Does surround make sense for 2 channel material?

An AVR may be a good choice for several reasons, even though you are currently using 2.0 for music. The main reason would be for bass management if you add a subwoofer. I highly suggest adding a subwoofer. It will make your music sound even better. I personally don't like surround sound listening modes for music. I have a 7.1 system but I still listen to music in 2.1. I do like blu ray concerts and I do like to listen to surround sound while viewing these concerts. They are encoded with either a DD5.1 or a 5.1 DTS HD Master Audio codec. Thats different than taking a 2 channel source and using a listening mode in an AVR to simulate surround sound. I don't care for that. It just doesn't sound good to me.

If you want surround, get multichannel recordings. Synthesized surround is, imho, unacceptable. That said, there are lots of great multichannel classical recordings (www.sa-cd.net) and more coming on SACD and on BluRay

So, while all AVRs will do the surround synthesis, some better than others, you should get one and get real surround.

My Denon AVR 391 does a good job doing stereo into 5.1 synthesis using DPL IIx imho.

When I had a Denon receiver I was always a big fan of Matrix sound. Pro Logic II is ok. I really don't just sit down and listen to music much anymore. When I did I mostly did 2 channel. DTS NEO:6 is alot more subtle than Pro Logic II. Pro Logic II seams a little too invasive as it heavily steers the sound into the center and surround channels as where NEO:6 just seams to give the sound more ambiance without messing with the original 2 front channels hardly at all. Pro Logic II is alot more dynamic, but I think in the end it was too costly to overall sound quality. IMO Pro Logic II is best used for material that is specially optimized for it. Video game consoles like the Wii and PS2 really benefit from Pro Logic II.

I like the way my Yamaha receiver puts 2ch into 5.1 with its '7ch stereo' mode.

It gives me a lot more adjustment than what NEO:6 or Pro Logic II has, as you can change how much the center or surround speakers are used from 0 to 100 in 1% increments.

I have just the right amount of ambience and depth that I like, without it sounding obvious that sounds are coming from the surround speakers.

So yes, I enjoy and listen to most of my 2 channel music in 5.1.

Even better is multichannel SACD's or blu-ray concert disks. Most music I buy these days is either one of these.

One thing that might be worth considering is that a center channel can help with that "head in a vice" feeling when you try to stay in the 2 channel imaging sweet spot. Put another way, there is the possibility for more than 1 listening position which is pleasing.

As Kal mentioned above, you need to think about MC music. There is a significant number of records available of mostly classical music.

Synthesized surround (either DPL or NEO) does or does not work for any specific records. Some of them sound very good after processing. For others it either does not make much of a change or make things worse. But it is better to have this options, considering that receiver/processor often have additional nice features, like bass management and room response correction.

If you go to a surround system, are you prepared to use identical speakers? Or at least very well matched speakers? It is crucial for good music performance, whether discrete or surround processed.

And be prepared to spend a lot of time over months and months tweaking and fine-tuning to achieve a seamless surround sound that doesn't even sound like a surround sound... but merely a superbly wide and deep and focused stereo experience.

Don't be one of those people that try a mish mash of speakers and listen to one CD with NEO:6 and decide multichannel is no good.

With some work it can be very very well worth it.

This is an excellent point worth repeating, so I am. Over the years, I've been through it all:

Nice fronts, center, and nice surrounds from different lines--Awk!

Nice fronts, center, and inexpensive surrounds from the same line--hmm, okay.

Nice fronts, center, and inexpensive surrounds and rears from the same line--bigger hmmm, better OK.

Nice fronts, center, and inexpensive surrounds and rears from the same line plus a basic sub crossed over at 100 Hz--getting there but some localization and cohesion problems.

Matching fronts, center, surrounds, and rears with a sub crossed over at 80 Hz--shocking step up in overall cohesiveness, but bass is uneven throughout the room.

Matching fronts, center, surrounds, and rears with two subs crossed over at 80 Hz--even more improvement with some of the bass issues reduced.

Matching fronts, center, surrounds, and rears with four subs crossed over at 40 Hz--ooh-la-la, Monsieur, your beautiful companion has arrived!

All of this is accompanied by a serious effort to position the speakers properly, calibrate the system, and treat the room if only subtly. There's a really big payoff doing this with a multichannel music only system that can be more gratifying than an HT. As the surround matrixing capabilities have improved over the years--and they've improved a lot--people have panned them when in fact their mix of speakers was inferior to the process or their lame disregard of set up and calibration were responsible for poor results.

You never have to buy a multichannel disc to reap the benefits from such a music system. That said, there's some compelling surround material out there that should be heard as the artist and engineer intended.

A few thoughts...

I disagree with Kal's labeling of surround processors as "synthetic." I'll get to that in a bit.

If you think about it, you have four real choices with a spectrum joining all of them. You can use two speakers and allow the room to do all of the surround processing (by bouncing sound here there and everywhere), you can use multiple speakers with room treatment and allow a surround processor to do surround processing, you can use two speakers with room treatments aimed at removing reflections entirely, or use multiple speakers with a surround processor and NO room treatments such that both room and processor are doing processing.

I'll start with the premise that the last two options are no good. If you only have two speakers, you need the room to create a reverberant field or else it sounds like dry crap. If you have a surround processor, it is logical that you believe it does a better job than the room and thus should treat the room accordingly, not to mention that having both going on is certain to yield less than optimal results. Sadly many people choose exactly this option.

That leaves the first two. In reality you don't want an extreme of either so the choice is whether to use multiple speakers and a dedicated processor, and that in turn dictates how much and what type of room treatment is required. I'll say here that both can give good results so you need a short list of pros and cons to help you choose.

Two speakers means you need very well designed room treatments with a near perfect mix of absorption reflection and diffusion. The size of the room is critical and placement of equipment and listeners is critical. You can create a very pleasing and convincing result for a single listener and the room/system will be of little use for movies or other multichannel sources.

In a multichannel system you still need room treatments but the goals are a bit different and easier to achieve in a variety of room sizes. Placement of equipment and listeners is not as critical and you can create a sweet spot large enough for multiple listeners. The room will work equally well for all sources and is ready for easy migration to other other surround techniques, speaker locations, etc as technology progresses.

Which to choose? If you have any desire to use true multichannel sources, your choice is made IMO. If it is only ever ever going to be for two channel then it may come down to cost, ease of implementation, and ultimate achievable sound quality. The cost is difficult. If you happen to have a great room then for the cost of analysis and proper treatments you could have a great two channel system. But you probably won't know until you have invested months to years in trial and error, measurements, various equipment, treatments, and room layouts to find out. With multichannel you have a non trivial upfront cost of multiple matching speakers, room treatments, and a good surround processor. The formula for how to treat the room is well established, and though it is recommended to take measurements along the way it is more tolerant approach to variances in treatments and speaker locations.

So what is the ultimate quality? That brings us back to Kal's comment. Good surround processing does not synthesize anything, it merely extracts ambient sounds in the recording using clues contained in the recording (whether placed there intentionally or not), and directs those sounds to the appropriate speaker. IMHO, this surround processing is done better than any room can achieve, though as said above it requires the investment in good matching speakers and a good processor. The soundfield is stable for multiple listeners, comb filtering is reduced, dynamic headroom is increased, and the ambient field is more realistic. IMO of course.

I knew someone would rise to the bait.


All extreme cases chosen, I assume, for easy illustration.


OK.


Done, thanks.


Not always and that lack of reliability is critical. The huge variability of results obtained from the different extraction/synthesis algorithms indicates that none (or all but one) is really extracting the ambient sounds accurately. In fact, which one I prefer (and I do use them at times) varies from source to source.


Of course. OTOH, IMHO, discrete multichannel simply trumps it, logically and subjectively.

No assumption necessary - as I noted that there was a spectrum joining all of these extremes, and later specifically said "in reality you don't want an extreme of either..."


I agree, absolutely. And once you set up your system to properly reproduce multichannel music, it is no longer optimal (or even "good enough" imho) for stereo reproduction. So the question is what to do with that vast library of two channel source material? If, as we agree, discrete multichannel is a worthy goal and you design your room and system appropriately, then you have already made the choice of how to BEST playback two channel source material. The other option is to have a separate room and system for dedicated two channel listening. Some people have done that.


I don't argue that in theory. But the corollary is true... each change in room size, diffusion/absorption/reflection amounts and location, speaker and listener positioning, changes how the room processes the sound as well. Which of those is correct? And in the end, which is easier to manipulate until you get a satisfactory result? Do you like the same room for jazz that you do for rock? If not, how do you easily change that? Seems that people are willing to swallow the idea a "one best solution for all material" when allowing the room to do the processing, but find this either impossible or as a negative when electronics does the processing. Why? Personally, I find that the ability to easily tweak settings (and a GOOD surround processor does allow a wide latitude in how you want it to process the sound) is a plus, not a negative for going full-time multichannel. In fact, with some processors you can have selectable memory settings for jazz, rock, movies, etc.

Something else in favor of surround processors is that they are always advancing. Once your room is "perfect" it's kinda hard to improve. And again, I think the "perfect" room is usually only a match at best for good surround processors of TODAY, rarely better, and often much much worse. With processing algorithms getting better, your multichannel room can continue to get better and better over the years. At some point, I don't see that this will even be a rational argument any longer (sort of like cable debates of today... there are still debates, but one side is simply irrational).

All discrete MC recording are "synthesized". It is just done at mixing stage. I suspect that often algorithms similar to DPL and NEO are used for that purpose.

Not with the music I listen to. If you listen to acoustically-produced discretely-recorded music, regardless of the number of channels, such is not necessary. If you listen to studio-synthesized music, regardless of the number of channels, you have been listening to such for a long while, so you must like it.

Since I find two-channel source material permanently compromised in requiring the spurious ambiance of the listening room, it is no biggie for me to say that filling in what it takes to make stereo enjoyable is acceptable, whether electronic or physical. It does not necessarily mean creating additional channels.


If the room's acoustics are set to suit good multichannel, it is sufficient for all good multichannel.


Red herring. No one would put a symphony in a small club. No one would put a jazz trio in a stadium (without electronics). But that has NOTHING to do with reproducing all in the same room if the multichannel recording has, as it should, captured the ambiance of the original performance venue.


Toys.


Well, I know which side I stand on.

Well Kal, while I think you and I agree on a great many things in principle, I guess I'm just worried that your advice isn't especially helpful to the OP. He specifically stated that he listens to two channel CD and FM source material. Your advice... throw all that away, replace your library with multichannel discrete. Aside from the problem that much of his library (and all of the broadcast material) may not be available in discrete multichannel, your advice still leaves his question in the general sense unanswered. If most discrete material is in 5 or 5.1 format, is it heresy to have a 7.1 channel system? Are wides or heights forbidden until they are explicitly encoded on a disc? If the disc has wides and you have highs, what to do then?

He's asking if playback channels need to match source channels in the general sense, and the answer is a resounding no. It hasn't been the case for a couple of decades, and is less and less the case as technology gets better and better. I do respect your opinion here... discrete is the reference; it's just that it isn't possible for much material, isn't practical to try to match your system configuration to the format for each recording and as recording standards evolve, and doesn't address the real problem of what to do when you already have a big library.


No, you misunderstand (and I was not at all clear). Do you prefer the same playback room for both jazz and rock. You answered that elsewhere in your post. We agree on the room side here - a room suitable for multichannel is good for all multichannel, whether extracted or discrete. My question was directed I suppose into the great void... towards those who think a lively room is a good surround processor for two channel playback. I've never understood why "they" think a fixed room is perfect for all recordings, yet a "fixed" algorithm can't do an equally good job on all recordings. The problem is in the recording - not the algorithm. In your case you dismiss both ideas... if it isn't discrete, you won't have it. Not everyone is as limited in their tastes, so the question remains important to a great number of people.


Hardly. I'm not talking about the old synthesized reverb "jazz" and "stadium" modes prepros and avr's of decades ago used (and are still in many today). I'm talking about the flexibility modern processors give in tweaking the parameters of DPL, or especially something like Trifield or Logic7. How aggressive the steering is, levels of steered content, etc. can all be user tweaked. If you believe these extraction algorithms have different results on different recordings, you have have a couple of easily selectable presets which may give satisfactory performance on a wide variety of material.

Accurate to what? Before discrete the multi-channel era, movie mixers used to monitor/check their matrix encoded 2-channel mixes in surround to insure proper decoding. If you could mimic those results, as the THX system attempted to do, then you were getting some semblence of accuracy to the original intent.

There is no similar convention with music recordings. I doubt most music mixers have ever heard their stereo recordings in surround, let alone mix them with surround compatibility in mind. So there is no reference to be accurate to.

Without a reference, I'm left with preference. As you said, no two surround processes yield the same results, any more than two room correction systems yield the same results. Recall that Sean Olive's AVS thread about his room correction comparo was all about preference; no claim that Harman's room correction was reproducing anything "accurately".

The same applies for me and surround processing. Variable results mean that I use more than one (depending on material). Fortunately, all it takes is pushing one button on the remote. Adjustablity allows me to customize the processing for my particular tastes, which lean towards the subtle (no one realizes I'm using surrounds until I turn them off).

In addition to variable results, surround processing isn't perfect. There is the occasional hiccup. But, for me, it's better than the alternative: listening to 2-channel music using only 2 speakers. How do I place 2 speakers so that I get a wide stereo soundstage AND a stable centre image AND wrap-around envelopment AND... well, you get the idea.

During the last 20 years that I've been listening to stereo LPs and CDs in surround, I've discovered that 2-channel music recordings have more than enough information to feed all 7 of my speakers. It may not be accurate (whatever that means), but it sure sounds pleasing.

I couldn't agree more!

This would have been helpful in your first post. We all could have written "Stick to 2 channels," and been done with it.

OK. I think my original reply revealed both that such reprocessing is generally easy and common and that my bias was against it.


Less so than going from 2.0 to 5.1.


I tend to ignore both but, since there are so very few of either, it is not an issue with me.


All things are possible but, I believe, it is quite a matter of preference but I have yet to be convinced by these ever-improving processes.


Sure.


Ah! Then I am on your side.


I do use Trifield from time to time and, sometimes, tweak it a bit but, in the long run, I always go back to stereo for stereo sources. Again, a matter of preference, as is the alternative.

As usual, Sanjay, your arguments are clear and reasonable, so I will not (cannot) counter them, except for the last line. I have yet to be pleased with such processing. Perhaps I would if I invested the setup effort and adjusted the parameters to suit the particular recording but, frankly, I cannot find the motivation to do that when I have, already, thousands of discrete, multichannel music recordings and more coming in every week.

Every time one of these threads pops up, I try out surround modes for 2 channel stereo, and every time I find it somewhat interesting for a bit. Then invariably I hear an a recording that sounds really bad and switch back to stereo. The switch back is so clear and right sounding that I find no use for the surround modes for 2 channel music.

For me, the positioning of the vocals in the mix just doesn't work with the surround modes engaged. I also do not like instruments bleeding into the surround speakers. Maybe I could fiddle with the settings for some of the modes, but stereo sounds so true to me that I don't find it worth the effort.

Thanks for starting the thread. It's an interesting topic.

I think this is key. If you have thousands of multichannel recordings, and more coming in each week, you might have a different outlook than some of us. I'll admit, I still listen mostly to music that was recorded for two channel. Different strokes for different folks.

Some receivers offer more tunability than others. Sounds like you need to increase the delay going to the surrounds to create more focused vocals and instruments.

Try setting the distance to your front L/R in the receiver slightly further than what they actually are, so it sends the signal to the front speakers slightly earlier to compensate. This is the same as increasing the delay to the surrounds.

Doing so will focus individual vocals and instruments down into a tighter spot... instead of being spread out over a larger area of space. (like if a singer's head sounds like it's 5 feet wide)

Pick a track with a strong center vocal and switch between straight 2ch and multichannel and see what happens to the voice. Multichannel should be capable of keeping it centered and focused... and perhaps only come forward in the soundstage. Individual instruments that were near center in 2ch should sound to come forward and slightly wider. Sounds that were near the extreme left and right should come further around to the sides.

Adjustments in the time domain are just as important as adjustments to balance etc.

Agreed. I do still have more stereo recordings than multichannel recordings.