Topic: V3 (< V3497) digital outputs out of phase relative to analog outputs

[Kindguy]

edit: I knew the minus t's would be coming my way... guess I'll think twice before offering trouble shooting help in the future.

Please don't let a jackass or two impact your posting - I, for one, appreciate your generally very informative posts.  Keep 'em coming.

Ditto T+

[Kindguy]

if my v3 is not under warranty anymore I have to pay Grace to have them put the digital out in phase with the input?  why should I have to pay for something that I didn't break in the first place?  am I in the wrong here thinking this should not be something we have to pay for?  otoh, I would gladly pay for it if it meant a renewel of my warranty, ala when the optical mod is performed.

thoughts?

Contact Jamie and ask if they'll cover it even if you're out of warranty, given the nature of the problem - a fundamental production flaw on their part.

here's Jamie's response.  it looks like the warranty question is a moot point since every v3 is still under warranty.

Hey Tim-

The first V3 (V3001) shipped October, 2002, thus every V3 in the field
is still covered by our warranty.  Cheers!

Great! Guess I'll just send mine in.

Scott, Sorry that sucks!

[Kindguy]

I wish Apogee would stand behind their product like Grace. Jamie If you look in here by chance thanks.

[mmmatt]

#1.  It is IMPOSIBLE to hear phase on a 2chan source unless they are out of phase with each other.  phase is 100% relative.
#2.  the best way to check for phase corectness is to mix two mono sources down to a single mono source.  If they are out of phase and they are identical sources. they will read -0- db (ie silent), if they are similar( like a stereo pair) the sound will be greatly diminished.  In phase it will be louder than the original sources.
#3.  In general listening the easiest ways to tell if your source (or playback) is wired out of phase is you will hear less bass responce in stereo than if you moved ballance over to just one speaker, and 2ndly the image will sound like shit. 

Clinton... I feel for you brother... you got the shit end of the stick but only if you were using the coax.  If you were using analog you were fine.  To my recolection you use a motu though and were probably using the coax (or optical) for your multi rig.  I had one of the R4's with chan 4 our of phase and it was a colosal pain in the ass before I sent it in for the fix.  It is really hard to hear phase when you are mixing multiple sources.  It really screwed me up when the neutrik combo xlr/1/4" jack turned out to be only out of phase on the xlr.  Talk about wasted time on the DAW...

Good luck guys, but 2 chan users you are fine.

Matt


*edit*  yes, in a multi/matrix situation you are going to muddy things up and lose parts of the recording... basically where things are the same between the wo sources.  Also if you want to hear what out of phase sounds like, unplug one of your speaker cables on your playback and reverse it.

[Genghis Cougar Mellen Khan]

Upon discovery we immediately corrected this issue in new production units.  Beginning with serial number V3497, all of the digital outputs are in phase with the input.  Any unit prior to this serial number will have inverted polarity on the digital outputs.

My unit is serial #V3167

Lets say I record the following:

source a: Mics > V3 coax > M1

source b: SBD > M1

If I don't reverse the polarity on the V3 source, by what this is saying the recordings would be 180 degrees out of phase from another?  I've never been able to matrix a SBD to a mic tape and been satisfied with the results, I always attributed the problem to not having the know-how to allign or deal with the drift correctly.  I've swapped the polarity before and the results were different, but I was thinking that it was the wrong thing to do. 

With one source being reversed, there is significant phase cancellation, or is that incorrect?  With speakers polarity, cancellation is obvious as hell, is this not essentially the same thing? 

If I sent files to someone recorded with the sources above, would anyone be willing to play around with them and see what they think?

[mmmatt]

Upon discovery we immediately corrected this issue in new production units.  Beginning with serial number V3497, all of the digital outputs are in phase with the input.  Any unit prior to this serial number will have inverted polarity on the digital outputs.

My unit is serial #V3167

Lets say I record the following:

source a: Mics > V3 coax > M1

source b: SBD > M1

If I don't reverse the polarity on the V3 source, by what this is saying the recordings would be 180 degrees out of phase from another?  I've never been able to matrix a SBD to a mic tape and been satisfied with the results, I always attributed the problem to not having the know-how to allign or deal with the drift correctly.  I've swapped the polarity before and the results were different, but I was thinking that it was the wrong thing to do. 

With one source being reversed, there is significant phase cancellation, or is that incorrect?  With speakers polarity, cancellation is obvious as hell, is this not essentially the same thing? 

If I sent files to someone recorded with the sources above, would anyone be willing to play around with them and see what they think? 

a mic is a very small speaker run backwards... it is essentially the same thing.  mater of fact, you can plug your headphones into your mic input and use them like a mic... sounds like crap but the pricipals are the same.    l > r phase is a lot easier to hear than l+r > l+r phase.  I guarantee this is what your problem was.  I have a full plate right now, but if you wanted to send the sources (or upload to my server) I would hapily take a look at them and mix a song or two for you.

When mixing 2 sources that are as vastly different as aud and sbd you won't get absolute cancelation, but the cleanest parts of the aud tape will be canceled by the similar parts of the sbd tape and you get crap.  all the reverbarance and "room" will be fine 



Matt

[cleantone]

I had a hard time SLEEPING last night. I couldn't stop thinking about how many recordings of mine are fuct. I literally almost always run a matrix. The last 2 track recording I made was in August I think. I rarely run V3> VX Pocket. Except at Matt Murphys Pub which is the only place I go with no real sound reinforcement. I think have made at least 200 recordings with the V3 and AES outputs. A good chunk of those were full multitracks in which I might use the V3 for bass and kick drum, or overheads, or a piano, etc... I do most often use them for a stereo pair of rooms mics. I don't even have stuff notated well enough to know for sure what channels ran through my V3 for my backed up recordings. I usually right everything down on paper but no text file and never have time to fill in the notes on a session.

It now makes a lot more sense as to why I would have such a hard time finding a place to line up files for a matrix. Obviously I wouldn't line up the audio to be proper with the polarity 180 degrees off. This means I have been finding somewhere that is a few milliseconds off and then EQing. What I mean is that I would match a positive compression for each track. So now knowing the V3's channel was actually negative rarefactions swapped I know I had been mixing them out of phase and off by some milliseconds. No wonder why I am NEVER 100% happy with my mixes.

Last night I opened an Apollo Sunshine recording I was working on. Actually I had finished it but luckily I had not erased the session. I found a section where the bass was feeding back a low drone and there was one kick drum hit. Nothing else. I soloed the V3 channel. I looped a small section and let it play many times. I inserted a phase inversion plug in. I would bypass the plug in when the 3 second sample repeated. I heard a minor pitch change. Not a half step but some amount of cent. I think I'll try to make a mp3 clip to post. It is true that a single stereo track is not "night and day" different with the positive and negative info swapped but it is definitly different. Even if it is only a mathmatical/microscopic difference.

I am REALLY REALLY REALLY unhappy right now.

As I and someone else mentioned... I wonder how long they knew and why there was no mention until we figured it out. Was it someones decision to say that thy should just fix the new ones and hope nobody notices?

If all my stuff was two track I wouldn't be that upset. I would just be wasting a shitload of media I have it backed up to and could fix a recording with minimal effort. Well I guess there is the fact that however many people have the origional improper recording anyway. The point is that I can't just flip the polarity of all of these matrixes and multitracks. 

[cleantone]

I just tried to recreate the experiment. Maybe it is my mind playing tricks on me. I grabbed one kick hit with the buzz drone behind it. I made a loop and repeated it four times. 1-2-3-4. Then I reversed the polarity on this and put them together. So the first 1-2-3-4 is reversed polarity and the second 2-2-3-4 if proper polarity. Obviously if there is a difference (which I think I do hear) it is VERY VERY subtle. I hear some change in tonality of the kick drum. Again, this is so very subtle it could be my brain tricking me even. Your gonna have to load these into a DAW for proper auditioning, unless you have another way to loop this section. I think it needs to repeat multiple times while you count out 1-2-3-4, 2-2-3-4; 1-2-3-4, 2-2-3-4. I am not really expecting everyone to hear what I might be but want to try anyway... just in case... I really think I can hear a SUPER SUBTLE difference. If your not a critical listener you I can easily see this sounding the same to you. I just tried on iPod headphones and could still hear it. So any playback system should work. Oh and quicktime worked okay for me. Loading it in a DAW will not hurt.

On another note I am surprised this thread does not have 1000 posts. Am I the only one who thinks this is a VERY BIG DEAL?

Another thing I noticed when testing yesterday is a very small latency between the alalog and digital outputs. Maybe it was my MOTU inputs but the AES signal was 2-3 milliseconds later than the analog. This of course makes perfect sense and should not have much impact on a multichannel recording. I supposed different cable length and different preamps would give the same minor offsets.

[scb]

I just tried to recreate the experiment. Maybe it is my mind playing tricks on me. I grabbed one kick hit with the buzz drone behind it. I made a loop and repeated it four times. 1-2-3-4. Then I reversed the polarity on this and put them together. So the first 1-2-3-4 is reversed polarity and the second 2-2-3-4 if proper polarity. Obviously if there is a difference (which I think I do hear) it is VERY VERY subtle. I hear some change in tonality of the kick drum. Again, this is so very subtle it could be my brain tricking me even. Your gonna have to load these into a DAW for proper auditioning, unless you have another way to loop this section. I think it needs to repeat multiple times while you count out 1-2-3-4, 2-2-3-4; 1-2-3-4, 2-2-3-4. I am not really expecting everyone to hear what I might be but want to try anyway... just in case... I really think I can hear a SUPER SUBTLE difference. If your not a critical listener you I can easily see this sounding the same to you. I just tried on iPod headphones and could still hear it. So any playback system should work. Oh and quicktime worked okay for me. Loading it in a DAW will not hurt.

On another note I am surprised this thread does not have 1000 posts. Am I the only one who thinks this is a VERY BIG DEAL?

Another thing I noticed when testing yesterday is a very small latency between the alalog and digital outputs. Maybe it was my MOTU inputs but the AES signal was 2-3 milliseconds later than the analog. This of course makes perfect sense and should not have much impact on a multichannel recording. I supposed different cable length and different preamps would give the same minor offsets.

i think it's all in your head

i bet if someone were to send you 2 files and 1 was reversed and you didn't know which was which, you woudln't be able to tell a difference

[Brian Skalinder]

On another note I am surprised this thread does not have 1000 posts. Am I the only one who thinks this is a VERY BIG DEAL?

I'm trying to hear the difference in the sample you posted, and I convince myself that I think I can.  But I bet I couldn't tell in an ABX environment.  I think it's a big deal, theoretically.  Practically speaking, for my 2-ch recordings, I don't think it's a big deal since I can't hear the difference.  But maybe my hearing/listening is deficient.  Still undecided about whether or not I'll go back and invert phase on all my recordings.

If I was multi-tracking, it'd be a bigger deal to me - not because I'd need to go back and fix all my recordings (the first time I multi-tracked the V3 with another source I noticed the phase was inverted), but rather because it would add an additional step in the post production process (inverting the V3 source).

[cleantone]

The fix to previously transfered/mastered recordings is just a reversal of both channels polarity, correct?

Yes that is right. Don't expect a drastic difference of course.

i think it's all in your head

The more I revisit it the more I am convinced I hear the difference. It is such a subtle change in tonality, or timbre, or something. I can hear the drone change, the cymbol, and the kick drum. I tried reversing the order and I hear the change happen opposite of what this file does. It REALLY is super super subtle.

i bet if someone were to send you 2 files and 1 was reversed and you didn't know which was which, you woudln't be able to tell a difference

That is a pretty safe bet. Accept for the fact that a note should start with a positive pulse and not negative. At least I think I'm right about that.

[mmmatt]

I just tried to recreate the experiment. Maybe it is my mind playing tricks on me. I grabbed one kick hit with the buzz drone behind it. I made a loop and repeated it four times. 1-2-3-4. Then I reversed the polarity on this and put them together. So the first 1-2-3-4 is reversed polarity and the second 2-2-3-4 if proper polarity. Obviously if there is a difference (which I think I do hear) it is VERY VERY subtle. I hear some change in tonality of the kick drum. Again, this is so very subtle it could be my brain tricking me even. Your gonna have to load these into a DAW for proper auditioning, unless you have another way to loop this section. I think it needs to repeat multiple times while you count out 1-2-3-4, 2-2-3-4; 1-2-3-4, 2-2-3-4. I am not really expecting everyone to hear what I might be but want to try anyway... just in case... I really think I can hear a SUPER SUBTLE difference. If your not a critical listener you I can easily see this sounding the same to you. I just tried on iPod headphones and could still hear it. So any playback system should work. Oh and quicktime worked okay for me. Loading it in a DAW will not hurt.

On another note I am surprised this thread does not have 1000 posts. Am I the only one who thinks this is a VERY BIG DEAL?

Another thing I noticed when testing yesterday is a very small latency between the alalog and digital outputs. Maybe it was my MOTU inputs but the AES signal was 2-3 milliseconds later than the analog. This of course makes perfect sense and should not have much impact on a multichannel recording. I supposed different cable length and different preamps would give the same minor offsets.

Don't let this eat you up clinton!  It is simple psysics... every action has an oposite and equal reaction.  the positive of the the source pushes the speaker out and the negative pulls it back in equaliy.  there is minimal time delay out of phase but that is it.  You are probably hearing the plugin.  A real-time anomilie of the DAW.  Burn off 2 2chan samples and play them back on a CD... you won't hear a difference.  Or let me put together a sample for you to listen to blind.  I'll do 4 sources with 1 being out of phase.
     Another thing too.  There are still phase diferences in omni mics, but they are slight.  Any times you were using omni's through the v3 it should be not enough to even mess with it.  Has anyone posted this on the Oade forum?  I think Doug would be able to settle this.
     I'm just stunned it has never been noticed in 3 years!
Matt

[scb]

if you want another way to test, reverse your speaker wiring in both speakers. see if you can hear a difference

[cleantone]

if you want another way to test, reverse your speaker wiring in both speakers. see if you can hear a difference

I understand what you mean by this. I need to hear it back to back since the difference is so subtle. The notion that I am hearing the plug in does make sense. I have a feeling that it's not the case though. I think if the plug in made a tonal change that was audible when recombining the sources they would NOT cancel out 100%. These did. That tells me that the polarity was reversed without change to the sound.

If you take a sample. Copy it and reverse the polarity. Line them up on seperate tracks and play them back at 100% equal volume they will cancel. If you then move one track by 1 millisecond, or add a tenth of a db in volume, or make ANY change at all the sound will come back. It wont be as if you are playing only one of the other but any minor change will not cancel 100%. So the fact that after using the "invert phase" plug in they cancel 100% leads me to think there was no tonal change. So the difference I hear is not from the plug in.

Scott, your correct that a blind test with my ears alone would not be able to tell the difference when I'm not comparing back to back, like this sample. That is strange in itself actually. So I pretty much agree that it is not quite worth the effort for people who have done two track alone. One thing that bugs me is that I have a lot of multitracks I have not mixed from as far as three years back. For the most part I do not know which channels had the V3 used on them. Far too many to remember and my paper notes would be tossed on my way out of the gigs. My ambient recordings have good notation and if and when I have the time I can revisit them to mix again properly. Probably with a substancial difference in end product too. I'm not sure I'll ever actually be able to make time for that though. Not to mention the fact that a lot of those matrixes are circulating. I do have a lot that are not as well though.

[mmmatt]

I think correct phase is very important.  Linn amps and pre-amps invert phase in some situations and you have to make sure you correctly flip your speaker leads as appropriate.  Different speakers will react differently.

It is simple psysics... every action has an oposite and equal reaction.  the positive of the the source pushes the speaker out and the negative pulls it back in equaliy.

I'm sorry but that's totally wrong.  We're not talking about a bouncing ball in a vacuum here.

Look at the waveform Cleantone posted and explain why it doesn't matter that the speaker cone is retracting and not pushing out on that first peak.

If all we listened to were sine waves, then maybe it wouldn't matter because those waves are symmetic.

there is a lot of discussion on this on the net.  My thoughts on this subject come from exaustive converstaions I had on the subject with an old friend who was (and maybe still is) a manufacturer of outdoor loudspeakers.  I am researching and I will post more info if I can find any two articles that agree with each other.  So far, thoes that sazy there is an audible difference say that it is dependant on frequency and amlitude as to if it is distinguishable.
   So far it appears as if, you are correct on this, in that it is not absolute, and if so I will appologise for my words and my adamancy.

Matt