Skip to main content

dda -2dbu (-10/+4 ref levels)

Member for

21 years
Hello,
I've recently purchased a dda interface console and am trying to figure out the best way to connect it to my 002r.
The direct outs and inserts on the dda are -2 dbu but the ins on my 002 and external adc are -10 or +4dbu.

Here are the work arounds I've come up with. Which would you choose
1. Set adc input to +4dbu and loose a little noise floor at the preamp
2. set ADC input to -10 and run the pre hot (less headroom)
3. Add transformer. I've noticed a transformer option for the direct outs of the dda. Is this where one would add a transformer to step up level from -2dbu to +4 dbu. If so how much will this effect sound. Will i have to spend top dollar on these transformers?

Any and all help is much appreciated
thanks

Comments

Member for

21 years

Member Thu, 02/15/2007 - 09:22
Kapt.Krunch wrote: -2dBu DOES seem an odd level.

- Kapt. Krunch

Allen & Heath does the same thing on some of their consoles. They somewhat arbitrarily about 1/2 of their console outputs as -2dBu and the other half as +4dBu, which is really annoying... Its beyond me.... but I think it has something to do with some vestige of [tape] recorders wanting to deal with -2dBu rather than the hotter 4 signal.... I dont know... Typically the so called "tape outputs" are the -2dBu outputs on some of the A&H's.... which really sucks if your not actually using "tape" (what year is this?).

Member for

15 years 5 months

Boswell Thu, 02/15/2007 - 09:53
It might look as though -2dBu is a compromise half-way level between -10dBV (-8dBu) and +4dBu, such that it could be made to work with either at a pinch without too many headroom problems.

In practice, the -2dBu level was probably driven by engineering and marketing considerations. To achieve direct outs and insert outs at +4dBu with a 20dB headroom (i.e. +24dBu clip level), you need at least +/-20V (preferably +/-24V) rails in your channel strip, giving a peak current of around 30mA into 600 Ohm loads. At the time these consoles were designed, this was a tough spec to meet while keeping distortion figures and thermal dissipation under control in a budget product. Halving the voltage from +4dBu to -2dBu, thereby quartering the power, makes a significant difference to the designer. It also put a marketing distance between these consoles and the fully-professional ones where +4dBu into 600 Ohm is standard on all outputs. Labelling the -2dBu outputs as "tape" outs softens the effect of a weaker spec.

Member for

15 years 11 months

RemyRAD Thu, 02/15/2007 - 10:12
Reference levels can be humongously confusing for most. Everybody's example as set above, is both correct and incorrect. How can that be? Because the Japanese have tried to reinvent the wheel their way!

For instance. The Teac/TASCAM 0 DB, -10 DB output reference level, is NOT referenced to .775 volts! Teac referenced their -10 DB output with reference to 1 volt! And so, their reference levels do not correlate to American/European reference levels properly.

As Boswell indicated, I would go for the headroom, especially for popular recording, where signal to noise ratio is not as big a factor as it is with fine arts classical music. Headroom is the difference between retaining the transients or clipping them, as inexpensive equipment so often does.

Head engineer (who's got the pot??)
Ms. Remy Ann David

Member for

15 years 5 months

Boswell Thu, 02/15/2007 - 11:17
RemyRAD wrote: The Teac/TASCAM 0 DB, -10 DB output reference level, is NOT referenced to .775 volts! Teac referenced their -10 DB output with reference to 1 volt! And so, their reference levels do not correlate to American/European reference levels properly.
That's why we write it as -10dBV, the "V" standing for 1 Volt. The original 0dBm was the level required to push 1milliwatt into 600 Ohms, which came out as sqrt(600/1000) Volts = 0.7746V r.m.s. When 600 Ohm terminations went out of fashion, the same voltage level was retained, but now into a relatively higher load impedance, so the reference voltage level was re-named 0dBu (u = unterminated). There is approximately 2dB difference between 0dBV and 0dBu, making the difference between -10dBV and +4dBu approximately 12 dB or a factor of 4 in amplitude.

Member for

15 years 11 months

RemyRAD Thu, 02/15/2007 - 13:12
Yes and that's why it's so confusing to most. But the unterminated "dbu", value is in fact generally terminated into at least 10,000 to 50,000 ohms. Just not 600 like the front end of the UREI 1176LN's.

You wouldn't believe how many people I've talked to, that couldn't understand the problem with the underwhelming sound they got, while patching their 1176's, LA3's, LA4's, into their inserts on a Soundcraft, Allen and Heath, Mackie or, any console that wasn't an API, Neve, SSL, Electrodyne, Sphere, Quad Eight, OK MCI also, etc.. And that's because those wonderful old limiters can really only be utilized in a control room that was intended for 600 ohm studio operation and many people here don't know that. And none of those inexpensive consoles inserts can be used with any 600 ohm audio stuff. Of course that's not true for the later non-transformer op-amp balanced input 1176/1178 silver faces, which work fine with your kids console.

I'm not even going to talk about the equipment that is supposed to have a 600 ohm resistor to terminate the output transformer, for a proper and flat response, like my Neves and API stuff. Of which there's lots of.

Unterminated at 51
Ms. Remy Ann David

Member for

21 years

Member Thu, 02/15/2007 - 13:51
thanks very much for all of your replys. I am learning so much. Really really great stuff
I realize i did leave out dBV when refrenceing -10. My mistake.

So so far everybody votes setting input of adc to +4dBu. I'm not a big fan of using attenuators on -2 dBu console feeding -10 input on 002 and adc becuse i have a lot of outboard gear that's +4 as well as all my outputs during mixdown. What sort of differences in performance are there between +4dBu and -10dBV? Are attenuators more transparent than transformers?My guess is yes. I havn't thought to much about running everything at -10dBV using attenuators. Any opinions on this?

To be honest I'm a bit worried about the extra noise that will be introduced with the level missmatch. I should have mentioned before that I do a lot of recordings with over 50 tracks and as many as 80 (thanks to quadzillas), compressing the hell out of room mics and such as well as some classical and acustic stuff. 6 dbs or so of noise per track really adds up when you get up there in track count.

Any comments on the transformer option. Am I way off here. Would the transformer option on the direct out of the DDA be able to step up level from -2dBu to +4dBu with the right transformer? Or is the transformer option only for sound perposes?

Member for

21 years

Member Thu, 02/15/2007 - 14:00
rambelam wrote:

To be honest I'm a bit worried about the extra noise that will be introduced with the level missmatch.

Any comments on the transformer option. Am I way off here. Would the transformer option on the direct out of the DDA be able to step up level from -2dBu to +4dBu with the right transformer? Or is the transformer option only for sound perposes?

Transformers = potential signal distortion but no thermal noise increase.

Resistor pads = thermal noise increase but no signal distortion.

Member for

21 years

Member Thu, 02/15/2007 - 17:39
thanks for clearing that up mises. Mises, when you say "potential distortion" is that to say there are transformers for this aplication that won't effect the signal enough for it to be noticable or negative? If so do you know aproximatly how much will I have to spend for this transormer?

I think i should clarify that the DDA console has an option to add transformers in the mixer itself at the direct out not somthing i would be plugging the dir out into.

I'm at a bit of a stand still not knowing if decent transformers are in my price range or if they're are the way to go.

If someone could fill me in to the performance differences between -10 dBV and +4dBu it would be much appreciated.


soon i will start a new thread with better questions now I have more info

thanks again everyone

Member for

21 years

Member Thu, 02/15/2007 - 18:47
rambelam wrote:

thanks for clearing that up mises. Mises, when you say "potential distortion" is that to say there are transformers for this aplication that won't effect the signal enough for it to be noticable or negative? If so do you know aproximatly how much will I have to spend for this transormer?

I think i should clarify that the DDA console has an option to add transformers in the mixer itself at the direct out not somthing i would be plugging the dir out into.

I'm at a bit of a stand still not knowing if decent transformers are in my price range or if they're are the way to go.


Transformer designs for audio, to be honest is not my specialty. I do digital communications theory, not really circuit design... but my knowledge of circuits is probably adequate enough for general advice.

Generally the more desireable transformers for audio are quite expensive (rediculously so... but thats the economics of market value).

I'm not even sure if there are any stock transformers out there specifically designed to convert from the -2dBu level up to +4dBu. If their were... I would buy them and I really wouldnt bother worrying about whether they were astounding quality or not.

Good transformers cost upwards of $120 - $200+ apiece (just from what I've seen)... so you couldn't realistically be putting one of those babies on every single channel on your board... though I suppose you could put it on just a few select channels... depends how many -2dBu outoputs you have.


I would not convert downwards to -10dBV... I'd step up to +4dBu. Just my little opinion on that. If I were going to pursue a route.... thats the way I would go, and I would stick with it... and just buy whatever it takes to make that happen. I think stepping down is not really a great idea unless somebody has a counteropinion on that. Most of your equipment wants to operate in that region anyway... so thats why I say go with that.


Did you mention something about your DDA console has an option for some transformers that you can put on the direct outputs (inside the console)? I would investigate what that costs. Forget about "custom" transformers.... Don't go there unless you want to seriously get into that game, and thats just gonna cost mucho bucks.

Member for

15 years 11 months

RemyRAD Thu, 02/15/2007 - 19:51
Your output is -2 at the insert and/or direct outputs largely because, most of the other devices, they intended for you to patch in, have input level controls. So it's not uncommon to take the insert and/or direct output into a compressor or other outboard device at that level.

Gain staging does not keep everything at the same levels. Levels are dropped and boosted throughout the internal signal path of all consoles, recorders, equalizers, etc.. So I think you should do nothing.

Your theories of the loss of signal to noise ratio on the order of +6 DB over 50 to 80 channels does not hold up. Good engineering is good engineering. We managed to do the same thing, all on analog tape without noise reduction for many, many years, with a signal-to-noise ratio far worse than the electronics in your unbalanced DDA console. So, 24 track wasn't just 24 track as we bounced and transferred many things that made some of our tracks, on the multitrack, over three generations down before the mix even started! Automation, noise gates and riding the gain is how we made our stuff sound good.

So again I say, LESS IS MORE and KEEP IT SIMPLE STUPID. You don't need resistive pads to convert to -10 and you certainly don't need transformers unless your audio seems anemic and needs more iron or, a balanced output? Electronics to boost the -2 output level is more electronics you don't need to feed your signal through. The output transformer option for your console is also not completely necessary unless, of course, as you surmise, they included a "step up transformer". It raises the level, when the windings on the secondary are greater than the windings on the primary side (which can also be run in the opposite direction).

API consoles all had transformer outputs that included 3 separate secondary windings. So you had a turns ratio of 1: 1 for a +20 DBm output or, 1: 2 for a +24 DBm output for, 1: 3 or a +28 DBm output drive capability. The other advantage is, as in my API 3124m units utilizes 2 separate isolated outputs from the same transformer. Great to use as a 1 x 2 active splitter or dual isolated outputs.

While Transformers can and do introduce distortion, the largest problem is the low-frequency distortion and the saturation of the core, that occurs with cheap Transformers. Good transformers can actually benefit your audio and/or color it in a pleasing way. Transformers are either specified as low-level microphone or high-level line devices. You wouldn't want to use a line transformer for microphone purposes and vice versa.

If transformers were bad, Neve consoles would never have cut a hit much less a demo. My audio through my Neves go through from 8 to 10 transformers from microphone input to bus output, except in the case of the individual modules which will only force you through 2. And you never heard anybody complain about that.

Tranny heaven
Ms. Remy Ann David

Member for

21 years

Member Mon, 06/16/2008 - 15:02
RemyRAD wrote: Ummm, you really can't run a classic 1176 on the inserts of most modern-day mixer boards. The originals had 600 ohm terminated balanced bridging inputs. Now all the latter models which were silver faced utilized a 10,000 ohm, balanced integrated circuit chip, transformer less input. Those you can patch on inserts of most budget consoles.
I can only disagree with that. Most of these "budget consoles" have the capability to deliver +18dBu into 600 ohms, which is plenty even for a "classic" 1176, unless you want 30dB+ compression.

Member for

19 years 10 months

Kev Mon, 06/16/2008 - 15:16
well I disagree

and it comes down to how much headroom you have chosen to have in those "budget consoles" OR the modern DAW

very difficult to get typical recorded program up into the sweet spot of things like transformered 1176 and LA2s

I have been asked a few times to mod these units to add just a little gain at the front end to suit the insert sends ... (or add a line driver to the send)
...
then PAD the output back a little to bring back into the insert return

I do think it is at the output that brings much of the sound people are looking for
so terminating and the PAD is probably the first thing to try.
The Gain Make Up Amp in these units often have enough to get you going but the risk is that you may get some noise with it. That's why finding 10db (or a little more) of signal at the front can help.

Member for

15 years 5 months

Boswell Fri, 02/16/2007 - 03:06
If somebody offered me 6dB more headroom at no penalty for most of the type of material I was recording I'd bite his hand off. It's one of the reasons that engineers run the gain trims lower and the faders higher on consoles with limited headroom.

So I would go with driving the +4dBu inputs of the ADC or 002r. You just might consider having a few inline XLR 6dB attenuators in your bag for the rare occasion that you come across very quiet material and need the extra dynamic range. That would be cheaper than splashing out on high-grade 1:2 transformers that you would scarcely ever use.

By the way, the thermal noise of a 600 Ohm resistor is about 129dB below +4dBu over a 20KHz bandwidth, so is not a factor.

Member for

21 years

Member Mon, 06/16/2008 - 16:23
Kev wrote: well I disagree and it comes down to how much headroom you have chosen to have in those "budget consoles" OR the modern DAW
very difficult to get typical recorded program up into the sweet spot of things like transformered 1176 and LA2s
It is your absolute right to disagree, but I know for sure that an 1176 is not expecting more than 0dBu to operate properly. When I installed the cutting room at Barclay Studios in 1976, we had a Neumann mixer, which ran on a single 24 V rail and used LF356's, hence offering only +16dBu output capability on the unbalanced insert points. The 1176's (with xformer inputs and 600 ohms constant-z attenuators) had enough gain to cope with that.

I do think it is at the output that brings much of the sound people are looking for so terminating and the PAD is probably the first thing to try.
I agree with that; The output xformer in the 1176 wants to see a 600 ohm impedance because it has a relatively high leakage inductance that makes it ring if not properly damped.

The Gain Make Up Amp in these units often have enough to get you going but the risk is that you may get some noise with it.
No, because the input attenuator is before the input preamp. Having less attenuation before the input preamp doesn't change noise.

Member for

19 years 10 months

Kev Tue, 06/17/2008 - 00:19
no no

I was referring to signals that might be for you, a very low signa,l with a great deal of headroom
and could be much lower than 0dbu

and I can see how a single 24v rail with around +16dbu unbal would also be quite low for a max signal

some people seem to want to see 20db+ of headroom


as for the gain noise / issue
to CRANK the gain make up amp can make noise for within the 1176

remember there are 1176's out there that were NOT the LN models
Low Noise

and I was referring to these things genericaly and including LA2's .... these things can be living being that have lived hard lives
and it is hard to find two the same

and for things like the Pendulum compressor +16dbu unbalanced would not get it into it's sweet spot and a PAD would be needed to bring it back into one of these desks

You seem to know enough about your gear to make your own choices and the comments here from Remy and I are not really directed at you.

IF all of this gets people thinking about how they interface there gear then this is a good thing.

While we are here on the 1176 ... the manual does give some good notes on how to interface it
.... read the manual

you can find PDF's on the net ... my site might even have one

Member for

19 years 10 months

Kev Mon, 12/28/2009 - 14:31
dpd wrote: ... what the rest of the world uses for a voltage reference: ... - ya' know?
too much history for that

Video is a 1Vpp world with 75ohm
but now with digital that doesn't seem so important anymore

the RF world has it's own issues

The 1Vrms has no regard for load
1 mW onto 600 ohms ... because that was the standard = .... you can google that

things change
digital interfacing had the chance to solve level issues as all is referred to 0dbFS
but
now we have clock issues

it's all fun

Member for

19 years 9 months

Davedog Fri, 02/16/2007 - 10:47
Boswell wrote: If somebody offered me 6dB more headroom at no penalty for most of the type of material I was recording I'd bite his hand off. It's one of the reasons that engineers run the gain trims lower and the faders higher on consoles with limited headroom.


I LOVE all you 'real' engineer types. This is great stuff here. GAFF STICKY THIS!

This Boswell quote I have hi-lighted is a key to gain staging for all you guys and gals with the Mackie/Allen Heath/ Soundcraft(Ghost and below) type of consoles.

I get what I need on my Ghost exactly like this though I do regulate the input voltage at a steady 128 vac. Its within specs so I dont worry about damage per se. The rail voltage comes up a touch and the pres get creamy and punchy while the EQ has an increased bite and the sweep is much more effective.Even surgical if need be.

I gotta agree with Boswell on this entire point. Take the headroom. Buy some outboard(or build) attenuation for needs...Like Remy....dont do nuttin. Your outboard likely has attenuation built into the front ends and makeup gains going out. While you might have to be more astute in your balancing act with this setup, I see no reason it should cause problems in recording materials and certainly with high track counts the added headroom would be to your advantage.

Did I mention that I LOVE you engineer folks?

Member for

21 years

Member Fri, 02/16/2007 - 11:34
Davedog wrote:

I LOVE all you 'real' engineer types. This is great stuff here. GAFF STICKY THIS!

Did I mention that I LOVE you engineer folks?

Now what fun would it be, being an [electrical] engineer, if we can't nitpick at little things like adding 449.7 nanoVolts of thermal noise (making the cumulative AWGN a whopping (-116) to (-122) below +4dBu.

Somebody has to worry about a few nanoVolts or microVolts of noise 8)

Member for

21 years

Member Fri, 06/13/2008 - 02:19
The actual reason why the operating level in many british mixers (and subsequently many japanese and US manufacturers) is -2dBu is that circuitry using commonly available IC's provides an actual clipping point of ca. +20dBu. Running the board at +4dBu internally would provide only 16dB of headroom, which was deemed insufficient, which is true in live applications, and that permeated to studio mixers. That's why they chose to run all internal signals at -2 dBu and interface with the external world via balancing/debalancing stages providing 6 dB gain/attenuation. The commonly used EBOS (Electronically Balanced Output Stage) take advantage of 6 dB increase of output capablity (typically + 26 dBu into a balanced load). Some high-end mixers used balanced inserts running at +4, but most of the products had unbalanced inserts at -2, in order to save circuitry.
Anyway, most of the outboard gear can make up for the difference. On a 1176, you just crank up the input and reduce output accordingly, on a 160, you just set the threshold 6 dB lower. For EQ, it shouldn't matter, except if it's very noisy.
In most cases, it will be transparent to the user; that was not the case when Tascam ran their mixers at -10 dBv (-7.8 dBu) from beginning to end.
You may have difficulties with a Fairchild 660 or 670, just as you would with a mixer that runs its inserts at +4, because the 660 was designed for 600 ohms input source AND 600 ohms output load (impeadance-matching), which no modern gear supports.
Most equipment nowadays is designed for impedance-bridging, i.e. low output impedance (10-100 ohms) and medium-high input impedance (5-50 kohms).
x