H4 has balanced neutrik-style connectors. I use mine regularly.
I thought you said that wasn't what we were talking about.

Frankly, I don't use passive DI's any more, not for fifteen years.
Your untapped schemo above looks perfectly fine, and it's arguably but of course with a bit more distortion and HF rolloff than a quality active solution offers.

Of the common configurations, I lean towards this one as being a bit more stable (center trimpot is optional), and I connect the tap to the sleeve of the unbalanced input.


I'll be sure and check a couple of my old passive boxes if I still have them, although that's more of a side discussion to the specific issue, as you described so eloquently above.

I don't use passive boxes anymore either, I have a bunch of Countryman Type 85 Active FET boxes. Most active DI boxes still use transformers at the output, the active part of the DI is used as a buffer to provide a VERY high impedance input (at least a megohm) on the 1/4" jacks to keep guitar pickup loading to a minimum.

I already conceded that either way works, John, and Countryman is top-notch equipment. I am agreeing with you!

In addition to attenuation, lifting the signal from the shield in some fashion is the one, overriding, essential issue for me, and hum is the biggest reason to use a direct box rather than a pot across an unbalanced line.

Whether the h1 sleeve gets connected to the cold side output or to a center tap, it's mission accomplished, as long as neither 2 nor 3 are connected to the audio shield and pin 1 is not shunted..That goes double if you are passing through phantom.

Did our foot of snow miss you guys yesterday?

Yes, the storm passed north of here, it's just cold, so I'm staying inside causing trouble on the Vegas forum. :)

Me too. Maybe we both needed some attention today.
Best as always.

The cheapo Nady passive DI has a center tap, and is just exactly the way I ran sound two decades ago, especially the cheap part (I suspect my Rolls DB25's were identical). For an unbalanced application, I get the signal from 2 and 1 with ground lift, but 2 and 3 is perfectly fine, too.

Either way, if the sound isn't cleaner from that $23 transformer-in-a-box than a potentiometer, I will buy it from the OP

Wow. Fascinating discussion. Far more than I expected guys -- and way over my head. I hope you enjoyed it. For my part, I will follow through on this in a practical way.

First, I ordered the PAC LC-1 Remote Amplifier Level Controller that John suggested. Cheap enough at $11 .. I should have it in a few days on Amazon Prime. If I get it on Friday, I'll post results on Saturday (California time zone).

I have no doubt it will attenuate the signal so the H1 can handle it --- but, as Musicvid points out, will it reduce the hum and the pulsing sound I'm hearing from the projector in the "silent" areas of the track or very low level areas of the sound track. We shall see soon enough. If results might be improved from there I'll go on to the direct box approach to see what that does.

I just ran this test using a Radio Shack attenuating Dubbing cable rated at 60db of attenuation. This cable has no way of adjusting the amount of attenuation.

Of course we know some of the hum and hiss is from the optical pickups. Running in a totally dark room with no fluorescents or dimmers is always suggested.

What %age of that noise can be controlled by a DI is anyone's guess, but I'm betting the answer is "some."
The clicks may be microphonics on the projector, and those can be hard to tame.
That would be better handled in Izotope.

Oh, you need to turn off all AGC on the Zoom, as well as on the projector if such a thing is possible.

There doesn't appear to be any clipping. You probably don't need a 60db cut but it's OK. There is some regular pulse noise, likely from the current peaks caused by the projector mechanics effects on the DC power supplies to the amplifier or EMF.

You're getting there.

I did turn off AGC on the H1 Zoom after doing a test on "auto". Makes sense that "auto" would not be the way to reduce hum in the low level areas of the track. I turned up the projector pot as high as I could without distortion and then backed it off a bit more, then brought up the H1 input gain to a comfortable level, maybe even a bit low for safety. I purposely choose a section of track with some silence and low level audio.

I never thought of using something like Izotope for the pulsing. I do have Nectar Elements. But not sure if that would be any good for reducing the pulsing.

I had a really quick go on the audio with Izotope RX, here's the result:

https://dl.dropboxusercontent.com/u/8437229/60db%20Attenuation%20cable%20sound%20test%20izo-5.mp3

Wow .. very impressive John .. I never realized such things were possible. That's incredible software but I see the basic version of RX is $350. I doubt I'd want to spend that much at the moment. .. But .. how is that done? Seems like sheer magic.

I see that the Variable Line Out has shipped. I should have it tomorrow. We'll see what that does. Gone for a couple of days -- back on this journey Saturday. THANKS EVERYONE !!

It IS sheer magic, probably Voodoo magic actually. Izotope RX is the single most impressive piece of software that I have ever bought, it does things that I would have considered utterly impossible just a few years ago. I sprung for the Advanced version and it was worth absolutely every penny. I do a LOT of audio and video restoration, in fact, that's my main business these days, I don't really shoot that much anymore.

As for what I did to that clip, I used the Declicker module to remove the periodic noise, I had to widen the click search width to about 3 milliseconds and play with the sensitivity of the algorithm. That periodic noise had significant energy up to about 700 hz and were spaced at 436 ms intervals.

Next I used the De-hum module and used very narrow filters to remove hum at 60hz, 120hz, 180hz and 300hz.

Finally, I used the De-noise module to remove the rest of the noise, the Izotope RX noise reduction algorithm can remove large amounts of noise with inaudible artifacts, it's the best noise reduction algorithm I have ever found by far.

Once you get the hang of it, Izotope RX is fast, I spent maybe five minutes working on your clip.

@ Paul

I'm always impressed by the tools available to fixed problems after the fact. Some of that noise could be prevented if you could find a qualified electronics technician to make some simple repairs to the projector. Here's the order that I would approach the effort as a hardware refresh.

1. Have someone replace the forty year old filter capacitors in the DC power supply of the audio amplifier.

2. Replace the coupling capacitors and emitter resister bypass capacitors (if they exist) in and around the signal path in the audio amplifiers.

3. Make your connection to the Zoom at the output to the first preamp stage as described above.

4. Look at the possibility of lifting the photocell from the projector preamp and using it to drive the Zoom directly. Source and input impedances must be considered. The battery operated Zoom would float electrically with respect to the projector electronics.

Your post production life would be simpler if you could make a friend locally with a soldering iron and curiosity about old audio circuits.

No one will understand better than me, if you have no taste for hardware tinkering.

I wonder how easy it would be to build one's own optical pickup to capture the audio from the film. Seems like a simple cheap LED IR emitter & sensor could do the trick, or even a visible LED and a CdS cell. Run a volt or so through it into a moderately small capacitor to filter out the DC, then into the line input of an amp. Mount the emitter/receiver in a small black box with a slit for the film, and fasten that to the projector either just before the film enters or just after it leaves.

Two things to consider: 1) Make sure the receiver is masked so that it only gets light through the audio track. The spot must also be very narrow along the long dimension too for better frequency response, just like the head gap in a tape player. 2) Sync will be off since the sensing is done in the wrong location, but that's certainly easy enough to fix in Vegas afterward as it will be a constant shift.

In any case it seems like it should be possible to create one with far less noise that the projector's circuitry.

Analog optical audio is noisy, period.
I designed and put together Kelly's project with a student, and the analog hiss was just as bad as what we heard on that projector clip.

You catch the downstream stuff (john_dennis' capacitor tips were first rate), and iZotope was apparently made for this stuff.

Be sure to document you finished technique and workflow so you can post it here.

Maybe John Meyer will see this and drop in to share his wisdom..

I called a friend of mine in Los Angeles that has worked in film audio for decades, he said that the optical audio track on 16mm film is only about 1.83mm wide and has a typical frequency response of +0/-6db from 100 to 5,000hz and a signal-to-noise ratio of 40db at best. Film scratches, dirt, dust, fading of the film stock, and even film grain are the limiting factors more than the electronics. He said that he's working on a system to scan the soundtrack using a high resolution optical film scanner and use digital image processing to clean up the image and recover the audio that way. He also asked if I had ever heard of Izotope RX and said that there is nothing better for cleaning up the audio from optical sound tracks.

I agree that beefing up the capacitors in the projector's power supply would at least reduce the hum, although Izotope RX can deal with that quite easily.

Would that be a mutual friend, John?
How's he doing?

Got the variable "pot" yesterday. I'm going to give that a try today. Keep in mind folks, I'm very much over my head on this discussion. But it is fairly easy to see how terrific the iZotope RX would work on something like this -- as a practical matter I would likely not refurbish the projector to clean things up that way (too complicated and expensive) .. but I will publish some results and the final workflow for the 16mm release prints I have to transfer -- maybe an hour and a half of film so far. I'm stuck doing what is practical for a low volume personal project.

But this has been an interesting discussion. The KINOGRAPH project might be of interest to some of you here (if you aren't aware of it). Part of this is an attempt to retrieve the sound from the optical waveform on the film. There is an example of a transfer at the end of this clip which is not nearly as good as what RX can do. Seems like even a scan of the optical track does not eliminate all problems. So far, I think post production using RX is the best answer (if one can afford it).

Personally, I'm just hoping this old Singer projector holds out long enough for me to make all my tests and then the actual transfers. I'll post current results later today.

Your projector actually sounded reasonably quiet compared to some I've heard.
And that's where izotope is at its best, with light to moderate noise. The results can be literally breathtaking.

With Grazie and John Meyer at the controls, they literally saved the big dance number in my production of JCS.

Sorry. Forgot to post the KINOGRAPH clip. Interesting.

Kinograph Project Film transfer

Ok. Did several tests with the LC1 Variable level control unit. Here's the best one. It works but I can't hear any improvement over the 60db attenuation cable with level control. Still plenty of hum and pulsing.

Perhaps a different combination of levels between projector, H1 unit and the LC1 would improve things but it does not seem so to me.