I thought i wanted a TSR-8

Kasey

New member
But now i've learned that it has DBX noise reduction as opposed to Dolby, which ive been told is better that DBX. Is it true that DBX is known to muddy the sound? I'd love a reel to reel but i'd prefer to not have to choose between tape hiss or noise reduction mud.
 
dbx type I professional, which is what's employed in the TSR-8 will not "muddy" your sound as it processes cleanly down to 50hz without any pumping or breathing artifacts.

dbx type II was the consumer - home version of the noise reduction and only processed down to 100hz because of the shortcomings of non-linear low frequency consumer equipment, the process was cut off at that point and the occasional pumping, breathing and overall muddiness could be detected on some machines depending on the recorded content...it wasn't a universal rule that you would always run into problems with type II, just that it was more likely to occur.

You have nothing to fear with the TSR-8's NR regime.

Cheers! :)
 
The dbx on the TSR/MSR machines is nicely done. I have no complaints with my TSR-8. :)
 
Beck said:
The dbx on the TSR/MSR machines is nicely done. I have no complaints with my TSR-8. :)
Heh - I've just bought two more :D ..it's a long story.
Hmm : B77, '32, A8, Uher, TSR-8, that's going to be 7 tape machines, 8 if you count the copicat. This is really going to make my eventual house move an interesting logistical exercise.

Anyway, at the risk of spamming the thread, this is my latest TSR-8 adventure for those who doubt its quality:


Ignore the lyrics, it will take too long to explain.
 
FALKEN said:
....whats a copicat?
http://britamps.mysite.wanadoo-members.co.uk/watkins/copicat/copi3.html
..this is the MK4. This particular one is branded 'Guild' - mine has the proper 'Watkins' logo on it but is otherwise more-or-less the same. The previous owner hotrodded it to add a 'motor off' switch which makes quite interesting sounds if you shut it down while it's operating.

I used it to provide the slapback echo on that recording.
 
How does noise reduction on Reel to Reel's work? as in... where is it placed in the chain within the machine? Is there noise reduction between the input signal and the tape? (that wouldnt make sense right? since the tape noise wouldnt be introduced yet) or between the tape and the output to the mixer?
 
Kasey said:
How does noise reduction on Reel to Reel's work? as in... where is it placed in the chain within the machine? Is there noise reduction between the input signal and the tape? (that wouldnt make sense right? since the tape noise wouldnt be introduced yet) or between the tape and the output to the mixer?
I've just realised that I don't really know. DBX-1 (which the TSR uses) and Dolby are double-ended noise reduction, which means they are effectively encoding the signal in a form which is less susceptible to hiss before it goes to tape and decoding it again on playback. In DBX-1 it's done with a compressor/expander pair, but exactly how this helps is unclear to me.
DBX-II is a single-ended noise reduction system, which means you stick it on the end at playback and it filters out the hiss (and probably everything else too).

In practical terms, yes, it sits between input and recording, but there is also a decoder at the other end - you get quite interesting results if you decode a non-DBX recording or forget to enable decode on playback (apparently some people use this effect as a poor-mans Aural Exciter). It's like Dolby-B on a cassette deck.

In the TSR-8 the signal routing is done internally. The main problem is if you forget to turn DBX on, which must be done each time the machine is powered on, because it's microprocessor-controlled and not a rocker switch like on the Fostex A8. The price of progress, eh?

On earlier Tascam machines like the '38 and the 80-8, you have to buy an external DBX rack, typically a pair of DX-4Ds or a single DX-8 if you can find one. The signal routing is a bit more complex with these, but although I have some I've never used them since I got the TSR instead of the '38, so I'm not entirely sure how the routing works.

If you're dead-set on Dolby, you could get 8 channels of that and 'forget' to turn DBX on - of course you'd have to manage the signal routing yourself.
 
jpmorris said:
I've just realised that I don't really know. DBX-1 (which the TSR uses) and Dolby are double-ended noise reduction, which means they are effectively encoding the signal in a form which is less susceptible to hiss before it goes to tape and decoding it again on playback. In DBX-1 it's done with a compressor/expander pair, but exactly how this helps is unclear to me.
DBX-II is a single-ended noise reduction system, which means you stick it on the end at playback and it filters out the hiss (and probably everything else too).

In practical terms, yes, it sits between input and recording, but there is also a decoder at the other end - you get quite interesting results if you decode a non-DBX recording or forget to enable decode on playback (apparently some people use this effect as a poor-mans Aural Exciter). It's like Dolby-B on a cassette deck.

In the TSR-8 the signal routing is done internally. The main problem is if you forget to turn DBX on, which must be done each time the machine is powered on, because it's microprocessor-controlled and not a rocker switch like on the Fostex A8. The price of progress, eh?

On earlier Tascam machines like the '38 and the 80-8, you have to buy an external DBX rack, typically a pair of DX-4Ds or a single DX-8 if you can find one. The signal routing is a bit more complex with these, but although I have some I've never used them since I got the TSR instead of the '38, so I'm not entirely sure how the routing works.

If you're dead-set on Dolby, you could get 8 channels of that and 'forget' to turn DBX on - of course you'd have to manage the signal routing yourself.
First off, dbx type II is a double ended nr system just like type I, the only difference is in the frequency range that it works within.

All of the Dolby nr systems are double ended as well with Dolby B being about the only one you can get away from using on playback if you don't mind the added hiss and high frequency boost. But, you're gonna be hard pressed to find many Dolby B nr units out there in the market with the exception of an old TEAC AN-60 unit which was only 2 channel so you'd need 4 of those puppies to have them in use on a TSR-8...good luck. ;)

Cheers! :)
 
Kasey said:
why is it necessary from input to encode? Theres no tape hiss yet correct?
What noise reductions systems try to accomplish is a preemptive strike on the noise which they know ahead of time will be created in the recording process so with Dolby based systems, they try and built up the signal in the areas where the hiss is known to reside and get it into a state where on playback, the signal is louder then the hiss and thus less audible. The follow up playback process is to now reduce the signal in an equal but opposite amount to what they were boosted going in so as to reassemble a reasonably flat outcome to what went in.

dbx based systems also employ a similar amount of boost and cut but only in the high frequencies and to manage the rest of the bandwidth, dbx also makes use of averaging or RMS compression and expansion to get the entire range of frequencies over the bar height of where the hiss resides and then do a reverse process on playback to reenact the original signal.

Where dbx excels over Dolby is that it actually preserves the original dynamic range of the original signal, provides broadband noise reduction in the neighborhood of 40db compared to 10 ~20db with Dolby systems, reduces print through effect as well as improves adjacent channel crosstalk! These are all very, very good things and especially so on narrower track width formats like the TSR-8.

Where dbx occasionally falls a bit short of perfection is in the way it encodes and then decodes what it just encoded. This is generally a subjective process that many users can totally live with in that it's really hard to remember exactly how dynamic something was or wasn't in the first place and as such, the slightly altered result is usually very tolerable.

Is any of this sinking in yet?

Cheers! :)
 
The Ghost of FM said:
First off, dbx type II is a double ended nr system just like type I, the only difference is in the frequency range that it works within.
Oops, my bad.

All of the Dolby nr systems are double ended as well with Dolby B being about the only one you can get away from using on playback if you don't mind the added hiss and high frequency boost. But, you're gonna be hard pressed to find many Dolby B nr units out there in the market with the exception of an old TEAC AN-60 unit which was only 2 channel so you'd need 4 of those puppies to have them in use on a TSR-8...good luck. ;)
Cheers! :)

I was thinking he could get some Dolby-A channels, although each one would probably cost a similar amount to a TSR-8.
 
jpmorris said:
Oops, my bad.
No problem. ;)



I was thinking he could get some Dolby-A channels, although each one would probably cost a similar amount to a TSR-8.
Dolby A cards would be a sweet solution for sure but, yes, they are a bit expensive as well as being available only in balanced +4db and also require very exacting calibration to work seamlessly with their host recorders.

dbx, to me at least, seems like the more practical alternative for us home recordists.

Cheers! :)
 
Kasey said:
But now i've learned that it has DBX noise reduction as opposed to Dolby, which ive been told is better that DBX. Is it true that DBX is known to muddy the sound? I'd love a reel to reel but i'd prefer to not have to choose between tape hiss or noise reduction mud.

The other alternative not yet mentioned would be to get a wide track machine and simply not use noise reduction. Wide track machines have S/N roughly somewhere in the 62-68 dB range, depending on various factors, with a dynamic range up around 80 dB or so (and of course a lot of resolution down below the noise without having to dither!). That's really not too shabby if you set levels carefully. Analog life is simpler when you don't mess with noise reduction.

For example, it shouldn't be too hard to find an Otari 1" 8 track these days. There's probably folks looking to sell. If you're anywhere near Nashville, MCI would make sense, since Randy Blevins is there and has tons of MCI stuff. There's a guy on the Ampex list right now looking to part out or otherwise get rid of a functioning AG-440C 1/2" 4 track, if that would be of interest. I just sold my 3M 1" 8 track about a month and a half ago to a fellow in Hawaii, of all places, but I still have a 3M 1/2" 4 track. No need to be thinking about noise reduction on either of those machines.

Good luck,

Otto
 
That's a worthy and good post, ofagen. ;)

The whole original selling points of the 1/2" 8 track recorders was tape economy along with the machines themselves also being considerably less expensive then the fully professional recorders of the day which offered far more robust construction standards in the mechanics and electronics.

The use of noise reduction systems became critical with those narrower track formats because of the limitations of those thinner tracks and all the more so as you add more tracks and the collective noise from them.

The recent tape manufacturing crisis we experienced has now also seen the price of blank tape rise pretty significantly and in the end, the mighty dollar always factors into the buying decisions that we have to make.

Many years ago, I had the opportunity to visit a smaller professional recording studio that did a lot of work for the film and television industry, mostly doing voice over work for commercials and narrative audio for documentary work.

They used a couple of different machines from Studer and Ampex which were 1/2" 4 track and 1/4" full track mono and there too, they used no noise reduction and didn't really need it as they kept their levels quite hot and also made use of compressors going in to keep things fairly steady for this specialized work that they did there. They worked in the way they did because there were very few tracks and because of the levels that they drove their recorders at. They also made extensive use of physically editing tapes and using blank leader tape in-between the actual speaking parts as a means of getting the dialog to line up with the needs of the production and to also keep the noise down to zero in-between the dialog lines. This was tedious, highly skilled work that many of us modern souls would shutter at if we were faced with a similar set of responsibilities.

When we look at other machines with more tracks, even in the professional realm of 1" 8 track or 2" 16 or 24 track, tape hiss can still be a very real issue we have to contend with if we don't pay very strict attention to our levels. The use of Dolby A and latter Dolby SR did factor into the equation of many professional studios that had the budgets to incorporate these NR systems onto their recorders. This was especially true and necessary as soon as we are talking about recording something different then straight out rock music that plays naturally at constant loud levels that would allow us to get away without using noise reduction. As soon as we want to record multiple tracks that have more dynamic parts to them, even with these wider format machines, noise and noise reduction systems will still be an issue we need to address.

Cheers! :)
 
I had considered 1" 8-track as a suggestion, but dismissed it since the tape is amazingly expensive and the machines are very rare. Also, I like semi-portable machines - the only one of those is the Brennell Mini-8 which is like hen's teeth even in its 'home country'.
You might be okay with a gas cooker-sized machine and paying >60 pounds a pop for tape, and if you are, the results should be above and beyond what you'd get from 1/2" with noise reduction. It's something I'd like to do myself if I could justify the cost.

(I now have the two 'new' TSR-8s - They have both been molested to some degree. Screws and washers nicked from them, the hex screws on the head cover replaced with crosspoints and so forth. Now I know what a knackered head looks like too :eek: Still, it should be an interesting project to get one of them up to spec. I'll keep the other for spares.)
 
The Ghost of FM said:
When we look at other machines with more tracks, even in the professional realm of 1" 8 track or 2" 16 or 24 track, tape hiss can still be a very real issue we have to contend with if we don't pay very strict attention to our levels. The use of Dolby A and latter Dolby SR did factor into the equation of many professional studios that had the budgets to incorporate these NR systems onto their recorders. This was especially true and necessary as soon as we are talking about recording something different then straight out rock music that plays naturally at constant loud levels that would allow us to get away without using noise reduction. As soon as we want to record multiple tracks that have more dynamic parts to them, even with these wider format machines, noise and noise reduction systems will still be an issue we need to address.

Cheers! :)

Ghost, I'm not sure what your actual experience working with these machines is, but I've spent at least the last decade recording almost all my projects on wide track machines up to 1" 8 track, and I honestly would say that noise was not an issue for me, and I'm kinda fussy. (edit: down below, I think I figured out a key difference: IEC1 eq!) Certainly the tradeoff between a little bit of hiss that is barely audible versus the advantages in the quality of sound, lack of dropouts and ease of maintenance and adjustment should be considered. And my projects are not all loud rock stuff, though some of it is. The one project recorded recently for another group was loud and therefore no problem as far as tape noise.

A few years ago, Bob Katz was suggesting that folks who wanted high quality sound for few bucks should forego cheap digital and use quality wide track analog (either 1" 8 or 2" 16 and forego noise reduction). He may have since decided that progress on digital gear makes that a viable choice as long as you watch every detail of the hardware and software you're using and don't do any more processing than is absolutely required.

2" 24 track does need noise reduction because each track has more noise and you have more tracks adding their noise together. Now, I haven't worked with 2" 16 track, but with proper track muting on an automated console, like I did on my 8 track with my little Mackie 1604, I'd think you could keep things in line.

As far as tape goes, I've still got 15 reels of 1" 3M 996 to sell, and I'm asking only $30 a reel for the almost unused reels and $40 a reel for the totally unused ones. If you run 15 ips IEC1 you can keep the high frequency noise down from what you get on NAB and still get about 32 minutes per reel.

Ah, that's right. That's probably a key factor in why I have had less problem with noise than some. The IEC1 eq curve makes better use of the high frequency dynamic range on modern tape. If you still use NAB eq, you get a lot more noise (lower S/N) in the 2K-6K frequencies where our hearing sensitivity peaks. Of course you also have more headroom, but it's probably headroom you don't need in that frequency range and with that darned NAB rec bass boost you have far less headroom where you need it down in the low end. I recommend that you look for a machine that can be set up for IEC1 eq when running 15 ips, whether wide or narrow track.

If you want portable, about the only large format wide track machine out there was the Stephens machine, and those are just about non-existent and parts are made of unobtainium, so you'd have to be braver than I to go that route. Mitch Easter was wanting to sell me his Stephens 16 track, but I just couldn't go that far out. I'm not sure if he ever sold it. But John used pro level unbalanced audio, so you saved the weight of all those input and output transformers. That's also part of why some folks consider them the best sounding decks ever. The transformers are a big part of the sound on those old machines like 3M, Ampex and early MCI. The Stephens decks had a cleaner sound because of the unbalanced audio, so if that's what you wanted Stephens was the way to go. I personally love the transformers in the path on guitar, bass and drum tracks. Most of the Stephens sound was just the sound of the tape. Anyway, a 2" 16 track in a console could weigh as little(!) as 100 pounds, and in a portable two piece rack it could be less, so they were popular in mobile trucks.

There are options for portable, small format, wide track machines. Mitch Easter has a cute little 3M M-23 1/2" 4-track in two portable cabinets that needs head relapping. I'm not sure if he'll ever actually sell it, but I think we had talked about $1000 or so, plus allocating $500 for JRF to lap the heads. The Ampex AG-440 series could be put into a rack, but they didn't have the kind of heavy cast deck plates the 3Ms have and aren't quite as solid and free of flex, so I wouldn't recommend them for portable use as much as the 3Ms.

Ahh, well, probably best to go the 1/2" 8 track route and get dBx type 1 NR. Or you could be a blasphemer like me and record some stuff on a 1/2" 4 track and port it over and record some more stuff on digital.

Good luck and have fun!

Otto
 
ofajen said:
Ghost, I'm not sure what your actual experience working with these machines is, but I've spent at least the last decade recording almost all my projects on wide track machines up to 1" 8 track...

Otto
Otto,

I currently have a 1" 16 track TASCAM MS-16 and a 1/2" 8 track, TASCAM 38, both machines equipped with Type I dbx.

I don't believe either of our feeling on the use of noise reduction are all the different when it comes right down to it. We both seem to saying that working without it is viable so long as we pay strict attention to levels.

As far as what one person's acceptance level of hiss is concerned, that might lay in the realm of personal preference and as such there is no one right or wrong answer in terms of it's necessity.

I do agree that the sound will remain more pleasing without it as any additional processing we throw in the way of the signal does present the chance to screw it up a bit more. The key question being what is better, less hiss with slightly altered fidelity or more hiss with unaltered fidelity?

To each his own. ;)

Cheers! :)
 
Back
Top