On Sat, 21 Nov 2015 08:52:55 +0100 (CET), F. Silvain wrote: USB MIDI has got the risk to cause noticeable MIDI jitter. Does USB MIDI provide galvanic isolation? Several times I read claims that USB MIDI connections aren't galvanically isolated. Standard MIDI connectors make usage of opto-couplers and shield of the MIDI DIN connectors is open. Regards, Ralf

On Sat, 21 Nov 2015 09:43:28 +0100, Christopher Arndt wrote: It's an issue as soon as sync is involved. Not necessarily for clock signals (only). Assumed you record the audio signal of a kick from a synth and _not_ at the same time a percussive bass that should be in unison with the kick, IOW you record this bass sound after you recorded the kick, then both MIDI events are affected by different MIDI jitter, regarding the mismatch the percussive bass and kick do cause an early reflection like effect. There is the argument that jitter is lower than inaccuracy of a musician, so it's impossible to notice. This is a half-truth, it depends to the kind of music you make and what you want to archive. Another question is, if there is or is not an issue regarding galvanic isolation. Regards, Ralf

On Wed, 23 Dec 2015 17:42:29 -0600 "Jonathan E. Brickman" <jeb(a)ponderworthy.com> wrote: There are a number of USB isolators on the market but for some reason none of them go to the High Speed spec, only Full Speed. However, that would be more than good enough for MIDI. -- Will J Godfrey http://www.musically.me.uk Say you have a poem and I have a tune. Exchange them and we can both have a poem, a tune, and a song.

On Thu, 24 Dec 2015 03:58:31 +0100, Ralf Mardorf wrote: ^^^^^^^^ serial (I'm sure I corrected the "typo" before sending :S) -- http://www.grundgesetz-gratis.de/

On Thu, 24 Dec 2015, Len Ovens wrote: Actually, much simpler would be to just take the optocoupler and add an inverter (or two) after to feed the computer. (basically a thru box) Make sure power comes from the receiver (computer in this case). -- Len Ovens www.ovenwerks.net

On Thu, 24 Dec 2015, Bob van der Poel wrote: Yes that is the spec. However, that spec also gives a part number for that iso-isolator. If builders are already using other opto-isolators which are outside the spec but are concidered ok if "it works", then if "it works" without any opto-isolator and by the way costs less, are you surprised that there are a lot of MIDI interfaces that are not to spec, but still work. How many buyers know the spec? How many of the buyers who know the spec measure to be certain? How many of those people take it back and demand a refund? (this MIDI interface is broken because there is an electrical path from the input pin(s) to the USB pin(s)) I think that any manufacture who got a lot of returns because they built something incorrectly would change but I don't think that will happen here. If you still have a PCI slot... find an old AI with a game port. It's better than a USB MIDI adapter. Keyboard controller (no audio out) USB direct is potentially better (higher speed), but may not be. Ground loops with computer audio can be a real pain. So far I have not had any problem with my PCI AI. USB has been a different story. I have been ok with a mic and an amplifier that is double insulated power (two prongs on the power plug). But as soon as I plug it into an amp that is grounded... every mouse move is heard... oh, that was a network transfer... and that was the sound of the screen saver. My PCI (D66) IF is plugged into a mixer that is grounded, but no problem. I have tried external power for the USB AI (no help) I need to try balanced cabling to see if that helps too... My d66 also has balanced i/o but I am not using/needing it :) Isolation boxes would work I am sure, but add up quickly costwise. -- Len Ovens www.ovenwerks.net

* + * * * + * * * + * + * ° ° ° ° ° ° ° ° ° ° ° ° ° (Stars and snow :D) Hi :) On Thu, 24 Dec 2015, Len Ovens wrote: A circuit with a processor is ridiculous. ;) On Thu, 24 Dec 2015 11:48:07 -0700, Bob van der Poel wrote: MIDI does work if the level is correct, with or without opto-coupler. The squares separating low and high levels are not that clear when using an opto-coupler, this is not an issue per se, but it's the cause that with an opto-coupler the signal is delayed and this becomes an issue for MIDI thru. The only advantage of an opto-coupler and a one side open shield is to avoid ground loops for audio signals, this is why it's required by the MIDI specifications. However, if the levels are ok, it works without opto-coupler too and without opt-coupler it's quasi better for the data transfer and it doesn't necessarily cause noise. Marry Christmahanakwanzika! Ralf -- http://www.grundgesetz-gratis.de/

Le jeudi 24 décembre 2015, 11:48:07 Bob van der Poel a écrit : MIDI uses an unipolar signaling which is very sensitive to electromagnetic noises, in particular those ground loop generates. That's why MIDI standard requires an opto-coupler. USB uses a differential signaling and can works without any galvanic isolation. -- Yves Guillemot

On Friday 25 December 2015 09:04:22 Ralf Mardorf wrote: Ralf et all; Having on several occasions, dealt with venues that were wired by color blind electricians, I look at galvanic isolation as a hard and fast rule where it can be done. Musicians with their own gear, routinely flip the polarity of ground switch on the front of the amp for least hum. And just as routinely break the 3rd pin off their power plugs because the venue is too damned old to have ever had a static ground in the duplexes anyway. Or if that worn out duplex has been replaced by a stage hand who never saw a copy of the NEC, the 3rd pin may not be connected to anything but the handybox its in. And too many times that has led to the whole stage lashup being hot enough to kill if a real ground is contacted. "portable" venues constructed on site by the usual group of "roadies" should be tested by a qualified test person and confirmed each time they are assembled, with the results posted where the talent can read it as they enter. Having opto-based galvanic isolation in the midi spec should not be skipped just because its USB with its differential signalling and doesn't stop working when theres a 3 volt+ AC potential to ground. The fact that this AC potential is even present, should be grounds that you write it into the performance contract that if his gear is defective, he can't use it on your stage. By that same contract, the smarter musicians who do know about such things can refuse to perform if the venues wiring isn't up to code. And he should still be paid in that event, as its not his fault. The quietest studio/control room lashup that ran on 60 hz power I've ever not heard, is actually powered by a big (5kw rated) 250 volt center tapped transformer, powered by the 127 line applied across the 250 volt terminals, with the secondary's center tap on its own deep earthen ground so the secondary's 1/1 line voltage from pin to pin was still 127, but it was a balanced 63.5 volt to ground on each pin. The equipment plugged in didn't know it was specially powered, but the board, which had single ended inputs, had a hum level that was about 90 db down from a live microphone input. Thats as close to never mind as you can get even with true differential inputs, and was about 2 grand lower in cost to build. Biggest problem? Convincing the electrician who installed the transformer, that it actually met code. It may not in some locales that write their own version of the NEC. Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene>

Hi there! Actually I am far less concerned by the connection, than I am the quality of the on board recorders to build a project. I do not want my piano to connect via USB. I established that last time I knocked on this door smiles. It is either going to be connected to the control surface I use for radio production, to the m-audio audiofile 2496 or to some other sound card. However if the on board sequencer produces something that can be played back in a fashion allowing it to work say with a composing package that has midi support, it might skip a step. so anyone have direct experience with pianos with their own recorder /sequencers built in? Thanks, On Sat, 21 Nov 2015, F. Silvain wrote: