Onkyo TX-NR818 "official" owner's thread discussion - Page 135

how do u know that it doesnt transfer the signal untouched? do u have any proof that u say it simply doesnt?

I have a question - my hdmi out "main" is showing that I have it set to 1080p and allows me to change settings if I want. In the OSD, when I go into setup to hdmi out, when I change it to hdmi out "sub" - the resolution is greyed out and says through. I have an Epson 5010 and just noticed this so does this mean that I have not been getting 1080p upscaled to the projector? How can I change the hdmi out "sub" to be set to 1080p? Thanks.

I ran tests with a Dvdo Edge and had it set to direct, which is supposed to not alter the signal and it altered not only 23.976 material but also 59.97 as well. The only work around is to use the sub out. I believe EVERY Onkyo receiver will do this and reviewers need to test for this from now on!!!!!!

This is very sad. As I mentioned in a previous post, the 818 was looking like a perfect AVR for me... but this significant video processing issue pretty much kills it (speaking as an AV enthusiast that uses 1080p/24 bluray content and requires anything between the source and the display to not screw things up). At best I need to consider the 818 as an audio receiver with a basic HDMI switch included (specifically no video processing or displaying of volume or anything like that). The 818 loses some glamor when viewed in this light.

Here is to hoping that Onkyo can fix it (and will fix it).

I agree about the reviews --- any professional reviewer should start thinking about how this will be checked in the next generation. I'll raise an eyebrow, and wonder if Onkyo has more control over reviewers than appropriate, if the new models have this problem and Onkyo isn't called out for it.

FWIW, I think it is kind of a big deal. I'm a bit surprised at the, more or less, lack of visibility this problem has outside this forum. If any bluray player outputted 24fps (not 23.9x), it would be considered DOA.

Thanks to all those that have helped to identify this problem.

should i just sell my 818?
hows the marantz 7005 or the denon 3313ci compared?

I do straight decode. I see DTSHD-MA and TrueHD on the panel.

Just got done doing more testing, With the settings of Direct and source it WILL pass a 59.94 frame rate unaltered though by choosing direct you lose all processing 1080i in 1080i out. Blurays are STILL being altered to 24fps no matter what settings I tried. If you own a Dvdo Edge and want to use the Main hdmi out of your receiver for OSD you can and the edge will fix the 24fps error, all you have to do is in the edge set the frame rate in the LOCKED position and it will output 23.97fps to your display. I'm thinking about jumping up to the 1010 model but I believe they all have this issue. we NEED models 1010, 3010, 5010 tested too.

What do you mean "I do straight decode" ?

How do I know? Simple. I test it. You should too, rather than rely on hearsay and theory.

HDMI sub = no VP, what goes in comes out. If the source is 1080i then 1080i will come out.

Well you were right in one thing the 24p was set to off , so i put it to on and watched a bit of "transformers-Revenge of the fallen" to my Sharp and it flickered when the movie started ( black screen) and nothing else but i will continue to do more testing during the next couple of days to switch between the tv and the PJ .

Menu pg 72. Under Listening Mode Preset, you can choose "straight decode" for trueHD and DTS-HDMA.
Or, you should just be able to cycle through the different Listening mode to get to it.

Amazing isn't it, months after we and Onkyo know the issue and we're still on the repeated user testing stage when Onkyo should have fixed it long ago.

It makes much more sense to bypass the Onkyo for video for anyone with a VP. I wouldn't want video to go through 23.98-24.00-23.98 changes: the numbers may be the same at the end but I can't just be sure if the extra frame is discarded correctly. I use the Edge for all video sources to TV and another HDMI from Onkyo to TV when I want to use the OSD. Since the Edge is doing the switching there should be spare HDMI ports on the TV. There's nothing in the video options that the Edge or the display (Kuro) can't do so I don't miss the HQV at all.

For what it's worth, here's data of a Behringer B2031P with XT32 on/off. Crossover frequency found by Audyssey: 40Hz.

Electrical:



Acoustical:

None. I've reported that to Onkyo, they promised to pass it to tech department but said it is hardly will be a fix for current models...
You can check power management of your TV if it can be disabled, but disabling it on mine doesn't change anything

So, strictly speaking, you also have this problem (the audyssey curve is all-above no-audyssey one below the found frequency) but it is not as extremely exaggregated as in my case (3 dB mostly with greatest 6 dB peak at ~34 Hz) and so, not as audibly noticeable and easily "correctable" with rising the crossover frequency.

Added:

Can you smooth the acoustical graphs, as I can see you have slow increase of target curve from LF to HF, at around 100 Hz it is -13dB and at 10 kHz it feels closer to -9dB. It is 4 dB difference, something very similar to what I have... What mic have you used for measuring, do you have individual calibration profile for it?
Edited by IgorZep - 1/6/13 at 3:55am

You don't seem to understand what Audyssey is doing. It's a room correction system. It tries to equalize the acoustical response so it matches a predefined target.
Differences in the curves with Audyssey on and off are caused by Audyssey's normalization process. This process tries to match the loudness of the equalized speaker to the un-equalized speaker. This is all perfectly normal and just how Audyssey works.

Do you expect it to work any different? How?

You don't seem to understand that equalizing acoustical response below the fall-off is nonsense! Every acoustics / speaker design engineer knows this - you can only attenuate in this range, not boost, or you get into trouble - the distortion and clipping, both electrical and mechanical (with the speaker cone). It is even mentioned in Audyssey FAQ...

There is no boost. Audyssey is simply adjusting gain. Of course if you look at this:



compared to that:



...it might look like a boost to you. In fact it's not. I've simply shifted one of the curves by 5dB.

What can be observed in my measurements is some ringing of the MultEQ filter which I'll have to discuss with Chris.

You are confusing two different topics. Boosting below the speakers resonance frequency is nonsense. But, the in-room -3dB point has nothing to do with the speaker's electrical and/or mechanical properties. It's virtually impossible to determine the resonance frequency of a speaker by measuring it's in-room response at the listening position.

The shifted one obviously will have different perceived loudness (it is the weighted average mathematically). Ok, in your case you can claim it is normalization (but really it isn't - both curves as is are weighted and normalized to the same base perceived loudness already, the level sets the gain of the uncorrected one to the reference perceived loudness, audyssey corrects everything to the same loudness), as it is within 3dB in your case, some precision tolerance can be claimed also. In my case - you can't even claim any of that, as if you move the basic curve to the highest point of audyssey graph above roll-off frequency - it will obviously already off in level (audyssey will be quieter as it is quieter in every freq. point that produces reasonable sound pressure), but still the boost will be higher than that... Isn't it normalization bug? And if you start to talk about perceived loudness of frequencies bellow the lowest frequency - then it would be total nonsense as there is nothing to perceive there except the distortion at any reasonable level.

Some graphs to illustrate the loudness normalization process.

1) The blue and purple curve is the left preamp output with (blue) and without (purple) Audyssey after calibration.
2) Then I introduced a shelving filter that boosts everything above 400Hz by 10dB, recalibrated and measured again.



Audyssey set the crossover for case 1) to 40Hz and for case 2) to 60Hz. Gain for 1) is +6.5dB and for 2) -3dB. The difference is 9.5dB. This is the curves for case 2) shifted upwards by 9.5dB:



Obviously Audyssey uses the standard method of reference level calibration and adjusts gain so the steady-state response in the 500-2000Hz band matches. It's debatable if this is the correct method of matching the loudness of speakers placed in an acoustically small rooms but it's undeniable how level calibration is described by SMPTE.
Edited by markus767 - 1/6/13 at 12:57pm

This is not a hard and fast rule, and there are a fair number of highly-regarded subwoofers that break it. The resonance of a subwoofer is based on how cone area, and suspension compliance interact with box volume. Distortion rising at low frequencies is almost entirely based on Xmax or linear travel. If you understand these terms you will see that they are all independent of each other.

It is possible to build a speaker system with a high Xmax driver whose cone area and suspension compliance works with box volume to create a system resonance that is in the 30-50 Hz range. As long as you don't require cone motion in excess of Xmax the system has low distortion with admittedly lower efficiency below system resonance, even down to 10-20 Hz. Modern subwoofer speaekr drivers can have voice coils that can take the heat and clean amplifier power has become a very cheap commodity This makes overcoming the lower efficiency below system resonance completely feasible and even attractive when a small sized system is desirable (which is just about all the time).

However, if you are talking so-called full-range or tower systems, their drivers often don't have the cone area and driver Xmax to play this game, and then your concerns can be highly justified.

Just another situation where all generalizations are false! ;-)

Ringing is pretty much normal I think as we do not care much about acoustical response in this range - there is really no "considerable response" here anyway, and in case of sub - it will be way much louder anyway making this range indistinguishable from other speakers. It just saves processing power and not trying to fine-tune what is unimportant to audio quality.

We are not talking about specialized subwoofers here, that has a way more power in their amps, a way more area of speaker cones with larger excursion by design, we are talking about common case with satellites. As Audyssey can't assume they are capable (and it pretty much knows that the amplifier is pretty limiting factor also) it should not in any way try to extend the frequency range below the frequency the speaker can play normally (and it is documented that it is not trying to do this by design). As the measured -3dB point is only what he knows - it uses that.

10 dB boost in my case cannot be explained reasonably by any normalization process whatever it can be as any other point in my graphs is not higher than 6 dB boost. So, it boost "unplayable" range considerably more than anything else.

Again, do the measurements the way I did and your data will be meaningful. The way you did it doesn't result in conclusive data - you didn't measure all calibration points. By leaving the mic at the very same location you would get meaningful data.

By the way, when did you measure the preamp out with Audyssey off? Before calibration or afterwards?

In respect to the normalization - the graph alone is meaningful by itself, I explained million times already that any average, independent of the weighting technique cannot average some set of values higher that the highest value in the set, so, if you tell that what I see in my graph as boost is the new normalized value - it cannot be that just by this reason. So, whatever and however was measured, the defect of the mic, the room specifics, or anything cannot produce that if there were no bug. As there is a bug you cannot say for sure that the level difference at low frequencies you see in your measurements is from the correct normalization process and not induced by the same bug I am experiencing, just in lesser extent, but you might be right. There is some probability, but without making the whole math calculating it - nothing cannot be said for sure. At least in your case it doesn't look so horrible and hardly affects sound quality.

The data with 1 position will be exactly the same, small variance will not explain anything, at least for this boost problem. I have speakers at surrounds that are closed boxes, placed obviously differently, they have exactly the same problem, and the greatest boost is at 15-50Hz still. I think it would boost more if it allowed to, but it seems that there is something limiting the amount of boost to 10dB. 50-100Hz range response is ringing. And that is with speakers detected as 100Hz!!! Everything except the room and mic is different, the result is the same.

Surely all the measurements are after the calibration. No any changes in between except the Audyssey Off/Music switch.

It seems that the new mic. will arrive soon, will try next week with that one.

does the onkyo 1009 have the 24 bug?