IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Calibrating azimuth on VCR: Can any software report or visualize phase
SuspiciousLizard
post Jul 27 2012, 00:06
Post #1





Group: Members
Posts: 2
Joined: 26-July 12
Member No.: 101790



I'm capturing VHS home movies on a VCR with stereo linear audio heads. The tapes I'm currently working with have only a mono track, but I get occasional distracting hitches in the audio that are only present in the right/left captured channel and not the other, so I need to collapse the audio into a single mono channel in software to fix them. (Well, I have to normalize the channels to equal volume first, since their relative volume depends on how the tape was recorded.) This means I have to worry about sound cancellation from phase differences. I'll probably end up correcting the audio phase in software anyway with Stereo Tool, but I'd like to start out with a properly aligned VCR at least.

As a result, I'm now trying to calibrate the azimuth for my VCR's linear audio heads, and it's a huge pain. I've done it before by comparing channel differences with a pretty high-latency method: Capture part of a tape with Virtualdub, export the audio to .wav, import to Audacity, normalize the channels separately, split the stereo channels, invert one, rejoin the channels into stereo, then collapse the channels into mono for the difference between the two channels after normalizing. Then turn the azimuth screw a bit, do the whole process over again, and compare the two differences to see whether I made it better or worse...lather, rinse, repeat. I've spent all day doing this before, and I eventually got my VCR aligned to my satisfaction (the difference track was very faint), but I've recently found that it's all out of alignment again. I'd hate to spend just as much time doing it this way all over again, so I'm looking for better tools.

Is there any Windows or Linux software that can visualize or otherwise report phase differences in real-time (or with some reasonable latency)? I would love to have something that worked on line-in audio, so I could turn the screw and get real-time feedback, but even having something that worked on a .wav file would save me a lot of time. ProPhase seems to do this, but it's MacOS only, and I'd like to find something I can actually use. Thanks smile.gif

This post has been edited by SuspiciousLizard: Jul 27 2012, 00:14
Go to the top of the page
+Quote Post
mzil
post Jul 27 2012, 01:19
Post #2





Group: Members
Posts: 499
Joined: 5-August 07
Member No.: 45913



I think Cool Edit Pro will let you visualize a Lissajous pattern showing the relative phase of the two channels and you can tweak your azimuth from that, but I'm not sure.

This may be of help:
http://open-reel.blogspot.com/2009_03_01_archive.html

There are also software programs that let you turn your soundcard into an oscilloscope, which also can show a live Lissajous pattern, I guess.

I've never used any, nor an oscilloscope, but here's a random example I googled up:

http://www.multi-instrument.com/

This post has been edited by mzil: Jul 27 2012, 01:29
Go to the top of the page
+Quote Post
splice
post Jul 27 2012, 04:22
Post #3





Group: Members
Posts: 119
Joined: 23-July 03
Member No.: 7935



Do it in real time, by ear...
Back when cassette players were common and often required alignment, I made a headphone extension cable with no ground wire. (Headphone cable plugs/sockets have 3 connections - left, right, and ground. Connect only the left and right connections.)
This will result in only the difference between the two channels being heard in the headphones. Feed the VCR signal into your PC and monitor it via the headphones. Adjust the input levels to match the two channels (minimum signal in the headphones). Now adjust the azimuth for minimum signal.


--------------------
Regards,
Don Hills
Go to the top of the page
+Quote Post
Soap
post Jul 27 2012, 04:27
Post #4





Group: Members
Posts: 1005
Joined: 19-November 06
Member No.: 37767



Splice, can I assume that technique depends on my ears having decently matched responses?

This post has been edited by Soap: Jul 27 2012, 04:28


--------------------
Creature of habit.
Go to the top of the page
+Quote Post
pdq
post Jul 27 2012, 12:01
Post #5





Group: Members
Posts: 3372
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



QUOTE (Soap @ Jul 26 2012, 23:27) *
Splice, can I assume that technique depends on my ears having decently matched responses?

No, both ears will be hearing the identical sound, just 180 degrees out of phase.

The problem with this technique, however, is if the two levels are not level matched. Perhaps you could use the amplifier's balance control to first minimize the lower frequencies, which are affected little by azimuth mismatch, then adjust azimuth to minimize the higher frequencies.
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Jul 27 2012, 17:54
Post #6





Group: Members
Posts: 3647
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (SuspiciousLizard @ Jul 26 2012, 19:06) *
I'm capturing VHS home movies on a VCR with stereo linear audio heads. The tapes I'm currently working with have only a mono track, but I get occasional distracting hitches in the audio that are only present in the right/left captured channel and not the other, so I need to collapse the audio into a single mono channel in software to fix them. (Well, I have to normalize the channels to equal volume first, since their relative volume depends on how the tape was recorded.) This means I have to worry about sound cancellation from phase differences. I'll probably end up correcting the audio phase in software anyway with Stereo Tool, but I'd like to start out with a properly aligned VCR at least.

As a result, I'm now trying to calibrate the azimuth for my VCR's linear audio heads, and it's a huge pain. I've done it before by comparing channel differences with a pretty high-latency method: Capture part of a tape with Virtualdub, export the audio to .wav, import to Audacity, normalize the channels separately, split the stereo channels, invert one, rejoin the channels into stereo, then collapse the channels into mono for the difference between the two channels after normalizing. Then turn the azimuth screw a bit, do the whole process over again, and compare the two differences to see whether I made it better or worse...lather, rinse, repeat. I've spent all day doing this before, and I eventually got my VCR aligned to my satisfaction (the difference track was very faint), but I've recently found that it's all out of alignment again. I'd hate to spend just as much time doing it this way all over again, so I'm looking for better tools.

Is there any Windows or Linux software that can visualize or otherwise report phase differences in real-time (or with some reasonable latency)? I would love to have something that worked on line-in audio, so I could turn the screw and get real-time feedback, but even having something that worked on a .wav file would save me a lot of time. ProPhase seems to do this, but it's MacOS only, and I'd like to find something I can actually use. Thanks smile.gif


For openers, doesn't the VCR have analog audio outputs? Do your monitoring from there!

Real time PC-based oscilloscopes and other analyzers are out there. What is the test tape that you are using?
Go to the top of the page
+Quote Post
chrisz78
post Jul 27 2012, 23:05
Post #7





Group: Validating
Posts: 6
Joined: 23-May 12
Member No.: 100095



QUOTE (Arnold B. Krueger @ Jul 27 2012, 18:54) *
Real time PC-based oscilloscopes and other analyzers are out there.


Where? I have never found anything but huge, unwieldy and expensive audio workstation appliances that do this. Is there a freeware PC software or VST plugin that does e.g. Lissajous figures from stereo audio input?

QUOTE (Arnold B. Krueger @ Jul 27 2012, 18:54) *
What is the test tape that you are using?


It doesn't make sense at all to use a separate test tape if the task is (as it seems to be in this case) to transfer old amateur recordings with optimum quality. You will have to adjust the heads for every tape to mimic the azimuth setting of whatever machine the tape was recorded on. This is true not just for VHS video but also for compact cassettes and open-reel tapes, though the effects of misalignment are nastiest with the very slow linear speed of VHS tape.

Chris
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Jul 27 2012, 23:59
Post #8





Group: Members
Posts: 3647
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (chrisz78 @ Jul 27 2012, 18:05) *
QUOTE (Arnold B. Krueger @ Jul 27 2012, 18:54) *
Real time PC-based oscilloscopes and other analyzers are out there.


Where? I have never found anything but huge, unwieldy and expensive audio workstation appliances that do this. Is there a freeware PC software or VST plugin that does e.g. Lissajous figures from stereo audio input?



PC Oscilliscope software
Go to the top of the page
+Quote Post
Glenn Gundlach
post Jul 28 2012, 09:30
Post #9





Group: Members
Posts: 364
Joined: 19-April 08
From: LA
Member No.: 52914



QUOTE (mzil @ Jul 26 2012, 16:19) *
I think Cool Edit Pro will let you visualize a Lissajous pattern showing the relative phase of the two channels and you can tweak your azimuth from that, but I'm not sure.
<snip>


You're correct about CoolEdit Pro / Adobe Audition and Lissajous. Here's something you may not have thought of. When using X-Y display (Lissajous) to set azimuth you're assuming that the gaps of the 2 channels are collinear. If there is an offset between the gaps you can have azimuth perfect but channels displaced in time. On an X-Y display this will look the same a azimuth but in fact is different and if a large enough offset can actually damage the HF response by lining up the phase. Worse yet, you have no way of determining whether is was the record or play head (or both) that has an offset. Just one more thing to like about analog recording.

G
Go to the top of the page
+Quote Post
pdq
post Jul 28 2012, 15:22
Post #10





Group: Members
Posts: 3372
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



The OP's objective is not to maximize HF on each channel. The recording is mono material, but with dropouts in one or the other channel. What he wants to do is phase match the two channels and convert to mono without destroying the HF due to phase mismatch.

Edit: What he perhaps should be doing instead is using azimuth adjustment to maximize HF, then after digitizing, shift one channel to get phase matching between channels before converting to mono.

This post has been edited by pdq: Jul 28 2012, 15:23
Go to the top of the page
+Quote Post
chrisz78
post Aug 4 2012, 12:16
Post #11





Group: Validating
Posts: 6
Joined: 23-May 12
Member No.: 100095



QUOTE (pdq @ Jul 28 2012, 16:22) *
The OP's objective is not to maximize HF on each channel. The recording is mono material, but with dropouts in one or the other channel. What he wants to do is phase match the two channels and convert to mono without destroying the HF due to phase mismatch.

Edit: What he perhaps should be doing instead is using azimuth adjustment to maximize HF, then after digitizing, shift one channel to get phase matching between channels before converting to mono.


A mono compact-cassette or VHS recording will NEVER have an inherent time-shift between both channels, as it was originally recorded by a single-channel recording head covering both stereo tracks plus the small gap between them (unless you assume that there are recording-heads out there with a vertically curved or angled head gap!). Neither would a full-track mono open-reel tape played as two half-track stereo channels for digitizing purposes (or a half-track mono played as two quarter-track channels) behave any differently. You always get max HF when the orientation of the playback head's gap fully agrees with the recording head's, and that automatically gives you zero phase and time shift between the two stereo channels which are really just two opposite edges of one original mono track.

With an original mono recording subsequently dubbed to a second or further generation on stereo machines, the problem you describe might well occur, but usually the high-frequency content by then is so badly compromised that you won't find much difference between "max HF" and "zero phase" positions either, and going for zero phase/time shift always has the advantage that you can mix the two tracks together to improve s/n ratio: in-phase signal is doubled in the mix while non-correlated random noise is not, so you gain about 3 dB s/n ratio and have more leeway to boost the treble afterwards. You cannot really rescue an n-th generation dubbing made on misaligned tape machines anyway, so I'm assuming the OP is generally working with first-generation recordings where azimuth makes a real difference.
Go to the top of the page
+Quote Post
SuspiciousLizard
post Aug 17 2012, 00:26
Post #12





Group: Members
Posts: 2
Joined: 26-July 12
Member No.: 101790



After I read the first reply here, I realized that going by a Lissajous pattern wouldn't be as good as actually hearing the difference channel myself. I wish I had kept up with this thread though, because I never knew - as splice pointed out - that I could actually listen to the difference channel directly. I ended up using a VST server instead with an appropriate plugin that showed the difference channel and levels. It made it a cinch to adjust the levels and then the azimuth!

Unfortunately, the VCR's sound circuitry has gotten so wonky (likely bad caps or something) that aside from high noise levels, the audio levels are totally inconsistent for multiple captures of the same tape. Sometimes the levels are too weak, and sometimes they clip. It's the best VCR I have for video by far, but after all this, it looks like I'll be recapturing the audio with something else anyway and piecing them together...go figure.

Anyway, thanks everyone smile.gif

This post has been edited by SuspiciousLizard: Aug 17 2012, 00:27
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 26th July 2014 - 04:37