Forums | Search | Archives

All Forums  Third Party Plug-ins  Duplicating Analog Tape Compression

Author	Topic
kkastagon Posts: 21	Posted - Mon Oct 28, 2002 8:45 am  Has anyone been sucsessful with getting a textured sound similar to analog tape compression without sounding like distortion?
Syntrillium M.D. Location: USA Posts: 5124	Posted - Mon Oct 28, 2002 9:01 am  You might want to check out the DX Plugin PSP Vintage Warmer. It gets pretty darn cloes to the real-deal, with lots of adjustable parameters (both for Drive and compression, Knee settings, etc) ---Syntrillium, M.D. _________________
cassembler Location: USA Posts: 1	Posted - Mon Oct 28, 2002 10:30 pm  Magneto is pretty nice, though probably not the most acurate... Good for lots of stuff. _________________ "If I had known, I would have become a watchmaker."- Einstein loudcow.com
kkastagon Posts: 21	Posted - Tue Oct 29, 2002 10:11 am  Thanks for the info. I tried the demo for PSP VintageWarmer, and I am very impressed. It definitely gave my track a warm color, without losing clarity.
post78 Location: USA Posts: 2887	Posted - Tue Oct 29, 2002 6:25 pm  Apparently Crane Song has a VERY attractive piece, and at only $3500! _________________ Answer = 1. Probably.
andyeb Posts: 45	Posted - Wed Oct 30, 2002 3:41 am  Syntrillium, M.D. wrote: You might want to check out the DX Plugin PSP Vintage Warmer. It gets pretty darn cloes to the real-deal, with lots of adjustable parameters (both for Drive and compression, Knee settings, etc) ---Syntrillium, M.D. I'll second that recommendation; I use PSP plugins everyday and I think they are great! You may also want to check out the mixpressor and mixsaturator that come together in the MixPack also from PSP. If you like tape delays, you might want to check out their new Lexicon approved tape delay sims. I haven't tried them, but they seem to be getting good press. Andy
Sonicillusion Posts: 7	Posted - Tue Nov 05, 2002 5:01 am  Magneto is pretty good to me...if ya dont over do it.
scooter Location: Panama Posts: 2	Posted - Tue Nov 12, 2002 1:10 pm  Heya kids, just was wondering - can anyone tell me in relatively laymen's terms or "cyber - chimpaneeze" where all the free plug-ins are located (all links will be greatly appreciated). Anything and everything to help my electric guitar playing sound better...amp simulators - effects - processors and boxes...the works. it's going to be fun learning from you all Scootie
jonrose Location: USA Posts: 2901	Posted - Tue Nov 12, 2002 5:02 pm  A quick search of the forums (Search link is up above) would turn up several, but here's the one everyone here keeps quoting: http://www.thedirectxfiles.com/plugins.htm Have a look around the forums for a few more... Best... -Jon _________________
scooter Location: Panama Posts: 2	Posted - Wed Nov 13, 2002 8:28 pm  thanks so much - I appreciate it. Gonna see what werks and what stinketh ..... Scootie;)
kylen Posts: 290	Posted - Thu Jul 03, 2003 10:52 am  Bang! Zoom! I just landed here after a Google Search for tape compression and... Quote: kkastagon - ...analog tape compression... So, I really didn't have to leave home to find this !:) Quote: Syntrillium M.D. - ..the DX Plugin PSP Vintage Warmer... Well I just happen to have Vintage Warmer and I've wondered what all the fuss about tape compression was besides the warm tape saturation that I love. I recently read that actual tape compression, such as on a real 30ips machine, can remove or prevent Bad Transients for free without you having to do anything. Bob Katz (http://www.tcelectronic.com/media/bobkatz.pdf) characterizes these little devils as a short transient below 10 ms, something that the ear can't detect. If I can't hear it then why remove it ? Well for one thing I think it would affect my ability to get the stuff I can hear (good transients) to peak out at 0db or -0.1db or where ever I want to set peak according to the medium I'm using (CD or tape). This may be a slightly different intention of the Vintage Warmer '30 ips tape' preset but I'll see. I'm having a vision of the good transients (characterized as average RMS to Peak using CEP stats) swimming in a mellow sea of saturation approx 1-2db deep while the bad transients are getting their little heads chopped off rather abruptly without any mourning or long speeches or anyone missing them!:D I just need to work on my approach and settings and statistics to ensure I'm only wacking the bad transients. Anyone else using tape compression for this ? kylen
post78 Location: USA Posts: 2887	Posted - Thu Jul 03, 2003 11:06 am  Quote: Bob Katz (http://www.tcelectronic.com/media/bobkatz.pdf) characterizes these little devils as a short transient below 10 ms, something that the ear can't detect. Bob's got some good ears, but if he really can't hear these things, than perhaps it's time for him to step aside... I believe that the ear can receive spacial information from a delay in the range of picoseconds. _________________ Answer = 1. Probably.
kylen Posts: 290	Posted - Thu Jul 03, 2003 12:11 pm  Quote: I believe that the ear can receive spacial information from a delay in the range of picoseconds Oh, OK Post78 thanks. Maybe I should step back to an outer layer of the onion and review my assumptions!:) You're saying that, in this context, there are no bad transients. All transients of any denomination need to be managed as you normally would manage [rms to peak] headroom. You mentioned delays but it makes me think this could apply to the transients also. ...more research...thinking, experimenting...:D thanks, kylen
post78 Location: USA Posts: 2887	Posted - Thu Jul 03, 2003 3:13 pm  Quote: You're saying that, in this context, there are no bad transients. I'd say that anything so short that it can escape a fast compressor is safe to simply clip, and these too will be in the ms region. I think that the fastest signal able to be captured on a red book audio CD is about 45us (1/22050), and I doubt that todays loudspeakers can handle signals anywhere near that short - something to do with Newton's first law... So there's clearly a practical limit as to what we need to worry about. Perhaps it's time for someone to come in and double-check my comments, here. _________________ Answer = 1. Probably.
SteveG Location: United Kingdom Posts: 6695	Posted - Sat Jul 05, 2003 7:56 pm  post78 wrote: Quote: You're saying that, in this context, there are no bad transients. I'd say that anything so short that it can escape a fast compressor is safe to simply clip, and these too will be in the ms region. I think that the fastest signal able to be captured on a red book audio CD is about 45us (1/22050), and I doubt that todays loudspeakers can handle signals anywhere near that short - something to do with Newton's first law... So there's clearly a practical limit as to what we need to worry about. Perhaps it's time for someone to come in and double-check my comments, here. The relatively easy ones: Loudspeakers can handle transients rather fast than a CD can produce without any trouble at all - it's not too difficult to get a response you can measure past 35kHz. Any compressor using lookahead can capture all transients - by definition! And now the slightly harder one: What's the shortest ITD? (Interaural time delay). Why this? Well, it determines the shortest relative time differences the ear can detect. Let's use an angle of 4.5 degrees - which is quite a realistic minimum. Don't worry if you think you can localise better than this because a) it's unlikely that you can, and b) wait and see what the numbers are... 4.5 degrees is pi/20 radians, so using the standard formula gives (0.09 x ((pi/20) x sin (pi/20)))/344 = 0.000020906 Secs, which is 20.9 microseconds. 0.09 = distance between the ears in metres. 344 = speed of sound in air in metres per sec. [/list:7b1399de7e] Picoseconds are nowhere in it. And the worst one of the lot, because there isn't a simple answer... The shortest transient the ear can detect? That's an interesting one, and I think that Katz is, as usual, just repeating what he's heard elsewhere. It's worth looking at what caused this figure of 10mS to be the one put forward, because it's usually completely misinterpreted. This was the figure that was determined as being the minimum baseline for metering to respond to, and even that's not straightforward. If you look at the IEC standard for PPMs, you will find a number of risetime figures used as calibration references, and the most important one really is 3mS, because that is the figure at which the PPM will under-read peaks by about 4dB - which PPMs consistently do. You don't want to know about the guesswork that went into this... How it got refined into an international standard in this form is... interesting... It is also a much more realistic general figure to use for the ear's response - but the ear actually responds to much faster transients than Katz supposes, but like all other devices, the faster the transient is, the less you will notice it. And masking effects alter this very considerably, so the only realistic conclusion you can draw is that the ear has a transient response that is good enough to resolve signals that can make the ITD work reliably for a few degrees of movement - and it's a damn sight smaller number than Katz supposes it is - but it's not straightforward. But in the right conditions, he may be at least an order of magnitude out... _________________
kylen Posts: 290	Posted - Sat Jul 05, 2003 8:41 pm  Thanks Post78 and SteveG for humouring me, or if not that putting out a valiant effort to answer my 'Why is ther air' questions!:) I'd can't say just yet that I've kept up with all you've said but for now I'll put it in my notebook - make more notes, etc. I really appreciate the information and experience you guys offer. I'd like to say I'm at least on the same planet as you guys - this is Mars, right?:D Yes, I see that the lookahead processing of digital compressors eliminates the need to worry about not being able to manage transients. Unfortunately I have once again mis-diagnosed a piece of audio that is simply suffering from a bad mix (Where's Voodoo?). It's not a transient problem but a loudness problem. The VU (average loudness as seen on the PSP MixPressor VU) is jumping all over the place and I've been trying to solve the problem by adjusting the dynamics with a peak detecting compressor...argh!@#! The vocals are jumping out too much in places (get a visual of the singer raring back and letting loose). Both Vintage Warmer (for sure) and Ozone (I'm guessing) use peak detectors. So when I get the loudness of the mix adjusted properly then I'll be back here to think about dynamic ranges with transients some more!:) kylen
post78 Location: USA Posts: 2887	Posted - Sun Jul 06, 2003 1:06 pm  Quote: Loudspeakers can handle transients rather fast than a CD can produce without any trouble at all - it's not too difficult to get a response you can measure past 35kHz. That's impressive, I'd say. Certainly, it doesn't surprise me too much that a speaker can handle such frequencies, but can it go from dead stop to full amplitude (relatively) quickly enough to hit a 45us transient (or faster)? I'd think that it's own mass wouldn't quite allow it, but then again, sometimes I think incorrectly. [t:0c8de3bbef]Forgive me if I'm wrong, but wouldn't the result be akin to attempting to play back a square wave?[/t:0c8de3bbef] I'd like to try that one again, but I'm struggling to find an adequate analogy... The square wave comparison works for my point, but at the same time I realize that a real analog square wave is simply not possible, unlike this point. Please disregard. :( Quote: Picoseconds are nowhere in it. Okay, my mistake. I knew that milliseconds were simply not even close, and I guess I remembered things a bit wrong. One of the benefits of not just remembering an answer, but a formula, as you've nicely demonstrated. :) Quote: Thanks Post78 and SteveG for humouring me... I tried... _________________ Answer = 1. Probably.
SteveG Location: United Kingdom Posts: 6695	Posted - Sun Jul 06, 2003 5:02 pm  post78 wrote: That's impressive, I'd say. Certainly, it doesn't surprise me too much that a speaker can handle such frequencies, but can it go from dead stop to full amplitude (relatively) quickly enough to hit a 45us transient (or faster)? I'd think that it's own mass wouldn't quite allow it, but then again, sometimes I think incorrectly. [t:186baaaaa7]Forgive me if I'm wrong, but wouldn't the result be akin to attempting to play back a square wave?[/t:186baaaaa7] I'd like to try that one again, but I'm struggling to find an adequate analogy... The square wave comparison works for my point, but at the same time I realize that a real analog square wave is simply not possible, unlike this point. Please disregard. The square wave thoughts are in the right area... yup, air is an elastic medium (we've done all this before somewhere!), so no square wave, but we can create a transient wavefront, and to do this we have to rely on a combination effect from both, or all three, drivers in the cabinet. And yes, with the right tweeter, you can get an appropriate level to sit at the front of a fairly hefty impulse - but you wouldn't want to do it too many times, because it will start to do painful things to the cone materials involving detachment in places - this is all to do with mass and acceleration. There are complicating factors to do with directionality as well - you would only get the full effect on-axis. But loudspeakers can still beat a CD out of sight for frequency response. Linearity, distortion, etc are nowhere in it, but frequency response? Yes. _________________
DeluXMan Location: Canada Posts: 330	Posted - Mon Jul 07, 2003 4:22 pm  'characterizes these little devils as a short transient below 10 ms, something that the ear can't detect.' Bob Katz seems to be off by about X100. Isn't 0.1 ms. more like it? Anyway, you can hear a single sample impulse just fine, in headphones and presumably also in a non reverberant chamber [the only place to experience a real short term transient is without wall reflections like in a speacial chamber, but headphones stuffed with kleenex may come close...]. Another thing to note is that any such impulses/transients contain energy at all frequencies below 1/duration of the impulse, so to be unheard it must be masked, which may be where Katz is coming from.
SteveG Location: United Kingdom Posts: 6695	Posted - Mon Jul 07, 2003 5:38 pm  As I said earlier, it's complicated, and of course Katz is miles out - and misquoting the original idea completely. Originally this started out as a metering issue, and what was originally determined was that an overload of up to 10mS was supposedly undetectable by the ear. Personally, I don't think that this is true either, but that's where this whole misapprehension came from. The original PPM spec was a compromise between the ballistics of the meter, and the ability of the electronics to partly compensate for them. Initially, one of the reasons that PPMs needles fell back slowly was that since overshoot was pretty well under control, you could get the needle to the next transient quicker if it was still somewhere in the vicinity of it, rather than back at zero. It also happens to be a lot easier on the eyes! _________________
DeluXMan Location: Canada Posts: 330	Posted - Mon Jul 07, 2003 7:32 pm  Ahh OK - the clip is not heard, got it. :P Anyway, unless something exotic is going on with tape saturation i bet you can come close to simulating the 'perfect' tape formula with fast enough processing and super fast attack/release settings.
	Topic

Page: