Is HDMI to optical through a TV lossless?

[theevilsloth]

If I plug an HDMI Chromecast into my TV's HDMI in, then use Optical out -> DAC -> amp -> RCA -> passive speakers, am I losing any quality at the HDMI to Optical stage? I know that TVs have bad DACs, but since HDMI and optical are both digital, is my TV able to convert this without losing quality? Thanks.

 

[twsecrest]

Normal optical, 2.0 channels of 24-bit/96 or 24-bit/192k audio digital audio.
Or Bitstream, Dolby Digital Live, DTS-connect (compressed digital audio) 6-channels (5.1) 24-bit/48k digital audio.
There might be some improvements in Dolby or DTS that I'm not aware of?

HDMI 8-channels (7.1) of un-compressed 24-bit/192k digital audio.
But I'm sure newer HDMI has improvements.

 

[theevilsloth]

HDMI 8-channels (7.1) of un-compressed 24-bit/192k digital audio.

Ok, so it sounds like for my 2.0 setup, HDMI -> HDMI ARC with a AVR -> 2.0 would be essentially the same as HDMI -> Optical -> 2.0?

Also - stupid question, but since I have a trashy TV, is there a potential that my TV is somehow downgrading the quality between the HDMI in and optical out? Or not really?

 

[formdissolve]

Ok, so it sounds like for my 2.0 setup, HDMI -> HDMI ARC with a AVR -> 2.0 would be essentially the same as HDMI -> Optical -> 2.0?

Also - stupid question, but since I have a trashy TV, is there a potential that my TV is somehow downgrading the quality between the HDMI in and optical out? Or not really?

Old post, but will reply. Most Optical SPDIF implementations on TVs are pretty terrible and loaded with noise and jitter. If you're just streaming, it's already compressed enough to probably not notice a difference, but you might. If you're watching blu-rays or anything, it will also downsample from 24-bit to 16-bit before passing the PCM signal over optical to your DAC.

 

[dasdoing]

Most Optical SPDIF implementations on TVs are pretty terrible and loaded with noise and jitter.

I think the main issue is the clocking. my old audio interface would always complain there was no clock pairing possible. My Motu does achieve the pairing with the same TV though. I guess most consumer DACs will have problems and default to internal clocking....which will create high frequency artifacts

 

[Elitzur–Vaidman]

24 bit bitperfect? Generally, no. Auidbly lossless? Unless your TV is absolutely terrible, yes.

 

[antcollinet]

Old post, but will reply. Most Optical SPDIF implementations on TVs are pretty terrible and loaded with noise and jitter. If you're just streaming, it's already compressed enough to probably not notice a difference, but you might. If you're watching blu-rays or anything, it will also downsample from 24-bit to 16-bit before passing the PCM signal over optical to your DAC.

Noise? Over optical? How does that work? Jitter of course doesn't matter as long as it's not enough to prevent the DAC from syncing the clock.

Downsample from 24bit to 16? Are you sure? Why would it do that? It certainly does't on mine.

 

[voodooless]

Noise? Over optical? How does that work? Jitter of course doesn't matter as long as it's not enough to prevent the DAC from syncing the clock.

Downsample from 24bit to 16? Are you sure? Why would it do that? It certainly does't on mine.

There are quite a number of bizarre issues reported out there on various TVs. These do include noise and distortion, and there are receipts. I think the downsampling is confirmed as well on some models. The same goes for resampling, at times very badly done.

Why.. who knows? By now, Toslink is a legacy port, so having it working is probably a low priority.

 

[Elitzur–Vaidman]

Noise? Over optical? How does that work? Jitter of course doesn't matter as long as it's not enough to prevent the DAC from syncing the clock.

Downsample from 24bit to 16? Are you sure? Why would it do that? It certainly does't on mine.

In my experience (TCL 635 vs LG G2), lower-quality TVs will resample down to 16-bit, but proper 16-bit is fully audibly transparent, so I don't think its a real issue.

 

[AdamG]

OP has not been back since February.

 

[formdissolve]

In my experience (TCL 635 vs LG G2), lower-quality TVs will resample down to 16-bit, but proper 16-bit is fully audibly transparent, so I don't think its a real issue.

They resample anything that is HDCP copy protected. If they downsample and dither correctly, it's fine. For example, if you have a Blu-Ray player plugged into the TV, and play a movie with say DTS-HD Master Audio which is 24-bit, then send PCM data over optical from the TV to a DAC, that will be downsampled to 16-bit due to copy protection.

As for optical ports being noisy, JDS Labs tested a popular OLED series from LG (C8/C9) and found the port to be pretty poorly implemented. View all the spikes and the horrible performance of 81dB from the port itself.

Analysis: Audio Precision's APx555 has the ability to interpolate a digital stream into time and frequency domains to analyze how it would render if perfectly converted from digital back to analog (i.e., without a DAC in the test setup). We used the ASIO4ALL driver to observe SINAD of a 1kHz test tone reconstructed directly from the digital S/PDIF output:

 

[formdissolve]

There are quite a number of bizarre issues reported out there on various TVs. These do include noise and distortion, and there are receipts. I think the downsampling is confirmed as well on some models. The same goes for resampling, at times very badly done.

Why.. who knows? By now, Toslink is a legacy port, so having it working is probably a low priority.

They sell A LOT of sound bars that are Toslink only because not all cheaper TVs have eARC. In that case, SINAD doesn't really matter because the sound bar itself will be pretty poor.. but at least better than the stock TV speakers.

 

[antcollinet]

Analysis: Audio Precision's APx555 has the ability to interpolate a digital stream into time and frequency domains to analyze how it would render if perfectly converted from digital back to analog (i.e., without a DAC in the test setup). We used the ASIO4ALL driver to observe SINAD of a 1kHz test tone reconstructed directly from the digital S/PDIF output:

Interesting the resampling - I've leant something today.

But what is the source of the noise - obvioulsy not electrical noise - or are we seeing jitter distortion? Or is it just really bad resampling or sample rate conversion? If jitter, then is the analysis failing to do the jitter rejection that the PLL or ASCR of a DAC would perform?

 

[formdissolve]

Interesting the resampling - I've leant something today.

But what is the source of the noise - obvioulsy not electrical noise - or are we seeing jitter distortion? Or is it just really bad resampling or sample rate conversion? If jitter, then is the analysis failing to do the jitter rejection that the PLL or ASCR of a DAC would perform?

I'm not entirely sure, but it also has to do with very poor clock implementation. Nearly all threads about this topic stem from people using ESS-based DACs with LG TV's and noticing severe pops/clicks and dropouts which is due to ESS attempting to resync constantly as detailed in that link I posted:

"This behavior is the result of the ESS's otherwise excellent jitter recovery logic. The patented jitter eliminator must periodically unlock and re-sync because the digital stream is so poor. AKM and Wolfson S/PDIF receivers are not asynchronous, so they pass along any significant jitter from an optical source, thus, eliminating any chance of a momentary pause for an S/PDIF re-locking event. Re-locks are ordinarily quite rare--you should never encounter this behavior with default jitter rejection values. And under typical circumstances, ESS's asynchronous jitter elimination results in dramatically higher jitter performance than older DACs without this technology."

I first noticed it with an ESS-based Topping on my C8. I then swapped out the C8 board with a C9, and noticed the same issue. I pulled out an old Schiit Modi 3 with AKM and 95% of the issues went away (however there are still random pops and clicks on my current AKM DACs which do not occur on my other optical sources). It's just that LG (and most manufacturers) seem to not care anymore as a previous poster said, but it's still annoying to say the least.

 

[antcollinet]

I'm not entirely sure, but it also has to do with very poor clock implementation. Nearly all threads about this topic stem from people using ESS-based DACs with LG TV's and noticing severe pops/clicks and dropouts which is due to ESS attempting to resync constantly as detailed in that link I posted:

"This behavior is the result of the ESS's otherwise excellent jitter recovery logic. The patented jitter eliminator must periodically unlock and re-sync because the digital stream is so poor. AKM and Wolfson S/PDIF receivers are not asynchronous, so they pass along any significant jitter from an optical source, thus, eliminating any chance of a momentary pause for an S/PDIF re-locking event. Re-locks are ordinarily quite rare--you should never encounter this behavior with default jitter rejection values. And under typical circumstances, ESS's asynchronous jitter elimination results in dramatically higher jitter performance than older DACs without this technology."

I first noticed it with an ESS-based Topping on my C8. I then swapped out the C8 board with a C9, and noticed the same issue. I pulled out an old Schiit Modi 3 with AKM and 95% of the issues went away (however there are still random pops and clicks on my current AKM DACs which do not occur on my other optical sources). It's just that LG (and most manufacturers) seem to not care anymore as a previous poster said, but it's still annoying to say the least.

If noise is pops and clicks then fine - but that isn't really noise, just a broken implementation where the clocks can't sync.

I also doubt that is what the AP is showing, because it would be much higher level than what is shown there.

 

[formdissolve]

If noise is pops and clicks then fine - but that isn't really noise, just a broken implementation where the clocks can't sync.

I also doubt that is what the AP is showing, because it would be much higher level than what is shown there.

You'd have to go back to their analysis and verify the test case they set up is correct, which is something I can't say because I don't know how to use an AP like @amirm or others. I am also assuming that the TV is taking the 24 bit 48k signal in that test and poorly converting it to 16 bit, which is where that noise could be coming from. It's more of a nuisance than anything, because Amir has shown that even cheap Monoprice HDMI extractors can have 140dB sinad over optical (transparent).

To me, pops and clicks from constant re-syncing is incredibly annoying, so I had to move away from ESS for that reason until I purchase something like the HD Fury Vertex 2 for my video sources.

FYI, they also did the same exact test on a motherboard and it showed absolute transparency, but that might just prove the motherboard doesn't downsample:

 

[antcollinet]

You'd have to go back to their analysis and verify the test case they set up is correct, which is something I can't say because I don't know how to use an AP like @amirm or others.

Me neither - I'm certain they know better than me what they are doing. I'm just trying to understand. Which brings me to.

I am also assuming that the TV is taking the 24 bit 48k signal in that test and poorly converting it to 16 bit, which is where that noise could be coming from

I've just had a closer look at the picture. All that "hash" seems to be at or below -96dB - which is where I'd expect the noise floor to be for 16bit audio, so you could well be right. It doesn't look totally like I'd expect but it no big surprise if my expectations are wrong.