reopened session sounds different

[sean c]

One of my engineers has been working on an ongoing project and when he revisits segments that he has already mixed, they sound different.

We cannot pinpoint exactly what is wrong other than to say that volume fades and levels seem to be off. He recorded the mix onto a stereo track (via internal bussing). He then reopened the session the next day to revise something later in the timeline and re-recorded including the earlier part that supposedly did not change. When we compared the previous record to the new one they are noticeably different in sound and waveform appearance during the unchanged section.

While I did not watch the guy work the whole time he swears he did not change that segment and thought he had noticed similar behavior on many other sections and simply wrote it off as operator error.

It is a recently built system of 10.4.8 on a Dual 2.0 G5 w 2GB RAM, HD2 Accel playing a DV out of the Mojo. We haven't noticed this on any of our other systems (6.4.1, 7.2, 7.3.1) .

Anyone every heard of this?

Thanks.

[Andre Knecht]

Anyone every heard of this?

Well, another current thread alludes specifically to automation anomalies, which may or not be the case here.

Also, 2 GB of RAM is a bit on the small side for a dual G5 running OSX, ProTools and video. RAM is dirt cheap these days — consider doubling yours, at least.

IHTH.

[Richard Fairbanks]

Maybe this is what happened to you. I have found stereo panpots that were set fully left/right on one day, but the next time the session is opened the left channel was set to dead center for any stereo panpot that was not automated. I've only seen this with 7.3 software on PPC and Intel G5.

The other type of issue is a killer, and it has happened in one way or another for several software versions. It has been a problem only when a session is close to maxxed-out DSP use. I think what happens is that if you add a plugin or alter an input/output assignment, which causes a DSP shuffle, it is possible to have mis-assigned outputs and/or incorrect bus levels. I've learned that, when the reshuffle happens, the best thing to do is re-launch Pro Tools and re-open the session. I have been burned more than once with mix levels being wrong (they sounded right until the next opening). Sometimes the output assignments are wrong where one channel in a 5.1 assignment is always on, even though the signal is panned fully out of that channel. I cannot say with absolute certainty what happens in 7.3.1 cs4, as opposed to earlier versions, but I am certain I have run into similar trouble with 7.3.1 on my G5 dual 2.5.

[StudioRhythm]

I have had the abovementioned two issues with my Intel system as well (7.3.1cs4 on MacPro, 10.4.9).

I've had hard pans revert to center, and every now and then a DSP shuffle will cause an aux track or a TDM plugin on an aux track to stop passing signal, or pass signal at a greatly reduced level. If you can figure out what stopped working, disabling and re-enabling the plugin/aux will get it working again. Of course this is a little scary if it's a busy mix where you might not notice it right away....

[Rich Breen]

Maybe this is what happened to you. I have found stereo panpots that were set fully left/right on one day, but the next time the session is opened the left channel was set to dead center for any stereo panpot that was not automated. I've only seen this with 7.3 software on PPC and Intel G5.

Hi Richard - Geez, this one's been driving me crazy! Happens about once a month, and never fails to totally screw me up. BTW, for me it's been on 7.3.1 on a dual g5.

Has anyone logged this? Have there been any discussions about it? I just haven't gotten around to it, cause I have yet to find any rhyme or reason to it, but it's an easy thing to miss in a large mix, and can really bite you if you're not paying close attention...

As far as the mis-assigned buss issues - I've been running into it for a while, but only when I've got Altiverb in a session - I'm fairly convinced it's an Altiverb or RTAS issue. In one instance, it completely cleared up by turning ADC *on*, but in others it seems not to care.

Both these bugs are *so* intermittent it seems almost impossible to troubleshoot them, but it sure would be great to squash them once and for all.

best,
rich

[soundonsound]

Sean,

Did you notice that the problem happened after some major change, such as saving the session at a different sample rate? That's when it happened to me on HD 7.3.1 - Reopened it and the automation was different, session sounded much different.

[BGee]

I've had all of the above issues happen as well - the pan pots issue really threw me for a loop the first time it happened. I've gotten into the habit of making lots of notes in the comments fields on various tracks to check this or that bus signal/this plugin is passing signal/that stereo pan pot is left 100, right 100/etc so I remember to verify them all before printing. All of which is a bit of a nightmare to have to do, but I have little confidence that things are exactly as I set them to be on big mixes in Pro Tools these days. One other thing that routinely goes wrong for me, and I cannot for the life of me figure out how to reliably repeat this problem, but often when I am using busses for send to key inputs on plugins, typically compressors, the delay time of the key input relative to the overall compensated delay of the mix with ADC on often changes. For example, I will duplicate a snare track, nudge it a few ms to the left so the key signal will be a little in front of the actual snare drum, and send to bus 1, then set key input on a stereo RenComp or McDSP CB3 to receive key input to clamp down on guitars for a moment. Depending on the day, or the humidity, or I don't know what, sometimes I will notice that my guitars are no longer ducking at the right time, so I'll turn on the key input monitor on the compressor and discover that there is a massive time difference between the real snare and the key snare, much more so than I had originally heard, requiring me to readjust the duplicated snare region timing to get it right again. It happens sometimes and others not. Frustrating. ADC with key inputs is very sketchy.