Loading ...
Sorry, an error occurred while loading the content.

X-Fi SB-1095 input dynamic range ...

Expand Messages
  • wa5bdu
    I bought a Sound Blast SB-1095 external USB sound card a couple years ago after reading some favorable comments from some Softrock users. But I abandoned it
    Message 1 of 8 , Aug 7, 2014
    • 0 Attachment

      I bought a Sound Blast SB-1095 external USB sound card a couple years ago after reading some favorable comments from some Softrock users. But I abandoned it after having a lot of problems both with radio applications and trying to rip vinyl and record with microphones.


      The problem was a lack of gain in the input channels, both line and microphone.  Even with a good line level input and all sliders to maximum gain, the recording level would not approach 0 dB, actually falling short by 5 to 10 dB, IIRC  This resulted in music tracks that were way down in level and a diminished dynamic range with SDR software used with the Softrock.


      So a couple years later I'm researching sound card upgrades again and decided to give the SB-1095 another go.  Bought a new unit from Amazon.  So imagine my surprise when it acts exactly the same way!


      So I'm asking -- Am I doing something wrong?  Is this normal for this sound card?  Should I build an amplifier to go in front of the Line In jack?


      BTW, I've tried it with different computers, different recording and displaying software, different sources providing line in, etc.


      73-


      Nick, WA5BDU

    • warrenallgyer
      Hi Nick I just ran some tests on my SB-1095 and I can concur with your observations. On the Line input, with: Both input and output sliders at 100 Playback
      Message 2 of 8 , Aug 7, 2014
      • 0 Attachment
        Hi Nick

        I just ran some tests on my SB-1095 and I can concur with your observations.

        On the Line input, with:
        Both input and output sliders at 100
        "Playback" set to "Listen" to the Line input
        With 0 dBV in
         I get -12 dBV out. A note of caution here: I observed clipping only 2 dB higher, beginning at +2 dBV in on the line input.

        On the Mic input under the same conditions, I could get 0 dBV out with an input of -13 dBV. Again, however, the clip point was only 1 dB higher, at -12 dBV in.

        All measurements made with an Amber 3500 audio distortion analyzer.

        Warren Allgyer
        9V1TD
      • wa5bdu
        Thanks for making those measurements Warren. I attempted a few measurements and/or observations myself although I don t have your measurement instruments.
        Message 3 of 8 , Aug 8, 2014
        • 0 Attachment
          Thanks for making those measurements Warren.

          I attempted a few measurements and/or observations myself although I don't have your measurement instruments.  

          Using a signal generator and monitoring voltage with my scope and monitoring the recording level with the Audacity program, I see that at 1.1 Vrms, I'm right at the verge of clipping.  The bargraph meter also shows 0 dBm here so they match.  Recording sliders are full ON.

          I have a homemade test CD with a 800 Hz sinewave track, which I believe I made at 100% level ... I connected the line out of my Phillips CD player to line in on the SB 1095 and it shows about 6 dB down from full level and the graph of the plot shows the same with the peak voltage 0.5 of the maximum available span.

          Going a step further, I put in a commercial CD, dense rock song from the Rolling Stones and the line out produces an average value of about 0.3 of the maximum span with peaks hitting about 0.5.  In comparison with a typical pre-recorded WAV file, one would see about twice the voltage or more and 6 dB more average power.

          My point in all this is this:  Every home recording device I've used in the past 40+ years (reel to reel, cassette, CD, digital) has had sufficient gain on the inputs to allow the user to decide how closely to approach clipping and if in fact to allow peaks to clip on occasion, or even overload grossly.  The SB 1095 seems to come up well short of this.  I've seen this in Rocky with this card as well.  A shortage of dynamic range.

          I think my observations are a bit different from yours in that I'm just looking at the recording portion of the card and you are looking at the total channel from input to output.  Your notes of clipping may be from the output channels, which might have sufficient gain range to cause clipping when set at maximum.

          73-

          Nick, WA5BDU


          ---In softrock40@yahoogroups.com, <allgyer@...> wrote :

          Hi Nick

          I just ran some tests on my SB-1095 and I can concur with your observations.

          On the Line input, with:
          Both input and output sliders at 100
          "Playback" set to "Listen" to the Line input
          With 0 dBV in
           I get -12 dBV out. A note of caution here: I observed clipping only 2 dB higher, beginning at +2 dBV in on the line input.

          On the Mic input under the same conditions, I could get 0 dBV out with an input of -13 dBV. Again, however, the clip point was only 1 dB higher, at -12 dBV in.

          All measurements made with an Amber 3500 audio distortion analyzer.

          Warren Allgyer
          9V1TD
        • Alan
          Original Message ----- Subject: [softrock40] Re: X-Fi SB-1095 input dynamic range ... Nick, As usual I m confused. I m not too sure what tests you are doing or
          Message 4 of 8 , Aug 8, 2014
          • 0 Attachment
            Original Message -----
            Subject: [softrock40] Re: X-Fi SB-1095 input dynamic range ...


            Nick,

            As usual I'm confused.
            I'm not too sure what tests you are doing or what the results show.
            Let's talk SDR.
            Does it work on receive? Is there any lack of sensitivity?
            And TX? What power can you get out?
            Surely this device is used for SDR by quite a few? I do not remember any complaints.

            73 Alan G4ZFQ

            > Thanks for making those measurements Warren.
            >
            > I attempted a few measurements and/or observations myself although I don't have your measurement instruments.
            >
            >
            > Using a signal generator and monitoring voltage with my scope and monitoring the recording level with the Audacity program, I see
            > that at 1.1 Vrms, I'm right at the verge of clipping. The bargraph meter also shows 0 dBm here so they match. Recording sliders
            > are full ON.
            >
            >
            > I have a homemade test CD with a 800 Hz sinewave track, which I believe I made at 100% level ... I connected the line out of my
            > Phillips CD player to line in on the SB 1095 and it shows about 6 dB down from full level and the graph of the plot shows the same
            > with the peak voltage 0.5 of the maximum available span.
            >
            >
            > Going a step further, I put in a commercial CD, dense rock song from the Rolling Stones and the line out produces an average value
            > of about 0.3 of the maximum span with peaks hitting about 0.5. In comparison with a typical pre-recorded WAV file, one would see
            > about twice the voltage or more and 6 dB more average power.
            >
            >
            > My point in all this is this: Every home recording device I've used in the past 40+ years (reel to reel, cassette, CD, digital)
            > has had sufficient gain on the inputs to allow the user to decide how closely to approach clipping and if in fact to allow peaks
            > to clip on occasion, or even overload grossly. The SB 1095 seems to come up well short of this. I've seen this in Rocky with
            > this card as well. A shortage of dynamic range.
            >
            >
            > I think my observations are a bit different from yours in that I'm just looking at the recording portion of the card and you are
            > looking at the total channel from input to output. Your notes of clipping may be from the output channels, which might have
            > sufficient gain range to cause clipping when set at maximum.
            >
          • wa5bdu
            Surely this device is used for SDR by quite a few? I do not remember any complaints. Yes, that was my question to myself also. I m fairly sure the sound
            Message 5 of 8 , Aug 8, 2014
            • 0 Attachment
              "Surely this device is used for SDR by quite a few? I do not remember any complaints."

              Yes, that was my question to myself also.  I'm fairly sure the sound card doesn't have enough gain on its inputs to give good sound recording from typical sources.  I'm not sure how this affects DSP radios.

              It seemed to have a negative effect a couple years ago when I tried out my first SB 1095.  That is, not enough sensitivity or gain:  "It takes a strong signal to show up".  Now I have a new PC and I'm two years smarter.  ;^)

              So now I've dug out my only current DSP device, a Softrock Lite II and installed Rocky on my new PC.  The Softrock is configured to listen to the K3's IF output at 8.215 MHz. (Actual crystal is 4 x 8.191 MHz)  Not a lot on the bands today but I did some listening and I do hear quite a few signals that sound OK, FWIW.

              To get more quantitative, I used my Elecraft XG3 signal generator, connected directly into the Softrock card and set to 8.202 MHz.  With the XG3 set to 50 uV output which is standard S9, the signal is good and solid on Rocky's display.  It shows a strength of 55 dB.  Sound about right?  When I reduce the signal strength to 1 uV, I see the line on the spectrum display reaching 20 dB and the signal is barely audible, but it's there.

              Sounds like it's doing OK, eh?  I'm not attempting to transmit, BTW.

              73-

              Nick, WA5BDU
            • warrenallgyer
              Nick The X-FI is more than adequate for SDR use. It s gain and dynamic range exceed the usable values for any HF operations. My numbers of yesterday were off
              Message 6 of 8 , Aug 8, 2014
              • 0 Attachment
                Nick

                The X-FI is more than adequate for SDR use. It's gain and dynamic range exceed the usable values for any HF operations.

                My numbers of yesterday were off by 6 dB due to the use of a mono summing cable in the tests. Using a stereo cable and an internal tone generator application set to output 0 dBV, with the slider at 100%, I measure 0 dBV out. My internal generator cannot go higher than 0 so I was not able to determine the clipping point of the output.

                With the same output settings, and input of a 0 dBV test tone, input slider also at 100%, the output measures -8 dBV. Increasing the input level to +8 outputs 0. Both are very low distortion, less than 0.05%. Increasing the input level to +9 incurs clipping and a dramatic increase in distortion. So you are correct, the X-FI attenuates the input signal by 8-10 dB.

                The noise floor measures at -96 dBV which, with the maximum output level of +8, gives a theoretically impossible dynamic range of 104 dB for a 16 bit device. I suspect what is happening is the DAC is disabled with no output and the true useable dynamic range is in the 85-90 dB range, still excellent for anything short of professional studio recording.

                As a former broadcast television chief engineer I would never want to record produced material with less than 10 dB of headroom. If I am recording a live microphone I want that increased to 20 dB. In this regard it seems to me the X-FI got it right. It provides 8-10 dB of headroom for most input devices and this prevents over-driving and distortion. It has a low enough noise floor so that gain can be added on playback without objectionable noise.

                I cannot conceive of a situation where it would be a good thing to drive an input device into clipping. You can certainly add more input gain but my ears wince at the thought of recording anything other than a reference tone with only 1 dB of headroom.

                Warren Allgyer
                9V1TD
              • wa5bdu
                Well, that might be OK if you are assured of a constant input level you are wanting to record. Just set all sliders to 100% and take a break. ;^) In the real
                Message 7 of 8 , Aug 12, 2014
                • 0 Attachment
                  Well, that might be OK if you are assured of a constant input level you are wanting to record.  Just set all sliders to 100% and take a break.  ;^)
                  In the real world Wild West of consumer devices and field recordings, levels can be all over the place.  And I want an adjustment range that will accommodate them all.
                  For example, my Audio Technica turntable says it has an output of 200 mV but my Pioneer cassette deck's output is 500 mV.  That's 8 dB of difference. And the Pioneer says it has an input sensitivity of 100 mV, which would probably be plenty of gain for the levels I'm seeing.  Not so with the SB 1095.
                  So in the past couple of days I've built a little amplifier with a modest voltage gain of 3 (9.5 dB) and now I have all the headroom I need on the SB 1095.  I don't know that I need it for SDR usage, but I think it has the bandwidth required if I decide to try it.
                  73-
                  Nick, WA5BDU
                • warrenallgyer
                  Nick The levels you are citing for various components all represent the maximum level that component can produce or accept. The equipment is not designed to
                  Message 8 of 8 , Aug 12, 2014
                  • 0 Attachment
                    Nick

                    The levels you are citing for various components all represent the maximum level that component can produce or accept. The equipment is not designed to operate at these levels even though it can.

                    Most consumer equipment is designed to operate at a nominal level of -10 dBV. At this level the SB 1095 has 10 dB of headroom and that is appropriate for consumer audio applications.

                    Properly set up, a consumer audio recording system would accept an input with peak levels of - 10 dBV, record them, and when they are played back, play them at -10 dBV peaks. If you were to configure the system of the components you mentioned to do this then none of them would clip or distort and you would have made optimal use of your system dynamic range.

                    The trick is doing this with a wide range of program material. If you have low level program material and you adjust the input gain of your recorder to compensate then you have just destroyed the unity gain of the system.

                    This is why professional audio facilities include a mixing/monitoring console that allows them to monitor and control the levels inside the system, maintaining component headroom and dynamic range, while leaving the system gain at unity.

                    Apologies to the board for taking this thread off topic. Nick, I would be happy to discuss this further with you privately.

                    Best regards,
                    Warren Allgyer
                    9V1TD
                  Your message has been successfully submitted and would be delivered to recipients shortly.