Slit Loss equalization improvement

[Renato Bulhoes]

Hello friends.
I'd like to understand better the slit loss control within a preamplifier circuitry. Sometimes called slit loss correction or equalization. I've seen some kind of ' pre tone amplifier ' PCB boards everywhere throught internet; would it be a good (MIC ?) preamplifier for film sound ? And what range of frequency controls look for within those kind of adjustments for Dolby SR, A, mono... slit loss eq correction/optimization ? I read an early archived post on which Mr. Guttag mentioned something like a sin(x)/x filter as beeing a better alternative control, so I'd like to understand the relationship in between them. Thanks !

[Steve Guttag]

Mic preamps CAN work as far as getting the level amplified in a low-noise manner. Some gotchas include that a (solar) cell likes to be a "current source" when reproducing audio which means that input impedance should be very low. With some mic preamps doing better if a loading resistor is put across the cell's output to drive the impedance lower. Another gotcha is that it is a solar cell. It generates its own DC voltage and some mic preamps cannot deal with that as it will bias the circuit so some form of DC blocking may be required (e.g. a capacitor or a transformer...each coming with potential frequency limitations too).

Rather than me explain a well documented cause/effect relationship between slit loss correction, I'll refer you to many Dolby manuals that have it documented in their Appendix:



Screen Shot 2020-04-25 at 8.43.01 AM.png

[Renato Bulhoes]

Wow thank you so much, Mr.Guttag. I've been researchig this kind of documents. Thanks a lot.

[Sean McKinnon]

I have heard (and experienced) that with modern reverse scan red LED readers that slit loss correction usually isn't necessary and may be detrimental. What are your thoughts on that Steve? I always turn the slit loss all the way down then focus the optics. If I can get flat out to +/- 16k I leave it off.

[Stefan Scholz]

Firstly, slit loss may be considered a linear distortion, and therefore reversible to some extend, it should not be the proper approach. Even in an analog domain, informational content lost is lost can never be recreated without artefacts. A slit loss correction is a tool to correct an improper reproducer, where the slit height is too high. This might have been done on older projection gear to achieve higher output levels, at the time, when amplification was different, in early tube amp days, and before solid state pickup cells became standard. As projectors tended to have a very long practical lifespan, it was common at the introduction of bilingual Dolby tracks to find slit optics with a very limited hf response. These required EQ, even though the correct way would have been to replace them with better optics, which could have been done in some cases, but due to cost ($1800 per lens) would have prevented some sales
The proper way, to have a slit small enough to not give attenuation at the high end, but still not be in the region of optical interference from a slit is best done in a macro scanning setup. This does actually not require much compensation for the linear distortion.
I have never tried to use SLC in my setups. As stated, lot of bad optics existed, and the best way to remove them and replace with more adequate ones was possible, requiring less EQ, and therefore better overall results.
Good optics existed in 1985, and in 1995 the macroscopic reproducer was standard in new projectors, and for a fairly low price upgrade kits from different vendors were available, and this became my preferred method.
AFAIK my projector setup has the SLC set to 18 kHz, and we get a good linear response to 12.5 kc, which is the border printed tracks on color film normally can go for. Black and white can go to 16 kc.
I would not go the way of SLC these days, if required, I would try to locate a macroscopic rebuild kit for the sound head in use, or try to convert an existing one.

[Steve Guttag]

Sean,

The whole optical soundtrack is a minefield when it comes to quality. It's chief advantage is that it is duplicated at the same time as the picture, making it highly attractive when it came out and even until its later years. This was clearly long before there was any sense of "high fidelity", let alone "stereo," in its many forms.

The upper end of the frequency response is determined by slit height. As the period of the cycle approaches the slit height, the response will diminish (the same is true on magnetic tape reproduction and why playback heads have smaller gaps than record heads).

Slit loss (or HF EQ on a magnetic system) will let one get that last bit out of the system.

With optical sound, the way to get higher frequency response is to use a smaller slit. However, as the slit height goes down, so does overall light level and consequently, your overall level. To compensate for that, you can raise the gain but that raises your noise right along with it. Most soundheads used a 1-mil slit but there was a narrow slit option (Sankor made most of the slit lenses for US projectors and the two can be distinguished by either a white or red index line. The red lines are narrow slits. I, honestly, forget what their slit height is (either .7-mil or .9-mil) but they did improve HF response.

So, moving over to reverse scan...there are both the same dynamics as well as some new ones. The whole waveform/cycle versus slit height remains but now we have an optical component of a lens that can magnify the soundtrack to effectively attack the problem from the other end...make the waveform larger. The problem of light remains but with the modern reverse scan using a bright LED (first IR and then visible red), we have more options than a very dated exciter lamp design (think about it, there isn't even a reflector...half the light generated doesn't get sent to the film/cell). This design does allow for higher frequency response, natively from the system. Another piece to the puzzle is the little preamp on the reverse scan reader to pre-preamp the signal before hitting a typical sound processor. The light is also different besides the color. It is collimated. This shows itself in the form of a more pronounced noise if the likes of scratching are on the film. The straight light waves will catch those scratches with a more pronounced shadow. I quick and dirty fix for a bad print is to put some scotch "magic" tape on the LED to make a "diffuser" to scatter the light some.

So back to the question...do you use slit loss correction or not? The answer is, it depends. First, before you go adjusting it consider what you are using as a reference. Is it Dolby's CAT69P? I can tell you that those varied all over the place. So you may be adjusting to the response of the test film more than the response of the reader. We ended up using NTAV's pink noise film...it was ruler flat and yes, you can get a smidge more HF response with SLC on most soundheads. When adjusting SLC, you should pay attention to your X-Y (lissajous or "Lots of Juice", if you're 'merican). As you increase SLC, you'll see your phase go wacky so it is very important to match the settings.

If you haven't tried one, see if you can get your hands on a Panastereo CSP-1200. The optical preamp in them are significantly better than in a Dolby or even lesser processor. I'm serious. It will be a jaw dropping sort of moment as to just how good and tight you can get an A-chain with a reverse scan reader, good test film and that processor. Not only will the frequency response be flatter but the X-Y response will make you think that you've shorted the two together it is so tight. But you need good test film or, again, your are measuring the response of the film instead of the soundhead. With a Pana, you can get, again, a smidge of more HF.

But let's step back a bit and look at this. What are we trying to achieve here? For a traditional soundhead, we're hoping to get things up to 12.5KHz. That was high-fidelity in the day. Heck 4-track magnetic used 12KHz as a pilot tone for the surrounds so clearly they weren't recording anything up there for the surrounds.

We have a screen in front of the speakers and that takes a lot out of the track above 9KHz. And then there is the question of what is being recorded on the film to begin with...not just what they laid down at the dubbing stage but also as it went through the process of being made into a composite negative and then the high-speed printers to make the release print. Ever look at a soundtrack projected? They are not the most steady things. So what valuable information do you expect to be on the track up near 16KHz? Of that, how much do you expect to actually get to the audience's ears? It is mostly noise up there.

As such, I don't fault anyone for just turning SLC full-down and moving on. But, if you want to get that last little bit, then sure, you can use it but be sure you are doing more good than harm. For the record, I did always set it though on the CP650, I set both controls to "15" as I once found that seemed to get the best response out of the CP650 on a typical reverse scan system (both Component Engineering and Kinoton RSSD).