Onkyo TX-NR818 "official" owner's thread discussion - Page 187

Found the setting...since all my devices (TV, BluRay etc) are newer...I could use the auto detect setting for HDMI...works like a charm.

The ONLY reference you provided was about parametric filters used for EQ. You have provided no links that confirm problems in the Onkyo Digital Crossover implementation. We are talking about crossovers here. Those are HPF+LPF. I provided graphs that clearly show that both filters in digital crossovers are in-phase in the overlap region or elsewhere they would have impact on the frequency response (as pointed out in your reference article it would create a comb-filtering effect there, of phase-cancellation). This is to response saying that I provided incorrect information about in-phase and gain at the overlap region.
I stating my understanding based off my pretty limited knowledge of university mathematical analysis and physics. I do not need cites from unrelated application of different kind of filters to understand how HPF and LPF work. They can be easily done in a way that their phase characteristics are not affecting sound in a negative way. And for this they don't even need to be phase linear. If they would then any crossover would be absolute evil, but it is still generally considered than three way speakers are much better than two way and bass-managed config is better than full-range.
There are a lot of crossover filters in anyone's typical configuration. Based on the information you have and strongly believe - turn off bass management in your receiver, discard you current speakers and use only wideband speakers with no crossovers in the future. But please don't tell anyone they should follow you also. It is hard to do constructive discussion with a person who states that my high frequencies are harsh after Audyssey calibration because my subwoofer in incapable and I should replace it or add more subwoofers.
I newer told that the video fix fixed my problems with Audyssey. I only told that there was a sonic improvement after the update. I don't know the reasons (I don't have the sources, do you? I don't know if previous update was flashed with no hidden errors). But there was a difference, and not only audible, but also metrical one - the crossover frequencies were detected consistently higher. There is not so much difference in the electrical response, but I had no time to go deeper into measurements.

It is true that you will not get more power than the PSU can supply to the amps, but it is not the only variable in the equation, things are a little bit more complex than that and amps are not ideal too, so, under some circumstances you can get more headroom, or a kind of 'isolate' the headroom of High Pass section from the Low Pass section, and when the Low Pass section is already in the distortion range High Pass one could still be clear from the distorting effects of Low Pass section. As our ears are more picky to the mid-high frequencies produced from the High Pass section this could still be audible improvement. While technically there is no overall headroom increase.
You still have to provide some argumentation on why do you think there are problems with 'overlap' solution while there are benefit with the crossoverless speakers. You have provided no source supporting your idea that 'overlap' crossovers are problematic. None had even touched this topic.

Analog filters aren't "imprecise" either.

There is no such thing as "the" digital crossover. There are digital filters that behave just like their analog counterparts. This is probably what your graph shows. So nothing is gained when replacing a existing passive crossover with a digital one. Nevertheless there are digital filters that can do a lot more. FIR filters can change phase without affecting magnitude and vice versa.
But as I've said before, there are more things to consider when designing a crossover from scratch. It's not just the on-axis response that is important but also the off-axis response. No trivial topic.

You can't correct the on-axis response of a speaker without altering its power response. Besides that, one can't measure the free field response in a living room. There are too many reflection too early in time that contaminate the results.
Edited by markus767 - 3/9/13 at 2:41am

Can someone please confirm the above for me also is this only for 5.1/7.1/9.1 encoded content? I'm a bit more confused now because when I'm listening to a CD or FM radio etc and the listening mode is set to anything that support 9.1/9.2 playback all of my speakers are active. Thanks Guys.

They are. They are made from imprecise components, any analog component is imprecise, the capacitance of every other capacitor is different to the next one in the factory line... This is the main reason that building high-slope analog filters is so complicated - it is too hard to match every detail so the LPF+HPF would work together with no problems. Digital solves this problem as every virtual equivalent of the analog filter is 'ideal' one with precise and 100% repeatable characteristics.
The graph I posted shows that there is nothing lost in the frequency response of the ordinary speaker with analog crossovers when we add digital crossovers with overlap and do not remove analog ones from the speaker. And there is nothing more needed to match the filters except setting proper frequency value and enabling overlap. This is what was objected when I made that claim initially.
Are you expected a change in the frequency response from active bi-amping. Please... Frequency response improvements is not what is gained by active crossovers. It meant to solve different things. The out of band signal was filtered out from the signal before DACs, so it had no chance to affect in-band signal anywhere in the analog chain - this is what gained.
Sure, but we are talking about crossovers and bi-amping. My only word here is that if you do bi-amping then active filters are an improvement, both in case if you completely replace passive filters in speakers by digital ones in receiver (this is not easy, as you have said - you have to do the matching based on measurements or you end up in doing more harm than good), and also in case if you leave passive crossovers in the speakers (then the setup is very easy, but you still have excess component - the passive filter - in the chain of signal). The question what is better to use spare amps for more channels or to bi-amp some existing ones - is another topic and it is more in the personal preference territory.
Most of the talk was about the case when we leave passive crossover in speakers, then you don't design the crossover from scratch, you only help it to better filter out-of-band signal and do it even before the amps, so the out of band signal just newer reaches the corresponding amp and passive crossovers in speakers.
You don't have to You will be using these speakers in the same room in the same place after all, with all those reflections etc.
Why makers of the speakers don't do it and use free-field measurements is because they cannot optimize the speaker for every room on the planet and there are no standard to the rooms and the placement of speakers in them.

Does any of this have to do with the 818????????? Enough already!!!!!!!

Igor,

I won't respond to the many things you got backwards about crossover design, filters, signal handling and room correction.
If you want to get a glimpse of what it takes to develop a speaker crossover then this is a good start: http://www.linkwitzlab.com/crossovers.htm
About room corection: http://www.dirac.se/media/12044/on_room_correction.pdf

Let's concentrate on the 818 which is what this thread is about.

Do we know what filter characteristics the 818 offers?
Do you have a speaker with a crossover that matches those filter characteristics?

 

Yes it does IMO. FWIW, this is my take on it: The 818 has an apparently 'new' way to biamp speakers which is unique AFAIK in the AVR world, or certainly so at the 818's price point. Some people believe that this 'new' method of biamping is flawed because it requires (in one configuration) that the passive crossovers in the speakers are still used, others say that that this is not an issue because of the clever way the 818 works. There is now a highly technical discussion going on between Dstew, Markus and Igor as to the rights and wrongs of this biamping methodology.  In general, such a deep discussion of biamping, filters, crossovers etc would be OT in an AVR thread, but in this case, because of the unique implementation of the 818 biamping method, the topic is, IMO, entirely relevant to the 818.

 

I don't pretend to be able to follow all of the technical discussions in the last posts, but I am trying to because, once a resolution of the different POVs is obtained, or whenever I manage to form a viable opinion of my own (if at all), it will contribute to my understanding that the 818 is a truly remarkable pice of kit. Not only is it capable of being used as a pre-pro, not only does it have Audyssey XT32 (definitely unique at this price point), not only is it a very full-featured unit in its own right, but now it also seems to have a unique capability to 'biamp' conventional speakers (so long as they are equipped with 4 terminals). If the unit was Audyssey-Pro enabled, it would definitely be a candidate for me, if my Onkyo 5509 prepro ever needed replacing.

 

Personally, I am sceptical about the ability of the unit to 'biamp' (at least in the generally understood meaning of the term) while it uses passive crossovers in the speakers, but the discussion is very useful and hopefully some kind of conclusion will eventually emerge, if the guys can stay focused.

 

Just my 2 percent of a dollar.

I beg to differ, this discussion started out about the Bi-amping capabilities of the 818 and has turned into yet another bickering contest of Bi-Amping and the feasibility of such. These long posts are making it difficult for other members to get questions they have regarding the 818 answered because they wind up a few pages back in quick order.

 

It's totally fine for you to disagree with me. I've had no difficulty reading the thread personally. It's easy to skip the posts you're not interested in. Which questions haven't been answered - if you highlight them in a post of your own, quoting them, it will bring them right to the latest post in the thread?  

 

Some of us are finding the biamping posts interesting, as I have said. If you aren’t, just skip them. If you feel really strongly that they are OT, report the posts and if the Mods agree with you they will delete them. I doubt if that will happen in this case though. 

 

EDIT: I just skipped back a few pages and found only one post that hadn't been answered, which was posted only 20 hours ago anyway. I have made an attempt at a reply which will bring it up to date (I didn’t reply before simply because my honest answer to the poster's question is "IDK".  I suspect you are referring to your own unanswered post but that was made originally 6 days ago - and was unanswered long before the biamping discussion started. Maybe nobody knows the answer?  But it doesn’t seem correct for you to assert that "These long posts are making it difficult for other members to get questions they have regarding the 818 answered". 

Edited by kbarnes701 - 3/9/13 at 5:51am

 

I am not sure but I don't think you are missing anything. HDMI is an 'end-to-end' technology - if you break one end of the chain it fails. That's why disconnecting the PJ or switching it manually restores the CD sound - it has redone the handshake which you disrupted by not having the PJ on. Others might have a suggested workaround, or tell me I am wrong, but that's my understanding.

 

TBH I am not sure. I have still not gotten my head around the idea of using active crossovers in the preamp and passive crossovers in the speakers at the same time.  Amplifier headroom would be increased in conventional active biamping using a separate amp for the LF and another separate amp for the HF (as in active speakers) - but when the amps share a power supply I can't quite see how the headroom is increased.

20 years have passed, nothing changed
I have read this doc also. But thanks.
We are
Unfortunately no. OK, not everything. We know that they are very high-slope filters (I haven't reverse engineered exact numbers). You can choose the frequency - with 1/3 octave steps (a weak point IMHO, could be at least twice more precise, another weak point is that 250Hz is the minimal crossover frequency). Then you can apply overlap of 2/3 octave for use with ordinary speakers without removing crossovers. You can choose not to and remove them, or connect to your DIY crossoverless speakers. Then you are on the 'designing your own crossover' advanced territory. And there are more options to configure like trim level, phase (0 vs 180, useless option as it is equivalent to changing phase by reconnecting wires) and delay for every speaker driver (this is really useful one for matching HPF/LPF together with corresponding speaker drivers, there is no analog for passive filters as the delay problematic to achieve in analog, so phase-matching is used there traditionally, they are different tools in some sense, but in other sense phase and delay is the same thing and you can achieve same goals of matching driver's response in the crossover region with both).
Why do you need to? Express your concerns exactly, and what is the supposed usecase - 'overlap', or 'true' bi-amping with no passive crossovers?

It is not purpose of one crossover to be exactly like the other, it's purpose is to match speaker drivers response in the crossover region. In 'traditional' passive crossover speaker design world they also do other things like driver's level matching (dispersing the power to heat instead of sound pressure) and correcting the frequency response of speakers to some extent. For both we have other and better tools in 818.
Edited by IgorZep - 3/9/13 at 6:23am

There was not so much activity here last few weeks as the 24p issue was fixed But sorry to take attention away. Do not hesitate to say loudly HELP and repeat your question

I agree with kbarnes that this seemingly new tech is very interesting, I also agree that I too get lost when markus, igor and dstew explain their positions, but I do read them.

I started a thread about the 'Onkyo Digital crossover network' in hopes that some of the more adventurous folks with the xx10 AVR's had tried it but it quickly fell off the first page with no replies. I also Googled it to see if there was anybody with first hand knowledge (in the whole world) and their impressions--nothing.

As keith says, if you're not interested just skip over it as I do with the posts about hacking into the inner workings of the 818.

Edit: BTW, here are the Onkyo speakers designed specifically for the crossover network;

http://www.intl.onkyo.com/whats_new/2012_0902.html
http://www.intl.onkyo.com/products/speakers/speaker_packages/d-509/index.html
Edited by Patrick Murphy - 3/9/13 at 6:42am

So we have two filters with unknown characteristics that we could a) apply on top of an existing passive crossover or b) instead of an existing passive crossover. Besides the fact that this is only applicable to two-way speakers, phase and frequency response will change in unknown ways for both cases. Doesn't sound too attractive to me.

It's a toy to keep knob twiddlers happy. The lack of documentation is completely unacceptable.

If people want to get into digital crossovers, start here: http://www.minidsp.com

What characteristic exactly will help you?
Yes.
What is a problem with three way speakers? Use it between woofer and mid driver and leave the passive crosses between mid driver and tweeter.
How people can ignore everything and repeat the same thing even after it has been proven and explained wrong. It is not in both cases!!! Everything works 'out of box' for the a) case! Stop complaining.
For the b) case - this is as usual with any active bi-amping. You have the speaker box with drivers and measurement gear, and do the work matching drivers. Do you think there is no enough adjustments in 818 to accomplish this?

Isn't this what most modern AVRs are?
It is what keep companies like Onkyo and Denon in business.
There WILL be a new knob to turn every 6 months.

While I agree there should be at least some documentation.
(maybe this was some sort of prototype feature for a future model that somehow made it through in firmware?)

How true
There is quite enough documentation even in the user manual, at least to those who understands how to match LPF/HPF filters in the crossovers and so is able to do this. Also the link on the Onkyo site I've posted. I agree it could be better through. But for those who are not idle talking but doing something digital crosses are working already without any problems.

Igor
All digital filters have a negative impact to either phase, distortion or both. That fact is not arguable and you doing so shows me your lack of understanding on this topic. Crossovers consist of filters, you attempting to separate them from other filters is not correct. Use of filters in my system improves the overall sound despite their negatives, you claiming I advocate no filters is laughable and rude to be putting words into my mouth. You not caring about the characteristics of the 818's digital crossover shows that you are not taking a good approach to what you are trying to accomplish. Even your comments concerning analog versus digital are a bit alarming to me. Your goal is to reduce IMD, yet "Intermodulation is caused by non-linear behaviour of the signal processing being used." - wikepedia; so to reduce IMD you add more signal processing? My multiple college degrees also consist of a great deal of calculus, differential equations, and physics, but I don't need them to see the flaw in your approach to reaching your goal.

Your word is not going to be good enough to convince me given your lack of understanding on the principles involved, not to mention your rude approach. If you would like to advance the discussion, bring a technical document which supports your position into the discussion. You claim the reason you don't do so is because you are educated in math and physics? Have you ever read a white paper? Again not a good approach. Is the real reason you do not bring a document into the conversation which supports your position because no such document exists? Out of curiosity, do you work in a technical field and if so what do you do? As I've said before we are simply talking in circles as you've done nothing to show my concerns are not applicable in this configuration.

For others, your goal should be to get your system performance as close to your desired frequency response as possible without EQ, then use minimal EQ to further improve the sound if needed (and it is the vast majority of time). That is why acoustical treatments and speaker quality are the two most influential factors to sound quality. My room is not treated yet due to WAF and no dedicated room and I have to make the EQ compromise. I do employ Audyssey XT32 and a minidsp to improve my system despite what Igor claims my position to be.

I do think the conversation is on topic, but since we are talking in circles at this point without Igor bringing in any evidence to support his claims, it's got to be getting old. I know it is for me. I suspect my prayers for his next post will go unanswered.

+1 to each of markus's valid points. His understanding does exceed mine, yet when I do read on the topic it's amazing, all of markus's points are supported by the material. It is the kind of experience which causes me to take note of a forum members contributions.
Edited by dstew100 - 3/9/13 at 10:11am

Then continue here: http://en.wikipedia.org/wiki/Filter_(signal_processing)

Ok guys...I have read through this thread. And a few others.

Was hoping to get a top 5 things I wish I knew when I first bought the 818 list...

I am trying to wade through all this info and find the gems but it is going very slow.

Example: Video conversion..should I set it to 1080p or 1080p / 24. I have a panasonic VT...so it handles 24 very nicely..but I dont really know.

Anyway...if you guys came give me the "Top 5 things I wish I knew how to do when I got my 818" I think that would be extremely helpful to everyone.

Set it to 1080p.

I'm not sure what the 1080p/24 setting actually does (besides the obvious conversion of whatever comes in to 24p). Ideally it would IVTC, but it could just be decimation. Better to let the source device handle it, if it can.

What is negative impact for you? Change in phase on the measured graph? For me it is audible change. While it is proven in general that changing phase can produce audible impact, AFAIK nobody yet proven that realistic phase change done by the real world crossovers can create audible difference in the real room conditions. The think is that any reflection of sound in the room create much more phase distortion than any crossover do. And as pretty much nobody likes listening audio in completely 'dead' room...
Filters are different! They impact phase differently too! Compare the phase response of crossover and of PEQ filter - you immediately will see the difference!
So, finally, those words from you. You use filters in your system and they improve the sound for you! Despite the fact they change phase... So... Digital Crossovers also improve sound for me, despite you are not believing in them and somehow think they are so special and different from other crossovers.
What... I feel crazy... I care about the characteristics that are important! I have all of them! They are all described in the manual and repeated here by me in response to Markus! You care about something imaginary. I asked several times already - WTF characteristic you need to know? For me it is enough to know - high slope, in-phase overlap region, in-phase crossover region at the 0 delay settings, known possible crossover frequencies, separately configurable trim level and delay for LPF and HPF. Those are all tools that are needed to configure and match crossovers for any kind of real world application. If you designed crossovers yourself you would know that. Please read and do not repeat stupid things that are answered already several times.
Ha ha.. Your thinking they are equal and all analog radio components are ideal don't alarm me even a bit already... Have you ever had a capacitor or resistor in your hands? What primary characteristics they have? The precision is one of them Your inability to reason and inability to read what others are telling is already known to me. Don't be surprised I am rude.
Ha ha.. I cannot even know what to tell here... Are you really believing that digital crossover filter can cause IMD? It needs to be seriously broken to do so. No, I don't feel crazy anymore, I feel someone else is crazy. Please read about what is Linear Time-Invariant. Most digital audio filters, esp. crossovers are LTI, and as such they cannot produce any THD or IMD.
Don't be surprised I am rude after you tell idiotic things (see above) that you do not understand yourself.
Programming, if you would work with Java then you would probably hear also about JRebel.
The only your concern was phase, and you had so much concerns about it while using a lot of crossovers in your config already... including active digital and passive ones... As if you had so much problems with them. All of them shift phase in similar way.
Please... not again. I don't need frequency response if the price is the enormous distortion.
You are not alone, I too need to make compromises. And killing the possibility of IMD (one of the worst sounding form of distortion) from the source by using digital crossovers is a lot better than to think about imaginary phase shift that several researches were unable to identify any audible impact in real world conditions.
Works for me. If it doesn't work for you - report here. If you just think it will not work at all - it is up to you to provide evidence, but please, on topic, and topic here is 818, and so filters in 818.
How much to each invalid one?
Sure, Markus is great contributor here, no one objects that.

You want some links... you can read here about digital filters and their linearity, and here about audibility of phase shift done by different crossover filters. You can find more resources if you google it

I bought my 818 about 2 1/2 months ago and initialy it was working great. The day before the superbowl I turned the receiver on to play around with the settings and no sound came out of the speakers other than a clicking noise as i turned the volume up or down (this happened the day before I was to host a superbowl party, thank god for my old 605). I ended up taking it in for service and after 2 weeks I picked it up as they had replaced an amp, hdmi board and some chip ( or some other part, cant recall). Got home, turned it on and everything was sounding and looking great until I hit the NET button and noticed that I was missing a lot of apps (pandora, netflix, rahpsody just to name a few). I called it in to onkyo and spoke with 2 techs and they told me that they had never herd of this problem so they told me that onkyo headquarters would call me soon with a solution and for me not to worry. That was 2 days ago and this is driving me NUTS!! Ive searche on this forums and other forums and cannot find anyone with the same missing apps issue. So, before I call them back and flip my ****, I was wondering if anyone has had this problem or herd of anything like this. Ive tryed everything from resetting the unit to flashing the latest firmware with no luck. Am I missing something like a hidden sub menue or something really stupid that I am not doing? Any help would be greatly appreciated.

Fact: In the digital domain there is no loss due to filtering, mixing, or what ever.
You may not care for digital and that's fine, but your assertions are biased and incorrect, IMO.

What apps are missing, and what are there?

If you see eOnkyo and Radiko, your receiver thinks it's a Japanese model. If you see Simfy, it thinks it's European. The data used to determine this is on the HDMI board in the flash for the ARM processor - at least thats where it reads the data from. I'm quite curious as to what it shows: which apps do you see?

Are heights worth it? 9.2?

I cannot do wides. Does it make a big difference over 7.2?