Sunday, May 19, 2013

A nice audio report

I just finished a contact with a very loud Austrian station, OE3DIA on 10 metres, who took time out while working a string of stations to give me a complimentary audio report, quite unsolicited. It's good when that happens! The comment was "Very nice audio cutting through the QRM" I was using the K3 at 80 watts and the mike was one of those Heil mikes with the dual insert, set to "narrow". The K3 transmit audio equalisation is factory standard, in other words flat.

As it happens I had just been doing some audio comparisons between the KX3 and the FT-817. There has been a thread going on the KX3 Yahoo group started by a disenchanted American ham who claims that the FT-817 has punchier audio than the KX3. It's rubbish, to put it politely. The KX3 has a built-in speech compressor, while my 817 has an RF processor made by Joachim, DF4ZS (more details on my FT-817 page) built into the microphone. Without it there is just no comparison.

I recorded some audio clips so you can hear for yourself:
There is a bit of distortion on those clips which was not noticeable when listening on the radio. I think I might have a problem with my sound card.

I'm not sure if the difference are that noticeable in those clips, but when you look at the needle of the power meter the KX3 certainly has the more punchy signal.

Both the FT-817 and the KX3 were running off 13.8V and set to 5 watts output. I couldn't compare them on battery power as I don't have the charger board for the KX3 and the external battery pack (10xAA NiMH cells) I intended to use appears to be past it and the KX3 kept cutting out on voice peaks.

3 comments:

  1. Hi Julian, really, I like the FT817 mic without compression. I always get nice audio reports with this setting. 73 Paul

    ReplyDelete
  2. Nice work, Julian! I did a similar work-up on my Kenwood TS-590, if you're interested, take a look at let me know what you think:

    http://hamprojects.blogspot.com/2013/05/revised-audio-testing-of-kenwood-ts.html

    73,

    Richard, KK4JDO
    http://homprojects.blogspot.com

    ReplyDelete
  3. Hi Julian,

    I was interested in your comments WRT the FT817 vs. KX3 TX audio, and the discussions that have been raging one the KX3 Yahoo group.

    I’m one of those folks who originally flagged up the problem with the KX3 TX audio, with this post. number 17680 made on the 1st Nov 2012.

    >>>>

    Hi All,

    I've noticed a problem with the KX3 Speech processing / Compression, which I
    have feed back to Elecraft.

    My voice is very bassy, so I have to use quite a lot of equalisation to filter
    out the LF end of the TX audio and give my speech more punch.

    http://g8jnj.webs.com/speechintelligibility.htm

    I'm not 100% certain what is happening with the KX3. But I suspect that if the
    distribution of audio frequencies in my speech is altered by adjusting the
    equalisation. This messes up the algorithm being used for the split band
    processing. As it may not be optimised for speech which contains very little LF
    and a lot of HF components.

    The end result of adding more HF lift in conjunction with the same level of
    compression. Is that the compressor tries to boost the LF speech components back
    up. Which partially negates the additional equalisation, and also seems to
    reduce the average transmitted power.

    So for me adding equalisation to roll off the LF speech whist winding up the
    level of compression is counter productive. It seems more effective to have a
    flat frequency response and a moderate amount of compression. Rather than a
    heavily equalised frequency response and lots of compression.

    I've played with professional split band audio processing in the past, and I
    think Elecraft need to do a bit more tweeking and link equaliser settings to
    optimise each of the compressor frequency bands for optimum output power.

    <<<<<<

    This was followed up in post number 20392 in December 1012 made by Ignacy, NO9E. He compared the average output power with processing between several Elecraft rigs and suggested that the KX3 with full processing was about 6dB down on the K3 with full processing.

    Wayne, N6KR, from Elecraft has also commented in post number 28810 made on Sat May 18th.

    >>>>>

    Sorry to weigh in, here, but this isn't the first time the KX3 design team has
    heard that the radio could be punchier. We take input seriously, even anecdotal
    evidence, which can be just as valid as lab testing after multiple reports. We
    have an '817 to compare with the KX3, and we've already started looking into DSP
    enhancements to speech processing. This is the beauty of an SDR.

    All opinions and suggestions are welcome -- at least by us ;)

    Here's to civilized discourse.

    Wayne
    N6KR

    <<<<<<

    So the Elecraft design team also acknowledge that further enhancements can be made.

    I think it is very honest of Wayne to acknowledge this problem, and I have a great deal of respect for him and the company.

    Therefore I don’t see what the continuing argument is…..

    Except I do – My main gripe and that of a few others is the general ethos of the Elecraft Yahoo group. Where anyone who dares to raise an opinion contrary to that of the group (e.g. all things Elecraft are good, criticism is evil, and unbelievers must be punished) gets moderated. If you do a search in the group postings for message number 26520, 26466, 26464 amongst others you will be able to view the first few lines of text but the messages themselves have been deleted. Just check the bulk of postings for consecutive message numbers, and you will get an idea of the amount of moderation that goes on. I raised a valid point, which has now been accepted as an issue by Elecraft. But I certainly resent being moderated for following up my original point. Which I why you won’t find me posting on the KX3 Yahoo group any more – assuming they would let me :-)

    Regards,

    Martin – G8JNJ

    www.g8jnj.webs.com

    ReplyDelete