I am not experienced enough to argue specifics of this part of the debate, but I do see a connection to other arguments of sound quality...
Sealed vs Vented
In the box vs Outboard gear for the studio guys
How about some blind test results are has this been already done, to which the results would be, in my opinion...predictable
Sealed vs Vented
In the box vs Outboard gear for the studio guys
How about some blind test results are has this been already done, to which the results would be, in my opinion...predictable
I think we are starting to talk past one another, rather than seeking to understand each others point of view. Sasha has asked (to paraphrase) how can we be sure that the extra A/D and D/A steps in DSP filtering do not have a sonic signature. Scott L has expressed an opinion that (to paraphrase) those extra A/D D/A steps probably do have a sonic signature, and DSP is not required anyway, so why do it... Let's stick to answering the questions and making cohesive arguments. No need to attack them.
Scott L - With the state of your system as described, I would not change a thing. A highly optimized active system with analog filtering and discrete solid state amplification (which seems like what you have) is hard to put together, but having gone through the time, effort, and cost, why would you change it?
So for you, changing to DSP filtering would offer little benefit, but there would be risk... there is always risk when one makes changes to a highly optimized system...
I envy your source material. High quality 2 track and 4 track (!) tape is not the ordinary source material for most people. Congratulations.
For someone trying to build a new system today, the advantages of DSP are compelling. In my case, I wanted an active speaker system. 90% of my source material is CDs, and I am starting to collect high resolution digital recordings (24 bit 96kHz / 192 kHz). Although I continue to enjoy my vinyl, I am not buying any more of it. So for me and others in a similar situation, DSP is a great solution.
In my opinion, the hardest thing for a DAC to do is to take the stream of bits from a CD and make good music from it. The CD standard of 16 bit 44.1k is theoretically good enough, but just barely. It is a complicated challenge to devise a DAC which works well with 16 bit 44.1... even a switch to 48k is quite a bit less challenging.
The DSP filter I use runs at 24 bit 196 kHz for both the A/D and D/A conversion. It is a straightforward textbook exercise to get that to sound good, there is so much extra bandwidth and capability there. Since my source is 16 bit 44.1k which is converted to analog and sent off to my active speakers, I am unconcerned about the comparatively benign A/D and D/A process in the speakers.
Scott L - With the state of your system as described, I would not change a thing. A highly optimized active system with analog filtering and discrete solid state amplification (which seems like what you have) is hard to put together, but having gone through the time, effort, and cost, why would you change it?
So for you, changing to DSP filtering would offer little benefit, but there would be risk... there is always risk when one makes changes to a highly optimized system...
I envy your source material. High quality 2 track and 4 track (!) tape is not the ordinary source material for most people. Congratulations.
For someone trying to build a new system today, the advantages of DSP are compelling. In my case, I wanted an active speaker system. 90% of my source material is CDs, and I am starting to collect high resolution digital recordings (24 bit 96kHz / 192 kHz). Although I continue to enjoy my vinyl, I am not buying any more of it. So for me and others in a similar situation, DSP is a great solution.
In my opinion, the hardest thing for a DAC to do is to take the stream of bits from a CD and make good music from it. The CD standard of 16 bit 44.1k is theoretically good enough, but just barely. It is a complicated challenge to devise a DAC which works well with 16 bit 44.1... even a switch to 48k is quite a bit less challenging.
The DSP filter I use runs at 24 bit 196 kHz for both the A/D and D/A conversion. It is a straightforward textbook exercise to get that to sound good, there is so much extra bandwidth and capability there. Since my source is 16 bit 44.1k which is converted to analog and sent off to my active speakers, I am unconcerned about the comparatively benign A/D and D/A process in the speakers.
Good point, I think it is safe to say that sealed is better yet there are very few such commercial speakers, and it could be again the question of cost.
What if most of your files are 24 bit, either 96 or 192KHz, would not in such case your DSP filter be a bottleneck? Or is there something I do not understand? Would not in such case a pair of analog active speakers do better justice than DSP active speaker, even if you feed DSP speakers with digital signal to avoid AD?
I do not know what your thoughts are about 16/44.1 vs 24/96/192 but I find higher bit and sampling rate sounding better, given everything else the same. So even if we assume that DSP speakers employ very good DA and amplification DSP would not be good in this case?
But then the question is, what is considered a good analog active cross-over? Does anyone considers for example Bryston active cross-over used in large active PMCs for example a good one?
Certainly I hear a difference between a CD at 16 bit 44.1k and a high resolution file in 24 bit 96k. I am working to expand my music collection of high res files. But many of the worlds best performances were recorded many years ago: Miles Davis "Birth of the Cool" is what it is. I can get it on old original vinyl which has been played a thousand times, a lessor quality vinyl pressing from the 1970s, or the remastered CD. that is the choice... same for great performances by Kiri Te Kanawa, Pavaroti, George Szell conducting the cleveland symphony, Oscar Peterson, Art Tatum... its a long list.
If most of my files were 24 bit 96/192 k (which they are not), I would run the digital signal directly from my music server (which is currently a Windows 10 laptop) to the Hypex DSP filter / amp, and I would see how it sounds. If it sounded better than letting the stand-alone DAC convert to analog, then going through the pre-amp, then going to the Hypex for A/D - processing - D/A, well, i would then make a change. I don't want to guess which would sound better. I have been surprised in the past.
Some audiophile reviewers have found that a CD can sound better if it is converted to analog and then put through a high resolution A/D process and stored as a music file. not all CDs but some. Who knows, maybe it is true.
If most of my files were 24 bit 96/192 k (which they are not), I would run the digital signal directly from my music server (which is currently a Windows 10 laptop) to the Hypex DSP filter / amp, and I would see how it sounds. If it sounded better than letting the stand-alone DAC convert to analog, then going through the pre-amp, then going to the Hypex for A/D - processing - D/A, well, i would then make a change. I don't want to guess which would sound better. I have been surprised in the past.
Some audiophile reviewers have found that a CD can sound better if it is converted to analog and then put through a high resolution A/D process and stored as a music file. not all CDs but some. Who knows, maybe it is true.
no matter how much better the cardioid response get, it doesnt compare to what a perfectly treated room can acheive.
analog always win, passive xo and room treatment over DSP for me.
and its not like kii or D&D magically solve early reflections or even come remotely close to "control" the LF resonance.
its all a big marketing idea, while cardioid do help, it is NOT comparable to what a great treated room can achieve.
This should be an interesting read: http://drewdaniels.com/audible.pdf
and this: Hi-Rez Audio Distinguished in Blind Testing | Stereophile.com
or this: AES E-Library >> Sampling Rate Discrimination: 44.1 kHz vs. 88.2 kHz
and this: https://www.willhowie.com/wp-content/uploads/2018/09/wh.3Daudio.sample_rate.discrimination.v1.7.pdf
and this: Hi-Rez Audio Distinguished in Blind Testing | Stereophile.com
or this: AES E-Library >> Sampling Rate Discrimination: 44.1 kHz vs. 88.2 kHz
and this: https://www.willhowie.com/wp-content/uploads/2018/09/wh.3Daudio.sample_rate.discrimination.v1.7.pdf
Last edited:
Maybe a fully treated room wins, but not everyone wants a room with full of treatment panels, resonators, diffusers everywhere. I point mainly to the low frequency response, where proper treatment must be physically huge and sometimes that's very impractical.
I like the DSP sulution because of flexibility, if I want to try a different crossover just a few clicks and it's done, because every crossover slope/topologie is a compromise in some ways. If someone happy with his active analogue or passive filters that's good for him.
@hifijim
The Hypex Fusionamps runs at a very specific 24 bit 93.75kHz sampling rate AD-DSP-DA (and certainly for a good reason) and the processing works at 32-bit if I remember correctly. But how this all reacts to other sampling rates in the signal chain I don't know and it's probably a complex thing.
Maybe my ears, my room, my speakers or even my own mindset are the limiting factor but I don't really hear a difference what the higher sample/bit rate soundtracks are offer. I am happy with the good old 16/44.1, and maybe that's my problem, but I don't see it as a problem.
People often don’t understand why others like different things and not what they like.
Someone likes beer, someone else wine, and someone else likes a whole different kind of beer. Someone else doesn’t like any kind of alcohol at all.
Someone hear (or see) different things which others don't.
There is no ultimate solution which works for everyone.
I like the DSP sulution because of flexibility, if I want to try a different crossover just a few clicks and it's done, because every crossover slope/topologie is a compromise in some ways. If someone happy with his active analogue or passive filters that's good for him.
@hifijim
The Hypex Fusionamps runs at a very specific 24 bit 93.75kHz sampling rate AD-DSP-DA (and certainly for a good reason) and the processing works at 32-bit if I remember correctly. But how this all reacts to other sampling rates in the signal chain I don't know and it's probably a complex thing.
Maybe my ears, my room, my speakers or even my own mindset are the limiting factor but I don't really hear a difference what the higher sample/bit rate soundtracks are offer. I am happy with the good old 16/44.1, and maybe that's my problem, but I don't see it as a problem.
People often don’t understand why others like different things and not what they like.
Someone likes beer, someone else wine, and someone else likes a whole different kind of beer. Someone else doesn’t like any kind of alcohol at all.
Someone hear (or see) different things which others don't.
There is no ultimate solution which works for everyone.
Last edited:
Scott, with all due respect, the fact that you can't imagine it doesn't mean it isn't possible. There are many, many factors involved in getting a good audio reproduction, and analog or DSP xover is only one of them. Latching onto that one aspect makes you fall into the trap Bruno has warned for repeatedly:
"The road to hell: specify the design, accept the performance;
the road to heaven: specify the performance, accept the design"
You are on # 1 ;-)
Jan
Last edited:
For him it is. One can live a very happy life believing the earth is flat. As long as you don't set sail and discover for yourself. And that is perfectly OK.
Some people prefer mono over stereo.
Some prefer tube amplification.
Some prefer high efficiency loudspeakers.
Even Bruno's quote sounds like a preference to me. Because the assumption that everything is measurable hasn't been proven yet. However likely it is though, it's still a belief.
Not really. It is not specific to audio at all. It's a basic design philosophy.
Either you specify that the thing you are designing has to do, and you accept whatever the design turns out to be. That will give you the best chance that indeed it does what you want it to do.
On the other hand, you can specify the design ('it must use a 300B tube', 'it must not use DSP') and then you accept whatever performance comes out of it. It will guarantee that there is a 300B in it, but whether it sounds good or bad in your view is up to chance.
Part of the design brief for the Kii was something like: 'when placed closed to a back wall, the acoustics must not suffer compared to placement 4 ft from the back wall'. That leads to an option to modify the filtering processing depending on the position of the speaker. That in turn leads to the use of a DSP option as the required process options are very hard and expensive to realize in analog, if at all possible.
Another part of the design brief was 'it should be able to play digital music streams as well as analog sources'. There's your ADC an DAC coming in.
The point I am trying to make is that to design a successful product is careful design, not mentally locked in sub-optimizations. Statements like 'if it only had an analog xover it would sound sooo much better' is BS. The whole design would collapse.
I am aware that sometimes people specify the design ('it must have a 300B in it', 'it must not have DSP') and then are convinced it has the best sound possible, but that's a whole other universe of course.
Jan
Last edited:
I understand your point, but Scott is convinced that analog is better. He may even have worse measurement results, but he still likes it better, we don’t know what's going on because we haven’t seen any measurements from him yet.
He accepted the tool, which is the analog crossover for him.
Last edited:
Jan,
I don't see the option of "modifying the filtering processing depending on the position of the speaker" mentioned on the Kii website; it suggests that something could be done in situ.
Can you point to the place where we find that option?
All I see is that the loudspeaker features a cardioid polar pattern for mid bass/ bass to allow for close back wall positioning.
When so, this could perfectly be done without DSP as well.
C'mon - the Kii has drivers on the side of the box also - by these it is possible to shape the directivity. By shaping the FR (filters!!) on these side drivers, the dispersion can be controlled in different frequency ranges. This can never be done with a 3-way with 3 drivers on the front using analog filters. Or at least it would be very hard.
No wait - it's impossible..
//
No wait - it's impossible..
//
Last edited:
Right TNT,
but with drivers on the side of the box directivity shaping can be done with analog filtering as good as with digital filtering (DSP), or what am I missing??
Please understand me well: the choice to use DSP on modern loudspeakers is a logical one.
When someone however claims that something can only be done with DSP, and I don't see any proof of that, I am just questioning...
but with drivers on the side of the box directivity shaping can be done with analog filtering as good as with digital filtering (DSP), or what am I missing??
Please understand me well: the choice to use DSP on modern loudspeakers is a logical one.
When someone however claims that something can only be done with DSP, and I don't see any proof of that, I am just questioning...
IIRC there is a small switch at the back for that, they demonstrated it at the intro I attended, years ago already.
Edit: Boundary Switch, https://www.cm-audio.net/images/kii/three/kii-three-manual-en.pdf , pg 10.
Jan
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.