@kgrlee I have a whole playlist that I've made from music that tests the audio one way or another. It currently has 742 tracks and 65 hours of music. You're welcome 🙂
But seriously, just to pick few examples. First disks that don't have issues. Any cd by Dave Brubeck, like Ultimate Star Collection should sound flawlessly. They are always well recorded, have few instruments and one solo saxophone is not obstructed by other instruments. Occasionally I hear that the piano there is just a bit harsh, sounds like distortion.
Or any disk by Ibrahim Maalouf like Diagnostic is always flawless, the solo trumpet plays all registers but is not obstructed. Any good entry level HiFi system should play this flawlessly, you don't need high end.
But now take a disk with many instruments playing, like Charles Mingus from Atlantic Jazz Legends , when they play multiple saxophones and trumpets at once, suddenly creating a very challenging environment. If you system has distortions there that's it, it's not an "ultimate desktop".
Now Monteverdi - A Trace of Grace. It is recorded with microphones at a distance, there is air and many harmonics and reflections. Sometimes the trumpet sounds clearly distorted. There are also voices, at a good system they should sound present.
But usually when I want to test a new version of a crossover I just spin a random 50 tracks from those 742, that is usually enough. Don't want to hear the same music in the loop.
Thanks for this Andrew. I shall certainly check out the Montiverdi and Charles Mingus.
May I ask what amplifier and speakers you are hearing the trumpets distort?
May I ask what amplifier and speakers you are hearing the trumpets distort?
Shiit Bifrost 2/64 -> Aegir -> Dynaudio Emit. Also smaller distortion in my desktop prototype, but there I think I know what the problem is, woofer breakup - so it's fixable.
@kgrlee Dynaudio Emit 10 that I endure for the past 5 years is an entry level $800 speaker and it is an absolute piece of s**t like anything under $1k. For the best possible desktop we should try to reach as close as possible to $10k Genelecs but target desktop specifics and rather music oriented speakers than professional monitors like Genelecs are, as they are not for desktop music. I am not aware of any commercial speaker specifically designed for computer desktop in the $5k+ range that can challenge Genelecs but in different environment. I started this thread with Minimon Mini but was explained that it's not so good as I thought.
Dyns are definitely not overloading because my prototype DIY box I am tinkering right now without any prior DIY experience is already destroying Dyns by quality, details, range and distortions. I found that in my DIY distortions were simply woofer breakup and I am currently trying to design a 4 th order crossover to fix it. I already have preliminary results that distortions are barely audible when I push them to -25 Db. But I still clearly hear them and other people reported before that for distortions -40 Db is better, I need to cross 2 octaves before the cone breakup distortion and have -48 Db with 4 order at the critical point. When the main sound is 80 Db distortions must be below 50 Db, better 40 Db or it's audible.
A 4 rth order crossover is quite challenging for me so I'm not there yet.
Dyns are definitely not overloading because my prototype DIY box I am tinkering right now without any prior DIY experience is already destroying Dyns by quality, details, range and distortions. I found that in my DIY distortions were simply woofer breakup and I am currently trying to design a 4 th order crossover to fix it. I already have preliminary results that distortions are barely audible when I push them to -25 Db. But I still clearly hear them and other people reported before that for distortions -40 Db is better, I need to cross 2 octaves before the cone breakup distortion and have -48 Db with 4 order at the critical point. When the main sound is 80 Db distortions must be below 50 Db, better 40 Db or it's audible.
A 4 rth order crossover is quite challenging for me so I'm not there yet.
Last edited:
You don’t need $10k Genelecs for great sound. There are a lot of superb DIY designs on this forum under $1000 that I think can sound as good if not better. Genelecs depend on deep and massive DSP to get flat response. But the sound is not natural sounding, IMO.
In some 2 decades of DBLTs, the speaker with the best performance by far is a little 5 ltr two way. It came up top in some dozen DBLTs usually against MUCH larger & $$$ speakers. It has a lot of our ju ju but we had other speakers with the same ju ju which, though good, didn't beat this little one. As it is small, it has very limited LF and spl capability. Only a couple of recording engineers realised it had limited LF though they still liked the bass.
I wish I knew for sure why it was so good. About the only thing I think I can replicate is the LF tuning.
The reason I asked you to try your speakers with more power is that "distortion on trumpets" (especially lots of trumpets) is much more likely to be amplifier clipping than speaker distortion.
https://aes2.org/publications/elibrary-page/?id=3793
https://aes2.org/publications/elibrary-page/?id=2476
I wish I knew for sure why it was so good. About the only thing I think I can replicate is the LF tuning.
The reason I asked you to try your speakers with more power is that "distortion on trumpets" (especially lots of trumpets) is much more likely to be amplifier clipping than speaker distortion.
https://aes2.org/publications/elibrary-page/?id=3793
https://aes2.org/publications/elibrary-page/?id=2476
Alas, these days I'm a REAL beach bum and no longer an AES member. I lost all copies of my own papers in a couple of HD crashes this Millenium.
If any AES members have copies of these, I would appreciate backup copies too ... for fair use by original authors and those involved in the original work of course 😊
If any AES members have copies of these, I would appreciate backup copies too ... for fair use by original authors and those involved in the original work of course 😊
At least you can recover some of your work. Most of mine became property of the government, a good bit is classified.
BTW, have you heard Monitor Audio older gold-anodized AlMg dome tweeters? Kind of a well-burnished brassy texture on sax etc. that other speakers IME just don't get right. Mo Studio series has been my "reference" speakers since mid-90s when they were SOTA, ceramic-coated AlMg cones all 1st-order.
Hi @xrk971
Based on this application of a small full range, is the PC83-8 an improvement over the TC9FD1808 you know so well?
Thanks
Eric
It has higher reach in the top end - like a real tweeter. But… at the expense of a ringing at 14.5kHz (can be seen as peak in frequency response) that will have some sibilance. This can be seen in the step function:
However, it is not that noticeable as 14kHz is quite high and almost at the limit of my old ears (16kHz). Also, the overall response is not as smooth but some may like how “lively” this sounds. It’s really up to personal taste.
However, it is not that noticeable as 14kHz is quite high and almost at the limit of my old ears (16kHz). Also, the overall response is not as smooth but some may like how “lively” this sounds. It’s really up to personal taste.
Yes, one can’t really complain for a $13 fiberglass cone fullrange with a copper cap and decent size magnet. They sound sound quite nice. There is a 4in version that some people might like better.
I am going to solder up the filter as P2P and stick them inside the speaker boxes for a more portable setup upstairs.
I am going to solder up the filter as P2P and stick them inside the speaker boxes for a more portable setup upstairs.
@xrk971 no, this is crosstalk between inductors. You may install the XO outside.
Also cannot have them close to the woofer magnet. http://www.troelsgravesen.dk/coils.htm
Also cannot have them close to the woofer magnet. http://www.troelsgravesen.dk/coils.htm
Orthogonal axis - should be minimal. If the axis were parallel it’s a different matter. I see it is one of the configs where TG says there is crosstalk. It’s the same channel - I’ll measure and see if there is a big effect.