OK, I realise I'm paddling about in murky waters here, and not wishing to embarrassing myself further, what "system" hiss - noise - do you put with when recording Audio in Vegas? S/N ratio?
I am not sure that I understand the question, but hiss noise is normally generated in analogue circuitry. Tape recorders were a source of hiss as are high gain amplifiers. Signal to noise ratio is a bit ambiguous unless further qualified, because it depends on things such as the bandwidth of the audio channel and how you measure a signal such as speech or music.
Perhaps you could give some background to the question.
OK, I've set up Vegas to Record. I am getting low level PC hiss. Without my mic or preamp connected, and I am getting, as expected, low level "hiss" from the pc.
Reading the Vegas INPUT Track Preview, what low level do you put up with? -200db, -100db, -80db, -50db? And in this context what ratio of PC hiss to mic level do you accept?
For example, and in my crudest of layman's terms, if I am recording at -6db and have already established -60db of system hiss (this from the Vegas Track meter) I'd call that a 10:1 Signal to Noise ratio.
"For example, and in my crudest of layman's terms, if I am recording at -6db and have already established -60db of system hiss (this from the Vegas Track meter) I'd call that a 10:1 Signal to Noise ratio."
Actually that would be a 54dB Signal to Noise Ratio.
On my office PC running V10 about the same. From my Firewire 410 with line level inputs around 90dB.
Generally 60dB S/N is considered acceptable although for cinema you might consider it 90dB The theoretic limit is around 114dB and the best gear gets very close to that.
Also keep in mind that the sound "stuff" on mobos is just plain noisy, even just playing back nothing.
If I use the audio input (analogue to digital converter) on my MoBo, I get a noise level of about -75 dB. If I use my external USB audio device, I get about -78 dB, which is 3 dB better, and would be a noticeable but not dramatic improvement. A good external input device should do better than that, particularly if it uses 24 bit sample size rather than 16 bit as in my examples. These are for line input levels. A microphone level input is likely to be worse than that.
However, the noticeability and annoyance of the noise is dependent on its nature. The noise could actually consist of hiss, which of itself could be pleasant in contact (eg a waterfall), line hum, which may not be very audible but quite visible on the meter, or hash or buzz due to digital signals crosstalking into the analogue circuit, which might be quite unpleasant.
Not really, I spent a couple of years in my younger days working with very old school radio guys, knowing all about "Decibells" was vital to not getting my butt kicked.
Being logarithmic you add or subtract, not multiply or divide. So if you have a signal that's -40dBm and add 6dB gain then you get a signal of -34dBm. The third letter (in this case "m") denotes an absolute reference, turning a relative value into an absolute measurement of power.
You may also come upon units of measurement such as "dBA". They refer to sound pressure levels and the "A" or "C" denotes which frequency weigthing curve has been used. "A" is pretty much aligned to human hearing whereas "C" is unweighted.
If you wanted your signal to noise ratio as a power ratio the answer is not 10 but around 25,000 (10^5.4.). That gives you a practical idea of how hard it can be to achieve a good signal to noise ratio.
Bob.
ps. A bit of a worry that I can recall all that stuff from 40 years ago and have no clue where I put my car keys 5 minutes ago.
If there's an analog circuit in the recording chain, you could also try measuring the inherent noise level by shorting the analog input to ground. That should give you a measurement of the true internal noise contribution of the circuitry. (I bear no responsibility if you blow the card....)
I've been digitizing hundreds of audio tape cassettes and cursing some unknown person who recorded at a signal to noise ratio of about 15 to 20db. If I get one at a S/N ratio of 30db, I think it's good. Several runs of noise reduction produces a usable, if not pretty, result.
1. If you have a lot of hiss in your computer, it would be a REALLY bad thing to use a NR filter after the fact. It is one thing to do that when you have no other choice; it is a completely different, and totally wrong, thing to make that part of your normal workflow.
2. Hiss can often creep in from peripherals that are connected through the sound system. You can test this by going to your sound mixer and muting the playback of every single device other than the one from which you are recording. Years ago I found that a CD-ROM player was causing exactly the noise you describe. I muted it and was able to get on with things.
3. If you still cannot fix the problem, purchase a really good sound card. It does make a big difference, mostly on the recording side. IMHO, if you are serious about audio, you should always use a separate sound card.
4. Make sure you don't have ANYTHING plugged into your computer that has a power plug going into any other outlet than the one into which your computer is plugged. Use a power strip, if necessary, to make this so. Very often two power plugs on the wall, even if next to each other, can come from a different circuit. While ground loops (which is what you can get from using different plugs) usually result in hum, not hiss, you can get some really bad things depending on how everything is connected.
5. As for noise settings, the following is copy/pasted directly from my Radio Shack noise meter user manual:
Set WEIGHTING to weight the sound measurement
So, at least for the Radio Shack meter, "C" is the setting for full-spectrum, flat response, not A. And, for this meter, A definitely has nothing to do with telephone response. I use this meter all the time to calibrate home theater systems, and for that job, it is really important to use the "C" setting in order to make sure that I properly measure some of the standing wave issues which, of course, are primarily a problem at the really low frequencies.
Measuring the noise with an unterminated input is not a real-world test. With a 600ohm load, line level input on my Echo-Mia, (+4dB reference TRS input) I see around -92dB. A much higher noise if unterminated, which is typical for analog inputs.
FYI, As I recall, the Otari MTR-90 were only around -60dB or so, @15ips w/ Ampex 456 (250Nw/m).. With DBX or Dolby NR it would be10-15dBs quieter, but not w/o other issues.
"Measuring the noise with an unterminated input is not a real-world test.
Noted, but I don't have shorting jumpers or a resistive load laying around and I'm too lazy to make them for this effort. Still the measured noise was significantly quieter than the analog input devices I'm likely to use, -56 db for my tape drives.
@ Grazie
Though you don't have a "sound card", you possibly have some of the connectors for your audio circuits on your motherboard, somewhere.
On the system I use for dedicated audio mixing, recording from the Tascam DM3200 via Tascam's Firewire interface I can get around -92 db noise floor in Vegas from the mic pres. On my video editing system, however, using the P6T's MB-integrated Azalia audio, I can get only around -62db if I'm recording to a internal HD. If I record to G-RAID, the noise floor can be as terrible as -54db, and what appears to be hiss can, when amplified, be heard to be a high-pitched whine that modulates with the system accessing the G-RAID via eSATA. I wish Vegas was able to use AJA Kona's audio, but I haven't found a way to do that in WIN7/64 bit. So obviously my editing system isn't used for any real audio recording.
I don't have a sound card for my DVD/CD burner to be connected to.
I haven't had a CD/DVD cable attached to a sound card in this century, though system builders may feel compelled to hook them up if someone else does the system build. OK, when I said "sound card" I simply meant that the CD/DVD has to hook into something for when you are playing red book audio (i.e., audio from a traditional audio CD). All my computers have an audio cable going from the CD/DVD to the sound system (which in my case IS a separate card). This includes computers built in the past few years.
Actually, I am not sure whether it would even be possible to play an audio CD unless there IS a connection between the CD/DVD and the sound system. Since the audio on a red book CD is not in digital format, the device cannot simply transfer bits through the IDE or SATA cable.
This is why you have to "rip" a CD in order to convert the audio to WAV format.
So, I think that if you want to play audio CDs, then there has to be an analog connection from the CD/DVD to the audio card or motherboard audio system.
So, did you actually try doing what I suggested??? Did you mute ALL the devices in your mixer control and then re-try the recording???
Sometimes I wonder why I bother to post in forums to try to help people when those people ignore what I suggest. Obviously I am not always correct, but if people don't even try to respond ...
As far as system "hiss" that is acceptable, it kind of depends on what the material is and how much "hiss" is audible. To me any "hiss" is annoying but with software tools like iZodope and others that forum members have recommended the hiss can be easily removed with little to no affect on the primary material.
Did you ever wonder why the small portable recording units don't have specs? (:
If you want clean source recordings you'll need to look into a quality dedicated audio interface with good specs. good ones are not cheap, but you can get decent ones for around $200-$300.
Some specs to look for or compare to:
Noise level, dB (A): -101.2 Excellent
Dynamic range, dB (A): 101.1 Excellent
THD, %: 0.0015 Excellent
IMD, %: 0.0086 Very good
Stereo crosstalk, dB: -100.3 Excellent
"Actually, I am not sure whether it would even be possible to play an audio CD unless there IS a connection between the CD/DVD and the sound system. "
John, it is possible to play an audio CD without an analog cable. [sloppy logic] I think it was part of DirectX a long time ago.[/sloppy logic] The data that comes from the disk is sent to the computer/CD drive data interface. That's how I've managed to get by without an analog cable for so long.
John Meyer: Hiss can often creep in from peripherals that are connected through the sound system. You can test this by going to your sound mixer and muting the playback of every single device other than the one from which you are recording. Years ago I found that a CD-ROM player was causing exactly the noise you describe. I muted it and was able to get on with things.
Well I would if could find such a control in Win7? Can you tell me where that resides in Win7 - John?
John Meyer: So, did you actually try doing what I suggested??? Did you mute ALL the devices in your mixer control and then re-try the recording???
As I said above. I don't see where the control for the CD/DVD player resides. All I can find are the controls for the Speakers. Should I physically go into the PC, locate the audio cable and remove it from the MoBo? Should I? There must be an easier way to test this?
As to using noise reduction, I agree, this is not curing the cause. I proposed this as a "inkling" to the level of hiss there is.
Still no clearer as to what to do. Getting below 90db would be real sweet.
Also I don't have to record anything. I can SEE the hiss in the meters. AND I hadn't connect the mic either. It's not as if WHEN I connect the mic, it is there, all the time.
"I hadn't connect the mic either. It's not as if WHEN I connect the mic, it is there, all the time."
Welcome to the horror that is mobo audio.
On the mobo you have digital signals at 5V radiating noise over a wide spectrum of frequencies. On the same board you have unbalanced anaog signal lines that carry a signal of 1/1000 of a volt.
For a signal to noise ratio of 60dB on thos lines the noise level needs to be less than 1/1000 times 1/1,000,000 which is an (American) billionth of a volt. Good luck with that.
For my money it was a no brainer. Buy an external sound box, in my case the now obsolete M-Audio Firewire 410. Balanced audio inputs with phantom power etc so I could use a good studio microphones, in general I've used the Rode NT1-A which is the quietest mic made, noise floor is 5dBA. In other words with the mic connected to good kit the noise is only 5dB above the threshold of human hearing. The Firewire 410 is now the Firewire 610 if you're interested.
There's now plenty of other external sound boxes that connect either via USB ot Firewire at various price points. My personal preference is for firewire for thse devices however I've used those single channel XLR to USB converters and the good ones work quite well. The Shure X2u gets good reviews if you don't want to spend the money on a full blown external box.