Question for the Sony guys: Does a very small amount of "Time Stretch" cause ANY audio degradation?

Comments

GlennChan wrote on 2/26/2004, 5:56 PM
I've done this myself, and the mini-disc recorder I used took about 30 minutes to go off 1 frame. However, your milelage may vary depending on a lot of factors like the recorder and temperature changes. Still... I think there is something wrong when you get out of sync that fast. It might have to do with sampling rate changes.

The thing to do is to stretch the audio to sync, but DON'T pitch shift. Pitch shifting will introduce artifacts you don't want. By slowing down the audio pitch will drop very slightly, but I doubt you will notice it.

Jay Rose has an article over at dv.com about sync. http://www.dv.com/dv_login.jhtml?_requestid=31315
Former user wrote on 2/26/2004, 7:30 PM
Cameras record at 29.97 also. They wouldn't record at 30fps unless you have some type of override.

Dave T2
riredale wrote on 2/26/2004, 11:02 PM
Videocurmudgeon:

When I do my documentaries, I often tape a complete choir performance, which runs about an hour. My plan, incidentally, is to do my next project entirely in surround sound. For handheld work, I will have two mics on top of the camera, one facing forward and one aft. The forward one will go on the miniDV tape, and the aft one will be recorded by a Minidisc unit velcroed to the side of the camera.

For performances, the two microphones will be separated and will be feeding two Minidiscs. The camera's stereo microphone will go on the miniDV tape and be used for emergency backup and master sync.

This is getting off-topic, but I've spent some time with my new surround-sound PC speaker setup listening to commercial DVDs. I've never had a surround system before, and it's interesting to see how and when Hollywood puts sound in the rear channels. To me it appears they are using the rear channels primarily for show. Sometimes there is simulated ambient sound, sometimes there isn't. Sometimes there's music, sometimes there isn't. Perhaps I'm a bit naive about all this (after all, they are making big bucks, doing this for a living, while I'm just hacking around), but it seems to me that it would be really fun to put ambient sounds coming from the rear onto the rear channels ALL THE TIME. But maybe in reality it will turn out to be too distracting.
RichMacDonald wrote on 2/27/2004, 10:35 AM
>it seems to me that it would be really fun to put ambient sounds coming from the rear onto the rear channels ALL THE TIME. But maybe in reality it will turn out to be too distracting.

Less is more. Remember, its *ambient* sound, its not supposed to cause you to turn your head around. Unless that's what you want :-)

Having said that, I always put something in the rear speakers...at least a reduced volume of what is going to the front left and right. Its personal preference from always listening to music with 4 speakers, but I like the sound to come from "inside" the room, rather than from the front wall. Especially music. It makes the location of the speakers disappear (you hear sound from "space", rather than hearing it come directly from the speaker), which gives the impression that they're better speakers than they may actually be :-) And having the "room making the sound" is closer to the live performance feel.

Riredale, I suspect you will not be at all pleased with the results of placing one mic forward and one aft. This is not a good way to capture "surround". Specifically, you will now be recording the main attraction in mono, while the rear channel will be mono and "wierd" because it is too dependent on the room and reflected colorations. Go to a live concert and actually listen to the reflections from the back wall: Not pleasant at all by themselves. So you'll be stuck with mono front and mono rear and almost no control in the studio. Better to use the two mics as front stereo. Get that right first. Then use reverb, echo et al. in the studio to mix your rear channel. IMHO, that'll sound much better, and give you far more options in the studio.

Alternatively, send the front stereo mics to a single minidisc -- cardiod/directional mics aimed at the stage -- record the camera stereo mics to miniDV as backup, and add a 3rd omni mic placed in the audience somewhere and hooked to your second minidisc. Plan to mix the 3rd mic as the rear channel, along with an effect-enhanced blend of the front mics.

Or forget the 3rd omni mic and use your camera mics to record the "rear channel". It'll take a lot of mixing in the studio and can be a crapshoot, i.e., it may sound horrible and nothing you can do about it. "Two mics can be a lot worse than one".

Did I mention thorough testing and evaluation long before the performance :-?

Pardon me if my response was long-winded or missed the mark. I seem to be doing that a lot lately...
riredale wrote on 2/27/2004, 10:47 AM
Rich:

Not long-winded at all. I gave the wrong impression in my earlier comment. I am using stereo microphones front and rear. So I am pulling in four channels.

You may be right about the ambient sound being a distraction. Nonetheless, when I was in college, I came across some fascinating papers about "total immersion," where a listener was enveloped in sound. Sounds originally coming from a particular direction were reproduced in the listening room by a speaker located in that direction. And so forth. In other words, the idea was not to create an artificial sound stage in your living room, but to actually immerse the listener in the sound field that existed at the live event. But again, it may be too weird for the typical listener.
Catwell wrote on 2/27/2004, 12:50 PM
Have you ever listened to a binaural recording. This is a recording made with two microphones in the ears of a dummy head. When you hear it through headphones the effect is amazing. Every sound is placed in perfect location to the original. It is the true immersive experience.
RichMacDonald wrote on 2/27/2004, 3:27 PM
Ok, got it now. As far as "ambient sound being a distraction" I was only thinking of the horror of a front-mono plus back-mono recording :-)

As for the "head-mics" post, has everyone heard the Joe Jackson album "Body and Soul"? As the story goes, Joe was sick of the studio sound, so he rented a concert hall and stuck two Neumann M-50s (I know nothing about this mike; just looked it up) 15 feet up in the air just where your ears would be if you were a giant with an extra wide head :-) Basically they just played and kept the mix to a minimum. IMHO, its one of the great live recordings of recent time. Anyway, if you haven't heard it, look for it.
plasmavideo wrote on 3/8/2004, 1:51 PM
I just came across something pertaining to this in another forum.

A mention was made that capturing you MD audio through the camera a/v inputs across firewire would cure the time difference. Probably worth a look!
riredale wrote on 3/8/2004, 4:18 PM
Hmmm...

I don't see how it would have any difference, except... Well, if the camcorder was the source of the difference between record and playback speed, then, yes, I see how copying the Minidisc audio over to miniDV first, then playing it back into the PC via firewire would, in effect, give the same "delay" to the Minidisc audio. I'll try playing with this later tonight if some time is available.
SonyEPM wrote on 3/9/2004, 8:41 AM
"Does a very small amount of "Time Stretch" cause ANY audio degradation?"

Technically, yes, always. However, our timestretching is really good, so tiny amounts should be at worst barely noticeable. In most cases even highly trained ears would have a hard time hearing the dif on a small stretch of 16/44.1 or 16/48, but it does vary with content, amount stretched, stretch mode used, resampling quality used, bit depth and sample rate used.
riredale wrote on 3/9/2004, 2:11 PM
Plasmavideo:

Thanks for the interesting idea. Last evening I went back and tested some more.

To recap: A few weeks ago, I did a test to see if camcorders and Minidisc recorders could stay in sync for long periods of time. I set up a Sony TRV8, a Sony VX2000, and two Sharp MD-MT15 Minidisc recorders in parallel. I did a few handclaps at the beginning and end of a 30-minute period.

After bringing the files into Vegas and putting them on a timeline, I noted that the two camcorders were exactly in sync. The two Minidisc recorders were exactly in sync with each other, but they lagged the camcorder recordings by 15 frames, which would mean 1 frame every two minutes. Not a big deal, since Vegas lets one shrink the length of audio clips, but still an extra step.

On Plasmvideo's suggestion I went back to one of the Minidisc recordings and this time I fed that recording (real-time) onto a miniDV tape in one of the camcorders. Finally, I brought that miniDV copy of the Minidisc recording into Vegas.

Sure enough, the recordings are now in perfect sync, even after 30 minutes. So the implication is that Sony miniDV camcorders (or at least these two) playback very slightly faster (0.03%) than they record. I don't, however, think it will be my plan to record my Minidisc audio onto miniDV before bringing it into the PC, since it means it will take twice as long to load audio. And of course for short clips the diffference won't be noticeable.
craftech wrote on 3/10/2004, 5:07 AM
riredale,
I recently did something similar in that I fed my minidisc deck's RCA outputs into my DV deck's RCA inputs (I could have just as well used my miniDV camera). Then I used VidCap 4.0 to capture it via firewire. That resulted in a perfectly synched audio track and a blank video track which I simply deleted after bringing it into Vegas.

On a different note:
I have also done the same thing to create an audio CD. In that case, I left the video track on the timeline which Vegas ignores when creating an audio CD.

John
Jay Gladwell wrote on 3/10/2004, 5:11 AM
Very interesting indeed!

My question would be: Does this process diminish the quality of the sound? If so, can the human ear hear the difference in quality?

J--
craftech wrote on 3/10/2004, 5:27 AM
My question would be: Does this process diminish the quality of the sound? If so, can the human ear hear the difference in quality?
--------------------------------------------------------------
Short answer....................No.

Long answer. I you have a very high quality soundcard, you might get a better wave file going through that instead. Not worth the synching issues IMHO.

I still didn't get a single response to this post however:

http://mediasoftware.sonypictures.com/Forums/ShowMessage.asp?MessageID=260279&Replies=0&Page=2

John