Tape Saturation

[amz-fx]

Anyone know of some frequency graphs of tape saturation in action?  When you feed in a 1k sine wave, what do you get out?

My Google-Fu is rusty and I cannot find very much in the way of details. This is strange since there are many vst plugins that claim to do tape saturation. Are the developers testing it and then keeping the info private, or just making up some basic distortion and calling it tape saturation?

I have a tape machine I could use, but it has been sitting around idle for so long that I am afraid that the pinch roller has become gummy... I have already replaced it once because of that.

Anyone got some good data to share?  The slim info that I found so far seems to indicate a soft bend into odd harmonics as the tape signal is increased.

I have a couple of vst tape plugins so I might run them into an analyzer and see what they do.

Best regards, Jack

[snk]

There are several factors.
The "tape" sound can be many things (and many tape machines can be very transparent!). It's not only saturation : the dynamic factor is also to take into consideration, so feeding a 1k sine wave wouldn't be enough to get a good picture of what can happen with tape processing.
Also, I'd say that the whole "recording to tape" process can define the overall "tape sound" : pre-emphasis and post-emphasis (boosting the treble while recording, and boosting the bass during playback). It will change the frequency spectrum a bit, as well as adding further soft saturation, smoothing transients, etc.
The tape speed will also create more or less treble damping and bass bump.

[snk]

I have gathered some links and resources on the matter, I will share it here in a little while.
You may check the free and open source plugin ChowTape, which is very nice and features a comprehensive user manual, with hysteresis saturation and bias pictures, as well as typical compression curves.
https://chowdsp.com/products.html

[snk]

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwi5ndyP-43zAhXuzIUKHdJVA40QFnoECAMQAQ&url=http%3A%2F%2Fwww.ant-audio.co.uk%2FTape_Recording%2FLibrary%2FEQ.pdf&usg=AOvVaw3hKddzqLiCvM0gXVvaz-3S
http://www.reeltoreel.de/worldwide/A77EquipmentReview.htm
https://www.tapeheads.net/showthread.php?t=20460
https://richardhess.com/notes/2007/08/26/tape-speed-track-width-frequency-response-and-dynamic-range/

[snk]

You may also check this article :
http://zadvitsky.com/index.php/studio-devices/tape-saturation-compressor.html

[amz-fx]

You may also check this article :
http://zadvitsky.com/index.php/studio-devices/tape-saturation-compressor.html

That is a very old project from a site called LHX2. You can see the archived version here:

https://web.archive.org/web/20100426181146/http://home3.netcarrier.com:80/%7Elxh2/

There are other interesting projects on the site, and everyone should download them while they are still available.

Thanks for your input, snk.

Best regards, Jack

[GibsonGM]

The marshall preamp looks kind of interesting and simple (tho lots of stages!). Have you built any of those, Jack?

[amz-fx]

Mike,

I breadboarded a couple of projects from that site many years ago but I cannot positively remember which ones. I am reasonably sure that the tape saturation project was one of them, which is why I immediately recognized it. The Fender amp was the other, I believe.

The site was quite popular in years gone by, and I would bet that there are several members here who have built some of the projects. The projects got mostly good reviews.

regards, Jack

[GibsonGM]

Given the simplicity, probably worth BB'ing that.  Curious how a tape saturation effect would sound. I always enjoyed playing thru a tape echo for just that reason!

[ElectricDruid]

just making up some basic distortion and calling it tape saturation?

Ummm, I wonder?...Which would *I* do?? 

[bowanderror]

The LXH2 tapesat circuit is just a Ge diode ladder with pre-emphasis & de-emphasis to mimic an RIAA-type EQ curve. Having built it, it sounds decent, but I wouldn't call it the sound of "tape saturation". Probably more useful for drums or buss saturation than for guitar. For guitar, you'll also have to add some gain to actually start hearing an clipping.

The DIYRE 15IPS is based on the same idea, but it's simpler & I think it sounds better. DIYRE also have an interesting guide on modding the circuit for the inherent frequency response of different tape machines, so you can play around with quite a bit on the filtering side.



There is also a discrete MKII version if that's your thing.

[bowanderror]

Not exactly what you were looking for, but I found the DIYRE TM79 to be much closer to the sound of actual tape. It uses coupled transformers with a heavy load & MOSFET body diode soft clipping between them. It sounds REALLY nice on the low end, and still manages to have usable sounds from subtle to heavy saturation. There is also pre-emphasis & de-emphasis, so plenty of room for mods.



Not a cheap build though as the transformers are ~$5 a piece.

[bowanderror]

My b for posting a million times, but this is probably closer to what you're looking for: Tape Machine Frequency Response

Most of the machines have both 15 & 30IPS lines, but they're all at the same input level, so no info on dynamic response.

[R.G.]

Tape saturation is a variant of the inherent signal limitations of magnetizing ferromagnetic materials. If you look at a nominal B-H curve and consider the H axis to be the input and the B axis to be the output, the shape of the nearly-linear low-H/low-B will jump out at you, as will the slow bend-over of the B response to increasing H excitation.

There are derivatives, integrals, hysteresis and such involved in the math, but think of H as the input and B as the output, scaled to voltage. Transformers do this too, but unlike tape decks, they don't have ultrasonic bias stuck in to minimize the hysteresis-based distortion, and as such they don't have a good way to make saturation distortion (more) independent of frequency, so with transformers, you get less ferromagnetic distortion as frequency goes up.

Pretty much, it's a very soft-kneed distortion that's symmetrical around zero, and as such produces mostly third-order distortion until it becomes pronounced or way overdriven. Using a string of diodes or other waveshaping gets to a more frequency-independent distortion without some of the complexity needed to try and make transformer distortion less frequency dependent.

The actual tone will depend on the softness of the distortion knee as a percent of the linear/pre-knee region, and what kind of pre- and post-filtering you do. The softer and more symmetric the +/- knees, the more you'll get first third order until you push it harder into fifth- and above. The less symmetric, the more second-, fourth- etc. will get in and produce the results for the filters to act on.

[teemuk]

With all the (practically inaudible) nuances (e.g. tape wobble, tape head effects) aside tape saturation is really just symmetric and moderately soft clipping. Add a bit of band-pass and that's it.

I once examined tape saturation effect* to find out what the fuzz was about and was somewhat dissapointed to not find anything more exciting than that. Even frequency dependancy of the distortion is subtle. You can easily fake tape saturation with plain clipping diodes. If a simple shunt clipping diode circuit isn't soft enough just put some series resistance to the diodes.

Do note that, in practice, at the extreme settings you most likely overdrive your tape deck circuitry instead of actually saturating the tape more and more. IMO, that could have be the more interesting part. However, it's likely also the most inconsistent part since tape deck record circuits of 1960's, 1970's, 1980's and 1990's likely are somewhat different from one another.

*I recorded sine wave input at different frequencies - low, mid and high - to C-cassette tape, and at three different settings of overdrive indicated by the tape deck's monitor. At medium overdrive there was a slightly visible symmetric soft clipping effect, at higher overdrive evident soft clipping, and at extreme overdrive the tape deck amplifier itself was getting overdriven resulting into "hard" brickwall -type clipping. There was a slight band-pass effect visible that shaped the clipped waveform. The waveform images have for long been lost and the imagehost site I used for uploading them no longer exists. This was several years ago. I found no results that would have encouraged me to investigate this "tape saturation" effect further. It was just plain soft clipping.

[Mark Hammer]

Are we being too focused on the properties of the tape AS tape, and callously disregarding the role of the preamp and tape head in creating the sound we equate with "tape"?  After all, tape doesn't GET to "saturate" unless the input levels feeding the head, to be imprinted ON the tape, exceed what the tape can handle.

And just exactly what sort of tape are we talking about?  Is it 1/2" 2-micron chrome tape, run at 15ips, or 1/8" ferric dioxide cassette tape, run at 1-7/8ips?

Ultimately, what people equate with "tape saturation" is fundamentally attempting to pack X amount of signal on a medium that, under the circumstances, can only handle Y amount with any sort of audio bandwidth.

[amz-fx]

Thanks to everyone for your thoughtful contributions to this thread.

Best regards, Jack

[amz-fx]

Also, some reading from the 1950s:

Limiters and Clippers

regards, Jack