What makes tape sound better than vinyl ?

Actually, LP can have an edge over tape when it comes to HF, because it's not subject to saturation the way tape is. I'm not sure why you think LP can't produce the full frequency range in wide stereo, but it sounds like you're not using a good phono stage, so that may be your problem.

+1  The LP has much wider bandwidth than any audio tape format. Lower distortion and noise too.

The signal to noise ratio of Dolby SR is around 90dbs.

Right. Lathe cuts are better than that; so far the only pressings I've found that get into the range come from QRP, the pressing plant owned by Acoustic Sounds. Quite literally the playback electronics are the noise floor.

I've often wondered what Dolby S could do for the LP...

Conversion of electrical into magnetic and back to electrical is not the same as conversion of mechanical into electromagnetical. Ralph, you know it better than I do. As for if something is altered in any case - probably. To use an analogy, converting water into ice is not the same as converting water into wood or stone.

The analogy does not hold up. Yes, there are two different processes involved and they both do quite well. Saying that one is inherently better than the other strictly out of the process ignores the progress of that technology. However we do know the specs that are out there and they have been there for a while- tape has not been advancing as much as the LP in recent years as there are very little in the way of modern tape machine producers. The LP sector has new pressing machines, new tone arms, new turntables and so on and so enjoys the advantages that come with newer technology.

How do you compare these specs to whatever it is you are referring to?

30Khz bandwidth is good for tape- but the LP has had bandwidth beyond that since the advent of the Westerex 3D cutterhead (1959). Its electronics are bandwidth limited at 42KHz. The head can go higher than that but the bandwidth limit is there to prevent damage to the head due to the RIAA pre-emphasis (which is +6db per octave). The bandwidth has been there in playback since sometime in the early 1970s.

The distortion is comparable, depending on the playback. A well set up arm and cartridge will have lower distortion figures; a poor setup will be much higher. 

The noise floor of a **well-pressed** LP exceeds the best of tape even with Dolby S. Inferior pressings will be considerably noisier (note added emphasis).

With tape everything stays within the domain of electromagnetic energy, with vinyl there is a conversion of mechanical energy. Do you think nothing is altered in the conversion ?

@inna There is a conversion of an electrical signal to magnetic, and then that magnetic field is stored on the tape. Do you think nothing is altered in the conversion?

Tape heads and tape are non-linear. To linearize either one, bias is added in the form of a high frequency signal that is at once so high that it can't be recorded and will not interfere (hetrodyne; i.e. 'birdies') with harmonics of the signal to be recorded. The Germans sorted this bit out during WW2 (BTW you can see one of three known Nazi tape recorders at the Pavek Museum in St. Louis Park Minnesota. If you are ever in the Twin Cities as an audiophile its a must-see). There is a certain amount of harmonic distortion associated with recording; the amount varies depending on the permeability of the record head and the formula of the tape. 3% is not uncommon at 0VU.

Solid state machines need a bias trap to keep the bias signal away from the record head driver circuit- the head driver transistor can otherwise be saturated. Tubes are immune to this problem- IMO this is one reason why tube machines can make better recordings all other things being equal (which in practice they never are). 

In comparison to vinyl, tape has less bandwidth (on top and the bottom) and less dynamic range- it is also noisier unless the LP has been abused. This is why its practical to use an analog tape as a master for an LP- the LP encompasses the performance of the tape medium. The limitation of the LP is in playback, not record- and that is where one runs into distortion, mistracking and the like, which is highly variable depending on the setup and quality of the pickup apparatus. A good playback will have no mistracking regardless of the program material.

One thing tape can do quite well is out of phase bass. It this occurs in an LP, the groove walls can get so close together that they can knock the stylus out of the groove. For this reason, circuits exist to sense out of phase bass and then cause mono operation below a certain frequency. I have found that if the engineer spends enough time with the project, there is usually a way to master it without using the processing, but engineering time is expensive, so the processing is usually left in place (BTW, its a passive process). Out of phase bass is a problem with multi-tracked recordings if the engineer isn't careful.

Lately there have been a lot of 10.5" reel to reel titles becoming available; the advertised provenance being that they are dubs from working copies (there are a lot of Russian titles claiming this on ebay right now) and they are going for serious cash. I am of the opinion that if you get a title that is in fact actually dubbed from a legitimately good source using quality gear that the result is spectacular. But a lot of these dubs I've been seeing don't measure up to that- I would be not at all surprised to find that many of them are mastered from a CD; some of them I've heard are oppressive on top compared to the original LP.

f you are saying that transferring cd's or vinyl to tapes sounds better then you are saying that the transfer is not an accurate transfer and that the distortion that the tape is adding is one that pleases you. There is nothing wrong with that but you do have to admit that you just introduced a uphonic distortion.

Many years ago when I had CD and cassette in the same system, I used to tame the bright, hard, cold sound of CDs by recording them to cassette. That worked fairly well. Back in those days my theory was that the CD could not record the ultrasonic noise that was part of the digital experience (while my speakers could do it easily); these days my theory is that the cassette limited bandwidth and so was unable to reproduce aliasing (which is interpreted by the ear as brightness and hardness).

I'm certain that the cassette was not true to the original, but it rendered the CDs listenable, so it was useful :)

They are the same track width so they should sound the same....

FWIW Four track 1/2" is uncommon- 8 tracks on 1/2' is a more common format, used by lower end semi-pro machines.

Besides, what about 1/2" tape machines ?

They're awesome! Not many titles available, but you can record your own. The format is a bit pricey...

Many years ago I noticed that the tape seemed to do something nice to the sound of vinyl.

Turned out the reason was because I was using headphones and the turntable didn’t like the vibration from the speakers. So I went about the process of providing for a vibration-free stand, platform and turntable and then the difference vanished- in fact went in favor of the LP.

Stand: custom Sound Anchors on Aurios Pro bearings
platforms: UltraResolution Technology (n.l.a.)

Good vinyl can have as much as 70dB dynamic range on the outer edge.

Its easily higher than that (I run an LP mastering operation); but compression and tape are often the reasons why you don't see that in practice.

Direct to disc recordings will not solve playback problems, one of the reasons why it was virtually abandoned.

Nothing solves playback problems on the record side :)

The reason you don't see much in the way of direct to disk is simply because its really hard to do! Musicians have to play all the tracks on a side perfectly and that's assuming you don't overcut because a performer is playing 10db louder in record than he was when the levels were set!

So its a bit of a trick but when it all comes together its magic.