Avid Pre calibration question

[jbreher]

The readying gear for work saga continues...

I have a digidesign Pre. From my understanding, the Avid Pre is merely a cosmetic update of this unit.

AAR, the User Guide mentions the ability to calibrate each channel. It is however rather light on specifics.

There is an onboard oscillator. The manual states that this oscillator will -- after temperature stabilization -- output a 1.228Vrms sine wave out a balanced dedicated 1/4" TRS on the back panel.

There is also a means of feeding this oscillator to each channel (enable insert on associated channel when OSC is enabled).

Lastly, each channel has a Trim pot.

It is left as an exercise to the reader of the User Guide how to employ the above to actually perform a calibration.

My assumption: For standard cal procedure when all is working as it should (i.e., the OSC outputting the proper level), one should enable the oscillator, and assign it to each channel in turn, while monitoring that channel's output signal level, and twist the Trim pot until the channel outputs 1.228 Vrms (or whatever level you'd like to use for nominal, though I'm good with 1.228Vrms). Apply to each channel in turn.

^^ Question one - does that sound correct?

Next issue - my unit does not output 1.228 Vrms out the TRS when oscillator is enabled. It outputs 1.116 Vrms.

My assumption: While routing the 1.116 Vrms oscillator signal to each channel, I want to twist the Trim pot to yield a 1.116 Vrms signal on the channel output, in order to calibrate to 1.228 Vrms (in light of my out-of-spec oscillator output).

My reasoning: The Trim pot is probably a simple voltage divider after the gain stages of the channel. As such, output should be a linear function of signal level before the Trim pot. The oscillator signal is likely being injected after the gain stages, and before the assumed voltage divider Trim pot. The cal procedure would set the output signal to a fixed ratio voltage divider - this ratio would be identical between the oscillator output at the OSC TRS, and the output at the channel out, regardless of the actual oscillator output. Therefore, if my oscillator was outputting 1.228 Vrms as the manual states it should, then the output of the channel would be 1.228 Vrms.

^^ Question 2: Do these assumptions seem valid?

[jbreher]

OK, I'm floundering. Any help would be appreciated.

Seems my assumptions above on how one is to calibrate the Pre are incorrect.

First, while enabling the OSC and enabling INS on a channel does route the OSC signal to the output. However, using this routing, the Trim pot for that channel has exactly zero effect on the output as monitored on the DB25. Accordingly, from my naive viewpoint, the ability to route the OSC signal to each channel's outputs is useless for calibrating the Pre itself.

I can't imagine digi would have implemented a feature that has no use, so I must be confused about something. Perhaps its utility is to calibrate something (e.g. an ADC) downstream? But even so, such utility is hampered by the fact that the OSC level (on my admittedly possibly compromised unit) is somewhat below the manual spec. What am I missing?

Next, I envisioned/postulated the following cal mechanism. I used a TRS to route the OSC signal to each channel's line input in turn. I set the measured channel to line, and 0dB gain. I was gratified to learn the the Trim pot in this case did indeed affect the channel signal output at the DB25.

Figuring that 0dB gain indicated on the GAIN/PARAM display corresponds to an actual 0 dB gain from Line input to DB25 output, I tweaked the Trim pot to make the measured signal level at the DB25 equal to the measured signal level (when unloaded - maybe a small error introduced?) at the OSC TRS output. Or very close - approaching this level, channels stop changing level with additional pot rotations.

At the point that the pots stop affecting the output level, I measured the following unloaded Vrms levels:
OSC TRS: 1.184
Channel outs - 1.171 to 1.183, with the low reading being quite the outlier.

I'll consider this good enough for the work I'll be doing, unless I've made some error in procedure.

Anyone want to comment on how I went about this?

[jbreher]

Wrong subforum? Stupid question? Nobody has a clue?

[arche3]

Why do you want to calibrate the pre? I suppose we can start there.

Sent from my SM-G975U using Tapatalk

[LDS]

Originally Posted by jbreher

OK, I'm floundering. Any help would be appreciated.

Seems my assumptions above on how one is to calibrate the Pre are incorrect.

First, while enabling the OSC and enabling INS on a channel does route the OSC signal to the output. However, using this routing, the Trim pot for that channel has exactly zero effect on the output as monitored on the DB25. Accordingly, from my naive viewpoint, the ability to route the OSC signal to each channel's outputs is useless for calibrating the Pre itself.

I can't imagine digi would have implemented a feature that has no use, so I must be confused about something. Perhaps its utility is to calibrate something (e.g. an ADC) downstream? But even so, such utility is hampered by the fact that the OSC level (on my admittedly possibly compromised unit) is somewhat below the manual spec. What am I missing?

Next, I envisioned/postulated the following cal mechanism. I used a TRS to route the OSC signal to each channel's line input in turn. I set the measured channel to line, and 0dB gain. I was gratified to learn the the Trim pot in this case did indeed affect the channel signal output at the DB25.

Figuring that 0dB gain indicated on the GAIN/PARAM display corresponds to an actual 0 dB gain from Line input to DB25 output, I tweaked the Trim pot to make the measured signal level at the DB25 equal to the measured signal level (when unloaded - maybe a small error introduced?) at the OSC TRS output. Or very close - approaching this level, channels stop changing level with additional pot rotations.

At the point that the pots stop affecting the output level, I measured the following unloaded Vrms levels:
OSC TRS: 1.184
Channel outs - 1.171 to 1.183, with the low reading being quite the outlier.

I'll consider this good enough for the work I'll be doing, unless I've made some error in procedure.

Anyone want to comment on how I went about this?

That sounds as good a game plan as I have heard. Have you tried taking a measurement with the gain set higher? Maybe +12? To see how well the Pre actually maintains those calibrations in real world use.

Originally Posted by arche3

Why do you want to calibrate the pre? I suppose we can start there.

Sent from my SM-G975U using Tapatalk

Ambisonic recording is a big reason why you would want the preamp calibrated. You need closely matched gains on all channels for the decoder to work properly. One of the upsides of the PRE is that Pro Tools will automatically link channels based on the track type you create. 1st order ambisonics = 4 linked channels.

[jbreher]

Originally Posted by arche3

Why do you want to calibrate the pre? I suppose we can start there.

So metering is consistent with actual signal levels throughout my system.

[arche3]

Originally Posted by jbreher

So metering is consistent with actual signal levels throughout my system.

How are you using the pre? Microphones? Summing mixers? If microphones I don't see the need to calibrate. If bringing level up from summing mixer the metering is not needed imo as all that matters is the adc level into protools. You can calibrate the input on the audio interface for pairs of mic pre outputs of the digi pre.

I'm not even sure what the trim pot does on the mic pre units.

Sent from my SM-G975U using Tapatalk

[the.engineer]

Sorry if I’ve missed this...have you measured the output with a volt meter to double triple check it’s outputting 1.228V?

[albee1952]

Good suggestions above. Here is how I would try it(not knowing any better). I would create a session with an AUX track sending(use the Signal Generator plugin) a 1KHz tone out any output NOT in use for monitoring. Patch that(with a physical cable) to each input of the Pre and feed each channel to a separate track in Pro Tools. Put those tracks in Input mode and watch the track meters. Compare against each other(adjust trimmers on the Pre for identical readings). And, compare against the output level of the signal gen track. This might not be technically "perfect", but certainly can result in matching levels


I also would likely look at the Pre levels and choose a target level that matches the highest number of channels that match before you tweak


Someone more technically savvy than me can likely figure a way for this to be done to actual technical accuracy

[jbreher]

Originally Posted by LDS

Have you tried taking a measurement with the gain set higher? Maybe +12? To see how well the Pre actually maintains those calibrations in real world use.

Good idea. I'll do that.