Dolby or dbx: which allows more saturation?

[shedshrine]

My Tascam porta 05, which has Dolby HX pro noise reduction, took a spill in the late '90s and hasn't worked since. Nonetheless, it's recordings live on, and seem to have a more tapey sound than its 424 mkII dbx brother.

Any definitive answers? I know when I slam the 388 too hard, the dbx tracking suffers (which can be a neat effect, but I digress). I can't remember if dolby is more forgiving, and therefore accepting of higher levels.

Hope everbody had a merry happy joy joy.

 

[shedshrine]

whoops, didn't know hx pro is not considered noise reduction.

"Dolby HX Pro
Not a noise reduction system, Dolby HX Pro makes it possible to record loud musical passages with fewer high-frequency losses and less distortion. It is available in better cassette decks and is also widely used by the recording industry to improve the quality of prerecorded cassetes."

 

[shedshrine]

Not trying to get caught up in semantics. It' obviously more important to be pressing record on your well maintained 4 track. I ask this post's question as I answered someones question about the need for a compressor with a four track, and wanted to provide the best answer.

Thanks,
M

 

[shedshrine]

elswhere on the web (stevehoffman forum), though doesn't speak specifically to saturation:


"The sad truth is, the better format/technology does not always win. Just look at that Dolby vs. dbx battle in the 80's. dbx was far superior to Dolby, but it lost the war due to inadequate marketing."


"I thought it was mainly due to marketing. dbx was not incorporated into consumer cassette decks until it was late in the game. As such, most pre-recorded cassette tapes were almost never dbx-encoded. Once Dolby had established the insurmountable lead, the battle was over. I think this example is very similar to Mac OS vs. MS Windows.

dbx failed because:
1) it was rarely marketed outside of Technics and TEAC
2) it required excellent quality tapes (usually metal or high end high bias) matched to the deck - i.e. given the 2:1 "compansion" errors in the deck or tape were magnified.
3) was not backwards compatible with existing cassette players. i.e. Dolby B was similar sounding on non-Dolby players. dbx sounded like AM radio on non-dbx decks. I had a dbx walkman (Panasonic RQ-J20X) and was fantastic. Too bad my tapes wouldn't play on anyone else's boombox or walkman or stereo unless it had dbx. I switched to Dolby B for compatibility.
4) Prerecorded tapes could not use dbx unless they a) used far better tapes and shells b) slowed down the duplication speed

It was doomed to fail for mainstream. It did sound great, though."

 

[A Reel Person]

Fortunately,...

I've been happy with dbx NR on cassette, myself being a Technics and TEAC user from early on.

I agree that dbx failed in marketing terms, but was technically & sonically superior to Dolby despite it.

MO is that Dolby/B/C would never actually kill the hiss when engaged, but did other less desirable things to the sound. On the other hand, dbx completely adulterated the sound, but did amazing things in actual NR and headroom extension, the likes that Dolby could only wish for.

I'm pretty sure both dbx & Dolby/B/C NR will mistrack when pushed out of it's normal operating range. Dolby might be more forgiving, (but IMO why bother?)

For my money, I'll stick with dbx.

 

[shedshrine]

(..scrambles to give Dave as many rep points as possible)

 

[Torin Krell]

I agree with you Shedshrine as to why dbx was not as successful as dolby.

HX Pro is an additional circuit added to most of the "newer" and better cassette machines. It varies the amount of bias to reduce high frequency saturation. Bias is a very high frequency signal (usually 40khz+) added to make the the recording process more linear and therefore reduce distortion. When there are a lot of high frequencies in the program material these high frequencies actually act as additional bias which lowers the high frequency output of the cassette tape. HX Pro senses this and reduces the bias signal in proportion to the highs in the program material thereby extending high frequency response and headroom.

Dolby is more pleasant sounding when overloaded. Dolby S with HX Pro is of course superior to dolby B & C and the professional and expensive dolby SR is the best of all.
In my experience DBX rarely sounds great when overdriven except for unusual low fi effects.
Personally if I wanted to overdrive a track I would rather use the sync capable track where I could disable the noise reduction whatever it is. If you are driving a track hard tape noise should be minimal anyway.
This is easy to do on many multitrack machines including the 424 and 388.

 

[cjacek]

Yeah, I love properly calibrated dbx noise reduction. It makes narrow track sound better than it should. Who would have thought that cassette or 8 tracks on 1/4" tape can sound really, really good? dbx makes that possible and I agree with the above, that dbx wasn't given it's fair shake at the market 'cause it really does sound better than any other format, when levels are respected.

 

[Bobby Darko]

true true true.
I choose hiss over dolby and dbx over both any time..

 

[ofajen]

A good question.

The shortcomings of Dolby noise reduction were enough that I abandoned their use on my Nakamichi machine shortly after buying it in 1981 and I never made a two mix on it (starting in 1983) with any noise reduction. I prefer and expect a consistent noise floor and consistent dynamics, even when monitoring with headphones on signals recorded direct from synthesizers and drum machines, and Dolby has a hard time with that.

The shortcomings of dbx were one of several reasons why I abandoned narrow track machines in the mid '90s, and I've used wide track machines ever since. Once I started using IEC1 eq at 15 ips on the wide track machines, I never had any more need for noise reduction.

Cheers,

Otto

 

[Beck]

A couple thoughts to add to the good info so far...

First off… I am an NR advocate and satisfied user.

Unfortunately, you can’t get significant conventional tape saturation with DBX because it treats the entire signal (broadband), is always working and adds about 30dB of headroom. But IMO, DBX is warmer and fuzzier than Dolby unless you push it too hard; then it presents harsh and non-musical distortion like digital when pushed above full scale.

Dolby (A and B), on the other hand, treat upper frequencies (selected bands) and disengages when the signal is hot enough to mask tape hiss, so it literally moves out of the way and allows tape to saturate as if there is no NR. Dolby C is more complicated in that it has anti-saturation and spectral skewing features, which later evolved into Dolby SR (Spectral Recording).

As already mentioned Dolby HX Pro is basically a dynamic biasing feature. Since high frequency content in the music itself helps bias the tape, HX Pro adjusts by backing off bias to the same degree that high frequency content is contributing bias. HX Pro is a very cool system, and although it is not NR it does contribute to a better dynamic range by expanding headroom. Both DBX and HX Pro prevent saturation by design.

Personally, I’m a DBX fan when it comes to multi-track NR, but here again it’s going to depend on the type of music or specific instruments you’re recording and your equipment. I can say it works exceptionally well on a lot of the later (mid 80’s - 90’s) Tascam stuff. Tascam R&D worked closely with dbx Incorporated in the 80’s to fine-tune it for their machines. Tascam DBX, particularly on products like the 246, 238, TSR-8 and MSR-16 is really a different DBX. YMMV with other products and/or outboard DBX units.

I like Dolby C as well (for mastering to half-track), but same situation as above… it’s not implemented well in every device that has it. Outboard Dolby C is hard to find because Dolby didn’t license it as a stand-alone option for very long. I use the Sony NR-500 Dolby C unit and can highly recommend it. The Nakamichi NR-200 is also an excellent stand-alone unit with a good rep; it has both Dolby B and C. (By the way, I recently saw a Sony NR-500 go for $9.99 on eBay. I couldn’t believe it. If I didn’t already have two I would have bought it. People just don’t know they’re out there or what to look for. It should have fetched $100.00+ easily… kinda sad.)

No NR scheme is transparent, but I’ve found them indispensable in the work I have done over the years. But then I’m very particular about tape hiss and have been at war with it from my early recording days.

Calibration is critical and I’ve always been obsessed with keeping my machines in tip-top, including having the NR optimized. Tape selection is also critical in having success with NR. Except for a couple times when I ran out of tape, I’ve never used anything longer than C-60 length tape in a cassette portastudio in over 25 years, and always TDK, Maxell or Denon high bias… SA & SA-X, XLII and HD7, respectively. I keep the machine clean, and regularly degauss. I also demagnetize brand new tapes right out of the box before use… that goes for open-reel and cassette.

Yet, other competent recordists that are just as picky and have as much if not more experience don’t like NR, so there are plenty of legitimate reasons not to use it… except IMO on cassette multi-track, which would never have been invented if not for NR. They aren’t really designed to be used without it. If you have a cassette multi-tracker that allows you to disengage it, please don’t. I know there’re a lot of folks out there these days that see it as an option because there’s a switch. The switches are really to make the device more versatile… things like listening to consumer cassettes at normal speed that weren’t encoded with any sort of NR, or an incompatible NR… most commonly Dolby B. You can’t play those with DBX engaged, hence the switch. Track 4 or 8 on a cassette multi-track is commonly switchable so you can stripe it with sync code. Other than that the good normally far outweighs the negatives of even sloppy NR.

Another option is the single-ended “Smart gates.” These would best fit with a half-track mastering rig rather than multi-track. My favorite is the Behringer SNR-202 Denoiser (older model made in Germany and the only piece of Behringer equipment I own) It only works on playback, so you can record without NR and “Denoise” the tape during the mixdown phase, or from analog master to digital at the last stage.

In summery:
- Objections to NR of any kind are understandable because they aren’t created or implemented equally. Many DBX systems cause pumping and breathing artifacts and Dolby C as implemented on some multi-tracks can sound shrill and is vulnerable to mistracking. Dolby S and SR do better in most situations.
- Tape selection and proper calibration is often the difference between NR working as advertised and not working at all.
- Personally speaking NR has been a blessing for me
- IMO Tascam has done a better job than any other manufacturer to make DBX work
- Outboard NR units I can recommend: dbx 150X (DBX Type I), dbx 140X (DBX Type II for cassette or consumer ree-to-reel), Sony NR-500 (Dolby C), Nakamichi NR-200 (Dolby B & C), Behringer SNR-202 Denoiser (AKA Studio Denoiser MKIII).

Hope that helps

 

[ofajen]

Yet, other competent recordists that are just as picky and have as much if not more experience don’t like NR, so there are plenty of legitimate reasons not to use it… except IMO on cassette multi-track, which would never have been invented if not for NR. They aren’t really designed to be used without it. If you have a cassette multi-tracker that allows you to disengage it, please don’t.

As you can tell, Tim and I generally make different choices as to whether or not to use NR, but we definitely agree on this point. For a while back in the '80s, I used a very nice 4 track cassette multitracker made by Audio Technica. I think it was something like an RMX64. A very nice mixer and 4 track double speed cassette deck. There's just no way around the need for using NR on those multitrack cassette machines, because of the noise buildup if it's disengaged. Similarly, I was never really satisfied with the sound from my Tascam 388 without the dBx NR on and I don't think I did any tracks without the dBx, except on the sync track. That's part of why I went to wide track open reel machines. Operationally, I am glad to have machines that simply need to be calibrated to spec without NR with the option to add saturation (or not) in a very controllable way by setting the record level.

Cheers,

Otto

 

[Tim Gillett]

Noise reduction really started with Ray Dolby's Dolby A in the mid 60s but interestingly was not initially applied to cassettes but to pro grade studio reel to reel machines. So one of the first albums to be recorded with Dolby A was The Beatles Sgt Peppers. The machine mainly used was a Studer 1" 4 track.
The Dolby A gave a broadband noise reduction of 10db which today seems modest, but then on 1" 4 track the tape noise probably wasnt too bad to start with.
But later on, Dolby B and then C and S were applied primarily for cassettes. The problem was partly that the NR was applied to cassettes, notoriously unreliable for azimuth stability, and used by mostly home users not used to the rigours of regular, even daily tape machine calibration as in good recording studios.
Dolby puts not less demands on regular calibration but more. For the professional using machines designed for regular calibration it was fine. They just had to be careful with calibration. But for home users using cassettes, and on often machines with no easy way to user calibrate,even if they had the skills, it was problems just waiting to happen. In general terms the same applied to dbx.

Years ago I recorded a fair amount of location material to cassette with Dolby B because it was to be high speed duplicated direct to multiple cassette copies. Dolby B gave a good improvement in speech clarity especially useful when taking into account possible losses in high speed dubbing, and knowing most people would not replay with Dolby decoding anyway.
But if strictly needed later on, any of these original tapes can be carefully transferred in real time and correctly decoded, making proper allowances for any misalignments, and slight losses due to demagnetisation.
I have other live material recorded to cassette with dolby C and that, I can testify more than Dolby B, can go out of tracking very easily, and seems to always need some corrections in playback.

I have a high regard for both Dolby and dbx but the bottom line is the replay signal must be pretty close in level across the useable bandwidth to the original input signal for it to track properly. I think the vast majority of problems experienced with mistracking will have been due to poor record/replay linearity, not the NR circuitry itself.

With NR you never got anything for nothing. What might have been an acceptable non linearity without NR could be deadly with NR. It required more of the machine and its maintenance, not less. For many home users on a limited budget I think that was setting them up for failure. Is it any wonder so many just came to switch the NR off?

Cheers Tim

 

[Ritzy]

Excellent thread and great info guys! I think I may have to print this one out eventually and add it to my 'tome of wisdom' (basically a folder were I keep this kind of cool info)

I also demagnetize brand new tapes right out of the box before use… that goes for open-reel and cassette.

Just wondering, how does one go about demagnetizing tapes?


Also (and veering off on a slight tangent here), when tracking to cassette, does anyone "play out" and then rewind new cassettes before recording onto them? This would be to even out any tension irregularities, etc. in the wound up tape which could possibly cause slight fluctuations in tape speed during recording. Is this worth it, or wouldn't you bother?

 

[Beck]

Just wondering, how does one go about demagnetizing tapes?

I should have said degauss or erase technically speaking, although it's the same process. Erasing tape with a hand degausser removes any little pops and clicks that may have been picked up in later phases of production, transportation and storage. It also makes the noise floor as low and uniform as it can get.

Also (and veering off on a slight tangent here), when tracking to cassette, does anyone "play out" and then rewind new cassettes before recording onto them? This would be to even out any tension irregularities, etc. in the wound up tape which could possibly cause slight fluctuations in tape speed during recording. Is this worth it, or wouldn't you bother?

Every time since I can remember going back to the 70’s at least… but I fast-forward and then rewind each new tape before use. (I first started doing that with 3/4" U-Matic video cassette in the TV studio.) Yes it’s worth it, and this is one of those “Little things” I’ve talked about here now and then that helped me get excellent results with modest equipment. Stuff like this became a routine that I still do automatically when using cassette.

 

[Torin Krell]

Like Ritzy I always fast wind and then play a cassette tape all the way through before recording on it with my portastudio. For really critical recordings I play them through twice and clean the tape path before recording.

Several years ago I saw charts of tape shedding which indicated that even polished quality tapes have the most shedding in the first few passes. Playing a tape through results in less shedding after the recording was made and also results in a far more even tape pack that maximizes speed stability. Cassettes of course also have some mechanical parts inside which essentially make them part of the transport. A little "exercise" before a recording smoothes the slip sheets and other mechanical parts for better performance. Playing a tape through a couple of times before recording also significantly reduces level variations and dropouts that significantly add to noise reduction misstracking