As an actual (non-music) engineer with a large studio, I can explain "latency", but there's no simple explanation. Anyone who says "hardware/analog only has no latency" is also theoretically incorrect, since everything takes some amount of time. Sound through air is relatively slow and audio through cables is limited to under the speed of light, which means although orders of magnitude smaller, there is still "latency" with everything. However in the case of sync, I think "jitter" is a bigger problem itself than latency. The "real problem" comes down to modern computers absolutely suck for doing anything requiring precise timing. They may be hundreds of thousands (far more?) of times faster than an MPC or Atari ST, but the way they are built and designed they are not able to accurately send out a MIDI pulse at a fixed time. I guess a simplified way to put it, is in the case of an Atari ST or an MPC, the "software" has complete control over the CPU. It is doing nothing else. The MIDI in also has a direct interrupt to the CPU. The second an event is received (or, based on the clock, which again has full control over the and the timing of everything) the "program" is interrupted/transmits immediately, and instantly handles the event. The reason it can do this is more or less because it only has to and only does ONE thing, run a sequencer. Modern computers are quite far from real time. You have multiple cores now, running thousands of threads, splitting the CPU time between sources, and also requiring synchronization. You have a bunch of devices over the PCIe link (like USB controllers or MIDI cards), which are NOT truly synchronous (Cubase ASKS for a note to be transmitted, which is then queued, and transferred in a burst, eventually to the device - it is not sent directly to hardware immediately). It is almost the same concept as why we have "audio buffers" - audio is not being transmitted in "real time". The interface is reading 44100 values a second to its output. Cubase promises to deliver the next 128 samples before it has run out of new samples to read. Those 128 samples are processed as a chunk and then queued to be sent out. This is why if you set your buffers too low you get dropouts - the computer isn't able to process the chunk 64 samples and get it to the hardware before it runs out of samples to read. You can consider modern computer MIDI "chunked" too, Cubase is constantly asking the notes to be sent - which goes through a long and non-synchronous process before it actually makes it to the device that should be sending it. Sometimes it takes more, sometimes it takes less time - but since it can't be "buffered" and sent BEFORE the device needs it, it has to be "real time", which it just can't do well. This is why jitter is a bigger problem for modern computers - the CPU is busy with something else? Your clock pulse gets delayed by 2 ms until it's free. This is how solutions like the E-RM and InnerClock generate a stable clock pulse from the audio out - the audio out is precomputed and buffered, and will always end up being sent at a constant rate by the interface. MIDI "can't be" (MOTU "tried", it should possible but not simple, and would require a tightly integrated hardware+software solution, which I guess no company has ever felt the need to hugely invest in). In the case of older gear, yes, everything in existence still has latency, which can be more or less depending on how well it's designed. When the MPC receives a MIDI signal, even though the CPU is interrupted immediately and it is handled immediately, it still takes a fixed amount of time for the CPU to decide what to do with it, and a fixed amount of time for the unit to start triggering a sample to be played. Usually older units don't suffer as badly from "jitter" for that reason - they're designed in one way for one thing and therefore that one thing often takes a constant time. They're not busy handling hundreds of background tasks before processing the event. However, the fact remains (again, depending on how well its designed), if the CPU is busy handling the playback of multiple other sounds, it's possible that something gets briefly interrupted (i.e. jitter). But, unlike a modern PC, when the entire hardware is dedicated one task, that can be controlled through design - unlike a PC, where you end up with whatever combination of everything in the system ends up interacting with each other (yes, one really shitty device could affect the timing of everything else, which is why back in the day it was wise to use a DPC spike tool if you couldn't set buffers low enough - bad WiFi drivers could be eating up lots of "time" delaying the audio from reaching the interface fast enough). In conclusion, PCs suck for accurate timing/music *because* they're so much more complex and advanced. Older music things have MIDI directly wired into their CPU. Also, like some people already said re chaining MIDI devices - DON'T do it. Each device will introduce more and more latency and each device will be more and more "off" until the one at the end of the chain is pretty far from correct. Just get a MIDI patchbay - I have a 16 in 20 out. They will transmit MIDI from one channel to all 20 others with only a very tiny (comparatively) amount of fixed latency. I have pretty much everything in my studio on it, mostly so I can easily toggle between my synths getting the keyboard out directly, the MPC out, or the PC out. I even "try" to keep each synth on a different MPC out since multiple MIDI channels will cause latency against each other (with a patchbay, it's just pushing a button to change the routing). That is exactly why the "professional" MPCs have 4 MIDI outs. If I remember correctly, MIDI is capable of transmitting 1 event per millisecond - which means if you are using all 16 channels, and each channel is playing a note at the same time, the 16th channel will have 15ms delay between the 1st channel. This is why you'd want to evenly distribute parts between the outputs.
Yeah, long read but spot on. The best we can do is run a dedicated machine with as few services and apps, use latencymon to verify real-time capabilities, and hope for the best.
Yeah, long read but spot on. The best we can do is run a dedicated machine with as few services and apps, use latencymon to verify real-time capabilities, and hope for the best.
Thanks for being honest about your sync issues. People have made me feel stupid for years because I have so many devices that each have different latencies and I can never figure out how to get them all to sync correctly - Ive always wondered why my setup is continually frustrating me and everyone else’s seems to work flawlessly. The usual explanation is “it must be your DAW...” and it usually is! I’ve been looking for a no BS master clock and I’ll give this a try. As of now, I’ve been going one on one device by device because I don’t trust my computer to do anything properly even with my current midi interface.
Brother, we've all been there. Let me guess, if yer like me, you've even tried multiple brands of expensive cables and various lengths as well. Zee Zjermans have done it once more with this secret weapon. And... It has a few more tricks up it's sleeve besides what it's known for as well..
A lot of people dont believe or cant hear this type of midi jitter/latency but its def real. I use a similar system that iv been using for about 10 years and it was life changing in the studio with lots of midi gear/chains. Mine is the Sync gen but its the same type of concept and there are some other similar solutions. expensive but worth it if you need this type of workflow.
I will say for a long time it never really “bothered me” but then I heard a really stable clock signal when making music with just an MPC as the master clock going to a few things. I’ve basically been chasing that but with a DAW. Once you hear it it’s hard to go back
If you can make do with just 1 midi clock (or use a thru device) then check out the expert sleepers usamo. Same tech as the erm multiclock and just as stable ALOT cheaper tho 🤔
Man I am **so** glad I'm not the only one who spends a crazy amount of time checking every transient while I'm editing because I don't trust my settings.
Latency is such a fluid issue, it’s almost always a problem and the source can always change. It is infuriating and a real inspiration killer a lot of the time. My main achievement within quarantine has been building an Ableton template that is all in sync, finally got there and I can sleep at night now.
For the solo problem: Someone has probably mentioned it already, but if no one has: you can put a track in ‘solo safe’ mode, that way it won’t mute when any solo is pushed. Control-Click on the Solo button of the channel you want to put in Solo Safe.
This is why we have Midi Patch Bays. Latency will always occur when you Midi In / Out / Thru from one unit to another. Connect all your Midi gear In / Out separately to a few patch bays. AMT8's or Motu MTP AV's will do it with no latency.
For anyone worried about this problem and don't have enormous chains of gear, check out the MIDI Solutions Quadra Thru. It's cheap and simple, I got it when I was I first building my setup to mitigate MIDI chain issues before they came up. Eventually I got an interface with enough MIDI outs but that Quadra Thru still sometimes comes in handy.
Hey Ricky, just want to let you know I love your content. Your videos are nerdy as hell and I love how educational they are. Its awesome to see all the intricacies of a hybrid setup that ill probably never have the money or time to set up myself. Keep it up man, your videos make my day.
wow thanks for making this clear. its the first time I've actually understood what the hell the ERM clock is actually doing and how it would possibly fit into my setup. TBH the latency thing is one of the things I find very discouraging about working with outboard hardware. I am primarily working ITB, but with a tool like this maybe I will be less hesitant to invest in more hardware in the future.. great vid, as always. TY!
DUDE!! Thank you!! This is what I have been looking for too. In short, I am sending midi to DMX to control my pixel LEDs along with controlling 4 synths. It’s been maddening to get all to sync perfectly. I’ve seen the ERM, but like you had sticker shock. I am going to buy now! Seriously bro, thank you! Love your videos.
I enjoy your content. It’s nice to see you drilling into a subject with such granularity. There’s nothing more frustrating than to sit through or scan a ‘how to’ video, only to discover it’s mostly general info but without the how/why’s, which always leaves more questions than answers. 🙏
Hi man! 😜 Great video as usual! I’ve had mine for several years now, and I’ve been preaching how good it is since the moment I set it up. Like you my ever growing hybrid studio has put me through the mill time and time again, but hay it’s taught me more about midi diagnostics than I ever would of done, so there’s that! 😂 Seriously though I totally agree with the price point, as it’s the most important piece of gear I own. Tried several scenarios, but settled on usb from Live to the Deluge, then dyn midi onto the Multiclock, then onto a Kenton midi thru 25, and out from there to everything apart from my 3 main synths. That lets me run the entire studio with midi from the DAW, midi controllers or the Deluge, and gives me transport control over the drums, Hydra, Iridium and Pro3, and more importantly, transport and sync control over their individual arps and internal sequencers, which is where the main delay issues seem come from in my experience so far. The Pro3 and the Matrixbrute for example were horrendously poor for keeping arps anywhere near the one without this piece of kit. In my situation at least, I use the swing setting knobs to add some groove to the drum midi chain. 😍 10/10 for the ERM Muliclock from me. 😉
Are you still running this setup? I'm trying to get back to into making music with my Deluge, Hydrasynth, some monosynths and iPad. But I guess I grew tired of the limitations and the annoyance of using a new system (to me). So are you still using this setup and does it work for you? Would I need to buy the ERM Multiclock to get the best result?
Trust is an important word in music making. If a setup you've worked on for ages suddenly shows issues, you can hit a wall creatively - even mentally at times. Good you found a solution for that one. Although expensive, the result will definitely pay off.
The steadiness of midi clock via devices like this is due to audio streams being prioritised and delivered accurately, as opposed to USB midi being delivered whenever the OS gets around to it (jittery mess).
@Knobcore is, of course, entirely correct. As an old git, I've *used* and heard great MIDI timing and it's something modern tech doesn't have (with rare exceptions like Sequentix / MPCs etc. Computers now are appreciably worse and less snappy than decades ago because they are doing far, far more behind the scenes. But then, don't just take my word for it: www.eejournal.com/article/why-your-computer-is-slower-than-a-1970s-pc/
@@JyotiMishra So true. I cannot stand the timing of modern DAW's. My old Mac IIcx running Cubase in the 90s sounded better than Live. I just discovered Reaper and its timing is better than most. I think I'm going to start using it soon. Definitely buying this E-RM as well. I love the timing of its swing. Reminds me of the Future Retro Swynx that I have.
@@exteriorsigns Good luck, Bud M! If you haven't already, please check out the Sync Gen II from Innerclock. I love this page of theirs -> www.innerclocksystems.com/litmus
You're diverse man. I was just watching your pocket operator tutorials and ran into some latency setting up my minilab 3 within Reason just now. Big thumbs ups for these videos and the knowledge. I appreciate it.
I will never have enough thanks to give to you for the tip of delaying the e-rm plugin track. I've had the multiclock for years now and only today I realized this would work wonders for me. Thank you a million times will still not be enough!
Hey Ricky! Found ur channel a month ago and I’ve watched like a 100 vids. U have helped me so much, inspired me. I don’t even have analog synths and just started getting into music production. Just wanted to say thanks for everything. Keep it going :)
You are on the right "track" Ricky.. There is a reason studio's have expensive clocks to sync everything. Not only for midi but even more if you use multiple audio interfaces. You could also use a low budged tactic. If it sounds good it's good don't zoom in on the waveforms!
Sometimes my lazy ass just considers that ... use the PC as a multitrack tape machine. Just play all the rest of the parts live over the beat. So what if it's not on the grid - as long as you're not editing based on the grid. But then you have to worry about plugins being synced to the beat too ... and then you're kind of limited/have to work extra hard to start moving things around. It would be "nicer" just to have perfect sync, but hey, when they recorded to tape they had no damn grid and I'm pretty sure all of the classic hiphop albums sound fine!
One thing that's probably worth mentioning is that the bandwidth limitations of MIDI itself introduce latency, and because it is such an old standard those limitations are very significant. For example if you transmit a chord consisting of 4 simultaneous MIDI notes the 4th note will actually be sent around 4 milliseconds later than the first one. While 4ms is unlikely to matter for most purposes if you're trying to transmit a lot of note and cc data AND clock down a daisy chain of devices you'll end up with problems even using amazing quality gear like the Multiclock (I have one too and I love it!)
_MIDI itself_ doesn't really have bandwidth limitations, it doesn't "know" anything about the data transport. It's the hardware layer running the MIDI protocol that limits the bandwidth, so the 4ms in your example may happen on a typical 5-pin plug connection, but not on a MIDI connection over LAN, or a virtual connection with basically zero latency.
@@DiAL033 Literally the first sentence of the MIDI 1.0 specification is "The hardware MIDI interface operates at 31.25 (+/- 1%) Kbaud, asynchronous, with a start bit, 8 data bits (D0 to D7), and a stop bit.". I appreciate that this applies only to MIDI over DIN but this video is about MIDI over DIN! You're 100% right though, I could have been clearer :)
It's basically repurposing some old technology to solve a problem. We'd use audio-rate signal laid down as audio to the edge track of tape on reel to reel. It would contain dumb-FSK, smart-FSK, or time-code. We'd then feed that into something like an Opcode Studio 3 to read. Tape itself had its limits with wow and flutter. Another point to consider is that using MIDI THRU to daisy-chain equipment together has its limits. You shouldn't do that if you care about timing, especially with older gear. Again, we'd use something like an Opcode Studio 3 or Studio 5 to get a direct run from an MIDI interface to a hardware instrument. Don't daisy-chain. Yeah, someone is going to respond "it works fine for me." Good. But it's a crapshoot, especially with more than 2 pieces of gear.
m B so I don’t daisy chain. I go from Logic to MOTU micro lite. Each device gets is own midi cable. Yet shit will lose sync from time to time. How do you fix that. Also VCV or Blocks will lose sync. The ERM looks like it would fix that.
I love my ERM Multiclock! It solved all my problems and made my hybrid setup a seamless and painless experience. After many years of pain, it changed my life.
In the last 16 years that I have used my setup in "hybrid mode" I have never had any issue with desyncs or offsets, cleaning up or whatever, and I have used legit anything I could get my hands on. At some point my BCR2000 was my midi interface and it worked perfectly.. Ableton Live (and I think FLStudio as well? not sure..) has seperate midi latency compensation dials on every midi in/out track that it can find. I would have several chains going and set the latency compensation for each of these chains and then all I had to worry about was the audio latency. With my crappy audio interfaces it's always been ~48 ms, so I tap that into the start offset of the samples and presto. That, at least, worked for me. I think what we can all pull from this is that it REALLY depends on what your workflow is. If your approach is much more performance/jam alike then syncing issues are much more problematic but if you sketch/compose everything it becomes no issue that something's a bit later or earlier. Never before commented on your videos but I thought it was about time. I like the videos you make and I secretly watch them while I am supposed to work. You rock!
Realize that midi is serial so the different ports allow you to work parallel! I recently bought a secondhand unitor MKll, It made me very happy to have 8 midi ports all working parallel controlled by the DAW midi master Clock.
Thanks! The E-RM is nice, but I had a lot of problems with syncing from MIDI to my modular. So what worked for me to get a solidly syncing hybrid setup (DAW/MIDI/CV Eurorack), was the Eurorack module Expert Sleepers ES-3 in combination with an ES-5. The ES-3 connects to any optical ADAT out on your existing interface, e.g. RME Fireface UCX to sync your modular to your DAW and send CV to control your modular (but there are also other Expert Sleepers modules to connect your DAW via USB, SPDIF, Audio). The working principle is the same like with the E-RM: it syncs from audio - so the latency that was introduced within the DAW is exactly compensated for, which is best if your DAW is the master. With the additional ES-5 expansion, that is directly connected to the ES-3, I send MIDI from up to 8 independently controlled audio tracks to 8 MIDI DIN outs which connect to my MIDI gear (mono jack adapter to MIDI/DIN). This gives you independent control over up to 8 MIDI outputs + modular sync from your DAW: There is a dedicated expansion module with MIDI/DIN outputs (ESX-8MD), so you don't need the adapters, but I had the ES-5 in place, already. With e.g. an Ableton template, you can make your live easier and start all of your hybrid projects with sync setup in place :-)
This video is exactly what I needed! thank you I've got an E-RM multi clock but I thought the learning curve was a bit steep as I'm new to MIDI. I got stalled on there not being a 4/4 option. 6:23 MIDI router description was very helpful. 11:34 great explanation of why the audio signal input in port 1 15:00 sounds like unipolar vs bipolar (on Moog gear) 18:00 I think the shuffle mode is for live performance stay safe 😃
It is not only about latency. I work with Logic Pro X and it´s midi-clock comes at the end in the computer processing hierarchy. So there you have additionally to deal with a very wonky/jittery clock especially the more happens in the arrangement. With the concept of ERM-MC using sample accurate audio-clock you are getting a very stable clock-timing. Now I have a very tight and stable grooving analog-DAW-kombination. Yes it was hard to spend 500 bugs for it. But at the moment the only way to fix this profound timing-problem. Thank you Ricky for this video!
I used to have no problem until I got into modular and semi modular, because I played one synth at a time with external instrument. I was forced to learn more. Still learning. Videos like this really helps.
It is amazing. This is amazing. To what you were doing, I would go DAW click to the Quadra-thru, and from there, send all the outs of the quad to the four hardware synths... an octatrack, analog rytm, and then, ... drumroll... a squarp pyramid. it serves as the sequencer for synths without sequencers... That squarp/synth clock chain from one of the quad outputs would go first into the arturia keystation midi, for its arpegiador, then out to a midi merger (same brand), and finally to the squarp input. This allowed me to have the arp of the keystation when I needed it, or just normal keys from the larger grand midi keyboard underneath it (hence the merger). But having the E-RM instead of the crude Quadra thru would be even more dope.
If you wanna solve the midi clock and midi latency problem forever, you made the first step with buying the ERM! Now make the second step and add a MioXL. Use the ERM only as a MidiClock distribution, if you have some analog clocked device (as example Eurorack gear) you can generate perfect analog clock with the ERM to! The MioXL (mio10 is also very good) is used for MidiNotes and every instrument is on a separate output. Don‘t chain midi devices! Then filter every incoming clock message in the MioXL except the one stable from the ERM. And you never will care about this latency shit again. If you don‘t have enough ports, just add another mio. We have gear for thousands of dollars, but the tools that really help have to be cheap? It‘s like buying a lot of cool cars and putting some 15 year old tires with no grip on them.
The iConnectivity MioXL is such an underrated device but for $500 you get 22 MIDI ports (8 DIN in/out, 4 DIN out only and 10 class compliant USB), ability to patch any input to any output, filtering, remapping, ethernet connectivity, etc.
Some tips for people having similar problems: - minimize the length of your midi chains (2 devices max) - most MPCs have multiple (up to 4 on the 3000) midi outputs, use them all - anything that is time critical should be first after the master midi clock: drums, bass, arpeggios - pads can come later in the chain since these sounds are not as sensitive to latency - I highly recommend the iConnectivity mio10 midi interface and midi router - highly flexible, you can connect two computers to it, has an USB midi host for midi keyboards etc. - it's 300 euros much cheaper than what was presented here - you can pair it with a Quicco mi.1 (50 Euro) bluetooth-midi adapter to control all connected synths from an iPad wirelessly, ex. PatchBase If you do it right you will end up with minimal latency, I have 10+ vintage drum machines and 6 rack synths and never needed special clocking. Not saying it's not a cool piece of gear, but you can get way more connectivity for 300-350 Euros. Make sure you really need the clocking options before spending 500 Euros on something like this.
MIDI chaining adds almost no latency at all - if the MIDI Thru is implemented correctly... If it is a hardware Thru socket that follows the MIDI spec, then it is a copy of the MIDI In data, so you get the latency through the opto-isolator and one or two logic gates for the buffering, which adds very little time. 'Soft' Thrus take the MIDI In and process it inside the main CPU/Microcontroller (or a dedicated Microcontroller for MIDI I/O) , then send it back to the MIDI Thru scoket, and this can cause latency. The main cause of the problems with MIDI that are often attributed to latency are poor cables and poor opto-isolators (old gear used to use cheap (=slow) opto-isolators), which degrade the edges of the currents that carry MIDI data. But the cabling is often the problem: many MIDI cables are not screened twisted pair (as specified in the MIDI spec), but are twin screened audio cable (which isn't designed for carrying data, and has other parameters wrong as well). The bandwidth of many MIDI cables is often poor as well - and this also degrades the edges of the currents that carry MIDI data. Run MIDI data through several slow opto-isolators and low bandwidth MIDI cables in series (a chain) and you get a MIDI signal that has edges which are not clean and sharp any longer, and this is what often causes many of the problems that are described as 'MIDI latency', and ultimately it can cause MIDI data errors. MIDI cables that meet the spec are not easy to find - I just spent 3 months trying to source a special-purpose MIDI cable from various music cable providers, and the only cables that came close to the MIDI spec were designed for carrying AES/EBU and SP/DIF data signals. Cables designed for audio are different to cables designed for MIDI: data cables need to have higher bandwidths, lower capacitance per meter, and different characteristic impedances. Most of the DMX cables that I was offered didn't meet one or more of the required specs. In general, MIDI cable testers check the continuity of the cable - not any of the important parameters. More info on screened twisted pair in my blog at: blog.synthesizerwriter.com/2020/01/quick-and-easy-midi-clock-sync-between.html
@@willdahbe sounds like you completely missed the point. Martin is saying, instead of running out a dropping 500€ on this box, rule out the cables as the possible source. If you buy this or any other box and you have shitty cables, you're still likely going to have problems
Quick tip about that solo channel thing.. I think, when you output the audiosignal to another output then the main output of your soundcard, you can output that audio signal and use “solo track” buttons with no problems. You’ll set that up like you would setup a sort of monitoring channel. (like dj monitoring).
When I use midi with my daw, I don’t record multiple instruments at one time. I’ll record each part one by one. One way to mitigate all the cables is using a patch bay.
If you’re a well planned artist that has an exact image of the final result in mind before recording, that is awesome. Good for you. I go back and forth with my stuff all the time.. but then again I don’t have enough stuff to have serious timing issues anyway.
Same here. I also try to record in as early as possible, so I commit to the sound. IF it doesn't work, then I go back. But the "record early commitment" part pretty much revolutionized my music making few years ago. No more "I just tweak this one thing... and never get anything done" thing, where I was for sooooo long time. Now I actuelly get s**t done. :D
Krmpfpks What I usually do is use vsts first, to hear what kind of hardware sound I want to use with each track. I only have a few hardware pieces. I like it like that because I don’t have to go through to many processes to find a decent sound.
Sauce_Aux even if you set everything up at once you still have to manage each piece of gear. If you have a patch bay for midi it wouldn’t be to bad to setup every time.
Preach it!!!! My synths are all over the place and there is no one adjustment that fits them all. Back in the day they had a very expensive device called the "SRC Friend-Chip" similar concept but ERM so more advanced.great explanation.
Very cool video thanks 🙏🏼 and I love this erm is is a life saver. The price for this unit is actually compensated with a tons of hours I saved with arrangement issues.
Dude ! SO glad you’ve posted this video it’s like the universe confirming I have to buy this ! Like you , I have had ALL the same insane frustrating problems recording hardware to audio and the after tweaking nightmare ! I have considered buying this for a while but yes the price for something that doesn’t make a noise was off putting , like you, say gonna close my eyes and hit buy now after your endorsement lol keep up the great posts stay safe 😊
FWIW I have all my outboard hooked up through an iConnectivity Mio10 which - so far - has solved this problem for me and is quite a bit cheaper. It isn't a clock generator (as far as I know?) but I've never had a problem with how it distributes clock signals. I bought it originally to stop daisy-chaining midi devices but it's helped in those terms too.
yeah, it seems that shorter chains is really important. i wonder if using one of the hardware sequencers as the master clock and using the mio to distribute the midi could give you the combo of stability and consistently low latency?
Awesome video as always. I use the SND ACME-4 similar to Multiclock but i went with the SND because of the swing and division on switches. Very hands on but both are good options. Keep up the good work ur doing. Loving it!
In like ‘85-‘90 the workflow was MIDI-units, Atari ST and tape, tied up with linear time code. Or, a Linn 9000 with a clock track from tape, essentially the same as your clock plugin. Daw not invented yet, we just heard that timing was off. We then understood that Midi (5-pin only option) is an 8-bit serial protocol. Meaning that ch 1 is sent first, then ch 2 etc, and each byte must be complete before being passed on. A through port will process from input in the same manner, inducing more latency. And the 8-bit logic was really slow. Result: Never used Midi through anywhere, got an 8 in/8 out interface, and used only ch 1 for all drums/percussives, via the separate interface outs, say port A to D. Other keyboard stuff can use ch 2-16, maybe via some through ports, but not sharing a physical cable/port with anything beat-related, say ports E-H. Nor would we send any controller data (pitch bend etc.) through the ports used for drums. Lacking sophisticated clocking equipment, Midi clock was only used from sync box (time code) to the sequencer. We never got to the point of using the internal sequencer in a drum module, instead just sending Midi notes form the main sequencer. Very anecdotal, but it was what it was. Today, we just use daw and plugins, not equally fun and inspiring maybe, but timing is never an issue... 👍😅
This video made me think, that Ricky you could be interested in Blokas Midihub, which is a intuitively programmable 4x4 + usb midihub. Not saying it would be a replacement for what you own, but it could well suit your needs. Just got mine, which I pre-ordered/backed and just as an example here are two things I was easily able to achieve with it: I'm now able to play my Digitakt as a polyphonic sampler, with my master keyboard. I'm also able to play my Volca Drum with my master keyboard so, that all the notes are mapped to the corresponding pitch values (as for some odd reason Korg didn't make it read midi-notes out of the box...). Great video once again!
I've never had any luck using ableton midi clock with external keyboards. It's always a little bit too late or little way too late. Gave up. 😑 I love it when you back track cuz got hiped up saying good things about it. Makes me realize there's still ppl who gives honest reviews out there. 👍
I just got fed up with computers and bought the MPC X and the MPC ONE for travel. And I was just about to pull the trigger on the all plugins FL studio bundle until I seen what those new MPCs can do. The way I see it it they solve 3 problems in one. DAW, MIDI Controller and audio interface all wrapped up into one. Granted only the X has phantom power, but there are workarounds for the others. The way I see it, a standalone unit dedicated and made for nother but music has to be much better for music production than any laptop. Granted you can't use third party plugins in standalone but there are workarounds for that to. And with everything they have built into those things, it's really a non-issue. I also bought the Roland JDXI to make my own MPC extension's. Can't wait to have endless sound creation capabilities with a bunch 9f knobs to twiddle with. Lol
Thank you for your thoughts. Obviously, you don't have too much sluggish and/or jittery MIDI gear around or gear with different sluggishness. If that would be the case, a tight clock cannot help, because the latency does not come from the clock, but from the MIDI latency and jitter from the devices themselves. I wonder, if your setup just cuts off some of the inbuilt MIDI latency for your hardware devices instead of coming them in late for this same amount of MIDI latency. So the click isn't tighter itself, but the delay doesn't disturb you so much anymore. In some Keyboard magazine issue from the 90s, there was an exhaustive comparison of MIDI latency and jitter for different devices - some workstations with up to 30 ms MIDI response plus irregular bidirectional jitter of aditional 10-15 ms and this not being a constant, but dependent on voice usage and MDI controller use as well. This would be something you cannot fix with a tight clock, but you can set a predelay, if you know the typical MIDI delay of your particular device. Fortunately, MPCs are quite tight and most modern gear is too.
Another case of GREAT timing on topics. I have a need for this exact type device. To finally complete the hybrid fusion in my studio. Definitely relate to the MIDI chains with my MV-Modules setup. I'll have to take a serious look at the ERM. Keep up the great work bro!!! Peace...Netm8kr
Part of the issue could also be your daisy chains. Try using a midi splitter. Kenton makes a 1:24 midi splitter. Master clock goes in the in. Then an out per device.
Just avoid these midi chains and use a stern configuration. That means each hardware device gets its own midi in and midi out connection to the midi interface. Just buy something like the M-Audio Midisport 8x8 and you do not have these problems.
YES! Was thinking the same thing. I don't ever use Thru. That's something is just a holdover from the days before multi-port midi interfaces (80's). Give each device its own cable and you can adjust latency per port as needed.
Amen. There are almost no reasons that speak for daisy-chaining and tons that speak for star-network. On the other hand it depends on the length of the chain, so if there's just one thru hop, it's okay I guess. (Which, coincidentally, is exactly the maximum amount of thru hops I have in my setup ;D )
I have a Roland A-880 MIDI patch bay for just this reason. It's 8 in x 8 out, with a button matrix on the front for making connections. Super easy to set up your default & then change things on the fly. It can merge ports and send one source to multiple destinations. It's ancient and for some reason, they stopped making them and I still haven't seen anything that comes close. That's why I have a backup. They show up on eBay from time to time. Totally worth it.
I would never ever use clock coming out of a computer. Ableton slaves super good. My solution to this midi issues with other gear was a machine called Boss RC 505 and have gotten two more since then. The clock of that looper is the most solid thing. I have the clock coming out into a Midi Solutions midi merger box where my keyboard also goes in. That way i get to play my synths set to their own midi channels and have everything stay in time. I can start and stop everything from the looper and it’s spot on everytime. The cool thing about this is that you can play the first phrase of something in your head and retain the feel and the tempo and everything in your studio will follow that. So you never have to worry about trying to start a new idea with specific tempo in mind. Just keep it natural. As soon as you close that first loop the clock sets and starts. I am surprised that these RC 505 or even the RC 202 are so solid when it comes to clock.
Really appreciate the video I've been doing a lot of ableton work but haven't quite got to routing things together. Glad to have this to consider when setting up!
Great video explaining that ERM device. I've always stopped playback and trimmed off latency from the beginning of the recording. Though that only works OK with drums, not other melodic sounds.
@RickyTynez I think all your midi problems came from the fact that you use the midi thru on some of your midi device to daisy-chain them. If you could connect them directly to their own channel I believe you wont have any problems. And in this way you could still use any of them to send clock or to controll any other devices (provided they support that). Here is the solution I want to go with when building my new studio: 1 x "MOTU MIDI Express XT" MIDI interface - 8 MIDI ports in, 8 MIDI ports out (as each port has 16 channels - that means 128 MIDI channels out, 128 MIDI channels in - can controll up to 128 MIDI devices simultaneously). 2 x "Kenton MIDI Thru 12" multipliers/splitters connecting them to 2 MIDI outport ports on the MOTU MIDI interface. I will have around 24 MIDI devices. To avoid any timing dissimilarities/problems, I will connect all my MIDI devices through MIDI DIN cables to the 2 Kenton multipliers (better said than splitters). No daisy chaining, no MIDI over USB and MIDI Din combination! Only MIDI DIN 5-pin cables! On the way back to the MOTU's MIDI input ports I will put 3 "Kenton merge 8" mergers (all 3 of them will handle 24 devices) and connect them to 3 of the MOTU MIDI input ports. All connections with the same cables brand, model and lengths. The audio will be through audio balanced cables straight to the patch bays - no audio-through-USB! In my Cubase Pro I will create a template with 1 track for each midi device (each on its own midi channel). This should work perfectly. Every MIDI device will have its own track inside Cubase, its own channel and will be able to send clock and controll any other device. Perfect setup for creativity. If you still have problems with your MIDI connections, I suggest to you to take in consideration this setup. What can go wrong? If some of the MIDI devices are slower than others in processing/sending the data, as each one is on its own channel and track, this can very easily be compensated for inside Cubase on each separate track.
I’ve used a midi sport 2x2 for a long time with success and paired it with an old Roland A-880 midi patchbay to route midi to all the synths. This avoids a long thru chain. I remember some older devices like the Yamaha A3000 sampler being sensitive to the midi channel which added latency if they were on a higher one.
Yeah... that's why I've used DP for like ... 30 years. Rock-solid MIDI timing (it's roots are a MIDI sequencer) w/ a couple micro lite MIDI interfaces along w/ Volta for my CV/Gate and clock. No lag. It's wonderful.
The main issue I have is the audio is delayed when it’s recorded back in my DAW. I need to offset it negatively independently for each device because there all different. But the way your saying you have your set you are setting the negative offset globally in your DAW. This can already be done so the ERM doesn’t seem any better than a MIDI hub and setting the offset in the DAW.
As someone who has worked on the development of a few hardware synths, your problem is in the MIDI signal protocol in general. MIDI doesn't have a fast data transfer rate (only 31250 bits per second). If we convert that into bytes per second, it's 3,906 bytes per second. Well, how much data does MIDI use to communicate anything? Your typical MIDI message be it note on, note off, volume, CC, etc is 3 bytes. That now gives you 1,302 commands per second. Still fast for most music and setups. It's really a case of the more things you have on the line, the bigger bottleneck it's effectively creating where devices have to wait for other data to be sent while waiting for their own commands. This is because MIDI can only send commands 1 byte at a time. So devices have to just wait for their commands to show up after everything else that doesn't go to them before they can execute them. A device to have a dedicated MIDI port for each device in your chain is the best way to go as you showed with hardware you used in this video. In that configuration everything can be sent data as fast as possible, thus eliminating any unnecessary traffic on each device's data line. I hope that helps explain it and wasn't too overly technical ^^;
These things are a revelation. I winced at a fat wad of cash for the syncgen pro when it came out but they are life changing.There is no substitute for these kind of devices in contemporary computers. It’s not latency, that can be managed with offsets. Jitter is the problem. USB packets midi clock on a janky way and no one seems to have been able to solve it. I spent years fighting the problem...transients drifting. There was a good explanation on the inner clock website including tests of various outboard gear -classics and modern- and how how they all have there own limitations to being synced...hint. The old MPCs do very well
this is why it‘s so important to master your instrument, or you will end up with ... more stuff. you don‘t have to do everything by youself. share. form a band.
I would like a band yet sometimes humans can be passive aggressive/troublesome,and I am NOT the one to play games with anyone...I USED to have a partner I DJed with,and he was responsible for losing most of the record collection back then,All I got was 'sorry',so I vowed to make my own studio,own my own narrative,and plus,I rarely sleep,so I can just get up and record what,how,and when i want to...freedom is addictive,and it allows me to work out new ways to do be more simplistic...
I'd love to be able to justify this. I have a similar hybrid setup just with different gear. My midi gear all goes through 2x ESI M8U, usually with Akai Force as the midi master. I've a laptop and a desktop, both running Ableton with different plugins installed. For the most part I rarely have issues, the odd waver in timing but thanfully not often at all and everything usually locks in time well. I'll keep this mind though as I'm sure its absolutely rock solid.
So let me get this straight? with 4 midi ports routing 64 channels, does that mean you can run midi clock from ableton, through an audio interface, through the E-RM Multiclock to keep the stable clock while using the 64 channels to route all kinds of midi data such as sequencers, CC parameter change signals etc without the need for something like an iconnect midi4+ or similar midi hub to route all kinds of midi signals between devices? 😲
It's a little more expensive, but the RME PCIE interfaces use their own tweak on midi processing and have excellent results. IMO it's the best way to go with a hybrid studio. You can get any converter you want on lightpipe (even use apogee symphony on Windows) AND you also get everything this box does. 32 channels of adat i/o and two midi i/o or more depnding on the PCIE you get. Faster than any other interface besides the presonus quantum thunderbolt. Works perfectly with the Expert Sleepers eurorack lightpipe interface. Better resolution than midi to CV.
Knobcore yep! Pcie is natively compatible with processors and the motherboard requires no extra thunderbolt chip to route things. You can also use passive cables. Thunderbolt cables are active and also require a chip. Also get this, I picked up a $2,500 broken audiophile receiver for $10. It couldn’t put out volume, but what it could do was take adat input and go analog put through the preamp output. So fucking mastering grade DAC conversion for $10 because someone didn’t want to pay for a repair! Lmao!
Knobcore oh absolutely. I bought an apogee symphony when the mkII came out since everyone was offloading theirs. It’s gone up in value by maybe $500-600 since then. Sometimes I don’t understand people’s obsession with the latest and greatest. It’s not like a record from 2008 sounds like it’s coming from a gramophone. The Ad/da improvements are still happening, but these companies aren’t NASA research labs. New models, at the ad/da level are barely improved, if that. And companies, like Apogee, that are owned by hardcore investment firms (I mean apogee is like 2% of Engaged Capitals portfolio) in some ways are just repackaging the same tech, giving it VERY subtle but pleasing eq adjustment, and then calling it the “mastering version”. I mean, is it not a coincidence that the first year they are owned by a private equity firm, that they re-release an already existing 2x6 symphony mkii module with a “mastering” tag? Is that not the exact move a private equity company would do? Mastering quality is to music production as Audiophile grade is to home hi-fi...
I have learned the hard way to never use Midi Thru. Synths crash ... not often ... but sometimes. And then the 2nd device in a chain crashes and you loose everything. Of course this happens only during live situations. I use a star system and a dumb midi thru box that by design cannot crash. My audio and midi is set up so everybody can talk to anybody if wanted but is still kept independent enough that if a synth crashes only that one synth crashes. It took me years working towards this setup. And it still can be improved
Excellent video. I actually found there was no latency when I have a sequence going but it’s when I play something in real time it’s when I started to notice latency.
Love your vibes brother! I've experienced actual physical pain trying to get my machines to make sweet love to abletons clock ha. i may have to invest in this thing. thanks for the video bro and happy sunday x
@Ricky Tinez What about capturing the internal clock of the MPC 3000 and sending that back out for every other pass? Does the Multiclock allow you to capture the clock of external gear and allow you to add that to an audio track in your DAW? Ever heard of Innerclock System's Sync Gen 3LX?
this information is a discovery for me that I am new in electronic music production , you are an innovator in this time coexistence between the digital (daw) age and the analog age, thank you very much for sharing your knowledge I follow your channel closely, regards desde Argentina (sorry me english is very rough)
As an actual (non-music) engineer with a large studio, I can explain "latency", but there's no simple explanation. Anyone who says "hardware/analog only has no latency" is also theoretically incorrect, since everything takes some amount of time. Sound through air is relatively slow and audio through cables is limited to under the speed of light, which means although orders of magnitude smaller, there is still "latency" with everything.
However in the case of sync, I think "jitter" is a bigger problem itself than latency. The "real problem" comes down to modern computers absolutely suck for doing anything requiring precise timing. They may be hundreds of thousands (far more?) of times faster than an MPC or Atari ST, but the way they are built and designed they are not able to accurately send out a MIDI pulse at a fixed time.
I guess a simplified way to put it, is in the case of an Atari ST or an MPC, the "software" has complete control over the CPU. It is doing nothing else. The MIDI in also has a direct interrupt to the CPU. The second an event is received (or, based on the clock, which again has full control over the and the timing of everything) the "program" is interrupted/transmits immediately, and instantly handles the event. The reason it can do this is more or less because it only has to and only does ONE thing, run a sequencer.
Modern computers are quite far from real time. You have multiple cores now, running thousands of threads, splitting the CPU time between sources, and also requiring synchronization. You have a bunch of devices over the PCIe link (like USB controllers or MIDI cards), which are NOT truly synchronous (Cubase ASKS for a note to be transmitted, which is then queued, and transferred in a burst, eventually to the device - it is not sent directly to hardware immediately). It is almost the same concept as why we have "audio buffers" - audio is not being transmitted in "real time". The interface is reading 44100 values a second to its output. Cubase promises to deliver the next 128 samples before it has run out of new samples to read. Those 128 samples are processed as a chunk and then queued to be sent out. This is why if you set your buffers too low you get dropouts - the computer isn't able to process the chunk 64 samples and get it to the hardware before it runs out of samples to read. You can consider modern computer MIDI "chunked" too, Cubase is constantly asking the notes to be sent - which goes through a long and non-synchronous process before it actually makes it to the device that should be sending it. Sometimes it takes more, sometimes it takes less time - but since it can't be "buffered" and sent BEFORE the device needs it, it has to be "real time", which it just can't do well. This is why jitter is a bigger problem for modern computers - the CPU is busy with something else? Your clock pulse gets delayed by 2 ms until it's free. This is how solutions like the E-RM and InnerClock generate a stable clock pulse from the audio out - the audio out is precomputed and buffered, and will always end up being sent at a constant rate by the interface. MIDI "can't be" (MOTU "tried", it should possible but not simple, and would require a tightly integrated hardware+software solution, which I guess no company has ever felt the need to hugely invest in).
In the case of older gear, yes, everything in existence still has latency, which can be more or less depending on how well it's designed. When the MPC receives a MIDI signal, even though the CPU is interrupted immediately and it is handled immediately, it still takes a fixed amount of time for the CPU to decide what to do with it, and a fixed amount of time for the unit to start triggering a sample to be played. Usually older units don't suffer as badly from "jitter" for that reason - they're designed in one way for one thing and therefore that one thing often takes a constant time. They're not busy handling hundreds of background tasks before processing the event. However, the fact remains (again, depending on how well its designed), if the CPU is busy handling the playback of multiple other sounds, it's possible that something gets briefly interrupted (i.e. jitter). But, unlike a modern PC, when the entire hardware is dedicated one task, that can be controlled through design - unlike a PC, where you end up with whatever combination of everything in the system ends up interacting with each other (yes, one really shitty device could affect the timing of everything else, which is why back in the day it was wise to use a DPC spike tool if you couldn't set buffers low enough - bad WiFi drivers could be eating up lots of "time" delaying the audio from reaching the interface fast enough).
In conclusion, PCs suck for accurate timing/music *because* they're so much more complex and advanced. Older music things have MIDI directly wired into their CPU.
Also, like some people already said re chaining MIDI devices - DON'T do it. Each device will introduce more and more latency and each device will be more and more "off" until the one at the end of the chain is pretty far from correct. Just get a MIDI patchbay - I have a 16 in 20 out. They will transmit MIDI from one channel to all 20 others with only a very tiny (comparatively) amount of fixed latency. I have pretty much everything in my studio on it, mostly so I can easily toggle between my synths getting the keyboard out directly, the MPC out, or the PC out. I even "try" to keep each synth on a different MPC out since multiple MIDI channels will cause latency against each other (with a patchbay, it's just pushing a button to change the routing). That is exactly why the "professional" MPCs have 4 MIDI outs. If I remember correctly, MIDI is capable of transmitting 1 event per millisecond - which means if you are using all 16 channels, and each channel is playing a note at the same time, the 16th channel will have 15ms delay between the 1st channel. This is why you'd want to evenly distribute parts between the outputs.
jesus christ man, this was such a good piece of information
Yeah, long read but spot on. The best we can do is run a dedicated machine with as few services and apps, use latencymon to verify real-time capabilities, and hope for the best.
Yeah, long read but spot on. The best we can do is run a dedicated machine with as few services and apps, use latencymon to verify real-time capabilities, and hope for the best.
Amazing, thanks for laying this out so clearly
excellent read.
Thanks for being honest about your sync issues. People have made me feel stupid for years because I have so many devices that each have different latencies and I can never figure out how to get them all to sync correctly - Ive always wondered why my setup is continually frustrating me and everyone else’s seems to work flawlessly. The usual explanation is “it must be your DAW...” and it usually is! I’ve been looking for a no BS master clock and I’ll give this a try. As of now, I’ve been going one on one device by device because I don’t trust my computer to do anything properly even with my current midi interface.
Brother, we've all been there. Let me guess, if yer like me, you've even tried multiple brands of expensive cables and various lengths as well. Zee Zjermans have done it once more with this secret weapon. And... It has a few more tricks up it's sleeve besides what it's known for as well..
A lot of people dont believe or cant hear this type of midi jitter/latency but its def real. I use a similar system that iv been using for about 10 years and it was life changing in the studio with lots of midi gear/chains. Mine is the Sync gen but its the same type of concept and there are some other similar solutions. expensive but worth it if you need this type of workflow.
I will say for a long time it never really “bothered me” but then I heard a really stable clock signal when making music with just an MPC as the master clock going to a few things. I’ve basically been chasing that but with a DAW. Once you hear it it’s hard to go back
If you can make do with just 1 midi clock (or use a thru device) then check out the expert sleepers usamo. Same tech as the erm multiclock and just as stable ALOT cheaper tho 🤔
“I’m not a scientist, I’m just a regular nerd.”
🤣
Man I am **so** glad I'm not the only one who spends a crazy amount of time checking every transient while I'm editing because I don't trust my settings.
Trust your ears not your eyes. If it sounds good it’s ok.
I didn't understand a word, but I'm glad you know what you're doing
Latency is such a fluid issue, it’s almost always a problem and the source can always change. It is infuriating and a real inspiration killer a lot of the time. My main achievement within quarantine has been building an Ableton template that is all in sync, finally got there and I can sleep at night now.
Wish I had enough synth friends to justify a homie midi port hahaha
Zeus' Penis? 😂
@@josue.guevara that took me longer than I'd like to admit to get 😄
Me too xD
Facts
One of the best purchases I’ve made for my hybrid setup. Solved so many timing issues that were coming from the computer’s latency.
Why do I love this channel so much, especially when I use complete different gear and rig.
Because of his attitude. He's not the "HEY THIS IS JASPER COMING AT YOU ONCE AGAIN WITH ANOTHER QUICK TIP TO ROCK. YOUR. WORLD!"
Haha right!! Exactly, plus we all carry the same addiction too.
Been using it for about 4 years already. It's been a live saver in my setup syncing up al generations of Midi gear. :)
Me too. Can’t do without.
What about modern MIDI gear with USB, do those have latency as well?
@@FuZZbaLLbee hell yes they do. I got a midi fighter twister and it has hella jitter just by itself
By far best tutorial and walk-through/review channel for hardware peeps. Thank you once again for a very informative video looking forward to the next
For the solo problem: Someone has probably mentioned it already, but if no one has: you can put a track in ‘solo safe’ mode, that way it won’t mute when any solo is pushed. Control-Click on the Solo button of the channel you want to put in Solo Safe.
You are amazing. Thank you to you and anyone else with the suggestion!
"don't hack me" 😄😄
Ricky, I just love your charisma 😄
This is why we have Midi Patch Bays. Latency will always occur when you Midi In / Out / Thru from one unit to another. Connect all your Midi gear In / Out separately to a few patch bays. AMT8's or Motu MTP AV's will do it with no latency.
Exactly. That's the way to do it.
@@sK3LeTvM1 Ditto … I’m running 2 X Motu Midi Express XTs … Zero issues
Your intro vid is the shit. Smooth, crispy and not 30 seconds long👌
you can configure the output of the multiclock plugin as an External Audio Effect to exclude it from the soloing behavior
Awesome.
this is why i love the comment section! thank you soooo much for this.
Cool tip ! Thanks 🙏🏼
Weworktomorrow also there is a trigger pulse signal that you can download from their site if you want to trigger your modular
Yuuup, internets put me onto this tip as well and was just about to post. Saves a whole lot of grief.
For anyone worried about this problem and don't have enormous chains of gear, check out the MIDI Solutions Quadra Thru. It's cheap and simple, I got it when I was I first building my setup to mitigate MIDI chain issues before they came up. Eventually I got an interface with enough MIDI outs but that Quadra Thru still sometimes comes in handy.
Hey Ricky, just want to let you know I love your content. Your videos are nerdy as hell and I love how educational they are. Its awesome to see all the intricacies of a hybrid setup that ill probably never have the money or time to set up myself. Keep it up man, your videos make my day.
Thank you for the kind words!! it means a lot, and im glad you relate to the nerdy side of things haha
wow thanks for making this clear. its the first time I've actually understood what the hell the ERM clock is actually doing and how it would possibly fit into my setup. TBH the latency thing is one of the things I find very discouraging about working with outboard hardware. I am primarily working ITB, but with a tool like this maybe I will be less hesitant to invest in more hardware in the future.. great vid, as always. TY!
DUDE!! Thank you!! This is what I have been looking for too. In short, I am sending midi to DMX to control my pixel LEDs along with controlling 4 synths. It’s been maddening to get all to sync perfectly. I’ve seen the ERM, but like you had sticker shock. I am going to buy now! Seriously bro, thank you! Love your videos.
I enjoy your content. It’s nice to see you drilling into a subject with such granularity. There’s nothing more frustrating than to sit through or scan a ‘how to’ video, only to discover it’s mostly general info but without the how/why’s, which always leaves more questions than answers. 🙏
Hi man! 😜 Great video as usual! I’ve had mine for several years now, and I’ve been preaching how good it is since the moment I set it up. Like you my ever growing hybrid studio has put me through the mill time and time again, but hay it’s taught me more about midi diagnostics than I ever would of done, so there’s that! 😂 Seriously though I totally agree with the price point, as it’s the most important piece of gear I own. Tried several scenarios, but settled on usb from Live to the Deluge, then dyn midi onto the Multiclock, then onto a Kenton midi thru 25, and out from there to everything apart from my 3 main synths. That lets me run the entire studio with midi from the DAW, midi controllers or the Deluge, and gives me transport control over the drums, Hydra, Iridium and Pro3, and more importantly, transport and sync control over their individual arps and internal sequencers, which is where the main delay issues seem come from in my experience so far. The Pro3 and the Matrixbrute for example were horrendously poor for keeping arps anywhere near the one without this piece of kit. In my situation at least, I use the swing setting knobs to add some groove to the drum midi chain. 😍 10/10 for the ERM Muliclock from me. 😉
Are you still running this setup? I'm trying to get back to into making music with my Deluge, Hydrasynth, some monosynths and iPad. But I guess I grew tired of the limitations and the annoyance of using a new system (to me). So are you still using this setup and does it work for you? Would I need to buy the ERM Multiclock to get the best result?
Trust is an important word in music making. If a setup you've worked on for ages suddenly shows issues, you can hit a wall creatively - even mentally at times.
Good you found a solution for that one. Although expensive, the result will definitely pay off.
The steadiness of midi clock via devices like this is due to audio streams being prioritised and delivered accurately, as opposed to USB midi being delivered whenever the OS gets around to it (jittery mess).
@Knobcore this sound ridiculous
@Knobcore " You can get similar performance out of an old computer with the right hardware" Care to elaborate? Very interested!
@Knobcore is, of course, entirely correct. As an old git, I've *used* and heard great MIDI timing and it's something modern tech doesn't have (with rare exceptions like Sequentix / MPCs etc. Computers now are appreciably worse and less snappy than decades ago because they are doing far, far more behind the scenes. But then, don't just take my word for it: www.eejournal.com/article/why-your-computer-is-slower-than-a-1970s-pc/
@@JyotiMishra So true. I cannot stand the timing of modern DAW's. My old Mac IIcx running Cubase in the 90s sounded better than Live. I just discovered Reaper and its timing is better than most. I think I'm going to start using it soon. Definitely buying this E-RM as well. I love the timing of its swing. Reminds me of the Future Retro Swynx that I have.
@@exteriorsigns Good luck, Bud M! If you haven't already, please check out the Sync Gen II from Innerclock. I love this page of theirs -> www.innerclocksystems.com/litmus
You're diverse man. I was just watching your pocket operator tutorials and ran into some latency setting up my minilab 3 within Reason just now. Big thumbs ups for these videos and the knowledge. I appreciate it.
as a old dj i love the shift function on the multiclock. you can shift every channel in time til the groove is right. for you of course
I will never have enough thanks to give to you for the tip of delaying the e-rm plugin track. I've had the multiclock for years now and only today I realized this would work wonders for me. Thank you a million times will still not be enough!
you rock! no one gets into detail on these much needed topics like you. thanks!
Hey Ricky! Found ur channel a month ago and I’ve watched like a 100 vids. U have helped me so much, inspired me. I don’t even have analog synths and just started getting into music production. Just wanted to say thanks for everything. Keep it going :)
You are on the right "track" Ricky.. There is a reason studio's have expensive clocks to sync everything. Not only for midi but even more if you use multiple audio interfaces. You could also use a low budged tactic. If it sounds good it's good don't zoom in on the waveforms!
Sometimes my lazy ass just considers that ... use the PC as a multitrack tape machine. Just play all the rest of the parts live over the beat. So what if it's not on the grid - as long as you're not editing based on the grid.
But then you have to worry about plugins being synced to the beat too ... and then you're kind of limited/have to work extra hard to start moving things around. It would be "nicer" just to have perfect sync, but hey, when they recorded to tape they had no damn grid and I'm pretty sure all of the classic hiphop albums sound fine!
One thing that's probably worth mentioning is that the bandwidth limitations of MIDI itself introduce latency, and because it is such an old standard those limitations are very significant.
For example if you transmit a chord consisting of 4 simultaneous MIDI notes the 4th note will actually be sent around 4 milliseconds later than the first one.
While 4ms is unlikely to matter for most purposes if you're trying to transmit a lot of note and cc data AND clock down a daisy chain of devices you'll end up with problems even using amazing quality gear like the Multiclock (I have one too and I love it!)
_MIDI itself_ doesn't really have bandwidth limitations, it doesn't "know" anything about the data transport. It's the hardware layer running the MIDI protocol that limits the bandwidth, so the 4ms in your example may happen on a typical 5-pin plug connection, but not on a MIDI connection over LAN, or a virtual connection with basically zero latency.
@@DiAL033 Literally the first sentence of the MIDI 1.0 specification is "The hardware MIDI interface operates at 31.25 (+/- 1%) Kbaud, asynchronous, with a start bit, 8 data bits (D0 to D7), and a stop bit.". I appreciate that this applies only to MIDI over DIN but this video is about MIDI over DIN! You're 100% right though, I could have been clearer :)
The last time I was this early, dinosaurs 🦖 were here
It's basically repurposing some old technology to solve a problem. We'd use audio-rate signal laid down as audio to the edge track of tape on reel to reel. It would contain dumb-FSK, smart-FSK, or time-code. We'd then feed that into something like an Opcode Studio 3 to read. Tape itself had its limits with wow and flutter.
Another point to consider is that using MIDI THRU to daisy-chain equipment together has its limits. You shouldn't do that if you care about timing, especially with older gear. Again, we'd use something like an Opcode Studio 3 or Studio 5 to get a direct run from an MIDI interface to a hardware instrument. Don't daisy-chain. Yeah, someone is going to respond "it works fine for me." Good. But it's a crapshoot, especially with more than 2 pieces of gear.
Yep, daisy chaining is a no-no. Built myself a zero latency 6-way MIDI Thru box for almost nothing.
Yea I was gonna say most people with old gear pre daw people say avoid diasey chaining if you can help it..
yep, no daisy chaining, that's it. No need to spend 500 bucks on such a device
m B so I don’t daisy chain. I go from Logic to MOTU micro lite. Each device gets is own midi cable. Yet shit will lose sync from time to time. How do you fix that. Also VCV or Blocks will lose sync. The ERM looks like it would fix that.
Wow, you just transported me back to my teenage years, name checking this technique. Great Stuff!
I love my ERM Multiclock! It solved all my problems and made my hybrid setup a seamless and painless experience. After many years of pain, it changed my life.
In the last 16 years that I have used my setup in "hybrid mode" I have never had any issue with desyncs or offsets, cleaning up or whatever, and I have used legit anything I could get my hands on. At some point my BCR2000 was my midi interface and it worked perfectly.. Ableton Live (and I think FLStudio as well? not sure..) has seperate midi latency compensation dials on every midi in/out track that it can find. I would have several chains going and set the latency compensation for each of these chains and then all I had to worry about was the audio latency. With my crappy audio interfaces it's always been ~48 ms, so I tap that into the start offset of the samples and presto. That, at least, worked for me. I think what we can all pull from this is that it REALLY depends on what your workflow is. If your approach is much more performance/jam alike then syncing issues are much more problematic but if you sketch/compose everything it becomes no issue that something's a bit later or earlier.
Never before commented on your videos but I thought it was about time. I like the videos you make and I secretly watch them while I am supposed to work. You rock!
Realize that midi is serial so the different ports allow you to work parallel! I recently bought a secondhand unitor MKll, It made me very happy to have 8 midi ports all working parallel controlled by the DAW midi master Clock.
I really love "it exacerbated the vibe". definitely going to start using that to describe distorted basslines
Ricky, you are the man. Pleasure to meet you at NAMM this year. Stay safe and stay blessed
Thanks! The E-RM is nice, but I had a lot of problems with syncing from MIDI to my modular. So what worked for me to get a solidly syncing hybrid setup (DAW/MIDI/CV Eurorack), was the Eurorack module Expert Sleepers ES-3 in combination with an ES-5.
The ES-3 connects to any optical ADAT out on your existing interface, e.g. RME Fireface UCX to sync your modular to your DAW and send CV to control your modular (but there are also other Expert Sleepers modules to connect your DAW via USB, SPDIF, Audio). The working principle is the same like with the E-RM: it syncs from audio - so the latency that was introduced within the DAW is exactly compensated for, which is best if your DAW is the master.
With the additional ES-5 expansion, that is directly connected to the ES-3, I send MIDI from up to 8 independently controlled audio tracks to 8 MIDI DIN outs which connect to my MIDI gear (mono jack adapter to MIDI/DIN). This gives you independent control over up to 8 MIDI outputs + modular sync from your DAW: There is a dedicated expansion module with MIDI/DIN outputs (ESX-8MD), so you don't need the adapters, but I had the ES-5 in place, already.
With e.g. an Ableton template, you can make your live easier and start all of your hybrid projects with sync setup in place :-)
Ricky Tinez’s content is some of the best content on the RUclipsz
This video is exactly what I needed! thank you
I've got an E-RM multi clock but I thought the learning curve was a bit steep as I'm new to MIDI.
I got stalled on there not being a 4/4 option.
6:23 MIDI router description was very helpful.
11:34 great explanation of why the audio signal input in port 1
15:00 sounds like unipolar vs bipolar (on Moog gear)
18:00 I think the shuffle mode is for live performance
stay safe 😃
It is not only about latency. I work with Logic Pro X and it´s midi-clock comes at the end in the computer processing hierarchy. So there you have additionally to deal with a very wonky/jittery clock especially the more happens in the arrangement. With the concept of ERM-MC using sample accurate audio-clock you are getting a very stable clock-timing. Now I have a very tight and stable grooving analog-DAW-kombination. Yes it was hard to spend 500 bugs for it. But at the moment the only way to fix this profound timing-problem. Thank you Ricky for this video!
You explained this so well. I own one but hadn't got across it fully. I have now thanks to this.
I used to have no problem until I got into modular and semi modular, because I played one synth at a time with external instrument. I was forced to learn more. Still learning. Videos like this really helps.
Being able to shuffle a TR-808 is so dope. Expensive when you don't own one, inexpensive when you do. Great vid.
thank you very much for a lot of clarification Ricky!
It is amazing. This is amazing. To what you were doing, I would go DAW click to the Quadra-thru, and from there, send all the outs of the quad to the four hardware synths... an octatrack, analog rytm, and then, ... drumroll... a squarp pyramid. it serves as the sequencer for synths without sequencers... That squarp/synth clock chain from one of the quad outputs would go first into the arturia keystation midi, for its arpegiador, then out to a midi merger (same brand), and finally to the squarp input. This allowed me to have the arp of the keystation when I needed it, or just normal keys from the larger grand midi keyboard underneath it (hence the merger). But having the E-RM instead of the crude Quadra thru would be even more dope.
If you wanna solve the midi clock and midi latency problem forever, you made the first step with buying the ERM! Now make the second step and add a MioXL. Use the ERM only as a MidiClock distribution, if you have some analog clocked device (as example Eurorack gear) you can generate perfect analog clock with the ERM to! The MioXL (mio10 is also very good) is used for MidiNotes and every instrument is on a separate output. Don‘t chain midi devices! Then filter every incoming clock message in the MioXL except the one stable from the ERM. And you never will care about this latency shit again. If you don‘t have enough ports, just add another mio.
We have gear for thousands of dollars, but the tools that really help have to be cheap? It‘s like buying a lot of cool cars and putting some 15 year old tires with no grip on them.
The iConnectivity MioXL is such an underrated device but for $500 you get 22 MIDI ports (8 DIN in/out, 4 DIN out only and 10 class compliant USB), ability to patch any input to any output, filtering, remapping, ethernet connectivity, etc.
thanks for writing this 👍👍🙂
Some tips for people having similar problems:
- minimize the length of your midi chains (2 devices max)
- most MPCs have multiple (up to 4 on the 3000) midi outputs, use them all
- anything that is time critical should be first after the master midi clock: drums, bass, arpeggios
- pads can come later in the chain since these sounds are not as sensitive to latency
- I highly recommend the iConnectivity mio10 midi interface and midi router
- highly flexible, you can connect two computers to it, has an USB midi host for midi keyboards etc.
- it's 300 euros much cheaper than what was presented here
- you can pair it with a Quicco mi.1 (50 Euro) bluetooth-midi adapter to control all connected synths from an iPad wirelessly, ex. PatchBase
If you do it right you will end up with minimal latency, I have 10+ vintage drum machines and 6 rack synths and never needed special clocking.
Not saying it's not a cool piece of gear, but you can get way more connectivity for 300-350 Euros.
Make sure you really need the clocking options before spending 500 Euros on something like this.
Would the iconnectivity midi4+ do the same thing?
Latency is the secret "mojo spice" that made 80s and 90s records tricky to re-create, I mean especially the groove or what is known as feel.
No, I think that was caused by something they used to use in the past called a ‘musician’.
it's the total opposite
MIDI chaining adds almost no latency at all - if the MIDI Thru is implemented correctly... If it is a hardware Thru socket that follows the MIDI spec, then it is a copy of the MIDI In data, so you get the latency through the opto-isolator and one or two logic gates for the buffering, which adds very little time. 'Soft' Thrus take the MIDI In and process it inside the main CPU/Microcontroller (or a dedicated Microcontroller for MIDI I/O) , then send it back to the MIDI Thru scoket, and this can cause latency. The main cause of the problems with MIDI that are often attributed to latency are poor cables and poor opto-isolators (old gear used to use cheap (=slow) opto-isolators), which degrade the edges of the currents that carry MIDI data. But the cabling is often the problem: many MIDI cables are not screened twisted pair (as specified in the MIDI spec), but are twin screened audio cable (which isn't designed for carrying data, and has other parameters wrong as well). The bandwidth of many MIDI cables is often poor as well - and this also degrades the edges of the currents that carry MIDI data. Run MIDI data through several slow opto-isolators and low bandwidth MIDI cables in series (a chain) and you get a MIDI signal that has edges which are not clean and sharp any longer, and this is what often causes many of the problems that are described as 'MIDI latency', and ultimately it can cause MIDI data errors. MIDI cables that meet the spec are not easy to find - I just spent 3 months trying to source a special-purpose MIDI cable from various music cable providers, and the only cables that came close to the MIDI spec were designed for carrying AES/EBU and SP/DIF data signals. Cables designed for audio are different to cables designed for MIDI: data cables need to have higher bandwidths, lower capacitance per meter, and different characteristic impedances. Most of the DMX cables that I was offered didn't meet one or more of the required specs. In general, MIDI cable testers check the continuity of the cable - not any of the important parameters. More info on screened twisted pair in my blog at: blog.synthesizerwriter.com/2020/01/quick-and-easy-midi-clock-sync-between.html
You can have the best cables, but won't even come close to the ease and flexibility of this box. Thanks for the info though!
@@willdahbe but I'll still have $600
@@willdahbe sounds like you completely missed the point. Martin is saying, instead of running out a dropping 500€ on this box, rule out the cables as the possible source. If you buy this or any other box and you have shitty cables, you're still likely going to have problems
@@GiveThemHorns completely got the point as indicated by my gratitude for the info given. piss off.
great summary of midi thru and the opto isolation in the spec!
Quick tip about that solo channel thing.. I think, when you output the audiosignal to another output then the main output of your soundcard, you can output that audio signal and use “solo track” buttons with no problems. You’ll set that up like you would setup a sort of monitoring channel. (like dj monitoring).
When I use midi with my daw, I don’t record multiple instruments at one time. I’ll record each part one by one. One way to mitigate all the cables is using a patch bay.
But then u gotta set up something every time you wanna use it
If you’re a well planned artist that has an exact image of the final result in mind before recording, that is awesome. Good for you. I go back and forth with my stuff all the time.. but then again I don’t have enough stuff to have serious timing issues anyway.
Same here. I also try to record in as early as possible, so I commit to the sound. IF it doesn't work, then I go back. But the "record early commitment" part pretty much revolutionized my music making few years ago. No more "I just tweak this one thing... and never get anything done" thing, where I was for sooooo long time. Now I actuelly get s**t done. :D
Krmpfpks What I usually do is use vsts first, to hear what kind of hardware sound I want to use with each track. I only have a few hardware pieces. I like it like that because I don’t have to go through to many processes to find a decent sound.
Sauce_Aux even if you set everything up at once you still have to manage each piece of gear. If you have a patch bay for midi it wouldn’t be to bad to setup every time.
Great video Ricky. Yeah, the Roland SBX-1 has swing but like you said most the stuff many of us use have not only swing but really good swing!
Preach it!!!! My synths are all over the place and there is no one adjustment that fits them all. Back in the day they had a very expensive device called the "SRC Friend-Chip" similar concept but ERM so more advanced.great explanation.
They have approved MIDI 2.0 specification.
I'm curious how the new MIDI 2.0 devices will perform.
The number of solutions to midi latency is as varied as the number of gear setups. Glad you got it sorted.
I appreciate you bro !! always good info. I'm facing the same problem and trying to rewire and adjust my studio with a hybrid system.
Very cool video thanks 🙏🏼 and I love this erm is is a life saver. The price for this unit is actually compensated with a tons of hours I saved with arrangement issues.
Dude ! SO glad you’ve posted this video it’s like the universe confirming I have to buy this ! Like you , I have had ALL the same insane frustrating problems recording hardware to audio and the after tweaking nightmare ! I have considered buying this for a while but yes the price for something that doesn’t make a noise was off putting , like you, say gonna close my eyes and hit buy now after your endorsement lol keep up the great posts stay safe 😊
Awesome video!
Yo! Thanks for tuning in brother! Fan of yours 🙏🏽
FWIW I have all my outboard hooked up through an iConnectivity Mio10 which - so far - has solved this problem for me and is quite a bit cheaper. It isn't a clock generator (as far as I know?) but I've never had a problem with how it distributes clock signals. I bought it originally to stop daisy-chaining midi devices but it's helped in those terms too.
Me too!
yeah, it seems that shorter chains is really important. i wonder if using one of the hardware sequencers as the master clock and using the mio to distribute the midi could give you the combo of stability and consistently low latency?
@@Kebin-Blebin yep, I see no reason why that shouldn't work
Same here, best value right now for midi connectivity.
Awesome video as always. I use the SND ACME-4 similar to Multiclock but i went with the SND because of the swing and division on switches. Very hands on but both are good options. Keep up the good work ur doing. Loving it!
In like ‘85-‘90 the workflow was MIDI-units, Atari ST and tape, tied up with linear time code. Or, a Linn 9000 with a clock track from tape, essentially the same as your clock plugin. Daw not invented yet, we just heard that timing was off. We then understood that Midi (5-pin only option) is an 8-bit serial protocol. Meaning that ch 1 is sent first, then ch 2 etc, and each byte must be complete before being passed on. A through port will process from input in the same manner, inducing more latency. And the 8-bit logic was really slow. Result: Never used Midi through anywhere, got an 8 in/8 out interface, and used only ch 1 for all drums/percussives, via the separate interface outs, say port A to D. Other keyboard stuff can use ch 2-16, maybe via some through ports, but not sharing a physical cable/port with anything beat-related, say ports E-H. Nor would we send any controller data (pitch bend etc.) through the ports used for drums. Lacking sophisticated clocking equipment, Midi clock was only used from sync box (time code) to the sequencer. We never got to the point of using the internal sequencer in a drum module, instead just sending Midi notes form the main sequencer. Very anecdotal, but it was what it was. Today, we just use daw and plugins, not equally fun and inspiring maybe, but timing is never an issue... 👍😅
Something therapeutic about listening to gear talk. Like hearing someone solve the problems that I wish were waiting for me at home
This video made me think, that Ricky you could be interested in Blokas Midihub, which is a intuitively programmable 4x4 + usb midihub. Not saying it would be a replacement for what you own, but it could well suit your needs. Just got mine, which I pre-ordered/backed and just as an example here are two things I was easily able to achieve with it: I'm now able to play my Digitakt as a polyphonic sampler, with my master keyboard. I'm also able to play my Volca Drum with my master keyboard so, that all the notes are mapped to the corresponding pitch values (as for some odd reason Korg didn't make it read midi-notes out of the box...). Great video once again!
I've never had any luck using ableton midi clock with external keyboards. It's always a little bit too late or little way too late. Gave up. 😑 I love it when you back track cuz got hiped up saying good things about it. Makes me realize there's still ppl who gives honest reviews out there. 👍
The MOTU microlite also shows up as separate MIDI devices in your DAW - 5 input and output devices.
Cool vid! You need a 6U rack for those Black Corp. synths! hehe
I have music contents check me out now
I just got fed up with computers and bought the MPC X and the MPC ONE for travel. And I was just about to pull the trigger on the all plugins FL studio bundle until I seen what those new MPCs can do. The way I see it it they solve 3 problems in one. DAW, MIDI Controller and audio interface all wrapped up into one. Granted only the X has phantom power, but there are workarounds for the others. The way I see it, a standalone unit dedicated and made for nother but music has to be much better for music production than any laptop. Granted you can't use third party plugins in standalone but there are workarounds for that to. And with everything they have built into those things, it's really a non-issue. I also bought the Roland JDXI to make my own MPC extension's. Can't wait to have endless sound creation capabilities with a bunch 9f knobs to twiddle with. Lol
Dope ferns
They are still computers just optimized for music
Thank you for your thoughts. Obviously, you don't have too much sluggish and/or jittery MIDI gear around or gear with different sluggishness. If that would be the case, a tight clock cannot help, because the latency does not come from the clock, but from the MIDI latency and jitter from the devices themselves. I wonder, if your setup just cuts off some of the inbuilt MIDI latency for your hardware devices instead of coming them in late for this same amount of MIDI latency. So the click isn't tighter itself, but the delay doesn't disturb you so much anymore. In some Keyboard magazine issue from the 90s, there was an exhaustive comparison of MIDI latency and jitter for different devices - some workstations with up to 30 ms MIDI response plus irregular bidirectional jitter of aditional 10-15 ms and this not being a constant, but dependent on voice usage and MDI controller use as well. This would be something you cannot fix with a tight clock, but you can set a predelay, if you know the typical MIDI delay of your particular device. Fortunately, MPCs are quite tight and most modern gear is too.
Another case of GREAT timing on topics. I have a need for this exact type device. To finally complete the hybrid fusion in my studio. Definitely relate to the MIDI chains with my MV-Modules setup. I'll have to take a serious look at the ERM. Keep up the great work bro!!! Peace...Netm8kr
Part of the issue could also be your daisy chains. Try using a midi splitter. Kenton makes a 1:24 midi splitter. Master clock goes in the in. Then an out per device.
Just avoid these midi chains and use a stern configuration. That means each hardware device gets its own midi in and midi out connection to the midi interface. Just buy something like the M-Audio Midisport 8x8 and you do not have these problems.
I have solved this with a pair of MOTU MIDI Express 128 units.
YES! Was thinking the same thing. I don't ever use Thru. That's something is just a holdover from the days before multi-port midi interfaces (80's). Give each device its own cable and you can adjust latency per port as needed.
Amen. There are almost no reasons that speak for daisy-chaining and tons that speak for star-network. On the other hand it depends on the length of the chain, so if there's just one thru hop, it's okay I guess. (Which, coincidentally, is exactly the maximum amount of thru hops I have in my setup ;D )
Chaining is fine for modules without any sequencer built it... but yeah that's a good idea for drum machines and synths with sequencers.
I have a Roland A-880 MIDI patch bay for just this reason. It's 8 in x 8 out, with a button matrix on the front for making connections. Super easy to set up your default & then change things on the fly. It can merge ports and send one source to multiple destinations. It's ancient and for some reason, they stopped making them and I still haven't seen anything that comes close. That's why I have a backup. They show up on eBay from time to time. Totally worth it.
I would never ever use clock coming out of a computer. Ableton slaves super good.
My solution to this midi issues with other gear was a machine called Boss RC 505 and have gotten two more since then.
The clock of that looper is the most solid thing. I have the clock coming out into a Midi Solutions midi merger box where my keyboard also goes in. That way i get to play my synths set to their own midi channels and have everything stay in time. I can start and stop everything from the looper and it’s spot on everytime. The cool thing about this is that you can play the first phrase of something in your head and retain the feel and the tempo and everything in your studio will follow that. So you never have to worry about trying to start a new idea with specific tempo in mind. Just keep it natural. As soon as you close that first loop the clock sets and starts. I am surprised that these RC 505 or even the RC 202 are so solid when it comes to clock.
Really appreciate the video I've been doing a lot of ableton work but haven't quite got to routing things together. Glad to have this to consider when setting up!
Looks great. Nice vid, thanks !
Great video explaining that ERM device. I've always stopped playback and trimmed off latency from the beginning of the recording. Though that only works OK with drums, not other melodic sounds.
@RickyTynez I think all your midi problems came from the fact that you use the midi thru on some of your midi device to daisy-chain them.
If you could connect them directly to their own channel I believe you wont have any problems. And in this way you could still use any of them to send clock or to controll any other devices (provided they support that).
Here is the solution I want to go with when building my new studio:
1 x "MOTU MIDI Express XT" MIDI interface - 8 MIDI ports in, 8 MIDI ports out (as each port has 16 channels - that means 128 MIDI channels out, 128 MIDI channels in - can controll up to 128 MIDI devices simultaneously).
2 x "Kenton MIDI Thru 12" multipliers/splitters connecting them to 2 MIDI outport ports on the MOTU MIDI interface.
I will have around 24 MIDI devices.
To avoid any timing dissimilarities/problems, I will connect all my MIDI devices through MIDI DIN cables to the 2 Kenton multipliers (better said than splitters).
No daisy chaining, no MIDI over USB and MIDI Din combination! Only MIDI DIN 5-pin cables!
On the way back to the MOTU's MIDI input ports I will put 3 "Kenton merge 8" mergers (all 3 of them will handle 24 devices) and connect them to 3 of the MOTU MIDI input ports.
All connections with the same cables brand, model and lengths.
The audio will be through audio balanced cables straight to the patch bays - no audio-through-USB!
In my Cubase Pro I will create a template with 1 track for each midi device (each on its own midi channel).
This should work perfectly.
Every MIDI device will have its own track inside Cubase, its own channel and will be able to send clock and controll any other device.
Perfect setup for creativity.
If you still have problems with your MIDI connections, I suggest to you to take in consideration this setup.
What can go wrong? If some of the MIDI devices are slower than others in processing/sending the data, as each one is on its own channel and track, this can very easily be compensated for inside Cubase on each separate track.
Was waiting for this 😁 very nice!!!
I’ve used a midi sport 2x2 for a long time with success and paired it with an old Roland A-880 midi patchbay to route midi to all the synths. This avoids a long thru chain.
I remember some older devices like the Yamaha A3000 sampler being sensitive to the midi channel which added latency if they were on a higher one.
Yeah... that's why I've used DP for like ... 30 years. Rock-solid MIDI timing (it's roots are a MIDI sequencer) w/ a couple micro lite MIDI interfaces along w/ Volta for my CV/Gate and clock. No lag. It's wonderful.
The main issue I have is the audio is delayed when it’s recorded back in my DAW. I need to offset it negatively independently for each device because there all different. But the way your saying you have your set you are setting the negative offset globally in your DAW. This can already be done so the ERM doesn’t seem any better than a MIDI hub and setting the offset in the DAW.
As someone who has worked on the development of a few hardware synths, your problem is in the MIDI signal protocol in general. MIDI doesn't have a fast data transfer rate (only 31250 bits per second). If we convert that into bytes per second, it's 3,906 bytes per second. Well, how much data does MIDI use to communicate anything? Your typical MIDI message be it note on, note off, volume, CC, etc is 3 bytes. That now gives you 1,302 commands per second. Still fast for most music and setups. It's really a case of the more things you have on the line, the bigger bottleneck it's effectively creating where devices have to wait for other data to be sent while waiting for their own commands. This is because MIDI can only send commands 1 byte at a time. So devices have to just wait for their commands to show up after everything else that doesn't go to them before they can execute them. A device to have a dedicated MIDI port for each device in your chain is the best way to go as you showed with hardware you used in this video. In that configuration everything can be sent data as fast as possible, thus eliminating any unnecessary traffic on each device's data line. I hope that helps explain it and wasn't too overly technical ^^;
These things are a revelation. I winced at a fat wad of cash for the syncgen pro when it came out but they are life changing.There is no substitute for these kind of devices in contemporary computers. It’s not latency, that can be managed with offsets. Jitter is the problem. USB packets midi clock on a janky way and no one seems to have been able to solve it. I spent years fighting the problem...transients drifting. There was a good explanation on the inner clock website including tests of various outboard gear -classics and modern- and how how they all have there own limitations to being synced...hint. The old MPCs do very well
this is why it‘s so important to master your instrument, or you will end up with ... more stuff.
you don‘t have to do everything by youself. share. form a band.
I would like a band yet sometimes humans can be passive aggressive/troublesome,and I am NOT the one to play games with anyone...I USED to have a partner I DJed with,and he was responsible for losing most of the record collection back then,All I got was 'sorry',so I vowed to make my own studio,own my own narrative,and plus,I rarely sleep,so I can just get up and record what,how,and when i want to...freedom is addictive,and it allows me to work out new ways to do be more simplistic...
I'd love to be able to justify this. I have a similar hybrid setup just with different gear. My midi gear all goes through 2x ESI M8U, usually with Akai Force as the midi master. I've a laptop and a desktop, both running Ableton with different plugins installed. For the most part I rarely have issues, the odd waver in timing but thanfully not often at all and everything usually locks in time well. I'll keep this mind though as I'm sure its absolutely rock solid.
So let me get this straight? with 4 midi ports routing 64 channels, does that mean you can run midi clock from ableton, through an audio interface, through the E-RM Multiclock to keep the stable clock while using the 64 channels to route all kinds of midi data such as sequencers, CC parameter change signals etc without the need for something like an iconnect midi4+ or similar midi hub to route all kinds of midi signals between devices? 😲
Always a pleasure to see your videos man! I got mine today so who better to check out than you I don't know man! Cheers!
Use an 'External Audio Effect' device for the soloing issue
"I'm a nerd, but I'm no super nerd". Priceless.
There's a max4live device called solo safe to avoid the problem you are facing.
Link?
It's a little more expensive, but the RME PCIE interfaces use their own tweak on midi processing and have excellent results. IMO it's the best way to go with a hybrid studio. You can get any converter you want on lightpipe (even use apogee symphony on Windows) AND you also get everything this box does. 32 channels of adat i/o and two midi i/o or more depnding on the PCIE you get. Faster than any other interface besides the presonus quantum thunderbolt. Works perfectly with the Expert Sleepers eurorack lightpipe interface. Better resolution than midi to CV.
Yep !
Knobcore yep! Pcie is natively compatible with processors and the motherboard requires no extra thunderbolt chip to route things. You can also use passive cables. Thunderbolt cables are active and also require a chip. Also get this, I picked up a $2,500 broken audiophile receiver for $10. It couldn’t put out volume, but what it could do was take adat input and go analog put through the preamp output. So fucking mastering grade DAC conversion for $10 because someone didn’t want to pay for a repair! Lmao!
Knobcore oh absolutely. I bought an apogee symphony when the mkII came out since everyone was offloading theirs. It’s gone up in value by maybe $500-600 since then. Sometimes I don’t understand people’s obsession with the latest and greatest. It’s not like a record from 2008 sounds like it’s coming from a gramophone. The Ad/da improvements are still happening, but these companies aren’t NASA research labs. New models, at the ad/da level are barely improved, if that. And companies, like Apogee, that are owned by hardcore investment firms (I mean apogee is like 2% of Engaged Capitals portfolio) in some ways are just repackaging the same tech, giving it VERY subtle but pleasing eq adjustment, and then calling it the “mastering version”.
I mean, is it not a coincidence that the first year they are owned by a private equity firm, that they re-release an already existing 2x6 symphony mkii module with a “mastering” tag? Is that not the exact move a private equity company would do? Mastering quality is to music production as Audiophile grade is to home hi-fi...
DUDE. I feel you on clocking. Plus. I really want a launchpad mouse pad!!!
I have learned the hard way to never use Midi Thru. Synths crash ... not often ... but sometimes. And then the 2nd device in a chain crashes and you loose everything. Of course this happens only during live situations. I use a star system and a dumb midi thru box that by design cannot crash. My audio and midi is set up so everybody can talk to anybody if wanted but is still kept independent enough that if a synth crashes only that one synth crashes. It took me years working towards this setup. And it still can be improved
Excellent video. I actually found there was no latency when I have a sequence going but it’s when I play something in real time it’s when I started to notice latency.
This dude is my fav RUclips music guy for sure learnt so much from this man
Love your vibes brother! I've experienced actual physical pain trying to get my machines to make sweet love to abletons clock ha. i may have to invest in this thing. thanks for the video bro and happy sunday x
@Ricky Tinez What about capturing the internal clock of the MPC 3000 and sending that back out for every other pass? Does the Multiclock allow you to capture the clock of external gear and allow you to add that to an audio track in your DAW? Ever heard of Innerclock System's Sync Gen 3LX?
this information is a discovery for me that I am new in electronic music production , you are an innovator in this time
coexistence between the digital (daw) age and the analog age, thank you very much for sharing your knowledge I follow your channel closely, regards desde Argentina (sorry me english is very rough)