low volume and hum using mic

[Xerlome]

I certainly didn't intend to make trouble here. I trying to understand something beyond my knowledge. I've made purchases based on recommendations here. I need to assess this best I can, not sure what I will purchase or not. It's not just an exercise. It is confusing when the experts disagree when I haven't the background to judge.

I don't know what to make of the 16 vs 24 bit issue, or the possible cable problem. The cable is the one pictured on page 13, 1st post (#121).

Possibly I made a mistake with the link, but the clip was page 10, 6th down, which I take to be #96. I'm set to first post on top, if that makes a difference.

I've had my computer examined at a shop a couple of times, and they say the USB ports are okay. They told me the XLR to USB cable I had was noisy, so I bought the Dual Pre and the XLR to XLR cable. It is better, if not perfect.

I appreciate all the help. I'm sure it is frustrating for those offering advice. I've been frustrated as well, working on the problem for well over a month, but the noise reduction does give me what I need, even if it is a half solution. I may (if i still can) to exchange what I have for some recommendations, though time, cost, and uncertainty of improvement are issues. Part of it is that, now that I can record voice okay, I hesitate to lose more time attempting to do better. Exchanging the mic would be to start over.

I don't want to trouble you people further if you feel you've offered what you can. I also take it that it is a challenge for you to tackle tough problems, and a benefit if you learn how to solve them in the process. I hope this is so.

[steve]

There is less quantization error at 24 bits than at 16 bits,

Yes. That is what is commonly referred to as the "digital noise floor".
For 16 bit audio, the quantization noise is 1/32768th part (about -90 dB)of the full-scale range.
For 24 bit audio, the quantization noise is 1/8388608th part (about -138 dB) of full-scale.

If the microphone pre-amp is correctly calibrated to hit 0dB digital at maximum "sound pressure level" and the "SNR" (signal to noise ratio) of the microphone is, say 70dB, then the quantization noise level is 20 dB below the microphone noise floor even with "only" 16 bits. In other words, the quantization error for 16 bit PCM is negligible IF the microphone pre-amp and ADC have been correctly calibrated in the design and manufacture (which they should be for a high-end USB microphone).

24 bit recording becomes very useful when dealing with systems that cannot be calibrated - such as recording from an analogue mixing desk with arbitrary input devices. The recording engineer only needs to set the levels within about 20dB of full scale to achieve the full fidelity of the analogue components. In this scenario, when using 16 bit A/D conversion, the recording engineer must be much more precise about setting levels because there is much less headroom, but full fidelity is still achievable if sufficient care is taken when setting levels. Professional studios much prefer to use 24 bit conversion because setting levels very precisely takes time, and studio time is more expensive than a 24 bit ADC.

[steve]

Check your facts (see fig. 24). There is a 6 dB loss: http://www.rane.com/note110.html

That article assumes that the output impedance on the balanced side is significantly lower than the input impedance of the unbalance input. That will almost certainly be the case when connecting active system components, such as connecting a mixing desk to an amplifier. In the case of (passive) dynamic microphones, the situation is somewhat different and other factors dominate - in particular, the input impedance is usually the most significant factor in determining the signal strength.

The Shure SM58 microphone has a rated impedance of 150 Ohms. Although SM58's are quite forgiving, the input load is still important because not only does it affect the absolute signal level, but also affects the frequency response of the microphone. This is because the electrical "load" produces a damping effect on the microphone diaphragm. The SM58 should be connected to a low impedance balanced input so as to have the correct amount of damping and noise rejection.

[steve]

The cable is the one pictured on page 13, 1st post (#121).

I guess that you're quite new to forums too - which is not a problem, very many people who post here have never used a forum before, so we try to make it a safe and helpful place and offer advice as required

If you look near the top left corner of each post, you will see a button that says "PERMALINK".
That button is a "permanent link" to that specific post.
If you click on that button, you may see the web page address in your web browser change slightly and will be something similar to this:

Code: Select all

http://forum.audacityteam.org/viewtopic.php?p=266256#p266256

If you then copy and paste that link into your reply, it will show up something like this:
http://forum.audacityteam.org/viewtopic ... 56#p266256

That then provides a direct link to that specific post (in my example above the link goes to your last post in this topic - try clicking on it to see it in action.

[kozikowski]

I certainly didn't intend to make trouble here.

Don't get discouraged.
Your requirements are perfectly ordinary, it's the solutions that tend to be difficult. You went right from concept to trying to produce a finished, polished, marketable work. People pay me to make good recordings, it's not as easy as the ads would have you believe. The forum is full of people in nearly the same boat you are.

"ACX rejected my work as being noisy and low volume."

And we do deal with other postings. I, for one, occasionally lose the thread of the thread, so to speak, on long postings. Particularly if there are multiple, similar postings.

So you should stick with it. We hate losing a poster.

I left you a Private Message. Click on (x New Messages) near the top of the page.

Koz

[kozikowski]

So to recap. You started out life with the little microphone USB cable which was insanely convenient, but didn't work very well. Graduate to a much higher end microphone digitizer which is larger, not quite as convenient, and works better, but still no award-winner. This one has controls on it and you have to take a little more hands-on effort.

When we left our hero, he was in discussions for which higher-end microphone digitizer was going to do a perfect job. Probably not many. We're still in the quest of good results without knowing what you're doing. This neighborhood is really popular, so there's no shortage of different manufacturers doing the same thing.

The next terrifying step is to a full, though tiny, sound mixer. Now there are no restrictions and there's no reason not to produce a perfect or near-perfect recording. But now you need to know how to drive.

That's the 10,000 foot (3000M) view in case it got lost in the weeds.

[Opinion]
I think you should go with what you got (assuming we can settle on a system of patches and repairs). After about the third audiobook you will have a terrific idea of how to improve the system and a much better idea if you even want to continue doing this. You haven't gotten to the point of script mistakes, voice corrections, editing and chapter compiling yet. Editing and post production take between three and five times the duration of the show [stunned silence].

Koz

[steve]

Thanks for the summary koz.

Xerlome, this would be a good time to remind us of what mic and pre-amp you are currently using

[chris319]

That is what is commonly referred to as the "digital noise floor".

Nope. Quantization error is not the same as a noise floor.

[chris319]

That article assumes that the output impedance on the balanced side is significantly lower than the input impedance of the unbalance input.

Still, there will be a 6 dB loss when an balanced source is wired unbalanced vs. balanced regardless of impedance.

[kozikowski]

Still, there will be a 6 dB loss when an balanced source is wired unbalanced vs. balanced regardless of impedance.

Unless it's balanced/floating. With no hard ground reference, grounding one side will just cause the signal on the other side to double.

If this were any of the active condenser microphones, you'd likely be right. Grounding one side will just kill that side. 6 dB down the toilet. But some of them float, too by way of Phantom Power Regenerators.

SM58s are pretty robust.

Koz