Program material for subjective evaluation

[radix]

Let's say I get a couple pairs of speakers into my room and measure them and EQ them with a DSP so things are as technically good as I can make them. I would then want to sit and listen to the speakers. How should one pick the program material to use in this subjective phase? I've seen the recommendations in Toole [sec 3.5.1.7], but it is pretty vague.

Thanks,
Marc

 

[Kal Rubinson]

I don't understand the problem. What do you like?

 

[radix]

I've seen one person (Mitch Barnett) recommend picking something bright, something dull, and something balanced to see if any are exaggerated. Toole recommends complex music, and lists 4 groups from most revealing to least revealing.

So, I am wondering, what is it about certain songs that lets one differentiate between speakers? Or does it not really matter much?

Sure, I could pick a few things I know well, but would those yield the most information in a comparison?

How do others pick tracks and why?

Marc

 

[Laserjock]

I've seen one person (Mitch Barnett) recommend picking something bright, something dull, and something balanced to see if any are exaggerated. Toole recommends complex music, and lists 4 groups from most revealing to least revealing.

So, I am wondering, what is it about certain songs that lets one differentiate between speakers? Or does it not really matter much?

Sure, I could pick a few things I know well, but would those yield the most information in a comparison?

How do others pick tracks and why?

Marc

I have a bunch of playlists on Tidal I’ve acquired from a bunch of different sources.
Do you have Tidal?

 

[radix]

I have a bunch of playlists on Tidal I’ve acquired from a bunch of different sources.
Do you have Tidal?

No, I have spotify and qobuz.

Anyway, I'm more interested in why you would use a particular song. What do you expect to hear from it, what differentiator or trait does it have that makes it worth including in an AB test.

If I am going to be comparing 2 or 3 sets of speakers, I need to get playlists of reasonable size.

Marc

 

[JSmith]

https://open.spotify.com/playlist/5R3Pzqu8SLyH9cw9WSXw91

https://hifigo.com/blogs/guide/18-b...l-performance-of-your-hi-fi-equipment-in-2020

https://www.musicindustryhowto.com/best-songs-to-test-speakers/

https://hifireference.com/tips-and-tricks/the-best-song-to-test-your-system/

https://www.cnet.com/news/the-audiophiliacs-top-music-tracks-for-testing-speakers-and-headphones/

https://www.lifewire.com/stereo-test-tracks-3134905

https://www.whathifi.com/au/features/10-best-songs-to-test-your-speakers



JSmith

 

[Laserjock]

I have Qobuz also. Just looked in playlists and searched “speaker” and a bunch of speaker test playlists came up.

Ultimate Demonstration Disc: Chesky Records' Guide to Critical Listening, available on CD from Amazon for $19.99 if you’re looking for certain things to listen for with explanation.

edit: typos

 

[Sancus]

How should one pick the program material to use in this subjective phase? I've seen the recommendations in Toole [sec 3.5.1.7], but it is pretty vague.

Figure 3.15 is literally a track list(same list in this Olive presentation pg7) which seems pretty specific to me, not vague. That section also explains the why: the best material is "complex and spectrally balanced", which is why pink noise performs very well. So basically you want material that covers as many frequencies as possible at once. That's what the research says, anyway.

Personally I would just use a couple tracks from the top of that list, pink noise, a couple tracks I listen to most frequently for enjoyment, and maybe one track with heavy low bass(if not already covered), and that's that.

 

[radix]

Figure 3.15 is literally a track list(same list in this Olive presentation pg7) which seems pretty specific to me, not vague. That section also explains the why: the best material is "complex and spectrally balanced", which is why pink noise performs very well. So basically you want material that covers as many frequencies as possible at once. That's what the research says, anyway.

Personally I would just use a couple tracks from the top of that list, pink noise, a couple tracks I listen to most frequently for enjoyment, and maybe one track with heavy low bass(if not already covered), and that's that.

Yes, Fig 3.15 is literally a track list, some of which fall into each of the 4 groups. So, some would be revealing and some less so. I am not deeply familiar with most of those tracks. I guess I was looking for more of the why than the what to help me pick tracks I know. Pink noise is pretty specific, and I know that one

So perhaps I simply need to pick some songs I know that fill out the spectragram.

Thank you

 

[JSmith]

So perhaps I simply need to pick some songs I know that fill out the spectragram.

I provided you with some links... have you looked at any of them? They are exactly what you wanted, in particular the HiFiGo one even tells you what tracks are to test what aspect of your system.



JSmith

 

[radix]

I provided you with some links... have you looked at any of them? They are exactly what you wanted, in particular the HiFiGo one even tells you what tracks are to test what aspect of your system.

I started at the bottom and checked out a couple. 90% of those songs I do not know, though there are a couple I could use (like the Saint-Saens). I did not get to the hifigo, I'll take a look at that.

Thank you for your help.

 

[beeface]

Personally I like to listen to what I ordinarily enjoy listening to. It's hard for me to will myself to listen to music I don't love just for the sake of demoing equipment.

A lot of typical audiophile showroom music isn't to my taste, but some of it I don't mind.

Harman have a playlist here that you might find helpful: https://artoflistening.harman.com/professional-reference-songs

 

[aac]

I use Lamb of God for test.

 

[Ultrasonic]

Let's say I get a couple pairs of speakers into my room and measure them and EQ them with a DSP so things are as technically good as I can make them.

This may be what you had in mind but in case not I'd just say that for me a vital stage of optimising EQ in at least the bass region is to listen to judge what I like best. There is no text-book 'perfect' to adjust to here. Only with this having been done would I be trying to make comparisons between different speakers.

As for what to listen to I would personally just pick examples I enjoy that covers my range of musical tastes, rather than worrying about listening with atypical attention to unusual tracks for things that you might never normally notice. The latter runs the risk of a purchasing decision being biased by the wrong priorities IMHO.

 

[JSmith]

The latter runs the risk of a purchasing decision being biased by the wrong priorities IMHO.

OP isn't using this as a purchasing decision from what I read... opening post says OP is buying speakers first, EQ, DSP etc. and then wants to know what is good to evaluate same from there. Certain types of sounds will be needed to achieve this, hence the request for a list of test tracks that test various aspects of the reproduction.



JSmith

 

[Frgirard]

Let's say I get a couple pairs of speakers into my room and measure them and EQ them with a DSP so things are as technically good as I can make them. I would then want to sit and listen to the speakers. How should one pick the program material to use in this subjective phase? I've seen the recommendations in Toole [sec 3.5.1.7], but it is pretty vague.

Thanks,
Marc

The voice of your family.

 

[Ultrasonic]

OP isn't using this as a purchasing decision from what I read... opening post says OP is buying speakers first, EQ, DSP etc. and then wants to know what is good to evaluate same from there. Certain types of sounds will be needed to achieve this, hence the request for a list of test tracks that test various aspects of the reproduction.

The very first sentence refers to 'getting a couple of pairs of speakers' into a room. I'd assumed pre-purchase but whether or not this is the case there is presumably still a plan to compare them to make some sort of choice between them.

 

[tuga]

Let's say I get a couple pairs of speakers into my room and measure them and EQ them with a DSP so things are as technically good as I can make them. I would then want to sit and listen to the speakers. How should one pick the program material to use in this subjective phase? I've seen the recommendations in Toole [sec 3.5.1.7], but it is pretty vague.

Thanks,
Marc

In my view test tracks should be able to expose shortcomings not sound good, and I definitely think that you should select tracks which you are familiar with.

I am not a fan of AB comparisons but if you choose to do it make sure that the output is level matched between different speakers. Also make sure that you position both speakers adequately (it is unlikely that two different speakers will perform best in the exact same spot).

In my experience one should not introduce more than one (unknown) variable so I would avoid comparing three speakers (or 3 of other equipment) in one go. Also avoid introducing changes to the system in the weeks prior to your testing sessions. I would listen to a new pair of speakers for a week or so, then return to the original speakers for a couple of days and only then AB.

It won't hurt if you listen to the selected tracks regularly before you get going with the testing.

Don't know what kinds of music you listen to but I suggest a track with lots of high-frequency content, a track with very complex sound (lots of instruments and things going on a the same time), one with female vocals with a bit of gritty sibilance, something with good separation and clear decay.

 

[radix]

The very first sentence refers to 'getting a couple of pairs of speakers' into a room. I'd assumed pre-purchase but whether or not this is the case there is presumably still a plan to compare them to make some sort of choice between them.

I have a room with a lot of obstacles and angles. It cannot be treated much, if at all (maybe I could get away with something right near the speaker). So I think I will need to break down and buy a couple speakers (say R3 or a jbl or a wharfedale etc) and really try them in my space. I'd leverage the 30-day return policy [1]. I'd likely pick a technically strong speaker (e.g. R3) and a more forgiving speaker (e.g. a jbl) and see how they do in situ after correction. Hopefully, I'd be really happy with one or the other and I can stop.

I previously had large towers in the room (B&O Penta mk 2, an active speaker designed in 1980s that I've had since new in 1994), which are currently out getting refurbished (the amps need re-capping, the mids need to be re-foamed). It is really very nice not having towers in the room, which is what is driving the effort to use bookshelves on two built-in cabinets. It might be this experiment fails and I need to go the F228be route.

Marc

[1] Online retailers make their money by not having B&M stores, so I have no problem buying and trying. That's their explicit return policy. And even some that have B&M stores still have the same no-question return policy. I've tried talking with my local B&M store (still about 30 miles away), but they fail to get back to me to schedule listening. If they actually provided me a service, I'd buy from them -- I'm not into exploit and buy online. Most of the local (and that's within a 100mi radius) "dealers" are actually home automation / high end HT installers and do not actually have showrooms.

 

[Ultrasonic]

I have a room with a lot of obstacles and angles. It cannot be treated much, if at all (maybe I could get away with something right near the speaker).

I was referring to EQ though, which you clearly are going to be using. What am I missing?