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

[JackoRoses]

Hi All,
First, let me apologize for our mistake in wiring the a/d converter in the V3.  It is a simple thing to check in the design phase of a product and we missed it.
If you are a V3 owner and would like to have your unit corrected please send it to us and we will quickly make the repair, test, calibrate, burn-in and return your unit to you.  There is no need to contact us ahead of time as we have posted a Return Authorization form on our web site at: http://www.gracedesign.com/support/v3_polarity_rma.pdf   Simply fill in this form and put it in the box with your V3 and ship it to us. 
While all V3s are currently under the original 5 year factory warranty we will repair this problem under warranty for the lifetime of the unit.

Now,  I think I should comment on the effects of this problem with respect to stereo and multitrack recording.  The fact that we missed this problem in all of our R&D listening tests and that we shipped almost 500 units without this problem being detected is a testament to how subtle absolute phase reversal is.  In fact, it was initially brought to our attention by a customer who happened across the problem while viewing waveforms on the computer screen.  While I personally have difficulty hearing the difference of polarity reversal, I do know people who claim that it is audible.  When pressed, however, they have not been able to give a description of the difference or weather it is “worse” or “better” sounding. 
That said, I fully agree with David Moulton's opinion that polarity of every signal involved in a recording should be “accounted for”.  As tapers and engineers we need to rely on our tools to behave in a predictable way and this includes polarity.
This becomes much more important in the context of multi channel recording as the audible effects of absolute polarity reversal when mixing with correctly phased signals is not subtle.  In the case of a soundboard matrix recording, if the room mics are inverted relative the the soundboard signal, there will be some cancellation when the two are mixed together.  The reason that some of you have reported difficulty with track alignment is undoubtedly because of this.  I sincerely apologize for any recordings that were rendered badly because of this issue.

Again, my apologies for the confusion here.  We are doing all that we can, as quickly as we can, to rectify this problem.
As always, our mission is to provide the best possible tools and support for capturing and creating the art of music.
Sincerely,
Michael Grace

This is 100% class thank you!
This is why I look forward to continuing my relationship with Grace.

[Brian Skalinder]

i'm sorry....i should have said polarity reversal.

Whoops, I meant polarity reversal, too.  Doh!!

re: skalinder's last post......

as an engineer once told me. 

waveforms mean nothing.  perception is everything

Exactly why this isn't a big deal for me - I perceive absolutely no difference between my V3 recordings regardless of polarity, as long as I have relative polarity across both channels.  My other posts are just my ramblings trying to comprehend *why* I don't perceive a difference.  Don't mind me...

[cleantone]

This is why I look forward to continuing my relationship with Grace.

How long you been with her? 

Does this mean that Grace design does not put a tone from an oscillator through each unit and look at it through a scope of some sort or a frequency analyzer on a computer prior to authorization for shipping?  It seems to me that the problem could have been seen then Huh? or maybe this test does not show the problem?  Just curious.

I don't know what they do obviously. I would think that sending an oscillator in the unit alone would not really alert you to this. Espessially if your not looking for this problem. Once the oscillator is on the tone being even and continuous you probably wouldn't be able to distinguish any "attack" to look for + vs. - . If you were testing for latency between the analog and digital outputs by recording into a DAW then you would/should notice.

[mbgrace]

Thanks for the response Mike.  Does this mean that Grace design does not put a tone from an oscillator through each unit and look at it through a scope of some sort or a frequency analyzer on a computer prior to authorization for shipping?  It seems to me that the problem could have been seen then

i know your time is limited but if you could go into more detail on how you test your products that would help.

edit:  or maybe this test does not show the problem?  Just curious.

This is a good question Brian.  In production the V3 goes through a battery of tests on our Audio Precision analyzer.
We measure the following parameters:
Analog-
Frequency response (including checking the high pass filters at each roll off frequency)
Distortion and noise
Intermodulation distortion
crosstalk
Gain accuracy at each gain setting
Trim control accuracy
MS decode accuracy
Input pad accuracy
EIN (equivalent input noise)
CMRR (common mode rejection ratio)

Digital-
Frequency response (at each sample rate)
Noise
Distortion
Crosstalk
single wire/dual wire functionality
ANSR spectrum accuracy

After all this each unit gets hooked up to a CD player via a custom box that converts the CD output to a mic level signal.  This allows us to listen to the unit with full program material to make sure it sounds right and that all of the controls function with no audible problems.  We do occasionally reject preamplifiers at this stage for problems that did not show up on the bench test.
 
There is no polarity test in our V3 production routine because the V3 has no wire in the signal path that could be connected improperly.  All of the connectors are pc mounted.  If the circuit board design is correct then all units would have correct polarity and there would be no need to test each unit for it.  Unfortunately I never verified the ADC pcb design for polarity in the R&D phase.

We do test the absolute polarity on other products where there is a potential for miswiring during assembly.

Oh, in case I missed the point of your question,  there are no measureable differences between normal and inverted polarity except for the 180º phase angle.

Michael

 

[OFOTD]

R&D problems happen.  They happen all the time.  A company owning up to the problems does not happen all the time much less at all.  This is why I have spent my money on Grace products and one of the reasons why I will continue to buy their products.  A company with people that care about their customer base in the first place as well as being positive members in the community that they help support.

Thanks Mike!

So what is the expected turnaround time if I were to send mine in to you guys in Boulder?

[JasonSobel]

So what is the expected turnaround time if I were to send mine in to you guys in Boulder?

the PDF sheet (Return Merchandise Authorization Form) that Michael posted in his first post of this thread states "Grace Desgin will repair your unit in approximately 3 business days and ship is directly back to you."

[BayTaynt3d]

I don't own a V3, nor do I know much about polarity, but all of this reminds me of an experience I had 6 years ago when I bought my first subwoofer. I pluged it in, had it kicking, and then started to look at the switches in the back. It had a polarity switch, and I was like WTF is that? I read up in the manual, and searched online a bit, and had an ah-ha moment about how freakin' important it was to be consistent with the red-black colors of the speaker cables from the receiver to the speakers. Doh! All of those younger years I had no freakin' clue how important that was. After playing around endlessly with my stereo after learning this, I found that (to my ears) it mattered more that they were all wired the same way (either red-to-red or black-to-red), rather than some stupid miss-mash like I had going before. I remember feeling like an idiot, but I finally understood why it mattered because I COULD HEAR it when the polarity was different between the speakers. It was also quite amazing when I started playing with that polarity switch on the subwoofer. Since it wasn't "red" or "black," I wasn't sure which direction to flip it once I had all of my speakers correct. So, I put my head right inbetween the sub and a speaker and flipped the switch back and forth. Ah man, the difference was NIGHT AND DAY -- it was like a noise-cancelation machine on the low end! Switch it one way, no bass at all coming from either place, switch it back and BOOM, BOOM, BOOM (at least it worked that way before I turned on the low-pass routing to the sub, which removed the freqs from the speakers altogether, but with the bass in the speaker and the sub). It was incredible how much it dampened the bass, it was so obvious it wasn't even funny. It really gave me a new appreciation for polarity that's for sure.

[Brian]

Muchas Gracias, Mike.  You RAWK \m/  Keep up all the great work!

[RebelRebel]

I vaguely recall reading a piece off-TS that discussed sonic impact of DC offset during the mastering process.  Don't think I'll go try to find it again, though. 

when an audio file  has DC offset, zero is not actually zero. So if you insert that into a good audio track, you get a click when it hits the wave file, and another when it stops.

The biggest problem is that this offset means one side of the wave will clip before the other one (if basically balanced waveforms). If the source is pushing close to 0dbfs, removing DC offset should be used to see if you get more headroom.

Most "prosumer" DC offset removal tools not only remove DC offset but remove good stuff too, and cause the audio to suffer.  a lot of instruments that involve force ( our breath) going largely in one direction to produce sound end up producing waveforms that look screwed up(ie skewed in one direction). (trumpet, trombone, oboe)but in all actuality are perfectly normal.

[wboswell]

Interesting article in Tape OP polarity reversal.  Got it in the mail today and had a chuckle.  Essentially, the writer explains how you can use polarity to your advantage when recording in a studio environment.

[RebelRebel]

#1.  It is IMPOSIBLE to hear phase on a 2chan source unless they are out of phase with each other.  phase is 100% relative.

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

that just isn't correct. With some instruments(percussion particularly) some people can definitely tell between correct and inverted playback. not just with stereo , either..


Sure most cannot hear it, but to say that it impossible is not accurate.

[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.

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

that just isn't correct. With some instruments(percussion particularly) some people can definitely tell between correct and inverted playback. not just with stereo , either..


Sure most cannot hear it, but to say that it impossible is not accurate.

yeah... we coverd that Ted

Matt

[Brian]

Most "prosumer" DC offset removal tools not only remove DC offset but remove good stuff too, and cause the audio to suffer.  a lot of instruments that involve force ( our breath) going largely in one direction to produce sound end up producing waveforms that look screwed up(ie skewed in one direction). (trumpet, trombone, oboe)but in all actuality are perfectly normal.

i'll never forget the first time i tracked a trumpet.  i freaked out when i saw the waveforms as i thought something was wrong.  weirdest looking waveforms ever.

thanks for the input on the DC offset.

[BayTaynt3d]

Most "prosumer" DC offset removal tools not only remove DC offset but remove good stuff too, and cause the audio to suffer.  a lot of instruments that involve force ( our breath) going largely in one direction to produce sound end up producing waveforms that look screwed up(ie skewed in one direction). (trumpet, trombone, oboe)but in all actuality are perfectly normal.

i'll never forget the first time i tracked a trumpet.  i freaked out when i saw the waveforms as i thought something was wrong.  weirdest looking waveforms ever.

thanks for the input on the DC offset.

Ah, man! Teddy and Brian Sax, where were you guys when I needed you most? You could have cleared up my recent problem in a second with those quotes. I thought my pre was screwed, until I did some more research and learned it wasn't my pre at all. LOL! 

See this thread here for what I'm talking about:
http://taperssection.com/index.php?topic=60987.0

-- Taint

[Brian]

Most "prosumer" DC offset removal tools not only remove DC offset but remove good stuff too, and cause the audio to suffer.  a lot of instruments that involve force ( our breath) going largely in one direction to produce sound end up producing waveforms that look screwed up(ie skewed in one direction). (trumpet, trombone, oboe)but in all actuality are perfectly normal.

i'll never forget the first time i tracked a trumpet.  i freaked out when i saw the waveforms as i thought something was wrong.  weirdest looking waveforms ever.

thanks for the input on the DC offset.

Ah, man! Teddy and Brian Sax, where were you guys when I needed you most? You could have cleared up my recent problem in a second with those quotes. I thought my pre was screwed, until I did some more research and learned it wasn't my pre at all. LOL! 

See this thread here for what I'm talking about:
http://taperssection.com/index.php?topic=60987.0

-- Taint

i saw that.  you didn't mention what instrument(s) you were recording int he first post so i figured it was some kind of error with your preamp.  sorry about that! +

everything cool now though?