I'm wondering what the consensus is on weight adjustment as one gets over 50 wpm. I notice that some ops apparently drop the dot/dash ratio to 1 to 2.5 or lighter from the standard 1 to 3 when using high speed. Perhaps its all subjective but for extremely high speeds (60-100 wpm) weight can make a BIG difference on copyability and clean keying. Any opinions on this? A corollary to this would be the big difference the keying envelope can make on ease of copy with some radios. There are some rigs, that shall remain nameless, that just dont key very well at extremely high speeds. Great if you like a challenge though!!!!
Excellent question and I just put up on the front page of our website a new TESTING POLL and 4 CW mp3 files sending just a short phrase from w4bqf's article, "QRQ or Copying CW over 70 wpm" I varied the weight and dot/dash ratio on each file slightly while sending CW at 75wpm. Let's take a survey and see which mp3 CW file everyone thinks is their choice for best sounding cw at that speed. I will give the details(which one has what weight and dot/dash ratio etc) of the files once there are enough poll results to be statistically significant. The files were made with FLdigi and AUDACITY.
Rick, I thought of an experiment that we might try and develop a consensus as you say.
Here is what I suggest to test for best settings on all the cw timing preferences at ones maximum qrq copy speed.
1. Download FLdigi NOTE: FLdigi is the only qrq cw sending program that will allow you to adjust every
parameter of CW timing from which type of RAISED COSINE waveform you prefer to incremental changes
in RISE and FALL TIME as well as weight and dash/dot ratio. At extreme qrq levels, a small change in
a CW timing option may offer a big improvement in qrq cw copy and the audible beauty of the cw note !
2. load a text file to send at ~1 to 2 wpm above your absolute max qrq cw copy speed: (RT CLICK on the blue screen and follow prompt) [NOTE: it is much easier to discern an increase in copy ability if you send just above your max while adjusting the timing parameters to see if any combination of timing options helps you copy better etc....]
3. Click on the Tx button or the T/R button and start sending the file. Then click on the TOP MENU TABS, CLICK ON CONFIGURE.. then CLICK ON MODEMS to bring up the options for FLdigi select the tabs as you see here:
then adjust the FLdigi timing settings to see which combination of weight,
EDGE SHAPE(HANNING or BLACKMAN) DASH/DOT ratio and EDGE TIMING(rise and fall time in milliseconds) sounds best to you... and let everyone that would like to participate in this experiment report their final results here in the comments section...
One of the more interesting things I have noticed is that once the CW wpm starts getting over 100 wpm, there is a percussion effect that starts to increase on top of the actual cw note. Also a low bass guitar like thumping also starts to take effect. The lower in rise/fall time you go the more predominant these 2 "extra sounds" become and will actually start to power over the actual cw tone, and you will hear more percussion and bass thumping than actual cw elemental notes.
I have found that for cw sent over 100 wpm, I prefer a much higher rise and fall time(my raised cosine edge preference is BLACKMAN waveform- between 8 to 10 ms on FLdigi)
Here is an example of the spectral purity of a file sent at 120 wpm on FLdigi at 10ms ( ~ 830 hertz )
Here is a spectrograph of 120 wpm speed with the only difference being the rise/fall time is 5 milliseconds
As you can see, there are low bass notes starting to get through ~30 & 70 hertz which are very audible using modern headphones... Also getting through are various lower freq "percussive frequencies" that are also very audible...There is a balance between the percussive effects, bass thumping effects, rise/fall time and cw pitch that come together and make adjusting all these cw timing parameters well worth the effort for best cw purity, sound, clarity and best all around settings for optimum, maximum copy and comprehension.
Here is a link to a short slide show that demonstrates the SPECTRUM difference of varying the rise/fall time
from 4 ms to 10 ms while sending cw at 120 wpm. As you will observe, there is a continued improvement in the reduction of low frequency components and edge noise as you go from 4 ms up to 10 ms on Fldigi.
RESULTS ARE IN: 18 VOTES TOTAL
FILE 1 = 50 % WEIGHT 2.8 to 1 dash/dot ratio 5.5ms rise/fall time
FILE 2 = 50 % WEIGHT 3.2 to 1 dash/dot ratio 5.5ms rise/fall time
FILE 3 = 38 % WEIGHT 3.0 to 1 dash/dot ratio 5.5ms rise/fall time
FILE 4 = 50 % WEIGHT 3.0 to 1 dash/dot ratio 5.5ms rise/fall time
- Dash/dot ratio of 3.2 to 1 , ties with default dash/dot ratio of 3 to 1 -
- all files received a significant amount of votes -