise reduction Continuing from my previous message -- I just loaded the current Firmware 2.02. I think NR works better now, but I'm not really sure. In the process however, I found that I was being distracted by the more "powerful" ranges of 3-x and 4-x. For SSB, the best setting seems to be "1-3". Before, I didn't think it was so effective. The 1 range maintains good volume equalizing between on/off. Higher ranges cause too much distortion. Maybe they are OK on CW, but any advantage seems to be not worth the time to continue experimenting. For now, I'm going to use "1-3" and forget about the higher ranges of NR. A good improvement would be if the NR processing could reduce automatically with signal strength, to avoid unnecessary distortion on strong signals. I trust we'll see some further improvement eventually but meanwhile, I can enjoy using it on low signal/noise ratio conditions and switching it off on stronger signals. Windy KM5Q _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Cloud Runner-3
Tony, I have AGC SLP set somewhere around 12 like you. That isn't the
problem. Distortion is the inevitable effect of DSP noise reduction. The higher the S/N ratio, the less NR is needed, so I end up turning NR on and off manually according to S/N conditions (on SSB). Someone else was asking for NR comparisons with other radios. I had a chance to compare K3 to a SGC-2020 ADSP2 (the late, advanced edition, highly regarded) -- NR is about equally effective, as is overall weak signal performance on SSB. However, the K3 has much more pleasant audio, with or without NR. I also have a ClearSpeach DSP speaker (purchased in '99), which reduces any noise that is steady-state, like background hiss. It's good, but K3 NR is substantially more effective. I still believe that there is room to improve the NR effect on steady background hiss. Nobody has commented on my description of my NR seeming to get worse over a few weeks (while using the same firmware). I've begun to doubt my own sensibilities. Windy KM5Q K3 #764 > Hi Windy, > I wonder if adjusting the AGC SLP would have an impact on NR > processing? I > have AGC SLP at 12 and when I need NT it is set for F1-2. Maybe a > higher AGC > SLP would cause less distortion with stronger signals? > > 73, > N2TK, Tony _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Dave Martin-3
I'd recommend a read of Lyle's description of NR in the FAQ and also in 'Operating Tips'. From this it's clear this sort of adaptive filtering is going to be most effective for CW.
I've been using it with F2-3 setting and it works fine for me (on CW). On signals other than very weak ones (which Lyle points out could be adversely affected by NR), the audio actually appears to increase a notch if anything. There is a small amount of distortion on the audio with the most noticeable being a softening of the cw characters which will make copying slightly more problematic in difficult conditions. Having said all that I still use filtering first and foremost before falling back on things like NR; and to comment on the suggestion that NR should remain on after switch off, I remember ages ago Eric or Wayne saying they specifically designed the NR *not* to remain on as folk were contacting them complaining their receiver had died when they switched their K3 back on and not hearing any background hiss - the NR is that good! Anyway it's only one button press to turn it back on... 73, Stewart Rolfe, GW0ETF (K3/100 145)
|
|
|
On Friday 20 June 2008 09:00:24 GW0ETF wrote:
> I'd recommend a read of Lyle's description of NR in the FAQ and also in > 'Operating Tips'. From this it's clear this sort of adaptive filtering is > going to be most effective for CW. > > I've been using it with F2-3 setting and it works fine for me (on CW). On > signals other than very weak ones (which Lyle points out could be adversely > affected by NR), the audio actually appears to increase a notch if > anything. There is a small amount of distortion on the audio with the most > noticeable being a softening of the cw characters which will make copying > slightly more problematic in difficult conditions. > > Having said all that I still use filtering first and foremost before > falling back on things like NR; and to comment on the suggestion that NR > should remain on after switch off, I remember ages ago Eric or Wayne saying > they specifically designed the NR *not* to remain on as folk were > contacting them complaining their receiver had died when they switched > their K3 back on and not hearing any background hiss - the NR is that good! > Anyway it's only one button press to turn it back on... > > 73, Stewart Rolfe, GW0ETF (K3/100 145) So I cannot understand why when I tap NR in CW mode I get the message N/A. I used to be able to use NR in CW mode. Charles - M0BIN _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
> So I cannot understand why when I tap NR in CW mode I get the message N/A.
> > I used to be able to use NR in CW mode. AGC must be enabled for NR to be available. 73, Lyle KK7P _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Cloud Runner-3
> Quick question, how about in data mode, there seems to be no nr avaiable then, is this normal ?
Yes. In data mode, the intent is to have as pure a signal as possible, so AN and NR are not available. RxEQ should be set to flat, and if using DATA A mode then TxEQ should also be set to flat. 73, Lyle KK7P _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by KK7P
On Friday 20 June 2008 18:19:10 Lyle Johnson wrote:
> AGC must be enabled for NR to be available. Thank you for the very quick reply Lyle. Someone must have switched off the AGC at the Radio Club and I hadn't noticed ! Been using the K3 for five weeks now and I'm still learning how to drive it properly :-) Perhaps the requirement for AGC to be on could be included in the NR paragraph in the FAQ ? (Belt and Braces) Many thanks and 73 from Charles - M0BIN _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by KK7P
Would it not be a good idea to not allow TxEQ and RxEQ in Data A mode?
Maybe it could be allowed by changing a setting in config if it were desired to use the equalisers for data but by default it would not. 73 Stephen G4SJP On 20/06/2008 18:31, "Lyle Johnson" <[hidden email]> wrote: >> Quick question, how about in data mode, there seems to be no nr avaiable >> then, is this normal ? > > Yes. > > In data mode, the intent is to have as pure a signal as possible, so AN > and NR are not available. RxEQ should be set to flat, and if using DATA > A mode then TxEQ should also be set to flat. > > 73, > > Lyle KK7P > > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
> Would it not be a good idea to not allow TxEQ and RxEQ in Data A mode?
It's on the list, but not high on the list. > Maybe it could be allowed by changing a setting in config if it were desired > to use the equalisers for data but by default it would not. My personal bias would be to always disallow it for DATA modes. Choices are good, but sometimes decisions are required :-) 73, Lyle KK7P _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Stephen Prior
Surely there is need for TxEQ - for example with WSJT mode to
compensate different levels from souncard and possible filter influence for all tones to set equal power-out. I do not remember if TxEQ settings is stored via mode or not. If not it should be other "item" on K3 wish-list on http://www.zerobeat.net/smf/index.php/board,25.0.html RxEQ per mode is there listed, TxEQ per mode maybe not (yet). 73! L. -dst- K3/10 #727 Stephen Prior napsal(a): > Would it not be a good idea to not allow TxEQ and RxEQ in Data A mode? > > Maybe it could be allowed by changing a setting in config if it were desired > to use the equalisers for data but by default it would not. > > 73 Stephen G4SJP > > > On 20/06/2008 18:31, "Lyle Johnson" <[hidden email]> wrote: > >>> Quick question, how about in data mode, there seems to be no nr avaiable >>> then, is this normal ? >> Yes. >> >> In data mode, the intent is to have as pure a signal as possible, so AN >> and NR are not available. RxEQ should be set to flat, and if using DATA >> A mode then TxEQ should also be set to flat. >> >> 73, >> >> Lyle KK7P >> >> _______________________________________________ >> Elecraft mailing list >> Post to: [hidden email] >> You must be a subscriber to post to the list. >> Subscriber Info (Addr. Change, sub, unsub etc.): >> http://mailman.qth.net/mailman/listinfo/elecraft >> >> Help: http://mailman.qth.net/subscribers.htm >> Elecraft web page: http://www.elecraft.com >> > > > > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
> Surely there is need for TxEQ - for example with WSJT mode to compensate
> different levels from souncard and possible filter influence for all > tones to set equal power-out. The TxEQ adjustment frequencies are set for voice operation. They are the same setpoints as used in the W2IHY 8-band EQ. They are probably not in the right places to compensate for whatever ripple might be in the subsequent crystal filters in the Tx Path. Also, such compensation would change with USB/LSB operation, and is not easily observable by the operator of the Tx, so might not be straightforward to set properly. Any mode that is designed for robust, or weak signal, HF data operation should have allowance for slight amplitude and phase variations, since that is a normal part of HF signal propagation. > I do not remember if TxEQ settings is stored via mode or not. They are not, nor are Rx EQ settings, at present. 73, Lyle KK7P _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
The best way is to try and check power levels on all tones used in
WSJT mode. In my case the playing with TxEQ help to compensate lover signal level from SoundCard at low tones. As such signals are driving mostly some VHF transverter system with some QRO PA which is during 1min period in 100% duty mode (like in FM mode), there is need to carefully check power level going out (drive level) to prevent some bad things :-( (mostly power shoud be reduced below nominal like in rtty or fm) As such setting is strongly dependent on soundcard type and settings there is no simple recommendation. One has to know what he is doing - but this is general rule with K3 :-) Maybe TxEQ and RxEQ will be accesible by external utility and profile will be possible to save/recall. 73! L. -dst- K3/10 #727 Lyle Johnson napsal(a): >> Surely there is need for TxEQ - for example with WSJT mode to >> compensate different levels from souncard and possible filter >> influence for all tones to set equal power-out. > > The TxEQ adjustment frequencies are set for voice operation. They are > the same setpoints as used in the W2IHY 8-band EQ. > > They are probably not in the right places to compensate for whatever > ripple might be in the subsequent crystal filters in the Tx Path. Also, > such compensation would change with USB/LSB operation, and is not easily > observable by the operator of the Tx, so might not be straightforward to > set properly. > > Any mode that is designed for robust, or weak signal, HF data operation > should have allowance for slight amplitude and phase variations, since > that is a normal part of HF signal propagation. > >> I do not remember if TxEQ settings is stored via mode or not. > > They are not, nor are Rx EQ settings, at present. > > 73, > > Lyle KK7P > > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Cloud Runner-3
Lyle,
You mention setting Rx equalization to flat for data modes. I set up RxEQ to make the waterfall uniform, flattening the slight residual spectrum colorization left over from roofing filter not quite being flat. It appears to work okay but now I wonder if I am helping or hurting the data decoding situation. It sure makes the waterfall look uniform. If what you say is true, it would be nice to have mode specific RxEQ and TxEQ, different for the wider band modes anyway = AM/FM/USB/LSB/Data/Data Reverse. I need a sticky note to record the different settings to flatten the background noise the way we are now. If the K3 Utility were modified to make EQ easy to change (like the Filter Configuration Utility) it would be almost as useful and save K3 memory. Mike Scott - AE6WA Tarzana, CA (DM04 / near LA) K3-100 #508/ KX1 #1311 ------------------------------------------------------------------------- In data mode, the intent is to have as pure a signal as possible, so AN and NR are not available. RxEQ should be set to flat, and if using DATA A mode then TxEQ should also be set to flat. 73, Lyle KK7P _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
«
Return to Elecraft
|
1 view|%1 views
| Free forum by Nabble | Edit this page |