Satin goes V1.1

[4damind]

- Independent Input gain would be very important. Some small button "unlink" for every instance, so the user can use it or not if grouped.

- and not to forget a "Automute Hiss" switch which mutes hiss if no signal is detected

[Compyfox]

O-kay that Grundik TK747 might prove to be troublesome. I can't get the EQ feel right - at least to me. And I don't have any adapters to measure my device currently with pink noise and sine waves.

Additionally, I could currently only reproduce the 7,5ips mode - if ever.


I wish I had a frequency plot of that device. Or a custom EQ. I'm trying to get something good looking with NAS and 7.5ips - but it doesn't want to click with me yet.


The FOSTEX one is looking good btw. Created three presets: Modern, Vintage and Broken Machine. Worked okay in tests with the Dolby A-mod NR. But again, I can't compare since I don't have that device anymore.

Though too tired for today to release anything new.

[sascha]

I used a digital meter post SATIN to find the correct noise floor value for my needs. In example: -60dBFS was -68,80 in SATIN.

dBFS is not quite suitable for noise
One could use VU, or RMS, but still a weighting factor is missing then (A, B, C). The fact that the noise texture changes with speed, bias and level makes it in fact more complicated.
[The old german EE joke: 'Wer misst, misst Mist' ]

I think I'll tackle the technical aspects of a Fostex 280 next (still have the manual). But I don't know how to recreate that frequency courve from 40Hz zo 14kHz... *hm...*

I used the 280 a lot in the early 90s. What a horrible machine (the crappiest eqs I've ever seen on a mixer). If there's one thing to model, then the immense w&f... I never managed to get clean guitar recordings.
Sure, Satin doesn't do 3.75ips, therefore the freq response and w&f won't be doable properly.
Anyway, try using 'flat' eq and vintage tape (which has more HF smear).


Concerning this - and probably more demand for LoFi delays and such -, I'll try and see what the model is capable of at lower speed. Chances are it gets sort of out-of-specs. Satin is a pure text-book approach to magnetic recording, we implement things like they were on paper (using a few corrective factors nontheless). That said, I've never tested for 3.75 or even lower speeds. Pre-emphasis might be problematic, at least experimental. The typical 'studio' speeds are fine with the 18k shelving filter I was using, but the lower you go, the more you'll be off the ideal response. With the current implementation, my goal was to serve with equal freq. response at all speeds, at least at nominal input gain (-18 (vintage) and -12 (modern)).

I know there's a certain discrepancy between Lofi-tape afficionados and the higher-level studio unit I wanted initially. I felt I needed to nail the 'perfect' studio tape sound first, as this was a long-term dream still unfulfilled by me for almost 10 years after I started my first tape project with next to non-knowledge.

[sascha]

The FOSTEX one is looking good btw. Created three presets: Modern, Vintage and Broken Machine. Worked okay in tests with the Dolby A-mod NR.

Keep in mind that Dolby C can't be emulated by any of the other types. As opposed to A with its fixed frequencies, C is using sliding-band filters, like the single one in B. The whole architecture is very different, and very intrusive, soundwise.

[Basspartout]

Just tried the demo and I am really impressed! Finally acoustic guitars getting really close to that late 70s/early 80s analog feel and sound, without losing attack or clarity or sounding muddy!

It's the 9th here in Berlin - can I buy it now?

[Urs]

Just tried the demo and I am really impressed! Finally acoustic guitars getting really close to that late 70s/early 80s analog feel and sound, without losing attack or clarity or sounding muddy!

It's the 9th here in Berlin - can I buy it now?

You can buy it already - link is in the first post. I'm going to sleep now, and tomorrow we'll do a new build, test it for a few hours and then post it on our website. Might be about 5pm.

And... two installs are no problem at all - it's a plain serial number.

[Basspartout]

Aaaaawesome Urs! Thanks a lot. I'm gonna sleep better tonight after I bought this fantastic plugIn.
Gut Nacht

[Compyfox]

dBFS is not quite suitable for noise
One could use VU, or RMS, but still a weighting factor is missing then (A, B, C). The fact that the noise texture changes with speed, bias and level makes it in fact more complicated.
[The old german EE joke: 'Wer misst, misst Mist' ]

Too late for today, but I might check again with DigiCheck tomorrow with the tape delay machines. I did shoot for an average signal of -64dBFS and -59dBFS in terms of noise.

I used the 280 a lot in the early 90s. What a horrible machine (the crappiest eqs I've ever seen on a mixer). If there's one thing to model, then the immense w&f... I never managed to get clean guitar recordings.
Sure, Satin doesn't do 3.75ips, therefore the freq response and w&f won't be doable properly.
Anyway, try using 'flat' eq and vintage tape (which has more HF smear).

I went by pure technical aspects, messed a lot with the head bump, pre-emphasis and gap width. Else "flat" and emphasis results in intensive lowend bump and a way too low highshelf. According to specs however, the frequency response of the Fostex 280 was from 40Hz to 14kHz. So it was more than with the flat EQ and used pre-empahsis.

This can however work with modeling really dirty 3,75ips machines.


I also do remember this machine being dirty (due to the EQ used), and lofi sounding. Depending on the settings while recording and playback. But it didn't have any strong wobbling - at least not to my rememberance. (according tot he manual, only 0,05%!)

I did manage to dial in a suitable noise floor (Hiss), the appropriate Crosstalk and also messed a bit with the Asperity setting. But I need to fine tune tomorrow.

Noise (pink) = RMS meter
Sine = peak meter


Of course I can't use Dolby-C yet, as it's not (yet) provided. But I found Dolby-A mod a nice alternative. It's an esque preset after all. Though using Dolby C could clear why this tape sounded so shitty.

Heck on playback with my still existing tape machines, I pretty much always turn off the noise reduction. More hiss, but no muddy sound.

Concerning this - and probably more demand for LoFi delays and such -, I'll try and see what the model is capable of at lower speed. Chances are it gets sort of out-of-specs. Satin is a pure text-book approach to magnetic recording, we implement things like they were on paper (using a few corrective factors nontheless). That said, I've never tested for 3.75 or even lower speeds. Pre-emphasis might be problematic, at least experimental. The typical 'studio' speeds are fine with the 18k shelving filter I was using, but the lower you go, the more you'll be off the ideal response. With the current implementation, my goal was to serve with equal freq. response at all speeds, at least at nominal input gain (-18 (vintage) and -12 (modern)).

Which is why I said: maybe a second version of Satin. Though a shift of the frequency response and center points of the EQ's is expected and normal.

I know there's a certain discrepancy between Lofi-tape afficionados and the higher-level studio unit I wanted initially. I felt I needed to nail the 'perfect' studio tape sound first, as this was a long-term dream still unfulfilled by me for almost 10 years after I started my first tape project with next to non-knowledge.

And that is more than appreciated. But if there's a tape delay built in, and a sand box type system - I can't help but start messing with settings and create the tape machine I'd like. And chances are, that is a crap Cassette Tape (which still has it's uses) or recreating my R2R collection in software form.


Either way... excellent work.
Would still love to see the delay presets as part of the regular release (which should be tomorrow, still - if all bugs are handled).

And I try to get more stuff done by pure "specs". Like I said - I have two machines prepared. Only needs finetuning.

[Compyfox]

Sorry for the double post, but Sascha made me completely nervous about the measurement thing.

So I went back to the RE-201 and Copycat presets. And rather than using a peak meter for measuring the noise floor (which is a masked pink noise), I used three different RMS meters (RME DigiCheck, bx_meter Standard/K-weighted and Cubase's built in Digital Meter). The nosie floor is now correctly at -64dB RMS (RE-201) and -59dB RMS (Copycat).

The presets are updated and uploaded to the same spot. The links in this thread remain the same. Just follow my sig.


INFO FOR Urs and Sascha:
I consider these presets final! If you want to implement them into the official release - use them!




REGARDING THE FOSTEX 280 AND GRUNDIG TK747:
I also fixed the noise floor and the crosstalk for the technical FOSTEX preset. Finetuning will follow the coming days, and an eventual release in here.

I'll see if I can find some adapter cables to hook up my old Grundig TK747 (which is not calibrated at all!). But this will take a couple of days as I have a busy week coming up.

[Breeze]

I have a FR (...you've probably already thought of this, but... ) :

It would be nice in the Delay mode if (a) modifier key(s) could be used to move all the same controls proportionately within the same group (ie.: delay sliders, Mod Rates, Mod Amts, Balance and Level).

Thanks for considering!

[micsue]

Tried out Satin its ok but they need to work on it more, like more presets for bass,guitars,snare,toms and overhead and work out the bugs.It sounds alot like Magnetic 2 by Nomad Factory.

[djanthonyw]

It sounds alot like Magnetic 2 by Nomad Factory.

[V0RT3X]

It sounds alot like Magnetic 2 by Nomad Factory.

[macmurphy]

alot .

a lot /spelling police

but it doesn't sound like magnetic 2. not a bit.

[Howard]

Would still love to see the delay presets as part of the regular release (which should be tomorrow, still - if all bugs are handled).

Hi Compyfox . I downloaded your RE-201 and Copycat delays today (latest versions) - with your permission, they will be included in the presets.