The best analog reel to reel studio tape recordings ~ 13 bit equivalent?

[hvbias]

Not sure if the time stamp will copy over to the forum's software. It's at 11:18 if it does not.

 

[Cbdb2]

No. Studer A827 64db s/n plus Dolby SR, 25db noise reduction is 89db. If you include the gains you get with soft clipping and a bit of tape compression another 6db. Pretty close to 16bit. Everything has to be perfectly aligned and calibrated for the tape, but it is possible.

 

[fpitas]

Tape got pretty good towards the end there. But it's a challenge to properly store it if you expect it to stay good.

 

[Tom C]

So, if you had such a maximally optimized tape you wanted to digitize, and you wanted a bit of headroom to ensure you capture everything on the tape, you’d need, say 18 or maybe 24 bit transfer?

 

[Cbdb2]

I said close to 16bits so16bits would probably be enough. The digital upgrade for those machines where all 16 bit (and 48/44.1 khz) and the major studios were buying them up like hot cakes even at 3 times the price of the analog machines. Not because it was a step backwards in SQ.

 

[Robin L]

So, if you had such a maximally optimized tape you wanted to digitize, and you wanted a bit of headroom to ensure you capture everything on the tape, you’d need, say 18 or maybe 24 bit transfer?

24 bit is now common, available on cheap gear. 24 bits is good for audio treatments, like de-noising.

 

[Cbdb2]

Now. Back in the day analog multitrack were still being used data was expensive. The first hard disk recorder I used got a hard drive upgrade. 4x 680meg drives at $6000 each! They could only play 4 tracks per drive and store 20 minutes per track. Going to 24bit would have reduced that to 13 minutes.

 

[MaxwellsEq]

Information can be recovered from below the noise floor in analogue channels. It's not an impenetrable concrete floor. So when someone states that noise is at -70 dB, that doesn't mean that you can't hear something at - 75 dB.

 

[Blumlein 88]

I think 13 bits is about right for all other than the very last rtr machines with noise reduction. Dolby A from 1965 gave 10 or 12 db noise reduction. Before then 13 bits is more than enough. SR was right at the end of the tape era.

 

[SoNic]

What people forget is that there might be 13-14 bits of analog THD+N on tape... but there were 16-24 channels (2" tape) to mix into final two channels.
That raises the dynamic available and lowers the probabilistic noise.
This way modern re-mixes of analog 24 ch into digital stereo domain could have close to 16 bit of actual information. Now, the actual music will never go that wide dynamic range, but the noise and distortion from the multichannel analog tape will be in-audible low.

 

[Blumlein 88]

What people forget is that there might be 13-14 bits of analog THD+N on tape... but there were 16-24 channels (2" tape) to mix into final two channels.
That raises the dynamic available and lowers the probabilistic noise.
This way modern re-mixes of analog 24 ch into digital stereo domain could have close to 16 bit of actual information. Now, the actual music will never go that wide dynamic range, but the noise and distortion from the multichannel analog tape will be in-audible low.

I don't think it works that way. If you sum identical signals the noise goes up by 3 db and signal goes up by 6 db which is a net gain of 3 db each time you double the signals. If you are combining or summing two tracks with different signals, the noise goes up by 3 db. The signal doesn't go up at all. So you increase noise 3 db each time you double the tracks being mixed together. So combining 16 tracks would raise the noise 12 db loosing you 2 bits of dynamic range.

 

[SoNic]

The signal on each track is mixed down into stereo, to fit into the same max signal envelope. So, the level of EACH track gets lowered, including it's noise and THD.
Sure, some tracks will have more contribution to the final mix, but none will be at 100%.
Also, for some sources, the thd noise it really doesn't matter - electric guitar, bass guitar, drums...

 

[Blumlein 88]

The signal on each track is mixed down into stereo, to fit into the same max signal envelope. So, the level of EACH track gets lowered, including it's noise and THD.
Sure, some tracks will have more contribution to the final mix, but none will be at 100%.
Also, for some sources, the thd noise it really doesn't matter - electric guitar, bass guitar, drums...

Still doesn't work. You can lower the level by 3 db lowering noise, but you also lowered signal. Then mixing different signals the signal level is the same and noise goes up. You don't gain any dynamic range this way, you loose it. You cannot add noise and have a lower noise level. Plus the medium you are recording to provides the limit on what you can achieve. The final two channel mix isn't going to have additional dynamic range or lowered noise levels beyond what the tape medium can provide.

If the device used for summing has a much lower noise floor than the medium which in the case of tape it will, you can mitigate the increase in noise with mixing multiple tape tracks, but you cannot increase beyond what the medium is capable of doing. Your final two channel mix can be no more dynamic range than tape is and most likely will be lower.

 

[Ported]

I concur.. having mixed many mulitrack songs.. even if the track is "blank" lifting the fader does add noise.. and the more tracks used etc..

 

[Tom C]

Random noise should always be additive, since negative values for noise would not be possible without inverting the signal.
Deterministic noise will always be additive, and harmonic distortion is deterministic.

 

[Ported]

I believe the aspiration for CDs was to better half inch tape running at 30ips and 12bit was shown to be enough but that required noise shaping etc so 16bit was adopted ...

 

[DVDdoug]

I didn't watch the video yet...

Digital and analog have different characteristics and limitations. As an analogy, you can't lower digital video resolution to make it look like VHS.

Quantization noise is similar to tape noise in-that it's most noticeable at low volumes and usually masked at higher volumes or with a dense mix. However quantization noise goes-away completely when there is no signal. Digital silence is at -infinity dB!

I've read that vinyl is equivalent to about 12-bits. I believe that's under the best conditions.... Obviously a lot of vinyl isn't that good. And the occasional "snap", "crackle", and "pop" is going to be a LOT more annoying to me so I'd probably prefer the 12-bit digital. (I've never listened to 12-bits. I've heard 8-bits and the quantization noise is audible.)

Quantization noise is similar to tape noise in-that it's most noticeable at low volumes and usually masked at higher volumes or with a dense mix. However quantization noise goes-away completely when there is no signal. Digital silence is at -infinity dB!

In order to find the "equivalent" I think you'd have to get a group of listeners and ask if they prefer the digital or analog. Assuming the prefer the digital you'd reduce the digital resolution until they consider the analog to be preferable.

I assume I couldn't tell the difference between a CD and a "studio quality" analog copy...

In the analog days there were usually multiple generations of analog tape (including overdubbing). I'd guess there were usually 3 or more generations. Of course you don't get generation loss with digital (assuming lossless digital).

I've heard "rumors" of classical recordings made directly to stereo tape with the record cut from the original tape. Also in the very early days of tape (1940s?) there were probably records cut from the 1st generation tape. Before tape, the disc master was cut directly. I think I have an old song on a CD that was made that way because I can hear the crackle. Or, or it's possible that the master tape doesn't exist anymore..

What people forget is that there might be 13-14 bits of analog THD+N on tape... but there were 16-24 channels (2" tape) to mix into final two channels.
That raises the dynamic available and lowers the probabilistic noise.

A lot of people don't realize that the same thing happens with digital. When you mix the resolution increases! But you don't get the generation losses.

 

[Cbdb2]

Still doesn't work. You can lower the level by 3 db lowering noise, but you also lowered signal. Then mixing different signals the signal level is the same and noise goes up. You don't gain any dynamic range this way, you loose it. You cannot add noise and have a lower noise level. Plus the medium you are recording to provides the limit on what you can achieve. The final two channel mix isn't going to have additional dynamic range or lowered noise levels beyond what the tape medium can provide.

If the device used for summing has a much lower noise floor than the medium which in the case of tape it will, you can mitigate the increase in noise with mixing multiple tape tracks, but you cannot increase beyond what the medium is capable of doing. Your final two channel mix can be no more dynamic range than tape is and most likely will be lower.

If mixing signals dosnt increase the level you could add overdubs to your mix, or explosions, forever without changing the level. Dosnt happen. Mixing signals adds the voltages. Simple math. Don't know how adding 2 music tracks together dosnt increase the voltage/signal.

 

[RammisFrammis]

Still doesn't work. You can lower the level by 3 db lowering noise, but you also lowered signal. Then mixing different signals the signal level is the same and noise goes up. You don't gain any dynamic range this way, you loose it. You cannot add noise and have a lower noise level. Plus the medium you are recording to provides the limit on what you can achieve. The final two channel mix isn't going to have additional dynamic range or lowered noise levels beyond what the tape medium can provide.

If the device used for summing has a much lower noise floor than the medium which in the case of tape it will, you can mitigate the increase in noise with mixing multiple tape tracks, but you cannot increase beyond what the medium is capable of doing. Your final two channel mix can be no more dynamic range than tape is and most likely will be lower.

Not all tracks are active all the time in a mixdown situation. Tracks with audio at only a particular verse or chorus are usually muted until needed. Like noted above not all tracks are near full volume at any point. I once watched a movie being mixed down and I was told there were over 200 tracks but the noise was inaudible because the tracks were used selectively and when a lot of them were used it was very loud anyway so the noise was buried.

 

[Cbdb2]

With music even the gaps between the snare hits on the snare drum track (etc.) are muted, before DSP gates where used.