The Secret Society of Lathe Trolls

tape · Joined: 25 Jan 2007 Posts: 90

This is probably a stupid question-
Does it make sense to eq your music according to the RIAA prior to cutting?

it looks like it boosts bass and dampen highs on playback...

studiorp · Joined: 01 Nov 2007 Posts: 44

Hi, your question is very important about cutting acetates or vinyl discs with the inverse RIAA equalization , this because RIAA filter when you cut your vinyl, decreases low frequencies and increases high frequencies.
When you will play your lp with your turntable and your amplifier, RIAA equalization will works with an inverse mode, so the high frequencies decreased and the low frequencies increases.
I have seen this motivation on Wikipedia :
"The main benefit of the system is that low frequencies, which would otherwise cause the cutter to make large excursions when cutting a groove, are much reduced, so grooves are smaller and more can be fitted in a given surface area, yielding longer playback times. "
Also if you don't utilize this system your lp will not be compatible with all RIAA phono preamplifier.

cuttercollector · Joined: 11 Jun 2006 Posts: 285 Location: San Jose, CA

Yes, you should apply RIAA EQ to your music and also equalize any big peaks from your head. It depends on what cutting system you have put together how much of this already exists in the electronics. Most pro systems have both of these things in the electronics. If not, or an older system that has some EQ but not RIAA, you will have to supliment what is there to get the desired result. One technique I have used is to cut pink noise about 10dB below peak level (but above the noise) and look at the playback on an RTA (real time analizer) hooked to a standard modern playback system with a reasonably good flat frequency response cartridge and preamp. Then, just adust the eq of your incomming signal for as flat a result as possible on playback.

JayDC · Posted: Fri Nov 02, 2007 4:40 pm Post subject:

so what your saying is 2 EQ's. One to Equalize the Head, and one to apply the curve.

Ah-ha..

Cool, so pump the pink noise through the head to an acetate, playback the acetate through some sort of wave form analyzer, then adjust the first EQ to flatten the peaks..

I bet this could make my setup sound even better.

#1
source -> EQ1 (head equalization) -> EQ2 (inverse RIAA) -> cutterhead..

or

#2
source -> EQ1 (inverse RIAA) -> EQ2 (head equalization) -> cutterhead..

Number one, or Number two?

cuttercollector · Joined: 11 Jun 2006 Posts: 285 Location: San Jose, CA

Well, I didn't mean to make it so complex. An RTA is the type of device used to "shoot rooms" as they say to flatten room response by playing pink noise through the sound system and adjusting the eq till it is "flat" by microphone measurement in the seating area. These systems can exist as stand alone or as software. If they have a pink noise output and a normal line input you can check the frequency response of any component or system. Record play response of an analog tape machine for example. One eq can do the job (which may or may not partially be done for you in the cutting electronics already).
You would feed the pink noise generator output into the eq where the normal program line level source would come in. Then to the disc recording amp and cutter in the same way as normal. Cut the pink noise on the disc at a moderate average program level. Hot enough to be above surface noise but don't blow up the system. Play the result back through a known good flat cartridge and phono preamp back into the Real Time Analizer. The resulting frequecy response curve is then what you want to correct to flat with the eq. So take note of how far up or down it is at various frequencies and apply the inverse of that curve to the eq. Like if it is up 6dB @ 4Khz apply -6dB to the 4 khz slider of the eq. RTAs are generally set up on 1/3 octave bands so it would be good to use 1/3 octave eq to set the system up. Anyway, once you have set your first compensation "curve" cut again and play back in the same way and see if it is closer to flat. It might take several cycles of cut/play/adjust eq/cut play before you get it in the balpark, even in the midrange. Just try to iron out any big peaks and make the overall response as flat as possible through the midrange. DO NOT apply excessive boost to the eq to counteract a badly drooping response at the high end. Your cutter may not respond that far out and you will likely blow it up if you push it!

tape · Joined: 25 Jan 2007 Posts: 90

well, it was not such a stupid question anyway -
thank you very much for the very good answers!

JayDC · Posted: Sat Nov 03, 2007 7:29 am Post subject:

Well, I didn't blow up the cutter, but did what you said... Found out that my cutter is pretty flat as is.. I boosted the high frequencies, and got a better response between 8K-10K, although with this boost I have to raise the volume of the amp in order to maintain the same relative loudness of the cut as before.

I was pushing about 4-5vac into the cutter, now it seems to be between 6-8vac. The cuts sound great, but I'm afraid that pushing that much current through the head for prolong periods of time may result in damage.

I wonder if there's a way to cool the head down, while making these loud cuts, in order to prevent damage. Thinking of possibly wrapping the head in flexible copper tubing, then pumping chilled water, or CO2 through it.

any who, I have recorded pink noise to a .WAV thats about 1 minute in length, for anyone that may need it.

tape · Joined: 25 Jan 2007 Posts: 90

Just a thought: why pink noise?
Would white noise not produce a more flat reference signal?

cuttercollector · Joined: 11 Jun 2006 Posts: 285 Location: San Jose, CA

Well, I refer you to the wikepedia artcle on pink noise which says in part about pink noise:

"There is equal energy in all octaves (or similar log bundles). In terms of power at a constant bandwidth, 1/f noise falls off at 3 dB per octave. At high enough frequencies 1/f noise is never dominant. (White noise is equal energy per hertz.)"

RTAs are set up on iso 1/3 octave centers so the filters measure energy in each individual 1/3 octave passband. If you then have a 1/3 octave eq on the same centers and adjust the energy in each pass band of the eq for flat response, the net result is , well, flat! If you apply this to any system the net result at least in theory is flat response.

It might be possible for example, to build a cutter head that would mimic the RIAA record curve exactly and so require NO eq to produce a net flat response cut when played back on a normal turntable/preamp that follows the RIAA playback curve.
Another example are crystal or piezo/ceramic playback cartridges which although limited (mostly by being made cheaply) in many respects, tend to have a curve when running into a high load impedance which mimics the RIAA playback curve, thus needing very little playback eq.

Regarding blowing up your head, if you do increase the cutting voltage and that increase is mostly at high frequencies you do run the risk of blowing it up. It's true with speakers too. You can't put as much level through something where you have boosted either bass or trebble frequencies to compensate for a native lack of response of the drivers. All the energy goes into trying to force the drivers to respond where they don't want to and just is dissipated as heat in the windings till something blows up.
That is why I said, focus on getting a flat overall midrange response, prefrably by _cutting peaks_ with the eq, moving the sliders below zero.
Be very careful about applying compensating boost (above zero) especially at the extreeme ends of the spectrum.
If this were possible to do to an extreme with boost and amplifier power you could make some crystal cutter in an old recordio flat from 20-20K and pump in enough power to make your cuts as loud as the latest hot dance records - _NOT_!

So, don't try to make your cutter more that it is. But I will say that one obvious sign is an increase in distortion. If you can hear the cutter "complaining" when you play back you are probably over driving it at those frequencies and are in danger of plinking it.

cuttercollector · Joined: 11 Jun 2006 Posts: 285 Location: San Jose, CA

A practical example.
When I tried this with my Rek-O-Kut head driven by an external power amp and eq, before I got it re-magnetized and rebuilt, I was able to get pretty nice sounding cuts, but at quite a bit lower level than most commercial recordings. The net response on the RTA after equalizing, was pretty flat (+/- about 4 dB) from about 60-8Khz, then rolled off slowly above that. There was pretty much no response above about 12Khz or below about 40hz. That's pretty typical of a non feedback cutter. After rebuilding I would expect the overall level to be quite a bit higher but not much greater frequency response. I think they use helium to cool the pro heads but I am not sure on this.

JayDC · Posted: Sat Nov 03, 2007 7:45 pm Post subject:

Okay, but how do they use the helium?

Do they discharge it directly on the coils, or cool the whole head down.

Would discharging the gas rapidly over the coils produce extra noise?

Also, could the frequency limitation be due to the mechanics, and not the electrics?

JayDC · Posted: Sat Nov 03, 2007 9:23 pm Post subject:

WoWOWoWoWoW!!!!!!

Flat response from 30hz-11Khz on a Presto 1-C, and it sounds incredible. Still cutting about 8-9vac peaking at about 29vac when hitting 10Khz at 3db/0. Recording level off the plate are at 10db/0 from the db meter on my pioneer DJM mixer. Don't see any smoke, so I assume its ok with the increased levels.

I clipped th 10K band a little on my last cut (there was a loud noise sweep that peaked 4db@10Khz), so it crunched for a second, but other then that it sounds sooooooo friggen great!..

At one point I thought I blew the head. During a test cut, about half way through the sound started to distort, and the volume drop a lot (from 10db/0 to about 1db/0). F*&k, I started to dismount the head so I could take a look at the coil, to see if I needed to send it to gib. Turns out, the vibrations from the head loosened the screw that holds the stylus. I disconnected the heating wire, and the stylus just dropped out.

I actually think I might be running out of amp at this point.

cuttercollector · Joined: 11 Jun 2006 Posts: 285 Location: San Jose, CA

Hmmmm, what is that hot enamled magnet wire smell?????

JayDC · Posted: Sun Nov 04, 2007 7:05 pm Post subject:

I don't smell anything yet. Do you happen to know at what temperature that wire will fail at? I have a digital temp gun to measure the heat.

Doug 6N · Joined: 21 May 2007 Posts: 66 Location: Washington

Hey Cuttercollector:

I'll email you sometime. It's good to have people like you on the board who can articulate an issue. I have the knowlege but am not good in the articulation department. Very Happy

Having previously worked with sound. I do own a RTA and Pink Noise Generator and often use it just as you describe. It's a handy gadget. Smile

Doug

JayDC · Posted: Wed Nov 07, 2007 10:20 pm Post subject:

It's official I need another amp. I'm using a 100watt amp, and I don't have enough gain. I've been cutting a few track with my new EQ settings, and I've found that I now have to rasie the volume of my source to the point it distorts in order to have enough gain to cut at a proper level. (gives me a little sizzle on the high end, due to the overdrive of my audio source)

So now I'm looking for a 200watt amp. This should give me enough room to peak the sound. It'll to allow me to drop my source level, and it should still should be loud enough to get me those 10db/0 cuts.

I think it will give me a little more clarity at high levels, on the high frequencies.

I'm looking it a Mackie FR-800, and its 205watts per channel. Was also looking at a crown, but I didn't like the I/O situation.

I think I might buy it.

Doug 6N · Joined: 21 May 2007 Posts: 66 Location: Washington

Hi Jay:

Me thinks there may be some other issue with amount of gain. As the 1C head can work quite well with a 10 watt amp. I'd say the 100 watt your using is a great plenty. Sounds like you must be losing signal. Possibly do to a bad amp but not due to it being only 100 watts and getting a higher wattage model.

Just my humble opinion Smile

Doug

JayDC · Posted: Thu Nov 08, 2007 4:26 am Post subject:

is that 10watts before or after the curve?

This 100watt amp is turned up to 11, and is showing an average of 5 volts, when my source signal is at -.05db. I would like to bump that average of 8 volts, like I did during the test, without have to raise my source above 0db (post inverse-RIAA). The digital clipping on the high end is sounding a bit sizzling. I'm thinking more gain on amp, less gain on source = clear sound. The problem is I have no more amp gain. I don't mind having a bit too much amp, I'll just keep it low.

cuttercollector · Joined: 11 Jun 2006 Posts: 285 Location: San Jose, CA

Voltage gain is not power output wattage. Sounds as if you might have need of more drive voltage (gain) from your program source to drive the existing amp to full output. It's all realative unless you establish a known standard. So you might need a line stage capable of more output voltage. You loose gain in the eq process if your average is more sliders cut than boosted, but some eqs have a overall output gain control to compensate. Also, is your presto a low impedance model or 600 ohm? A normal solid state amp with no output transformer will never deliver full rated power into anything higher than 8 ohms, much less 600. This all was somewhat masked by cutting at lower levels. Try something like a DJ mixer between the eq and power amp for more gain. I don't know what your amp is, but some amps have a rated voltage input for full power output. It can be anywhere from .3-.5V up to 3-5V depending on the power amp.
Voltage across head divided by head impedance gives current across head. Current across head multiplied by voltage across head gives power across head in watts. Have you tried your cutting amp and source across a normal speaker and listened to how loud it plays before the INPUT signal to the amp clips? I don't think the Presto heads will take more than about 50 watts peak before blowing up. The average should be MUCH lower than that.

JayDC · Posted: Thu Nov 08, 2007 3:14 pm Post subject:

The head I have is 15ohms. It was tweaked by Gib. YOur right, I guess I could just use a DJ mixer or something else to boost the gain before the amp, I was just trying to have the least amount of equipment in the signal chain, to cut down on the noise.