Does the Transporter outperform SACD (2ch)?

[Eric Seaberg]

OK, so I was off a bit on the high end. Did you measure the fall off
over 25kHz? I assume it kept falling off at 6 or maybe 12 dB/octave

It was really dependent on the gap in the poles of the record head. Smaller gap gave you more hi-freq. It really varied from machine-to-machine and manufacturer-to-manufacturer. I was happy just to get 22k out of the A800, but the roll-off was quite drastic above that.

And the S/N of this model was 7-10dB better than most anything else on the market, approaching 72dB in some cases (16-track/2-inch head stack) because the first pre-amp stage was literally inches below the heads.

The studio I built back in '83 still has that machine. I've offered to buy it from them but they won't sell it.

[MrStan]

BTW, Dolby 'A' or 'SR', which was designed for professional use, is TOTALLY inaudible when setup properly. Dolby 'B', the consumer 'cassette' noise reduction scheme, is VERY audible because the record/playback levels were never matched for proper compansion. The 'other' system you may have been thinking of was DBX, which didn't have the 'level-setting' requirements that Dolby did.

Sorry Eric, I'm with Pat here. There hasn't been a Dolby NR recording of a Female choir that I haven't heard the Dolby pumping in and out on the trebles producing a kind of modulation over the voices. I used to think it was the microphones until one day I had the opportunity to do a recording myself with and without Dolby A. Switch off the Dolby and the effect disappeared. It's very slight but audiable and the same effect is there but much more exagerated with Dolby B. If you really want to hear it try Dolby C.
Actually you can measure it. Effectively the Dolby adjusts the high frequency response of the recording and playback according to the content of the recording. If there are no treble sounds present then you can turn down the frequency response of the recording chain to loose the noise, as soon as high frequency appears you restore the equalisation using the recorded sound to mask the noise. Unfortunately the spectrum of the noise doesn't exactly match the spectrum of the material and with some material it appears to cause some type of intermodulation which is audiable. Put a frequency analyser on the output and it is possible to just see it.
The only advantage of Dolby A was to split the audio spectrum into 3 bands and deal with each band seperately which made it essential to align the equipment properly before recording. But is was still annoying once you noticed it was there.
Fortunately this particular problem disappeared when we entered the era of digital mastering.
Strangely enough I have heard a similar effect happen during demonstrations of "Audio Watermarking." This time it seemed to produce an extra "edginess" to the Brass instruments. Enough to be able to do a blind test as long as there was some brass content in the recording. That was a few years ago, they may have fixed it by now as I haven't heard it on any broadcasts.

[MrStan]

It was really dependent on the gap in the poles of the record head. Smaller gap gave you more hi-freq. It really varied from machine-to-machine and manufacturer-to-manufacturer. I was happy just to get 22k out of the A800, but the roll-off was quite drastic above that.

And the S/N of this model was 7-10dB better than most anything else on the market, approaching 72dB in some cases (16-track/2-inch head stack) because the first pre-amp stage was literally inches below the heads.

The studio I built back in '83 still has that machine. I've offered to buy it from them but they won't sell it.

Many studio multi-track machines employed pre-distotion circuits in their recording chains. The idea was that the type of distortion produced by magnetic tape was entirely predictable therefore an anti-signal could be generated at record time and the recording process would cancel this out and the distortion was reduced considerably. Easily by a factor of in excess of 10 times. Since this type of distortion isn't actually that audiable, most engineers actually aligned the machines to a higher recording level restoring some of the recording distortion but improving the overall signal to noise ratio. This process was also used on the Nagra portable 1/4 inch machines and these machines could achieve noise figures in the 60 - 70 db range while keeping the distortion below 0.5%.

The only problem was that the distortion was predictable for a given batch of tape stock. This meant that the machines distortion circuitry had to be realigned for each new batch of tape and this wasn't an easy setup. Not too bad in a studio but I was responsible for keeping over a 100 of these machines on the road and when a new batch of tapes was delivered to stores, life could get hectic. Especially in the mid 70's when the dreeded tape oxide shed happened. Something went wrong with the glue that stuck the oxide to the tape and the oxide was literaly stripped off by the heads and it clogged up the gaps. Nothing for it but to reject the whole batch of tape which was hundreds of rolls and change it which meant re-aligning all those machines again. This tape binding glue was used by a number of companies and many tape masters from that period have lost their high frequency sparkle and sound quite dull. When they are re-mastered onto CD you can easily hear the difference if you compare them to the old vinyl recordings. Digitally re-mastered doesn't always mean better! I wonder sometimes why they don't return to the vinyl masters and process out the noise.

Actually the recording process is more complex than usually assumed and the high frequency response is only dependent upon the head gap width during playback. Something very strange happens when you record and most recording heads have actually got quite a wide gap as this inversly effects the linearity or distortion of the record head. If it's narrow you can't generate enough magnetic field to fully magnetise the tape without overloading the head.

The recording actually appears to take place slightly outside of the head gap area and as far as I could detect the high frequency response was dependent upon the straightness of the trailing edge of the recording gap. If you put the gap under a microscope and imagine two lines bordering the trailing edge of the gap the distance between these lines appeared to determine the high frequency response and not the gap itself. The position of the recording relative to the gap appeared to be dependent on the recording bias applied. As the edge of a head was likely to be straighter than the gap itself, most recordings were actually much better than the resultant playback therefore old recordings often sounded better when played on a new machine as head technology improved. We had many recordings made on old EMI machines and they sounded a lot beeter when replayed on a Studer. Although you measured a roll off on the machine around 22k the recording was probably a lot better. I made a recording on a semi-pro machine which rolled off around 15k at 7.5 inches but playing it back on the test tape making machine (which just happened to be a modified Studer) producd a result that was actually pretty flat up to 20k and that was about the limit of the Studer playback.

If you have a machine with a combined recording head then it looks like it is impossible to have a head doing both jobs but in fact a second gap was made in the rear of the head and as far as the recording was concerned it appeared to use the total of the two gaps giving it good linearity whereas for playback it depended on only the front gap although the back gap did provide a little loss in sensitivity.

Analogue alignment was so much more challenging than digital and producing the alignment tapes to align the machines in the first place was a real challenge. How do you align a machine to make an alignment tape if you haven't made an alignment tape to align it with. It's an awful lot of trial and error. Yes you could buy them in from the recording companies but if you checked one against another they never agreed.

They couldn't even agree on the correct azimuth which is essential, making sure that the tape gap was aligned at exactly 90 degrees to the tape. It must be absolutely identical on the record and playback machines otherwise you stand to loose that precious high frequency, but it should also be at 90 degree to make sure you use the maximum width of the track and thus get the best SN ratio and if you were recording stereo then the two tracks on the tape head should be in exact alignment otherwise you introduce a phase error between the two channels. Actually the heads were never that perfect so it was up to the engineer to make allowances when he aligned the machine.

Fortunately I never had to deal with multi-track tape machines although I did have to maintain multi-track magnetic film machines and produce test films but due to the rigidity of the medium and the sprocket drive we never could get the same performance out of these.

How it ever worked so well was a real miracle but it did.

[pfarrell]

MrStan wrote:
> The only problem was that the distortion was predictable for a given
> batch of tape stock.

Sometimes it seems like its each tape, or each week or something.
Its like the tape is alive.

> Especially in the mid 70's
> when the dreeded tape oxide shed happened. Something went wrong with
> the glue that stuck the oxide to the tape and the oxide was literaly
> stripped off by the heads and it clogged up the gaps.

a dark period for sure.


> Actually the recording process is more complex than usually assumed
> and the high frequency response is only dependent upon the head gap
> width during playback. Something very strange happens when you record
> and most recording heads have actually got quite a wide gap as this
> inversly effects the linearity or distortion of the record head.

For some reason, the Studers at the place I rent for clients that insist
on 2" tape seem to have worse high end at 30 IPS than at 15. Which makes
no sense, higher tape speed is supposed to make it all better.




> Analogue alignment was so much more challenging than digital and
> producing the alignment tapes to align the machines in the first place
> was a real challenge.

Did you put 10 feet of generated signal on each tape to help some poor
guys five years later?


> Actually the heads were never that perfect so it was up to
> the engineer to make allowances when he aligned the machine.

True. Nothing is that perfect, and studios do a lot of stuff with tapes,
engineers get tired, punching in the same spot over and over is boring, etc.

> How it ever worked so well was a real miracle but it did.

Because its the music, not the technology.


--
Pat Farrell
http://www.pfarrell.com/