In most of the speaker reviews I've read, the reviewer will evaluate and comment on speakers based on listening to a bunch of music and maybe a movie or two.
I think this is probably not ideal for evaluating mid-bass because who knows what kind of post-processing has been done to the sound? Unless the reviewer has listened to the source material a million times using high-quality, neutrally-voiced headphones, he probably won't know what the song/movie is even supposed to sound like.
But deeper male voices (starting somewhat below 100Hz) are something that we are all familiar with and we can all probably immediately and instinctively tell if there are problems with its reproduction. And podcasts offer a jillion hours of deeper male speech, with many podcasts being recorded in studios using fairly high-quality equipment.
I was thinking about this because I listen to a couple podcasts frequently, both in my car and using high-quality in-ear monitors. I have also tried listening to the podcast on two different sets of bookshelf speakers: Ascend CBM-170s and Wave Crest Audio HVL-1s. Here are my listening notes:
In-ear monitors: crystal-clear reproduction, sounds like the podcaster is sitting next to me. There aren't any echos/reverberations that would indicate what kind of room he's in, how big it is, etc. The sound quality is high enough that I can tell if he has his window open because I can hear extremely faint traffic and ocean noise sometimes.
Car stereo: I have a stock stereo with no tweeters. The podcaster's voice sounds pretty natural, but flatter and less crisp, presumably due to sound reflecting within my car. I've never been able to hear noise from outside his window. Of course this may also be due to environmental noise while I'm driving.
CBM-170s: back to being very crisp and sensitive enough that I can hear the window noise again. But there is a touch of reverb in the lower frequencies that make it sound like the podcast is being recorded in a medium-size room with hard surfaces. Possibly because I have the CBM-170s in a medium-size room with hard surfaces.
HVL-1s (in the same location as the CBM-170s): the higher frequencies are pretty crisp but the lower frequencies have a fairly significant amount of distracting reverb. It's hard to describe the impression this gives. It's not that the podcast is being recorded in a small room. It's almost like the podcaster is too close to the microphone, or that the mic is picking up early reflections off a desk. Actually it sounds a bit like the podcaster is somehow inside the speaker enclosure.
I know that my IEMs are accurate thanks to measurements on InnerFidelity. Going from there, it seems like my stock car stereo is fairly accurate but simply not as good as the IEMs. And the bookshelf speakers are both adding reverb that doesn't exist in the recording. The CBM-170s are pretty listenable but I find the reproduction from the HVL-1s distracting and fatiguing and I have to switch to headphones after a couple minutes.
I have a pet theory that what I'm hearing from the bookshelf speakers is port output, since the output from ports is delayed by a wavelength, which basically means adding a ~10ms reverb to ~100Hz frequencies. The CBM-170 ports are tuned to ~50Hz which means output around 80-100Hz is going to be pretty faint, which would account for the slight impression that the podcaster is in a larger room. But the HVL-1s are tuned around 65Hz so there's going to be significantly more port output around the lower frequencies of the podcaster's speech.
Do you guys listen to a lot of (unprocessed?) male speech on your systems and if so, does it sound accurate?
