Acceptable Noise Level on XL1s? at DVinfo.net

May 7th, 2002, 01:48 PM

After recording a narrated sequence with a number of different mics, I was pulling my hair out trying to track down the obvious noise floor of everything that came from our XL1s.

In trying to pin this down, I even recorded a clip with the XL1s in record mode but NO tape in the transport and NO mic attached. It was fed directly via FireWire to our Mac for capture into Final Cut as a Voice Over.

That 30 second clip had obvious audible noise (essentially white noise). Using spectrum analyzer software, I measured the noise at -104 to -88 and it covered the full audio range from 1 to 24,000 Hz.

Is this an acceptable noise level for audio generated from an XL1s? Am I being unrealistic in assuming it should be better?

Canon has agreed to examine and repair the unit under warranty, but I don't want to go to that time and expense if there is nothing to be fixed.

Thanks.

Don Palomaki · May 7th, 2002, 04:00 PM

What was the configuration of the camcorder when you made the measurement? Power source, audio mode, audio input termination, audio settings? Any peaks in the noise profile?

What was the RMS noise level relative to peak record level?

Based on some tests I ran a while ago, you can expect the RMS noise floor to be on the order of 78 dB below peak record level for line input and on the oreder of 63 dB below peak record with mic level inputs. This is arguably better than the S/N of most field shooting environments.

If you need lower noise, suggest invest some real money in higher quality/lower noise DAT recorders, mics, and preamps.

May 7th, 2002, 04:39 PM

Thanks for the reply Don. The details I have are:

Power Source = AC Adapter
Audio Mode = 16 bit
Output Channel = L/R
Audio In 1= None
Audio In 2 = None
Mix Select = Variable

Audio 1/MIC Control – Input Select = MIC
Audio 1/MIC Control – Record Level = Auto
Audio 1/MIC Control – Level = 12 o’clock position
Audio 1/MIC Control – Balance = 12 o’clock position

Audio Termination = not sure I understand what you want but there was nothing connected to the XL1s except the AC adapter and the EVF cable.

As to the spectrum analysis, the reading I did measured 2048 points from 1 to 24000 Hz. There are no hills and valleys in sense that there are no patterns, all adjacent readings are vastly different. The single highest readings on the chart are:

1731 Hz = -88.4
5413 Hz = -88.6 (but also note that 5419 = -114
6749 Hz = -89.2
8814 Hz = -89.0

Given the numbers you sighted in your post, the noise floor I measured would seem to be quite acceptable – even better than normal.

Regarding your question about RMS noise level, I don’t know how I would go about measuring that.

Please note that I am very much a rookie when it comes to this type of audio stuff, so I appreciate you patience if I am not communicating in the correct terms or perhaps misunderstood your questions.

Thanks again for the help.

May 7th, 2002, 05:07 PM

Don,

In your post, you ask about the power source I was using for the camera during the audio tests.

When audio quality is critical, is there a preference or advantage to using either battery or AC power?

Don Palomaki · May 8th, 2002, 02:10 PM

Geneerally speaking, a battery will introduce less electrical noise on the powersupply bus than the A/C adapter, and should be a bit lower noise. Also, it reduces the chance for stray pickup by audio leads near the power lead by eliminating the power leads. Whether or not it makes a difference in your tests depends on lead dress and other factors.

With audio level in AUTO mode, the and nothing connected to the input, the preamps are at full gain and any noise present at the open circuit input is fully amplified. It is sort of a worst case for noise. Shorting the input should reduce the noise slightly, and using manual gain set at, say 12:00 should be even lower.

What are you using to measure the sound levels? I use the RMS (root-mean-square, an indication of the signal's energy level) reading from Cool Edit Pro's statitics window. Most volt meters are calibrated to read the RMS value of a sine wave (pure tone).

Also, if listening with earphones, any hiss will ususally be much more apparent than listening via speakers.

May 8th, 2002, 03:16 PM

Hi Don,

I see what you're saying about the gain with no input--that explains quite a bit. My logic of "no input therefore no noise" just doesn't hold water.

As for the measurements, I was using the Spectrum Analyzer feature in Amadeus II. It's a Mac OS X sound app that I like.

I am beginning to conclude that between your insights and those of another audio guru over at Ken Stone's site, that I am actually dealing with acceptable noise levels and simply need to use a noise gate or more precise video editing in situations where dead silence to noise is noticeable. And get rid of the headphones. As one friend to me, those will drive you crazy!

Thanks again for your most helpful responses.

May 9th, 2002, 09:31 AM

I have dealt with the same noise you are talking about and my solution was to use any external mics using line level instead of mic level. I achieved this by purchasing a SoundDevices MixPre and going out line level to the XL1s. Using line level compared to mic level is much quieter. I avoid using the MA-100 as well, I just feed XLR's to the MixPre and go out of the MixPre to the RCA jacks on the back of the XL1s. Just make sure those RCA jacks are then set to line level rather than mic level and you'll notice hardly any noise. I then keep the XL1s audio gain to about the 12 o'clock position in manual mode and use the MixPre to raise and lower volume levels not to peak over the -12db position on the XL1s.

Steve Siegel · May 10th, 2002, 09:16 PM

I have been grappling with what I think is the same noise you're talking about, Pixel. It's a low pitched hum that pervades all work. I learned that it is, or course, generated by the camcorder, is worse on a tripod or table top than when hand-held, and disappears completely if you take the microphone out of its clamp and hold it in your hand.
I will be building a little Rube Goldberg device to try to eliminate the noise. If it works, I'll post it. By the way, Lightwave Systems makes an isolator mount to do the same thing. It's $150.00.
Don, you were right about loosening the clamp reducing the noise, but it's not because the innards of the mic are less squeezed, but that the contact with the camcorder housing is lessened.

Steve Siegel

Don Palomaki · May 11th, 2002, 05:29 AM

Thanks for the report onthe screw. Folks have a tendency to want to tighten all screws. However, the mic mount screw should not be fully tightened, just tighten it enough so it holds the mic in position, but allow some wobble of the mic in the holder. Over tightening compresses the vibration damping material (rubber) and when compressed too much it no longer provides damping/isolation. About like bottoming a shock on a car.

May 7th, 2002, 01:48 PM	#1
pixel perfect Posts: n/a	Acceptable Noise Level on XL1s? After recording a narrated sequence with a number of different mics, I was pulling my hair out trying to track down the obvious noise floor of everything that came from our XL1s. In trying to pin this down, I even recorded a clip with the XL1s in record mode but NO tape in the transport and NO mic attached. It was fed directly via FireWire to our Mac for capture into Final Cut as a Voice Over. That 30 second clip had obvious audible noise (essentially white noise). Using spectrum analyzer software, I measured the noise at -104 to -88 and it covered the full audio range from 1 to 24,000 Hz. Is this an acceptable noise level for audio generated from an XL1s? Am I being unrealistic in assuming it should be better? Canon has agreed to examine and repair the unit under warranty, but I don't want to go to that time and expense if there is nothing to be fixed. Thanks.

May 7th, 2002, 05:07 PM	#4
pixel perfect Posts: n/a	Best Power Source for Audio? Don, In your post, you ask about the power source I was using for the camera during the audio tests. When audio quality is critical, is there a preference or advantage to using either battery or AC power?

May 7th, 2002, 04:00 PM	#2
Don Palomaki Wrangler Join Date: Sep 2001 Location: Northern VA Posts: 4,489	What was the configuration of the camcorder when you made the measurement? Power source, audio mode, audio input termination, audio settings? Any peaks in the noise profile? What was the RMS noise level relative to peak record level? Based on some tests I ran a while ago, you can expect the RMS noise floor to be on the order of 78 dB below peak record level for line input and on the oreder of 63 dB below peak record with mic level inputs. This is arguably better than the S/N of most field shooting environments. If you need lower noise, suggest invest some real money in higher quality/lower noise DAT recorders, mics, and preamps.

May 7th, 2002, 04:39 PM	#3
pixel perfect Posts: n/a	Thanks for the reply Don. The details I have are: Power Source = AC Adapter Audio Mode = 16 bit Output Channel = L/R Audio In 1= None Audio In 2 = None Mix Select = Variable Audio 1/MIC Control – Input Select = MIC Audio 1/MIC Control – Record Level = Auto Audio 1/MIC Control – Level = 12 o’clock position Audio 1/MIC Control – Balance = 12 o’clock position Audio Termination = not sure I understand what you want but there was nothing connected to the XL1s except the AC adapter and the EVF cable. As to the spectrum analysis, the reading I did measured 2048 points from 1 to 24000 Hz. There are no hills and valleys in sense that there are no patterns, all adjacent readings are vastly different. The single highest readings on the chart are: 1731 Hz = -88.4 5413 Hz = -88.6 (but also note that 5419 = -114 6749 Hz = -89.2 8814 Hz = -89.0 Given the numbers you sighted in your post, the noise floor I measured would seem to be quite acceptable – even better than normal. Regarding your question about RMS noise level, I don’t know how I would go about measuring that. Please note that I am very much a rookie when it comes to this type of audio stuff, so I appreciate you patience if I am not communicating in the correct terms or perhaps misunderstood your questions. Thanks again for the help.

May 8th, 2002, 02:10 PM	#5
Don Palomaki Wrangler Join Date: Sep 2001 Location: Northern VA Posts: 4,489	Geneerally speaking, a battery will introduce less electrical noise on the powersupply bus than the A/C adapter, and should be a bit lower noise. Also, it reduces the chance for stray pickup by audio leads near the power lead by eliminating the power leads. Whether or not it makes a difference in your tests depends on lead dress and other factors. With audio level in AUTO mode, the and nothing connected to the input, the preamps are at full gain and any noise present at the open circuit input is fully amplified. It is sort of a worst case for noise. Shorting the input should reduce the noise slightly, and using manual gain set at, say 12:00 should be even lower. What are you using to measure the sound levels? I use the RMS (root-mean-square, an indication of the signal's energy level) reading from Cool Edit Pro's statitics window. Most volt meters are calibrated to read the RMS value of a sine wave (pure tone). Also, if listening with earphones, any hiss will ususally be much more apparent than listening via speakers.

May 8th, 2002, 03:16 PM	#6
pixel perfect Posts: n/a	Hi Don, I see what you're saying about the gain with no input--that explains quite a bit. My logic of "no input therefore no noise" just doesn't hold water. As for the measurements, I was using the Spectrum Analyzer feature in Amadeus II. It's a Mac OS X sound app that I like. I am beginning to conclude that between your insights and those of another audio guru over at Ken Stone's site, that I am actually dealing with acceptable noise levels and simply need to use a noise gate or more precise video editing in situations where dead silence to noise is noticeable. And get rid of the headphones. As one friend to me, those will drive you crazy! Thanks again for your most helpful responses.

May 9th, 2002, 09:31 AM	#7
Guest Posts: n/a	I have dealt with the same noise you are talking about and my solution was to use any external mics using line level instead of mic level. I achieved this by purchasing a SoundDevices MixPre and going out line level to the XL1s. Using line level compared to mic level is much quieter. I avoid using the MA-100 as well, I just feed XLR's to the MixPre and go out of the MixPre to the RCA jacks on the back of the XL1s. Just make sure those RCA jacks are then set to line level rather than mic level and you'll notice hardly any noise. I then keep the XL1s audio gain to about the 12 o'clock position in manual mode and use the MixPre to raise and lower volume levels not to peak over the -12db position on the XL1s.

May 10th, 2002, 09:16 PM	#8
Steve Siegel Major Player Join Date: Nov 2001 Location: Albuquerque, New Mexico, USA Posts: 840	I have been grappling with what I think is the same noise you're talking about, Pixel. It's a low pitched hum that pervades all work. I learned that it is, or course, generated by the camcorder, is worse on a tripod or table top than when hand-held, and disappears completely if you take the microphone out of its clamp and hold it in your hand. I will be building a little Rube Goldberg device to try to eliminate the noise. If it works, I'll post it. By the way, Lightwave Systems makes an isolator mount to do the same thing. It's $150.00. Don, you were right about loosening the clamp reducing the noise, but it's not because the innards of the mic are less squeezed, but that the contact with the camcorder housing is lessened. Steve Siegel

May 11th, 2002, 05:29 AM	#9
Don Palomaki Wrangler Join Date: Sep 2001 Location: Northern VA Posts: 4,489	Thanks for the report onthe screw. Folks have a tendency to want to tighten all screws. However, the mic mount screw should not be fully tightened, just tighten it enough so it holds the mic in position, but allow some wobble of the mic in the holder. Over tightening compresses the vibration damping material (rubber) and when compressed too much it no longer provides damping/isolation. About like bottoming a shock on a car.