{"id":675,"date":"2026-01-18T17:39:15","date_gmt":"2026-01-18T22:39:15","guid":{"rendered":"https:\/\/guitar-faces.com\/blog\/?p=675"},"modified":"2026-01-18T17:39:15","modified_gmt":"2026-01-18T22:39:15","slug":"when-your-daw-freezes-but-your-cpu-is-fine-a-musicians-introduction-to-latencymon","status":"publish","type":"post","link":"https:\/\/guitar-faces.com\/blog\/thoughts\/when-your-daw-freezes-but-your-cpu-is-fine-a-musicians-introduction-to-latencymon\/","title":{"rendered":"When Your DAW Freezes but Your CPU Is Fine: A Musician\u2019s Introduction to LatencyMon"},"content":{"rendered":"
\n
\"\"
LatencyMon monitoring latency in Reaper on Windows 10, 2026\/01\/18<\/figcaption><\/figure>\n<\/div>\n\n\n

Latency \u2014 learning what to do about it<\/h2>\n\n\n\n

Step by step, in by inch<\/h3>\n\n\n\n

I assumed I understood audio latency on computers. Buffer size? Check. Sample rate? Check. CPU load? Keep it low, freeze tracks, print effects. I absorbed Kenny Gioia’s Adjusting Recording Latency (Loopback Test) in REAPER<\/a> video, made the adjustments, and upgraded to an interface (Rubix44) with better monitoring abilities.<\/p>\n\n\n\n

And yet\u2026 every so often \u2014 on machines that should have been fine \u2014 I\u2019d get the dreaded symptoms: a DAW freeze, a sudden glitch, or that heart-stopping moment where audio just\u2026 stops responding. <\/p>\n\n\n\n

The CPU meter wasn\u2019t pegged. Disk activity looked normal. Nothing obvious was wrong. I chalked it up to \u201cWindows being Windows\u201d and moved on. And that worked\u2026 until it didn\u2019t!<\/p>\n\n\n\n

The Hidden Kind of Latency<\/h3>\n\n\n\n

What I’ve learned is that not all latency lives inside your DAW. There\u2019s another layer \u2014 deeper, quieter, and far more insidious \u2014 called interrupt latency. This is about how quickly your system can respond to real-time events: audio buffers, MIDI messages, USB traffic, network activity.
When that layer misbehaves, your DAW can freeze even when everything looks fine.<\/p>\n\n\n\n

Enter LatencyMon!<\/h2>\n\n\n\n

Download LatencyMon<\/strong> from Resplendence Software<\/a><\/p>\n\n\n\n

What is it?<\/h3>\n\n\n\n

LatencyMon is a free diagnostic tool that doesn\u2019t measure audio performance directly. Instead, it watches how your system handles interrupts and deferred procedure calls (ISRs and DPCs). If those spike too high, real-time audio suffers.<\/p>\n\n\n\n

At first glance, LatencyMon can feel\u2026 unfriendly. Acronyms, scary red text, references to drivers you\u2019ve never heard of. But with a little context, it becomes an incredibly empowering tool.<\/p>\n\n\n\n

The Big Mental Shift: Chronic vs. Event-Driven Problems<\/h3>\n\n\n\n

Here\u2019s the most important thing I learned: A system that runs clean for hours is not \u201cbad.\u201d A system that occasionally spikes has a trigger. This distinction changed everything.<\/p>\n\n\n\n

If LatencyMon reports consistently high numbers, you may have a chronic configuration issue. But if it runs happily in the green for long stretches and then suddenly explodes? That\u2019s almost always event-driven. Events that contribute to latency include:<\/p>\n\n\n\n

    \n
  • Sleep\/wake.<\/li>\n\n\n\n
  • Display changes.<\/li>\n\n\n\n
  • Network activity.<\/li>\n\n\n\n
  • USB power management.<\/li>\n\n\n\n
  • Audio devices you\u2019re not even using.<\/li>\n<\/ul>\n\n\n\n

    Once I understood this, LatencyMon stopped feeling accusatory and started feeling like a detective\u2019s notebook. And my system stopped freezing. Mixes I worked on ‘last night’ started running just the same the morning after. <\/p>\n\n\n\n

    The Usual Suspects (and Some Surprises)<\/h3>\n\n\n\n

    In my case, the biggest offenders were not my audio interface or my DAW. They were things like:<\/p>\n\n\n\n

      \n
    • Display audio drivers (HDMI \/ DisplayPort pretending to be speakers)<\/li>\n\n\n\n
    • Unused onboard audio<\/li>\n\n\n\n
    • Virtual audio devices I wasn\u2019t actively using<\/li>\n\n\n\n
    • USB power management quietly re-enabling itself<\/li>\n\n\n\n
    • Network and ACPI activity during power state transitions<\/li>\n<\/ul>\n\n\n\n

      Disabling Intel Display Audio and unused Realtek audio alone dropped my worst interrupt spikes from tens of milliseconds to well under 50 microseconds. That\u2019s not a tweak. That\u2019s a revelation!<\/p>\n\n\n\n

      A Word About \u201cGood\u201d Numbers<\/h3>\n\n\n\n
      \n

      After selectively disabling devices I didn\u2019t need during recording, my system settled around 30\u201340 \u00b5s peaks. At that point, I stopped chasing numbers and started making music again.<\/p>\n<\/blockquote>\n\n\n\n

      LatencyMon doesn\u2019t give you a single pass\/fail score, but rough guidelines help:<\/p>\n\n\n\n

        \n
      • Under 100 \u00b5s ISR\/DPC peaks: excellent<\/li>\n\n\n\n
      • Under 500 \u00b5s: solid<\/li>\n\n\n\n
      • Over 1000 \u00b5s: worth investigating<\/li>\n\n\n\n
      • Tens of thousands of \u00b5s: something is seriously misbehaving<\/li>\n<\/ul>\n\n\n\n

        After selectively disabling devices I didn\u2019t need during recording, my system settled around 30\u201340 \u00b5s peaks. At that point, I stopped chasing numbers and started making music again. Which, frankly, is the goal.<\/p>\n\n\n\n

        Less Is More (Especially on Windows)<\/h3>\n\n\n\n

        Every enabled audio device is something Windows feels responsible for managing, even if you\u2019re not using it. You probably have some of these in plain sight:<\/p>\n\n\n\n

          \n
        • Webcam microphones.<\/li>\n\n\n\n
        • USB mics.<\/li>\n\n\n\n
        • Virtual routing drivers.<\/li>\n\n\n\n
        • Calibration \u201cvirtual devices.\u201d<\/li>\n<\/ul>\n\n\n\n

          None of these are bad. But they don\u2019t need to be present during a recording session. I now think of my system in modes:<\/p>\n\n\n\n

            \n
          1. Recording mode<\/strong>: only the interfaces and drivers I need<\/li>\n\n\n\n
          2. Meeting mode<\/strong>: webcams, virtual routing, USB mics<\/li>\n\n\n\n
          3. General use:<\/strong> everything else<\/li>\n<\/ol>\n\n\n\n

            And I have a laptop at the ready to offload 80% of modes 2 & 3. <\/p>\n\n\n\n

            Disabling devices<\/h3>\n\n\n
            \n
            \"\"
            Recording mode<\/strong>: unused devices disabled. VB-Voicemeter is a virtual device I use to stream mixes in Zoom\u2014a routing utility that doesn’t control hardware. <\/figcaption><\/figure>\n<\/div>\n\n\n

            Disabling devices (not uninstalling them) takes seconds and removes entire classes of background activity. Open Device manager –> Sound, video and game controllers. <\/p>\n\n\n\n

            LatencyMon as a Learning Tool, Not a Stress Test<\/h2>\n\n\n\n

            Used poorly, LatencyMon can make you paranoid. Used well, it teaches you how your system actually behaves. The key is not to panic over single spikes. Watch for patterns. Correlate spikes with actions. Change one thing at a time. Stop when the system behaves. And most importantly: don\u2019t chase perfection. Chase stability.<\/p>\n\n\n\n

            Why This Matters for Musicians<\/h3>\n\n\n\n

            We tend to blame ourselves when technology gets flaky.
            \u201cMaybe I pushed the buffer too low.\u201d
            \u201cMaybe this plugin is unstable.\u201d
            \u201cMaybe I need a new machine.\u201d
            Sometimes the problem isn\u2019t musical or computational at all. It\u2019s infrastructural. Understanding that \u2014 and knowing how to verify it \u2014 is deeply calming.<\/p>\n\n\n\n

            A Practical Takeaway<\/h3>\n\n\n\n

            If you\u2019ve ever stared at a frozen DAW wondering what just happened, learning to read LatencyMon \u2014 even a little \u2014 is worth your time. <\/p>\n\n\n\n

            It certainly was for me.<\/p>\n\n\n\n

            \u00a7<\/strong><\/p>\n\n\n\n

            <\/p>\n","protected":false},"excerpt":{"rendered":"

            Latency \u2014 learning what to do about it Step by step, in by inch I assumed I understood audio latency on computers. Buffer size? Check. Sample rate? Check. CPU load? Keep it low, freeze tracks, print effects. I absorbed Kenny Gioia’s Adjusting Recording Latency (Loopback Test) in REAPER video, made the adjustments, and upgraded to […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[5],"class_list":["post-675","post","type-post","status-publish","format-standard","hentry","category-thoughts","tag-vanquishing-doubt"],"_links":{"self":[{"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/posts\/675","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/comments?post=675"}],"version-history":[{"count":8,"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/posts\/675\/revisions"}],"predecessor-version":[{"id":685,"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/posts\/675\/revisions\/685"}],"wp:attachment":[{"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/media?parent=675"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/categories?post=675"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/guitar-faces.com\/blog\/wp-json\/wp\/v2\/tags?post=675"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}