Although I can trust myself to not run at 0 DBFS, I'm not sure other operators can be so trusted
Mark,An ADAT output can not be at +24dBU. That would be a voltage reference. In the case of this particular console that output could be set to utilize 24 bits (default) or it could be set to 20 bit or 16 bit. You would need to monitor this stream with a meter that could tell you what bits are being used to know how fully the stream is being utilized. Also, it would appear that you are limited to a max 48kHz sample rate by using ADAT out of the console and into the EQ since I don't see an SMUX setting for the EQ. I could have missed it though, not familiar with that piece.Lee
Maybe it was the levels you were seeing on meters, maybe it was not. It certainly was never above 0dBfs. There are no bits left.
What are you using to measure? If you are using a meter that averages then you were not reading real peaks, if you were using a meter that reads RMS then you were not reading peaks, if you were using a digital readout on the console then we have no idea what the actual meter ballistics are. You need to utilize a real peak reading meter if you want this exercise to be effective. Until then you are not comparing like readings or values.Lee
I'm going to disagree with this. A digital console designer can choose any numerical value they want to represent what is metered as 0 DBFS.
I am using the consoles metering. I agree that it is not perfect, but with a sine wave or pink noise as the stimulus, and at a constant level, I don't think the ballistics are as important as with dynamic music. Unfortunately I don't have any other meters available and the 01V96 manual doesn't seem to specify the exact type of metering, but I have ASSumed it is a peak meter, and agree that is a risk, but considering the steady state stimulus I don't think it is a huge issue.
And I'm going to disagree with you. It's true that what the meter shows for any given signal level is part of the design, but 0dBfs is defined as the point where you run out of ones and zeros. It is not up to the designer to change that fact. The whole point of using 0dBfs as a reference in digital consoles it so that the operator knows where their level is in relation to the brick wall that exists when the ones and zeros run out. In most pro gear "0", like the tone you use for setting up record decks is -20dBfs. That is the standard broadcast "0" tone level when setting digital to digital levels from consoles to decks. using -12dBfs as your maximum intentional level means you peaking at about 8dB above "0".Mac
...You do have other meters available if you have a measurement system. You can choose to measure a direct output from various points within the mixer by sending it to an AES output and metering the actual value on a peak meter. Pretty sure you could do this with SMAART, I do it with SpectraFoo all the time.Lee
I agree that the designer's chosen number used to "represent 0 DBFS at the output stage" is likely the max value the DA converters can handle without introducing distortion, but it would seem imprudent to not include some amount of numerical headroom above 0 DBFS in the processing pipeline so that. for example, a compressor or reverb that needs to multiply, carry the intermediate result, and then divide wouldn't overflow (clip, wrap around, or generate a NAN) on the multiply-carry even when the input signal was way below 0 DBFS.
The internal processing of the console is not the same as the output convertors. The output convertors are at best 24bit, the internal is at least 32 for most consoles, some higher. Once in the digital domain not exceeding that stage's capabilities is what is important, but 0dBfs is defined as the end of ones and zeros, and it happens whenever that stage runs out of headroom. Your fixation on optimizing your sound system around this spec is misguided. Optimize your system with real operating levels, which should be at least 12dB below 0dBfs.That is all. I'm through banging my head on the wall.Mac
Page created in 0.13 seconds with 23 queries.