1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

Adat output louder than analog output ?

Discussion in 'Recording' started by rboyart, Mar 3, 2010.

  1. rboyart

    rboyart Active Member


    I have a strange problem and i need some help.
    I'm using a DIGI003 with SSL XLOGIC ALPHALINK MADI-AX as A/D converter, and recently i discovered that my ADAT OUTPUT are louder than my ANALOG OUTPUT.
    I'm Protools, I sent 1K Hz/ 0 dBFS with a signal generator to ADAT OUTPUT 1-2[bleep] and then i plugged the OUTPUT of my A/D converter to ANALOG INPUT 7-8 and i don't have the same level, my input level is 3.5 dB more louder.
    But when i do the same thing with my OUTPUT ANALOG without using the A/D converter, i have the right IN/OUT level.
    So, i deduced that the problem come from the A/D converter but i don't understand it.

    Then, i tested another A/D converter and it's also weird, neither, i don't the same IN/OUT level if i use the ADAT OUTPUT,
    Could the problem be the ADAT OUTPUT of my Digidesign?[bleep]
    For the record, I'm using always the same cables.

    Has anyone else experienced this?


  2. Boswell

    Boswell Moderator Distinguished Member

    There is not a problem with the equipment. You need to understand levels. The ADAT output is digital, and digital signals are usually referenced to full scale (0dBFS). There is no analog level implied or associated with a digital signal, so when it is converted to analog by a D-A converter, the analog output will be determined by whatever output gain the D-A converter has been designed to use. As an example, if you put the same 0dBFS ADAT signal into an Alesis HD24 acting as a D-A and a Yamaha 01V96, you will get +19dBu out of the Alesis and +24dBu out of the Yamaha. Both are correct.

    With your Digidesign equipment, the analog full-scale of the DACs is set to be the same as the analog full scale of the ADCs, so connecting one round to the other should give the same amplitude. Use any other make or type of equipment, and the full scale will probably be different, so connecting that to the Digidesign input or output will give different results. All of this is correct behaviour, and you just have to plan your gain staging and headroom calculations accordingly.
  3. rboyart

    rboyart Active Member

    Hi, thanks for your quick answer! :)
    ok, so, i have to deal with my hardware i thought there was a normalization going on in the audio industry, but ok, i see there is no human understanding ;)
    I'm not very good about dBu level, but i try to maintain my audio signal paths in +4 dBu (1,228 volts) and i have calibrated my machines like this -16 dBFS=0 dBVU=+4 dBu, so when you tell me that the Alesis HD24 and a Yamaha 01V96 acting as a D-A , i will get +19dBu out of the Alesis and +24dBu out of the Yamaha, what exactly do you mean with that huge dBu level?
    So is normal, if I'm routing a signal out of Protools with ADAT OUTPUTs, to always think about what machine I'm going to and how many dB this machine is going to take or to add? that's correct?
    Thanks again for all your help,

  4. Boswell

    Boswell Moderator Distinguished Member

    You are correct in implying that level standardisation is much misunderstood.

    The standard professional nominal level for analog signals is +4dBu (1.228V rms). What varies is the headroom from this nominal figure to the point of clipping. So the Alesis I mentioned has 15dB of headroom and the Yamaha has 20dB, although both these pieces of equipment are specified as operating at a nominal level of +4dBu. And yes, it really does mean that a sinewave represented internally in digital form at 0dBFS would measure 6.9V rms out of the Alesis, 12.28V rms out of the Yamaha and 4.912V rms out of your equipment (which you say is calibrated to 0dBFS=+16dBu).

    Digital audio is referenced to 0dBFS, and because piping it between different pieces of equipment (via ADAT, S/PDIF, AES/EBU etc) is a lossless operation, it always results in the same levels (relative to 0dBFS) being received as were transmitted.

    If you were serious about level standardisation between all your equipment that performed A-D and D-A conversion, you might decide to reduce everything to the levels of the lowest-headroom device by inserting attenuators in the D-A outputs and amplifiers in the A-D inputs of any piece of gear that had greater headroom than the minimum. This would, of course, be crazy, but it indicates that you do have to think carefully about level interfacing when interconnecting any two pieces of equipment by an analog cable.
  5. rboyart

    rboyart Active Member

    OK, thanks Boswell!
    For the moment, I don't think if I'm going to standard all my equipments levels that performed A-D by inserting attenuators, but for sure, I'm gonna keep in mind all machine's levels in my head, before routing audio signals paths.
    Thanks for your useful advices, :)


Share This Page