h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Oh! Gnomon hi! What happened to Tony Curtis?

Sorry hadn't seen your posts above!

HELL

I got rid of Tony Curtis since the iCan site does not apply to people outside the UK.

HELL

Hiss, or pink noise, is random but of uniformly low amplitude as opposed to the sounds one wants to record which will have variations of amplitude. It's just one type of noise you get when using tape. There are other distortions produced with this method including rumble, wow and flutter from the mechanics of the system and noise introduced in the processing of the signal both during the recording and playback. Each component-every electrical connection, resistor, capacitor, transistor- which the audio signal passes through will add some measure of distortion.

Extremely high end audio equipment lacks all the bells and whistles of what many consider high quality. Typically there are no switches for multiple sets of speakers, no equalizers, no bass control, no treble, no balance-all these 'features' add distortion. $3,000 pre-amplifiers might not even have a volume control, power switch or transistors-there still exists a huge tube/transistor debate as to which is closer to perfection.

Regarding the passing of vinyl-it is a long, long way from death. There are audiophiles who argue incessantly about the quality of the various digital recording processes and how they all distort the original. If vinly were truly dead then you wouldn't find turntables [bare, no tonearm, cartridge or stylus] selling for $US 25,000; vibration-isolating feet for hundreds of dollars each; tonearms of fantastically expensive materials for $12,000; cartridges that go for $15,000 and styli [stylus aka needle] made of diamond, sapphire, ruby or some synthetic that can run $5,000 AND several hundred to replace at the factory after your kid plays DJ with his Barney record on it. It is not uncommon for an audiophile to break $100,000 for his 'record player'.

Tapes as cassettes or 8 tracks are/were a convenience that sacrifice(d) quality for portability. You also skipped over mentioning wire recording as far as history goes. This didn't last long and reel-to-reel tape was never a smash hit either, even though it was far better than cassettes or 8 tracks.

On the db [Dolby] methods it may be easier to comprehend this way. Let's call absolute silence 'level zero', and 'hiss' 'level 5', on a scale of 100. If your audio material's softest passage is 'level 3' it will be overpowered by the hiss. Hiss is constant, so let's split the full range of our audio by octaves and grab those over 1,000 Hz [not a specification, just an easy number for explanation]. Take these higher frequency sounds we want to record, and boost just them say ten times-leave the lower frequencies as they were. Now we're dumping this on the tape at ten times level 3 so we're at 30: pretty loud in comparison to the hiss as well as the virgin lower frequencies. Play this back on a non-db system and it can make your ears hurt-constant hiss is very annoying and tiring. Playback on a db machine will be amazing, as during playback the level 30 audio is brought back to 3, while the hiss that was 5 is now nearly absent. Dolby B was double the treatment of A, and Dolby C was another notch up from B.

Tape structure affects noise-the type of metal used was a factor in retention of the signal. Cheap tape used ferric oxide [aka rust], however it's ability to retain a magnetic pattern was poor. This means the recording level had to be high, bordering on distortion, in order that the playback be a reasonable image of the original. Higher quality tapes had more metal content, finer particles or a different metal oxide such as CrO2. Different metals retain magnetism differently, or not at all. If you use a metal that really 'hangs on to' an applied magnetic field then you can lower the recording level and the distortion along with it, and when you play it back your 'volume' is still there. Metal tapes are never really pure metal but polished mirror smooth-a uniform surface lessens distortion. Cheap tape will have a dull surface when new and can actually eat away at the parts it passes over like sandpaper does wood.

There are a hundred topics or more here, many untouched [ie: speakers]-each could see it's own detailed entry.

That's very interesting, Dr Jeffreyo, but what are you suggesting? Do you want Hell to include all this in the entry? I think that would be a bit extreme, since the entry is supposed to be an introduction.

Thanks for taking the time to read the Entry and for your constructive criticism.

"There are a hundred topics or more here, many untouched [ie: speakers]-each could see it's own detailed entry."

Exactly. This Entry is about 'Tape Hiss', and not about acoustic/recording/reproducing/suppressing noise in general. If you feel inspired to write your own Entries, go ahead. I would really be proud to see that my Entry has inspired someone to write an Entry.

In this sense it is my oppinion that including the detailed information contained in your comments, while useful and insightful, would move the focus of the Entry to other areas. In other words: I would like to keep this Entry short and focussed on 'tape hiss'.

Now specifically to your comments:

Other sources for noise: "Each component-every electrical connection, resistor, capacitor, transistor- which the audio signal passes through will add some measure of distortion." Right! But these distortions and noises are not 'tape hiss'.

Vinyl a long, long way from death. OK, if you see it that way. I think there are also many people who prefer to write with mechanic typewriters. Turntables still have an application for those scratching DJs, who use the turntables as an instruments. Even so, I think that considering a large, global, average scale it is safe to say that vinyl is dead.

wires, 8-tracks, and other recording techniques. I don't mention them, but I provide a link to another Entry that explores this in detail.

Dolby. I intentionally kept that short in this Entry. It would be too far out of the scope of this Entry. I was planning to do one on Dolby and Noise Reduction, but you see, that's like opening a can of worms. You'll end up having to write an Entry about Noise, and about 'random numbers'... I am working on it, actually. It seems like you could help.

tape structure... Yah. Again I chose to keep that at a minimum to keep the Entry focussed.

Cheers. And thank you very much!

HELL

Oh, and thanks Gnomon for stopping by

Hell, this entry is exactly technical enough without being overbearing. I think this is a great job and I don't think it will be long now....

I have not read all the backlog
I note you sat that noise increases linierly? when copying the tape noise increases by 3db which is logarthmic.


Lightman

OK, but your auditive perception is also logarithmic, so... linear is roughly in order, I think. I was not talking about dB's there, just as an amount of perceived noise.

HELL

Ah yes once again I seem to have run off on tangents while trying to provide information-which I hope was a little helpful at least.

Apropos 3dB...

3 = 10*log(x)
10^(0.3)= x
1.9953 = x

So... +3dB corresponds to a factor of 2 (and -3dB to 0.5). Strange how I didn't notice that before, but that means that increasing the noise by 3dB is the same as *doubling* it... This reasoning works fine for the first copy. Noise is additive. Add 1 and 1 and get 2 (a factor 2 in noise increase or +3dB). The next copy is 2 + 1 and gives a factor 3 (+4.77 dB). The next copy gives a factor 4 (+6dB)... And so on. You cannot simply add the dB's there. The good news is that while noise increases linearly with every copy, the signal to noise ratio decreases logarithmically. The difference between the 1st and the 2nd generation copy is huge (3dB) compared to the difference between the 15th and the 16th generation copy (0.3dB). Or, you will not notice much difference between the 15th and the 16th copy. (I am assuming an idealized experiment here) Of course there are other things besides 'tape hiss' that get added as noise every time you make a copy, and the loss of high frequency modulations, that make copies sound duller and duller...

Note: In my reply I said that you perceive it as a doubling. I have to take that back. *Noise* increases linearly. We should *perceive* a doubling of the noise (idealized experiment) every 2^n-1_th copy (or in the 1st, 3rd, 7th, 15th ... copy of the copy)

HELL

After a certain number of copies the deterioration of the frequency danymics is worse than the increasing noise, I think...


HELL

Hi HELL,

I'm really surprised you only just realised that 3DB down is half power. The points on a response curve where the gain is 3DB down from the peak are even called "the half-power points".


Awu

Hi again HELL,

I didn't intend that last post to come over in a bad way
The entry is of course very good.
It's just that the 3DB points are so special, and so often used (bandwidth is usually the range of frequencies on a bell curve that are bound on both sides by the half power points for example).


If you Google for "3DB down half power" I bet you'll find an example.

Awu.

No worries.

I know about all that...(UNK!) It's just that I was not thinking in terms of dB until that post appeared... Strange, no?

HELL

Your Guide Entry has just been picked from Peer Review by one of our Scouts, and is now heading off into the Editorial Process, which ends with publication in the Edited Guide. We've therefore moved this Review Conversation out of Peer Review and to the entry itself.

If you'd like to know what happens now, check out the page on 'What Happens after your Entry has been Recommended?' at EditedGuide-Process. We hope this explains everything.

Thanks for contributing to the Edited Guide!

Well done!

I think I can spare from from Saturday, so here you go!


Geggs

Well done!

Thanks everybody!

HELL