Re: HiMD Clipping Test results
Re: HiMD Clipping Test results
"Rob Danielson" danielson_rob
Thu Jul 13, 2006 11:21 pm (PDT)
At 12:10 PM +0800 7/14/06, Graham Evans wrote:
>Here are the tests requested Rob. imho they demonstrate that:
>the effect IS clipping.
>the effect is NOT related to PiP.
Hi Graham-- Confirmation enough for me that using Low Sens setting
with ECM lavs does eliminate or significantly reduce the clipping
encountered with moderately loud sources and the High Sens setting.
That the clipping is still present with the 9 volt Sound Pro battery
box is curious, but I'm not sure we can entirely rule out PIP, yet.
(Gosh, it seems like Sound Pro would be aware of this phenom,...)
That I don't encounter clipping with condenser mics on a portable
phantom supply using High Sens has got me puzzled. The spec sheet on
the Asahi Kasei AK5356 ADC/Mic Pre that Richard pointed-out could
provide some clues. Many thanks for your useful discovery; I'm a low
sens believer. Very useful to add Tr4 to the test. Rob D.
>I would appreciate hearing any further conclusions which anyone can draw
>from these results.
>If I didn't have advice otherwise I would assume that the low-sens
>setting introduced a pad. Since this is not the case I am inclined to
>think the clipping occurs at an A/D conversion stage, or perhaps in a
>digital pre-amp stage.
>Enough of my guesswork. Here it is:
>The files are around 480KB each. There are 6 of them. Find them at the
>"Nature Recordist File Depot" here:
>sound - pass
>The Microphone used throughout the tests was a dual-mono wired Omni
>pattern Audio Technica: AT943-SP (sold by Sound Professionals as
>"PREMIUM MINIATURE LAPEL MICROPHONE"). I believe this mic is also
>referred to as the ES943.
>All recordings for these tests were made in HiSP mode with the WAV
>conversion done by SS3.4 automatically on import. That is my current
>workflow so I just continued it. Not ideal I admit but plenty of
>fidelity to make useful results.
>As noted on previous email I noticed the clipping effect when people get
>a bit excited talking. For these tests it was easier and less
>embarrassing to reproduce the effect by singing. I have not normalised
>or manipulated the recordings just cropped for length.
>1. First I created the base recording on the HiMD in by singing into
>the mic at fairly close range. I used "Low-Sens" and manual
>mic-recording level of 15 to get a result with no clipping but still
>using the maximum headroom. tr1.wav
>2. Next I played the result back on the speaker trying to acheive a
>roughly "human" level of amplification. I did this by using the same
>mic distance as for the control recording, "Low-Sens" on the HiMD and a
>manual mic-recording level of 15 once again. I then adjusted the volume
>on the speakers to produce roughly the same levels on the meters as for
>the first recording. Notice a slight speaker rattle from my cheap
>speakers on one of the syllables. You may also hear my computer fan in
>the background too. tr2.wav
>3. Leaving the speaker volumes as set I then switched the recording
>setting to high sens and brought the gain down to 10. This was maximum
>use of headroom without causing clipping due to an excessive
>mic-recording level. It does however demonstrate the clipping effect
>discussed. The next result is further confirmation of that. tr3.wav
>4. Again leaving the speaker volumes as set I recorded well below
>maximum headroom to show how the clipping result is duplicated
>irrespective or mic-recording level (which was 8 in this instance). tr4.w=
>5. Exactly as for 4 but routing the microphone signal through a PiP 9V
>battery box (this is a Sound Professionals model which includes optional
>Bass Roll Off Settings - not used in this instance). The PiP box seems
>to have marginally attenuated the signal but has not effected the
>6. Finally, to show the negative case, I backed off the speaker volume
>and left the settings on the HiMD as per #4. I removed the PiP battery
>I think test 6 was not really neccessary.
The University of NSW School of Computer and Engineering
takes no responsibility for the contents of this archive. It is purely
a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way.
If you wish to get material removed from the archive or
have other queries about the archive e-mail
Andrew Taylor at this address: