Big Dog Little Dog

Why Your Mix Sounds Fine on Headphones but Nothing Else

I spent three years trying to fix mixes in the dark by turning the volume up. That turned out to be the opposite of the solution. The problem is that our brains do weird things when sound is loud. Frequencies we normally dismiss become obvious. Bass frequencies start affecting everything else in the room. The mix that sounded punchy and detailed suddenly loses its low end when the volume drops. The workaround is straightforward once you understand what you are actually doing. You listen at two completely different volume levels and pay attention to what changes between them. This is the Big Dog Little Dog method. Big dog is loud. Little dog is quiet. You are not trying to make the mix sound good at both volumes. You are using the difference between those volumes to identify problems that your brain would otherwise ignore.

What Big Dog Little Dog Actually Means

The technique has nothing to do with dogs. It is purely about volume reference points in your mix workflow. When you set your monitoring level and then adjust it significantly, certain frequency ranges behave differently. Human hearing is not flat. We hear more bass and more treble at high volumes according to the Fletcher Munson curves. This is a real psychoacoustic effect that every engineer has to work around whether they know the name or not. So you are essentially checking translation. You render a short section of your mix and play it back at a loud monitoring level that you have calibrated to feel natural for your space. Then you turn it down to a very low volume where the sound becomes almost faint. The elements that disappear or dominate at each level tell you what needs adjustment.

How to Do It Without Wasting Hours

First you need a reference volume. Most studios use 85 dB SPL measured at the listening position with pink noise through your monitors. That is the standard. You do not need an expensive meter for this to work. A decent SPL app on your phone can get you within a few dB if you calibrate it against a known source. Once you have a loud level locked in, pick something quieter. Around 60 to 65 dB is usually where the character of the mix starts to change in useful ways. Here is the actual process. Load your mix into your DAW. Bypass any master bus limiting so you are hearing the raw mix. Play a section that contains your full arrangement with all elements present. Set the fader to your calibrated loud level. Listen for thirty seconds. Note which instruments feel balanced. Then drop the level to your quiet setting without changing anything else in the chain. Listen for another thirty seconds. Now pay attention to what shifted. If the kick drum vanishes at low volume, your sub frequencies are likely too boomy or your phase alignment is off. If the vocals sound thin, the mids are probably crowded. If the stereo image collapses entirely, you have width issues that will cause problems on mono systems. I track these observations in a simple text file rather than relying on memory. The details fade quickly.

A Problem I Ran Into That Took Me Weeks to Sort

I had a mix where everything sounded perfect at my calibrated loud level. The bass was tight. The drums punched. The vocal sat right where it needed to be. When I dropped to the quiet level, the entire low end simply vanished. I spent roughly two weeks chasing this. I tried EQ adjustments. I tried multiband compression. I even replaced the bass synth patch entirely. Nothing solved it at the low volume. The issue was mono compatibility on the sub layer. I had been looking at the stereo waveform and it looked fine. The problem was that the sub bass on the bass track was subtly out of phase between the left and right channels. At high volume the ear picked up the body of the sound and the phase issue was masked by room reflections and speaker limitations. At low volume the phase cancellation became the dominant characteristic and the sub disappeared from the mono sum. The fix was checking the mono correlation meter. It read around 0.72 on that track, which is acceptable but not ideal for a sub bass source. I reversed the polarity on one side of the stereo bass track and the correlation jumped to 0.94. The mix held its low end at every volume level after that. I should have checked mono compatibility first. I wasted probably forty hours before I stopped assuming the problem was frequency balance instead of phase.

Where This Method Actually Fails

Big Dog Little Dog is not a complete mixing strategy. It will not tell you to automate a vocal entry or decide whether a snare sample fits the song. It is a diagnostic tool. You use it after you have already built the mix to a place where it sounds reasonable. If you try to mix primarily by switching volumes back and forth you will drive yourself insane and produce something that sounds flat because you are constantly chasing translation rather than making creative decisions. Another limitation is that it only catches problems that are visible through volume changes. Some issues are time-based. A reverb tail that clashes with a vocal phrase will sound the same at any volume. A rhythmic phasing effect from delay repeats is invisible to this method. You still need to listen critically at a single volume in a treated space. This technique complements that work. It does not replace it. The biggest failure mode is using it on unmixed raw stems. If you have twelve individual tracks that have not been blended yet, the volume test is useless. The ratio between elements is completely wrong and every frequency will seem problematic. Build your balance first. Then run the big dog little dog check to refine the translation.

Practical Tips That Actually Help

Use the same section of the song every time you run this check. If you switch between the verse and the chorus your reference point changes and you cannot tell whether a problem is consistent or specific to one section. I usually pick the hook or the main drop where all elements are playing together. Keep your monitoring environment consistent. A cold room changes speaker behavior compared to a warm room. I test in the same temperature window every session. The difference is subtle but measurable on sensitive speakers over a long session. Turn off any air conditioning that creates airflow across the monitoring position. Even a small breeze can shift perceived frequency response on open-back headphones or nearfield monitors. Take breaks between volume levels. Your ears adapt within twenty to thirty seconds. If you immediately switch from loud to quiet your hearing is still partially desensitized and the comparison is unreliable. Wait a minute. Walk away. Come back and listen fresh at each level. This habit alone cut my revision sessions roughly in half because I stopped making unnecessary EQ moves based on fatigued hearing.

Record your monitoring levels in your session notes. I keep a simple log with the date, room temperature, SPL settings, and any issues I noticed. Six months later when I come back to a project I can reconstruct my monitoring environment exactly. This matters more than people realize because a different monitoring level on a different day can make you think a mix needs work when it is already fine. Apply this at the mixing stage and the mastering stage separately. A mix might pass the test but the mastered version could introduce issues. Compression and limiting change the dynamic envelope in ways that affect how the mix responds to volume changes. Run the big dog little dog check on the final master before you send it out. It takes about fifteen minutes and catches problems that would otherwise surface as complaints from clients or listeners on phone speakers. The method is simple enough that beginners pick it up quickly. The application is where the skill develops. You learn to distinguish between a mix that genuinely has a frequency problem and one that just sounds different at low volume because of how human hearing works. That distinction takes time and repeated exposure to well mixed references across multiple genres. I recommend keeping a library of professional mixes you trust and running them through the same volume test occasionally to recalibrate your own listening judgment. Your ears drift. Regular calibration resets that drift.