STR-DN 1080 - Calibration setting not stored

[phil_merge]

Hi,

After performing the auto speaker calibration and selecting a setting ("Full Flat" for example), this setting is not stored. 

When I switch the receiver off and on again, the speaker calibration setting is ignored, although it still says "Full Flat" in the menu. The sound is as if it was set to "Off".

I need to manually set the setting to "Off" and to "Full flat" to hear the correct speaker calibration.

Thanks for your help!

Phil

[Beeper85]

Link to video

This video is of a kids cartoon. It shows the setting as in full flat and how it sounds. I then change it to engineer and then back. You can hopefully hear that only then is full flat is engaged. I then go to 'off' and sonically the setting 'off' is the same as the sound was when first turned on and the setting menu showed full flat as engaged (which it was prior to turning on and off).

[Beeper85]

Are you able to see the video? Whilst not the best quality, hopefully it should make it very clear as to what the issue is

[dutchice]

Hi @Beeper85,

Thanks for the video and even more for your engagement on the issue. I was indeed able to download your video. As I mentioned before already, sound issues require adequate testing and can not be determined by personal perception. Else the discussion will come down to the difference in what we hear instead of a factual conclusion.

The good news is that despite everything, I am expected to receive the test results either this afternoon or tomorrow morning. I will post back the findings as soon as I receive them, including the frequency spectral analysis which will serve as evidence on what exactly is the case.

Cheers,

Dutchice

[Beeper85]

Thanks

It will be interesting to see the published results... though I suspect we all know this is just going to show that my issue (and others who have this issue) are 'isolated instances'  

[Beeper85]

Perhaps if there is some way to factually and conclusively measure my receiver's behaviour that would be helpful?

If you can advise on a file to play via the receiver, and whether there is perhaps a spectrum analyser that I can run from my phone, you will be able to clearly see the fault (and obejctively so without any reference to 'what we hear') 

[dutchice]

Hi Ben and Philippe,

I just managed to have a look at the results myself. I admit that despite my initial scepticism the issue had made me very curious too. Let me give you a brief summary on how the test was performed.

First the FSS (Frequency Spectral Signature) of the 4 different calibration types (Off, Front Reference, Engineer & Full Flat) was captured after a full factory reset and a D.C.A.C. EX (5.1 sound calibration) performed with A.P.M. enabled. So all other specific device settings (DSD Native, D.L.L., EQ, etc.) were left at their respective factory reset default values. The FSS captures belong to a sample source of about 1 minute and 30 seconds of a Netflix 5.1 48 kHz title which was delivered from a TV connected by HDMI ARC. On the vertical axis of each FSS you see the frequencies ranging from about 100 Hz to 16 kHz and on the horizontal axis the timestamp is registered. The brighter the color in the FSS the higher the frequency response level measured in dB.

Fig 1: Frequency Spectral Signature for Calibration Type: OFF

Fig 2: Frequency Spectral Signature for Calibration Type: Front Reference

Fig 3: Frequency Spectral Signature for Calibration Type: Engineer

Fig 4: Frequency Spectral Signature for Calibration Type: Full Flat

Once the reference signatures were captured for all 4 calibration types, the steps to replicate the fault were performed according to your description and an FSS was captured to compare the signature with that of the 4 reference signatures.

Now the following FSS captured in Fig. 5 is the result of when the calibration type was left to Full Flat prior to powering OFF and ON the TV. The FSS of Fig. 6 belongs to the result of when the calibration type was left to Full Flat prior to disconnecting and reconnecting the mains to the AVR after 1 minute.

Fig 5: Frequency Spectral Signature post power cyclcing the TV with calibration type left to Full Flat in advance

Fig 6: Frequency Spectral Signature post disconnecting and reconnecting the mains only to the AVR after 1 minute with calibration type left to Full Flat in advance

I recommend you first download the figures to your desktop and then make a close side-by-side comparison. The first aspect you will immediately notice is that Fig. 5 and 6 have technically the same FSS. I invite you then to take a closer look at either Fig. 5 or 6 and compare it to Fig. 4 which is the FSS for calibration type Full Flat.

• You should be able to notice the bright spots occurring at about timestamp 0:50.0 & 1:00.0 for the 5 kHz to 6 kHz range. These are characteristic for all three figures 5, 6 and 4. You don't see these bright spots in Fig. 1 where the calibration type is set to OFF.
• You should also be able to notice that the frequency response level for the range between 100 Hz and 500 Hz is much more pronounced for the entire length of the sample on Fig. 1 compared to that of Fig. 5, 6 and 4.

In fact all 3 calibration types other than OFF try to expand and evenly spread the frequency response levels across a wider range of frequencies. Anyone can tell from analysing the figures that the Full Flat calibration type is maintained post a power cycle or mains reconnection of the receiver.

Obviously the results only conclude that the issue is not a general fault across all devices, so if you still strongly believe that your device is faulty then I suggest you contact support for a service repair.

Just bear in mind that all the digital sound processing features are integrated into groups of microprocessors on the AVR, therefor it is technically impossible for the device to have only one single malfunction on this particular process while no other features are affected. So if hypothetically your device might be malfunctioning, the root cause would come down to its microcode software and not hardware since all other features are performing fine. But then again, it is impossible that the microcode software is performing differently on your device compared to all other devices. My personal opinion is that you still won't be convinced that the device is functioning properly even after a service repair.

In response to your question of performing a test yourself at home, I certainly think that that should be possible. Indeed all you need is a quiet environment and a mobile phone. I am sure you could search for a frequency spectral analysis app or at least only do the sound recording with an app then perform the analysis on a desktop. Which ever sound sample you choose does not make much of a difference as long as you use the exact sample for each recording.

Let us know your thoughts!

Cheers,

Dutchice

[Beeper85]

Thanks for this - I do appreciate you looking into this in detail!

I do agree, from the FSS captured here, thugh quite hard to tell, the full flat setting pre and post power cycle does appear to be the same.

However, and I really wish you could hear this for yourself (if you are in the North West, I'd be happy to sort something out!), something is defintely 100% incorrect with how my receiver is acting.  I will try and capture FSS (and photos of the corresponding settings as displayed!) which will hopefully show visually the issue.  It is increidbly easy to hear though when listening - voice clarity in particular is instantly noticeable (my wife notices this and is not an audiophile!) where full flat / off after a power off/on is boxy and bassy and lacking the high ends (this will be a result of the speakers I'm running) v 'resetting' the full flat which then gives a much more balanced eq on voices.  

It is more than simple 'bias' in how I am listening that means I hear 'full flat' as far better once the option is reselected (even if showing it is seleceted after a power down) - it is only after much searching that I found this was happening and landed at this post after searching the symptoms.  

Phillipe, I would appreciate your views on this - I am concious I am not expressing myself well!

[Beeper85]

wondering if the best way to do this as a 'control' is to use a white noise sample and then take a spectrum analysis of this - wonder whether that will show any hot spots or uneveness, or at least a difference between full flat before and after power down?

[phil_merge]

Hi. I'm not at home this week.

I can make a proper test using a pink noise signal generator, a calibrated measuring mike and it's spectrum analyzer.

I often do this in recording studios and movie theatres (datasat ap25) as part of my job as recording engineer and studio designer.

I am used to calibrating monitoring setups.

Using a test signal will provide a better insight into the spectrum analysis.

I'll try to do this next week.

Cheers

Philippe

[Beeper85]

To keep you updated, I have quickly tried a white noise sample monitored with a linear frequency analyser.

You will be happy to hear that it clearly shows the issue - the setting ‘off’ gives a curve with a much higher peak dB reading for the bass, and lower dB reading for treble, and then full flat flattens the readings as would be expected.  

The read outs show the behaviour - after power cycle, the receiver in full flat mode (on display) is giving a read out that is the same as read when ‘off’ pre cycle, then giving a read out the same as before power off once the setting has been changed away and back again.

I will pull together images later.