thanks amir I hope when the 8500 arrives you can do those tests since it can turn off all the internal stages and the most logical thing is that it should affect the final noise less ... my pc only has hdmi I do not have a graphics card with toshlink but I will try a dac external that I have to connect by optical from the dac to the denon 8500 ... with all this say that I do not hear any audible distortion at the volume of 75db on the absolute scale, I do not know if the volume should be increased more greetings
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Denon AVR-X4700 AVR Review (Updated)
- Thread starter amirm
- Start date
thanks amir I hope when the 8500 arrives you can do those tests since it can turn off all the internal stages and the most logical thing is that it should affect the final noise less ... my pc only has hdmi I do not have a graphics card with toshlink but I will try a dac external that I have to connect by optical from the dac to the denon 8500 ... with all this say that I do not hear any audible distortion at the volume of 75db on the absolute scale, I do not know if the volume should be increased more greetings
I remember when Gene and the AH gang were panning Atmos and Dolby Atmos Enabled speakers (which I have and enjoy) before they even heard it. You can watch this video but be warned it's 16 minutes you'll never get back.
Here is the analog level diagram for the L/R channels for the Marantz AV8805. The level diagrams for higher level Marantz and Denon AVR's and AVP's all look like this.
This level diagram puzzles me because appears it shows that all L/R digital inputs, where no subwoofer is present, are reduced by 12dB in the digital domain. This sets the output at -32dBFS. 0dBFS is 2.0V This is reduction is an effort to prevent digital overs from the bass that is mixed into the L/R from other channels when a subwoofer is not present.
.4dB of this reduction is recovered in the analog filter after the DAC IC. 9.5dB of this reduction is recovered by adding 9.5 to the volume control setting. This addition would not be visible to the user. This leaves the signal at about -22dBFS (vs. about -20dB expected) at the nominal output level. This is relative to 2V output. See the attached level diagram.
At -22dBFS, the external volume control would be set at about 60.5 indicated on the AVR/AVP, and actually set to 70 internally due to the 9.5dB of added gain. 70 is 12.5dB below 0dBFS of 82.5.
Increasing the exterior volume control setting to 82.5 indicated for 0dbFS or 2V requires adding 22dB to the internal volume control setting. The -12.5dB actual internal setting of the volume control is increased by 22dB to 92. This means the volume control is set for +9.5dB of gain. The actual output is 2V, which normally requires a 0dB gain setting of the volume control, since the 0dBFS output of the DAC circuitry is normally 2V, but a setting of +9.5dB is actually required to makeup for some of the previous 12dB reduction.
It appears the gain available for feedback in the volume control internal circuit is cut dB for dB with actual settings of the volume control over 0dB gain. An approximate actual gain setting of the volume control to 9.5dB appears cut the feedback by about 9.5dB and this may increase distortion by about the same amount, that is, around 9.5dB or 3x.
Please take a close look at the level diagram since it is likely more clear than the convoluted text above. Also see if the above is on the right track.
Your thoughts?
Added:
The question that arises from the above is what causes the AVR/AVP to actually "think" that a sub is present or not in all situations. The measurements for the AV8805 and AV7705 were mediocre at 4V output, which is 2V output from the volume control with 6dB gain in the XLR buffer stage. Reducing output reduced distortion significantly. Is the excess distortion at 2V single-ended output, somehow due to the level changing process above?
This level diagram puzzles me because appears it shows that all L/R digital inputs, where no subwoofer is present, are reduced by 12dB in the digital domain. This sets the output at -32dBFS. 0dBFS is 2.0V This is reduction is an effort to prevent digital overs from the bass that is mixed into the L/R from other channels when a subwoofer is not present.
.4dB of this reduction is recovered in the analog filter after the DAC IC. 9.5dB of this reduction is recovered by adding 9.5 to the volume control setting. This addition would not be visible to the user. This leaves the signal at about -22dBFS (vs. about -20dB expected) at the nominal output level. This is relative to 2V output. See the attached level diagram.
At -22dBFS, the external volume control would be set at about 60.5 indicated on the AVR/AVP, and actually set to 70 internally due to the 9.5dB of added gain. 70 is 12.5dB below 0dBFS of 82.5.
Increasing the exterior volume control setting to 82.5 indicated for 0dbFS or 2V requires adding 22dB to the internal volume control setting. The -12.5dB actual internal setting of the volume control is increased by 22dB to 92. This means the volume control is set for +9.5dB of gain. The actual output is 2V, which normally requires a 0dB gain setting of the volume control, since the 0dBFS output of the DAC circuitry is normally 2V, but a setting of +9.5dB is actually required to makeup for some of the previous 12dB reduction.
It appears the gain available for feedback in the volume control internal circuit is cut dB for dB with actual settings of the volume control over 0dB gain. An approximate actual gain setting of the volume control to 9.5dB appears cut the feedback by about 9.5dB and this may increase distortion by about the same amount, that is, around 9.5dB or 3x.
Please take a close look at the level diagram since it is likely more clear than the convoluted text above. Also see if the above is on the right track.
Your thoughts?
Added:
The question that arises from the above is what causes the AVR/AVP to actually "think" that a sub is present or not in all situations. The measurements for the AV8805 and AV7705 were mediocre at 4V output, which is 2V output from the volume control with 6dB gain in the XLR buffer stage. Reducing output reduced distortion significantly. Is the excess distortion at 2V single-ended output, somehow due to the level changing process above?
Last edited:
valerianf
Addicted to Fun and Learning
This Marantz has at least a pre-out amplifier (NJU72343), as XLR connectors are available.
The subwoofer OFF side effect is puzzling.
May be in the future a subwoofer ON/OFF test needs to be added to the review.
The subwoofer OFF side effect is puzzling.
May be in the future a subwoofer ON/OFF test needs to be added to the review.
Excited about the SR6014 review - I've got the 6011 and 6012, love them. I hear 0 noise at reference level in my listening seat from 1/1.5/2 and 2.5m (yes I have calibrations for all viewing distances!). I do hear noise when I walk up to the speakers though - say 50cm away. This is with R300s and LS50's. Room has a background noise of around 25-35db according to my UMIK-1.
I'm sad about Audioholics, the chap that runs the place isn't the most likable out of all this. His behavior on the forums is worse - I saw some real silly posts on the lossless vs lossy debates and although the words he writes are conciliatory you can see his "liking" of negative comments on ASR on YT broadcasts true intent. Thankfully a Matthew Poes has been producing fantastic content that he can publish but actual Gene content has been very lacking.
I'm sad about Audioholics, the chap that runs the place isn't the most likable out of all this. His behavior on the forums is worse - I saw some real silly posts on the lossless vs lossy debates and although the words he writes are conciliatory you can see his "liking" of negative comments on ASR on YT broadcasts true intent. Thankfully a Matthew Poes has been producing fantastic content that he can publish but actual Gene content has been very lacking.
Amir found a big bug in their software that will have impacted a large percent of the consumers who buy those items. However, there are ways around it and Amir found that out too. Sites like AH are little help to consumers while ASR will continue to be a force for good and force manufacturers to raise the bar and eventually make all of our ears much happier.
It's a really interesting info about how these machines deal with level. Since no more than 24bits is found in sources and AFAIK the digital process is done at 32bits, that headroom is used when needed.
An interesting number is the DOLBY LIMIT LEVEL = 1.41Vrms. It could explain why many AVRs are optimized with that voltage as reasonable maximum limit.
One missed piece is the DSP which would be in the center.
I guess you are right, so when no SW is present the DSP applies an "extra" negative gain in the internal digital process to avoid clipping/overflow in digital domain when adding the LFE channel to the mains. So the NJU72343 input is programmed to apply different gains to the main channels depending on that, to reach the "normalized" analog level.
A possible reason for the detected clipping from the PCM hdmi input could be the Direct Mode. Let's say that DSP treats encoded signals (Dolby/DTS) with specific algorithms to deal with down-mixing in a proper way, and probably, those down-mixing algorithms aren't triggered when dealing with a direct 8 channels PCM input in direct mode.
Just guessing.
Last edited:
Gene’s now responding that he’s NEVER seen anyone send a multi-channel signal to a receiver for stereo listening, which makes me ask a REW question. Maybe @amirm knows the technology behind the REW software. Since I’m configured for multiple-channel via HDMI and I test just one speaker via REW, is it sending a NULL signal on the other six channels? Thus my SINAD would be even worse than the original tested result? Thanks.
It reminds me on iPhone 4 and Apple telling people they are holding it wrong. All respect to Denon for willingness to explore and collaborate measurements. It is a remark of a company that cares about it products and it customers. Everyone can have a bug or slip somewhere.
To be fair to Apple, they also showed that the issue happened with competitors. This issue is so far only happening with the 2020 Denon models.
Also, maybe Apple should rent out their 17 anechoic chambers to speaker companies
Last edited:
While certainly not a Gene apologist, I am not sure he's wrong....
Just to be super clear, I thought from your previous responses to similar questions, the "more noise"/downmix bug would only be an issue if the AVR is configured for 2.1, or 3.1 etc.. I presume. So if I understand correctly, as long as I configure the AVR, in the speaker configuration menu for 7.1 or more, like 11.1 in my case, I should be able to use HDMI and select stereo (2.1), or direct mode, right or wrong?
The thing is, even for people who use Audyssey, Denon does allow the unit to be configured for say 7.1 after auto setup and there is no requirements to physically connect more than 2 speakers or any speakers. It seems to me it is a matter of configuration matching, not much to do how many speakers are actually in use, if I understood right.
Gene likes to play fast with the truth. He's not wrong that MOST people won't run into this situation in the real world, however since Yamaha is not exhibiting this same downmixing issue, anyone who is running a 5.1 or 3.1 configuration for movies is going to experience a drop in SINAD when the Denon downmixes. This is almost EXACTLY what he said when he was arguing that All Channels Driven was not an adequate test because it doesn't reflect the real world (that was in 2006). Yet in 2019 he writes "The SR8012 was able to muster and impressive 96 watts/ch at 1% THD and 86 watts/ch at 0.1% THD with 7CH driven. This should be plenty of power for all but the largest and most demanding home theaters. " Ha! Suddenly 86 watts x 7 at 0.1% is plenty for MOST people, but some people might still need more power! This cracks me up as he was defending Yamaha receivers in 2006 that couldn't break 33 watts x 7, stating it was an issue you would never bother about in the real world. I guess the real world has changed.
I'm certain Gene is a nice guy, but he knows where his bread is buttered.
May be someone can summarize a list of possible conditions that would result in degradation in SINAD. From my own understanding, I thought is would only be a problem if the AVR is not configure for at least the same number of speakers as the number of channel contents in the HDMI input signal, whether all of the channel contents are active. As I agreed long time ago, it is a bug that needs to be fixed, but my question is, until it gets fixed, the fact that mine is also configure for 11.1 (okay 11.2 to be popularly correct), a when I use HDMI to play a stereo track or 7.1 track I do not change my speaker configuration, so will I be "bugged". I don't think so, but I guess I may be wrong.
The question I keep asking is about the key word "anyone". If I had only 5 speakers and 1 subwoofer, and if I play a 7.1 track on a BR movie, I would lost about 2 dB SINAD right? Now if there is a 5.1 track (English), and I play that track then I won't lose that 2 dB? It is a bug that is a bug, I know, but I think there are may be people who think they are getting screwed while waiting for a fix, when in fact, they can relax and enjoy their favorite movies as long as there is a 5.1 track in language of their choice.
That's assuming I understood the bug issue correctly, and it would be great if some member who is a language specialist can provide a list clearly stating the conditions when the bug would apply, and/or not.
haha me neither, and I think if everyone just sticks to facts, exercise due diligence in fact checking and be forgiving, then hopefully members like you are I who frequent both sited could continue to have fun without being Gene, Denon, or Amir apologists ever..
I am sure Gene could take it well if someone (who has enough respect for him) tell him to....(may be via PM..)
Honestly I don't know the answer. If your BluRay player sends the 5.1 track with 2 null channels, then you'll get the 2 dB SINAD loss. If the BluRay players sends the 5.1 track only you won't get the loss. Has anyone ever checked BluRay, or streaming devices, if they are doing this correctly? Another great reason why Denon needs to fix this issue.The question I keep asking is about the key word "anyone". If I had only 5 speakers and 1 subwoofer, and if I play a 7.1 track on a BR movie, I would lost about 2 dB SINAD right? Now if there is a 5.1 track (English), and I play that track then I won't lose that 2 dB? It is a bug that is a bug, I know, but I think there are may be people who think they are getting screwed while waiting for a fix, when in fact, they can relax and enjoy their favorite movies as long as there is a 5.1 track in language of their choice.
That's assuming I understood the bug issue correctly, and it would be great if some member who is a language specialist can provide a list clearly stating the conditions when the bug would apply, and/or not.
UplandShadows77
Member
- Joined
- Jun 26, 2020
- Messages
- 18
- Likes
- 2
Not sure what it's for but it's been "downloading" for about an hour now..
Edit: This is a NEW FW update, I updated to the one that came out a few week(s) ago already; not trying to confuse anyone.
Edit: This is a NEW FW update, I updated to the one that came out a few week(s) ago already; not trying to confuse anyone.
OP
- Thread Starter
- #640
The top level config of 11.2 or whatever has nothing to do with this problem. That is simply a trick to allow the amp to shut off or not.
I have shown the problem clearly here and it is precisely due to not actually enabling the speakers in the configuration. In this first case, I am enabling 5.1 configuration and there is still overflow indicating channel mixing is occuring:
Here, I went to 7.1 and now clipping goes away indicating we are finally getting pass through operation:
This means to get the best performance for stereo playback, you must give up channel mapping/mixing. You can't have both.
This is what my testing shows. And this is what is confirmed by Denon Engineering as our consensus.
