Someone ran the same movie through an Apple TV and then off the 4K disc, sound level matched the two with REW and a UMIK-1, and measured the result on the RTA. Average SPL curves first, peak measurements second. The gap came out at roughly 0.9dB. What caught my eye more than the number was what he had to do to get there: -20 on the AVR for the Apple TV, -26 for the disc. Six dB of difference in volume setting before anything was actually compared.
That is the part people skip when this argument comes round. Half the "the disc just sounds bigger" posts are really "the disc plays louder at the same volume knob position". A lossless track off a UHD disc has room a bitrate-capped stream does not, but the reference level a disc is mastered to is not necessarily what a streaming app hands your receiver after whatever sits in the chain in between.
I am not sure a 0.9dB difference on averaged SPL settles it either way though. RTA over a whole film mostly tells you the tonal balance survived the trip. It says nothing about what happens on a hard transient, and dynamics are exactly where I would expect a compressed stream to give way if it gives way anywhere. Gating the measurement to one scene with a real swing in it would be far more convincing than an average over a reel.
Still, credit for measuring instead of listening once and posting. If anyone wants to redo this properly, pick a scene, gate to it, and post crest factor rather than the curve. Has anyone here actually picked disc over stream blind, or is that difference always just volume?
That is the part people skip when this argument comes round. Half the "the disc just sounds bigger" posts are really "the disc plays louder at the same volume knob position". A lossless track off a UHD disc has room a bitrate-capped stream does not, but the reference level a disc is mastered to is not necessarily what a streaming app hands your receiver after whatever sits in the chain in between.
I am not sure a 0.9dB difference on averaged SPL settles it either way though. RTA over a whole film mostly tells you the tonal balance survived the trip. It says nothing about what happens on a hard transient, and dynamics are exactly where I would expect a compressed stream to give way if it gives way anywhere. Gating the measurement to one scene with a real swing in it would be far more convincing than an average over a reel.
Still, credit for measuring instead of listening once and posting. If anyone wants to redo this properly, pick a scene, gate to it, and post crest factor rather than the curve. Has anyone here actually picked disc over stream blind, or is that difference always just volume?
