Signal-to-Noise Ratio

[PHILANDDON]

I was comparing specs on analog 2 track recorders...am I correct that with SNR, the higher the number the better? Or is it not so simple?

 

[The Ghost of FM]

Yes. The higher the better.

Numbers around 70db are pretty respectable and if we're talking two track recorders, that should be clean enough to use without noise reduction so long as you keep your levels health and consistent.

If you are recording music with very quiet passages, you may hear a bit of hiss if your monitoring levels are quite loud. Otherwise, steady loud content, like rock or pop shouldn't be an issue at all.

Cheers!

 

[PHILANDDON]

Thanks Ghost. For someone who wants to TRACK to a quarter-inch, two-track deck rather than master-to-one, are there any models with great SNR specs to recommend. My material can be on the soft side.

 

[themaddog]

Even if you're recording "soft" music, it still should be tracked at a loud volume, right at or before 0 so it doesn't saturate. When Ghost talks about soft passes, he is talking about music that has a wide dynamic range, such as classical music. But, even in this case, the dynamic differences would only be apparrent when you mix everything all together, as when you track, everything should be recorded at a pretty loud (right before 0) level.

Does this make sense, or would you like me to clarify with an example or two?

-MD

 

[PHILANDDON]

No, it makes perfect sense. But for acoustic music..i.e voice and guitar...in a country/folk vein, would you agree that going over zero is a bad idea? Or am I being fussy.

Also, could you recommend some 2 track models with great Signal-to-Noise specs, especially for tracking.

THANKS.

 

[FALKEN]

Phillandon, I've done this with my tascam 32. Built a 2x1 mixer out of an altoids box with some pots and resistors...it was pretty fun.

 

[sushi-mon]

No, it makes perfect sense. But for acoustic music..i.e voice and guitar...in a country/folk vein, would you agree that going over zero is a bad idea? Or am I being fussy.

Also, could you recommend some 2 track models with great Signal-to-Noise specs, especially for tracking.

THANKS.

I just got a Fostex A-2. Which is two track 1/4 inch with 15ips. I dubbed some mixes down onto it, and also dubbed two bass tracks outta Ptulz. Sounded great. I have not tracked straight to it yet, but I am not getting bad hiss at all. I am sure that tracking to it should be no problem to get clean tracks without NR.

A 16 track 1/4 would be a problem with NR for sure. I personally want to get a 1/2 8 track or 1 inch 16 track. Use them for tracking, and master down to the A-2.

I have used Quantegy 499, sounded great, but the tape sheds. I am planning on having my machine set up for GP9, which is supposed to be very clean tape with great SNR capabilities.

I probably would like to try 456 for better saturation properties for tracking and GP9 for mastering, but, I would have to calibrate the machine between tapes. Not sure I want to go through all that hassle. So if GP9 gives me good tracking sound, then I will go with it all the way. Of course, if I had two machines I would set up the tracking one for 456 and the mastering one for GP9.

 

[MCI2424]

I was comparing specs on analog 2 track recorders...am I correct that with SNR, the higher the number the better? Or is it not so simple?

100% signal is best. 0% is normal for many decks that are powered off.

Somewhere in the middle is OK.

 

[PHILANDDON]

100% signal is best. 0% is normal for many decks that are powered off.

Somewhere in the middle is OK.

Uh, funny, I guess. Seriously, I'm tracking low level material without a compressor and I'd like to avoid hiss, so I was wondering which quarter-inch, two-track machines have the best SNR ratio's. Still wondering, if anyone knows. Thanks.

 

[evm1024]

Uh, funny, I guess. Seriously, I'm tracking low level material without a compressor and I'd like to avoid hiss, so I was wondering which quarter-inch, two-track machines have the best SNR ratio's. Still wondering, if anyone knows. Thanks.

I'll take a stab at this....

There are 2 cases that I see that may apply to what you want to do.

The first case is where the dynamic range of the material that you want to capture is greater than the dynamic range of the deck that you are using. It is presumed that the deck has the least dynamic range of anything in your signal chain.

Using the highest energy tape (+9) and recording hot (up to the limits of your record/playback electronics and head) may be usefull. (caution, extra mechanical stress on transport, etc)

In this case you will have to make a choice as to what you want to lose or you will need to add (dbx/dolby) a noise reduction system or change the dynamic range via compression.

Bassically you are screwed. You either distort the loug passages so as to keep the quiet ones out of the hiss or you put the quiet passeges inthe hiss and keep the loud out of saturation. Recording level does the selection.

In the second case your signals dynamic range is less than the decks dynamic range. THere is lots that you can do here.

You could use a quiet outboard mic preamp and feed the signal into line in. The goal here is to limit the noise introduced by the decks mic preamp. Of course the mic should have low internal noise and use balanced cable to the preamp. Keep the preamp to deck cables short.

Adjustments of the preamp gain and the deck gain should be done to locate the quietest combination of gains that will bring your signal to a reasonable level on tape. I would look for the gain in the deck that introduces the least internal noise (actually, has the best SNR) and then adjust the preamp for a good signal level on tape.

I suspect we are talking only a fraction of a dB here or there and either a lot of trial and error or some really nice (read $$$) test gear.

Of course your recording environment needs to have a SNR greater than the decks to be even worth trying to get the signal. So moving the source to an anechoic chamber and get the mics close.....

So what it it you are recording.....

 

[Beck]

Uh, funny, I guess. Seriously, I'm tracking low level material without a compressor and I'd like to avoid hiss, so I was wondering which quarter-inch, two-track machines have the best SNR ratio's. Still wondering, if anyone knows. Thanks.

What evm1024 said is good advice.

As for the actual S/N spec here's the story:

The S/N can be a little tricky because different manufactures use different methods. Most manufacturers measure on a weighted curve (A-weighted), which means it takes into account what humans can hear, rather than just what measuring instruments can detect. For example, a measuring device can detect noise above 20 kHz, but humans can’t hear it, and are more sensitive to some frequencies than others. An A-weighted spec will always be better than an unweighted spec.

Otari’s S/N measurements are unweighted, so at first they look better because their unweighted spec is as good as other manufacturers A-weighted spec. But there’s a catch – they only measure up to 18 kHz, rather than 20 kHz. This cuts out upper frequencies that would otherwise contribute to the unweighted noise measurement.

To complicate things even further, some companies give an average S/N spec, while others give a minimum. TASCAM specs were historically the most honest, straightforward, and on the conservative side. Their specs are the minimum you can expect from the machine and they are nearly always better than what’s listed.

I believe the Tascam BR-20 has the best S/N in a ¼” half-track, at 71 dB (A-weighted.) The Otari MX5050 lists 72 dB (30 Hz – 18 kHz).

The TASCAM 32, 22-2 and STUDER A810 all list 68 dB S/N.

The TASCAM 42, Fostex Model 20 and E-2 list 70 dB S/N.

But then there’s that darn “Specifications are subject to change without notice” disclaimer. For example, the TASCAM 42 listed a 68 dB S/N spec in 1983, but 70 dB a couple years later.

Not to mention typos – the early TASCAM 22-2 manual erroneously listed a 66 dB S/N spec, and was later corrected to 68 dB in the manual and brochures.

The bottom line in this long answer to a short question is that the format in general has a record/reproduce S/N of about 70 dB A-weighted (give or take) @ 15 ips, NAB equalization.

In practice the tape you use and your operating level @ 0 VU will make the biggest difference in the background hiss you can hear at listening levels.

And just when you thought that was it, there’s noise reduction, which changes everything.