AppleTV X - I am playing with something new

INTERESTING....

The 2nd developer beta of 18.0 ( 22J5305e ) clearly looks better. Its the first firmware I have seen that had a affect on picture. Not sure why its better, but, I spent some time with a 2nd unit on the first 18.0 beta and the unit with the 2nd beta was easy to pick out on a blind AB..

So thats exciting. A better pic is coming..

BUT.. its VERY Beta with some clear issues they need to work out. Like you can get stuck in Tidal and cannot get out. Even app restart does not help. You gotta remove and reinstall. There could be lots of other issues I am unaware of. So don't try 18.0 at home just yet. At a minimum wait for the public beta..

I should be the only one playing with a dev beta. A dev beta MIGHT brick a ATVX and I cant fix that. I am willing to sacrifice a ATVX. So dont do this.

Not sure how a better pic occurred..
 
  • Like
Reactions: DDgtt and BKDad
Not quite sure how you do a "blind" AB with video gear, but I get what you mean. :)

I wonder if Apple actually is taking picture quality more seriously. Or, maybe, other software development moved some stuff around enough that a new compile has the happy accident of better video.

(Insert fingers crossed emoji here)
 
Not quite sure how you do a "blind" AB with video gear, but I get what you mean. :)

I wonder if Apple actually is taking picture quality more seriously. Or, maybe, other software development moved some stuff around enough that a new compile has the happy accident of better video.

(Insert fingers crossed emoji here)

I doubt Apple had any idea about PQ. Maybe.. I think its a happy accident.. That maybe is not very visible with a std ATV.

By "Blind AB" that refers to a AB test done without my knowing which was which. I had someone else flip wires around and then I had to guess which was which. Its something I do ALOT of. Its the only way to really be sure of a mod or change. If I, or others, can pick it out in a statically meaningful way, then most likely there is a valid difference.
 
Apple hasn't shown much interest in picture quality to date. The last hardware "update" was a step back.

But, maybe there's some movement in the software:

Sequoia HDMI Passthrough

I'm not suggesting that the company decided to pursue higher quality video and audio. But, perhaps in order to offer a new feature they had to clean some stuff up. I won't complain. Of course, there's still time for them to backslide some more.

My somewhat unrelated experience with this sort of thing is that the Apple engineers pretty much all want to do really great work and provide great products. Their management often has other goals and are encouraged by Apple culture to pursue them. The engineers have lots of barriers put up around them, not only in terms of where they can wander, but also in the feedback they get from beta testers and users.
 
Apple hasn't shown much interest in picture quality to date. The last hardware "update" was a step back.

But, maybe there's some movement in the software:

Sequoia HDMI Passthrough

I'm not suggesting that the company decided to pursue higher quality video and audio. But, perhaps in order to offer a new feature they had to clean some stuff up. I won't complain. Of course, there's still time for them to backslide some more.

My somewhat unrelated experience with this sort of thing is that the Apple engineers pretty much all want to do really great work and provide great products. Their management often has other goals and are encouraged by Apple culture to pursue them. The engineers have lots of barriers put up around them, not only in terms of where they can wander, but also in the feedback they get from beta testers and users.

Hmmmm.... No passthru options yet visible on ATV audio for 18.0 beta. But that would be fun. Everyone has also begged for things like 96/24 and 192/24.. But no movement on that ever..
 
  • Like
Reactions: DDgtt
On a separate note,, as i know some of you have my Switch X in addition to the ATVX, it has some high level reviews in progress now. It turns out its replacing all the switches and routers for high end 2ch audio streaming that it has come up against, even setups costing a LOT more. It is doing for audio systems what a ATVX does for streaming video.

A switch X does improve things with the ATVX.

I updated the web site for the Switch X with features, interface screen shots and a video on how to config. This all goes over a lot more detail about the Switch X. The features link and the video link are on here. https://dejitterit.com/SwitchX/SwitchX.htm

Also I now also support AES67/Dante/Ravenna on it. You can get 16 channels of DSD256 thru it with Ravenna via Merging Hapi and Masscore.
 
  • Like
Reactions: DDgtt
@Xymox Any plan to give KScape another try now that the V is out? They say that it has some electrical improvements.

Howdy :)

Well... no.. There are a few reasons. Primarily that the ATVX is close enough for most people that a single platform for EVERYTHING streaming is good enough. Some say better. So the engineering work and support and distribution of a modded kscape just does not seem worth it. I keep hoping Kscape gets a ATV app that can play content. Then its the best of both worlds. Athe ATVX as a Kscape end point.

The second reason is that kscape will have a legal fit if I just crack the cover of one of their boxes. This is because of SERIOUS legal agreements they have with studios. Making a better version of thier box would bring massive, immd, legal peril. I know this for certain. I have asked.

I even offered to help them, free of charge, on a new player, but, was turned down.

The ATVX is at least very close and has a much better selection of material and a way bigger eco system and is Apple. Plus a ATVX is way cheaper to buy and way cheaper for content. I am watching DirecTV Stream right now for example. I was listening to Sirrus/XM earlier. I will most likely use Infuse later to play some local content off my Qnap NAS.
 
  • Like
Reactions: BKDad
@Xymox

Is your product "limited" if you will by the streaming companies capping their bandwidth?

Some 4KUHD physical media and KScape files are 100 mbs bitrate.

Whereas, Apple, Netflix etc... only getting 20mbs or so on a stream.

Also, what happens when Apple releases a Firmware update to the product?

Do you have to provide a "custom" firmware to X owners?
 
This is a question I get all the time. Its a understandable question.

I will cover it in a few ways.

A good analogy is digital audio. Something most people here know well. SURE DSD256 can sound great. 192/24 can sound great. But is it always the case that it sounds better ? Then there is a the question of the hardware. The hardware for digital playback REALLY matters, a lot. The server, the DAC, the digital cables, the switch, the clocking.. Its easy to argue that the hardware is very very important to the sound, more so then the audio format and bitrate.Digital audio hardware has had 30 years of hard core evolution where each part that makes up a DAC has been perfected. Along the way it became clear jitter mattered for example. A high end DAC internally looks more like a high end lab grade piece of Textronix or Agliant test equipment. The digital audio hardware has been refined over decades by listening and measuring.

It turns out that streaming and video systems have not had nearly any of this type of engineering evolution. HDMI based systems are nearly untouched by proper engineering. HDMI circuits in all gear are simply copy pasted from "reference designs" from a chip makers datasheet. This "reference designs" are the cheapest way that the system is known to pass HDMI tests. You open up any HDMI devcie, look at the data sheet for the HDMI chip and you will find a copy paste from the chip maker. If a mfgr deviates from this, then HDMI certification is VASTLY more expensive and takes huge amounts of time. So even super high end devices copy paste a very cheap HDMI circuit in.

This has led to a situation where since the dawn of HDMI that no one even tried to improve HDMI. In fact no one even looked to see if improvements had any affect on the pic/sound. Some people in the buis have played with this tho. Datasat for example. There were some very quiet mods made to the HDMI boards and this was very much a hush hush thing because it would break HDMI certification rules.

As it turns out... HDMI seriously sucks. It "reference designs" are pretty much abhorrent. WHile the resultant signal does pass tests, it produces a really nasty audio/video experence. Pretty much, including me, knew that. Then one day I modded my Oppo really seriously deep into the HDMI system. The picture popped off the screen and sound was just STUNNING.. HUGE jump.. At that point I knew HDMI SUCKED..

Like you I looked at these streaming services as garbage because of the low bit rate. I tried them all. They all equally sucked. Then one day a client forced me to to mod a AppleTV. It was IMMD a OMFG experience. The sound went from total garbage to a detailed image with depth and nuance, musicality jumped thru the roof. It was stunning the difference. Then i noticed the pic.. Whoa...

I spent months playing with all this. I applied every bit of the hardware model used in high end digital audio to a AppleTV. I jumped into the HDMI with $100,000 test equipment. I spent time working to dejitter things like the CPU and RAM and the buss that feeds the HDMI chip. I cleaned up the whole platform by huge margins. EVERYTHING I did I could see and hear. It turns out a HDMI platform is just as esoteric as a high end DAC/Server/Switch... Along the way i also ran into a number of things that really rocked my engineering world about some key things in modern digital electronics that no one else has noticed. While I wont say what one of the key findings was, lets just say the PWM switching regulators have a issue that no one seems to have noticed. I have discussed this with some trusted engineers in this industry and they say I should patent.

The result was so crazy, I decided to make and sell the units.

So to get back to your question.. I think cleaning up the HDMI and the entire hadrware platform is more important then the bitrate alone.

OK.. So.. Bitrate.. A still image takes no bitrate. Its when you have motion that you need bit rate. Video for consumers uses variable bit rate. This means when there is little motion, but rate is REALLY low. The more motion, the higher the bitrate. If you run out of "cap" on bit rate then the areas in motion will loose resolution which is not that visible to the eye/brain.

Bitrate is only as good as the material being transmitted. For example you have have DSD256 of a recording, but, that does not mean its a good recording that makes use of it. A lot comes into play when your talking picture quality, not just bitrate.

The AppleTV can do 600Mbps streams from alocal NAS. I have a lot of test material at these high rates. I also have post production clients using the ATVX and playing high speed streams and files. SURE high bit rate looks better, most when things are in motion, but the quality of the material most times is way more important then the bitrate.

So.. The results people have shared here on this thread doing A/B/C/D/E of all manner of device VS the ATVX is that ok, on the right material, in some scenes, a UHD disc or a kscape might look and sound better. BUT. The HDMI and hardware improvements make this a very fuzzy line.

BluRay is dying. Physical media is history. Sadly.. KScape is expensive and the content is limited and expensive. A AppleTV has EVERYTHING ever made on it already on some app somplace. The hardware improvements puts it right that with Oppo and kscape. For me, I still have my Oppo and a kscape system, but, neither are hooked up. The Apple ecosystem is all I need. I have a single source component. My display is calibrated to my single source. For me its a ideal solution.

I use it now for so many things. Right now I am playing the Sirrus/XM app and have it playing Yacht Music, hahaha.. And it sounds surprizingly good for Sirrus/XM..

There is a reviewer working on a stand alone review of it as a streaming audio solution. The multichannel audio guys love it too.

So.. Bitrate is not the thing to focus on.. Just like in audio. Its the hardware that really matters.
 
This is a question I get all the time. Its a understandable question.

I will cover it in a few ways.

A good analogy is digital audio. Something most people here know well. SURE DSD256 can sound great. 192/24 can sound great. But is it always the case that it sounds better ? Then there is a the question of the hardware. The hardware for digital playback REALLY matters, a lot. The server, the DAC, the digital cables, the switch, the clocking.. Its easy to argue that the hardware is very very important to the sound, more so then the audio format and bitrate.Digital audio hardware has had 30 years of hard core evolution where each part that makes up a DAC has been perfected. Along the way it became clear jitter mattered for example. A high end DAC internally looks more like a high end lab grade piece of Textronix or Agliant test equipment. The digital audio hardware has been refined over decades by listening and measuring.

It turns out that streaming and video systems have not had nearly any of this type of engineering evolution. HDMI based systems are nearly untouched by proper engineering. HDMI circuits in all gear are simply copy pasted from "reference designs" from a chip makers datasheet. This "reference designs" are the cheapest way that the system is known to pass HDMI tests. You open up any HDMI devcie, look at the data sheet for the HDMI chip and you will find a copy paste from the chip maker. If a mfgr deviates from this, then HDMI certification is VASTLY more expensive and takes huge amounts of time. So even super high end devices copy paste a very cheap HDMI circuit in.

This has led to a situation where since the dawn of HDMI that no one even tried to improve HDMI. In fact no one even looked to see if improvements had any affect on the pic/sound. Some people in the buis have played with this tho. Datasat for example. There were some very quiet mods made to the HDMI boards and this was very much a hush hush thing because it would break HDMI certification rules.

As it turns out... HDMI seriously sucks. It "reference designs" are pretty much abhorrent. WHile the resultant signal does pass tests, it produces a really nasty audio/video experence. Pretty much, including me, knew that. Then one day I modded my Oppo really seriously deep into the HDMI system. The picture popped off the screen and sound was just STUNNING.. HUGE jump.. At that point I knew HDMI SUCKED..

Like you I looked at these streaming services as garbage because of the low bit rate. I tried them all. They all equally sucked. Then one day a client forced me to to mod a AppleTV. It was IMMD a OMFG experience. The sound went from total garbage to a detailed image with depth and nuance, musicality jumped thru the roof. It was stunning the difference. Then i noticed the pic.. Whoa...

I spent months playing with all this. I applied every bit of the hardware model used in high end digital audio to a AppleTV. I jumped into the HDMI with $100,000 test equipment. I spent time working to dejitter things like the CPU and RAM and the buss that feeds the HDMI chip. I cleaned up the whole platform by huge margins. EVERYTHING I did I could see and hear. It turns out a HDMI platform is just as esoteric as a high end DAC/Server/Switch... Along the way i also ran into a number of things that really rocked my engineering world about some key things in modern digital electronics that no one else has noticed. While I wont say what one of the key findings was, lets just say the PWM switching regulators have a issue that no one seems to have noticed. I have discussed this with some trusted engineers in this industry and they say I should patent.

The result was so crazy, I decided to make and sell the units.

So to get back to your question.. I think cleaning up the HDMI and the entire hadrware platform is more important then the bitrate alone.

OK.. So.. Bitrate.. A still image takes no bitrate. Its when you have motion that you need bit rate. Video for consumers uses variable bit rate. This means when there is little motion, but rate is REALLY low. The more motion, the higher the bitrate. If you run out of "cap" on bit rate then the areas in motion will loose resolution which is not that visible to the eye/brain.

Bitrate is only as good as the material being transmitted. For example you have have DSD256 of a recording, but, that does not mean its a good recording that makes use of it. A lot comes into play when your talking picture quality, not just bitrate.

The AppleTV can do 600Mbps streams from alocal NAS. I have a lot of test material at these high rates. I also have post production clients using the ATVX and playing high speed streams and files. SURE high bit rate looks better, most when things are in motion, but the quality of the material most times is way more important then the bitrate.

So.. The results people have shared here on this thread doing A/B/C/D/E of all manner of device VS the ATVX is that ok, on the right material, in some scenes, a UHD disc or a kscape might look and sound better. BUT. The HDMI and hardware improvements make this a very fuzzy line.

BluRay is dying. Physical media is history. Sadly.. KScape is expensive and the content is limited and expensive. A AppleTV has EVERYTHING ever made on it already on some app somplace. The hardware improvements puts it right that with Oppo and kscape. For me, I still have my Oppo and a kscape system, but, neither are hooked up. The Apple ecosystem is all I need. I have a single source component. My display is calibrated to my single source. For me its a ideal solution.

I use it now for so many things. Right now I am playing the Sirrus/XM app and have it playing Yacht Music, hahaha.. And it sounds surprizingly good for Sirrus/XM..

There is a reviewer working on a stand alone review of it as a streaming audio solution. The multichannel audio guys love it too.

So.. Bitrate is not the thing to focus on.. Just like in audio. Its the hardware that really matters.
And the Emmy Award 2023 for Audio/Visual Engineering goes to......? :)
 
"that no one seems to have noticed"

It's hard to notice things when you're motivated not to look.

To be fair, most engineers are not driven by high performance considerations. They've been trained at a young age (career wise) that schedules and cost are everything and the only things. (I bet somebody has written a book about that, but, if so, I haven't discovered it yet.)

Power supplies and power systems in general are not given the treatment they deserve. A lot of engineering problems could be solved if people chose to pay attention. I've found that most practicing engineers don't really appreciate basic nodal and mesh analysis, which could explain why common mode currents and all that is ignored. My theory is that because those basics are taught to first year engineering students, they get lost. It's tough enough adjusting to college life that first year and making it to class... Anyway, power and its distribution are the foundations on which the other electronics are built. A crappy foundation gives less than great results.
 
And the Emmy Award 2023 for Audio/Visual Engineering goes to......? :)

Wahahahahaha...... Maybe if i cracked open a camera and stuffed parts in.. No awards for consumer gear. Well there are.. I did take product of the year back in 2003 in Stereophiles Guide To Home Theater and Perfect Vision for my projector :) Johnthan Valin gave me product of the year :) I am still proud i made the best CRT proj ever made :) Fun times..

Sadly, main stream magazines have a policy of not reviewing modded gear :( So while a bunch of well known reviewers have a ATVX they cant really cover it.


 
"that no one seems to have noticed"

It's hard to notice things when you're motivated not to look.

To be fair, most engineers are not driven by high performance considerations. They've been trained at a young age (career wise) that schedules and cost are everything and the only things. (I bet somebody has written a book about that, but, if so, I haven't discovered it yet.)

Power supplies and power systems in general are not given the treatment they deserve. A lot of engineering problems could be solved if people chose to pay attention. I've found that most practicing engineers don't really appreciate basic nodal and mesh analysis, which could explain why common mode currents and all that is ignored. My theory is that because those basics are taught to first year engineering students, they get lost. It's tough enough adjusting to college life that first year and making it to class... Anyway, power and its distribution are the foundations on which the other electronics are built. A crappy foundation gives less than great results.

Amen to that..

Yea engineering schools today are all based on how to use modeling software. Analog is not really taught because its not in much use. Most used in RF.

The modeling software can produce gear that works. So why should they think about anything ? True of math too.

Another issue is test equipment. Modern scopes are great for periodic signals. They sample and digitize and you can do all manner of math and statistics on waveforms. Which is awesome for SOME uses.. But you stick a $100k Tektronix MSO on a analog power supply rail and look at noise, there are no combos of settings that will let you really see random stuff well. So modern engineers simply cant see what is going on. You gotta have a ANALOG scope to really see whats going on. No sampling. You literally see EVERYTHING. So anyone who is working on modern electronics with a digital scope is missing a lot. A Tektronix 7904 with all the right plugins is the only way to work on gear. I spent 4 months on mine restoring and then modded it.

So a lot of modern engineers are blinded by digital test gear.

Since I was 18 I have had a different view of electronics. I am very into materials science. I am driven in my engineering by the flow of electrons thru the circuit and materials. Crystal lattice structure of metals. I am keenly aware of how a electron FLOWS thru everything, jumping from crystal to crystal.. How they flow to nodes and sum up like water in a pipe. So a power supply is super important. This is where the flow starts. If its sloppy, a sloppy flow will occur. This is true all the way thru every circuit, trace, wire and part. Eventually this flow creates a desired flow, in the ATVX case, HDMI. The diameter of the wire, how its made, the materials ALL matter. Where each part is physically located. I looked at a PC board and think flow. PWM regulators are nasty little beasts that chop up the flow and change the chop rate based on how much needs to flow thru them. It took some creative thought to really tame these little abominations. I have a lot of physics depth and materials science. Whats a resistor made of ? how is it bonded together, what shape is it, what are the leads made of.. Chips, what does the dice look like like, how many sections on it, what material and size are the wires that lead from the dice to the pins..

They dont teach that stuff in engineering even back in the day. I was fasicanted by what was REALLY going on in the materials. I consider the components my paints and the gear I make my engineering art.
 
Last edited:
  • Like
Reactions: Davidny
Dear Xymox,

Thanks for the insights!
Regarding: "A Tektronix 7904 with all the right plugins is the only way to work on gear.". Why didn't you go for a Tektronix 7904A? Which plug ins do you use?
Well,,, the A had a different on screen display system. This is strictly a passion issue, not a performance or use thing. But i liked the built in calibrator system which was deleted in the 7904A and I think the vertical amp in the 7904 had less noise.

The "Analog-ROM" character generator in the 7904 was a marvel of analog design. So I wanted to have that.

I think I have all the modules that are useful. I had to completely recap and do a bunch of refurb on them all. The scope and modules have EVERY transistor in a socket. after 50 years ALL these contacts are wonky, so, I had to remove EVERY transistor and clean every pin and then I used Stabliant on everything. So bringing a Tek 7904/A and its plugins back to new working condition is a HUGE undertaking. I also luckily came across a new never used replacement CRT and so the CRT one my main unit is nearly brand new.

After mods it was really nice...

I mostly use the 7A22 differential input plugins because i can go down to 10uV / div.. Plus they have handy filtering.. But are bandwidth limited. But I have a 7A19 if I really need bandwidth. I also have a 7A13 which is handy for doing DC offset measurements. For horiz I use a 7B92A and a 7B87. I have the special calibration plug-ins and also have other calibration equipment. So my 7904s are spot on in calibration. But I have other modules too I use. Like the 7CT1N transistor curve tracer. A 7L5 + tracking gen spectrum analyzer. I have a lot more stuff. https://w140.com/tekwiki/wiki/7000-series_plug-ins

I also have extensive Tektronix TM500/5000 series systems and plugins. Which I plug together into dedicated systems. I have a custom power supply analyzer I made up that can produce impedance spectra measurements easily. I can even record power supply noise and then play it back to test a power supply using the types of current noise it will experience. I can also produce all manner of current waveforms directly in circuit on a operating devvice and measure results. https://w140.com/tekwiki/wiki/TM500_and_TM5000_plug-ins https://w140.com/tekwiki/images/6/66/A-3183.pdf

In doing most of what I do, its all under 1Ghz and so this gear is the ideal way to do it. Modern digital gear sux.. IMHO...
 
Last edited:
Gee, I had to study that stuff...
Nice...

Today its all just computer modeling..

What got me into materials science is when I was 16 and I decided I would use a big ga wire to make for a better connection to my phono cartridge. Less resistance. Lower capacitance. To my surprize a large ga phono lead sounded terrible. Measured great, but sounded like ****. That caused me to eventually discover there is a exact cross section of metal crystal structure to transport a specific current. Then I discovered "skin effect" which is still really interesting to me. Then tunneling. Electrons take interesting paths thru metals and are affected in interesting ways. I still have a belief that the cubic crystal of copper/silver/gold jiggles just a little each time a electron jumps from one crystal to the next. That if you have a spectra of current, not just a single unvarying value, you need different ga wires to cover that. IMHO.

In the ATVX I do something fun with the power dongle cable. First of course the wire is not normal and I choose specific dia of each strand and the number of strands. I then use teflon insulation and use 2 of these in parallel.. There is a cap network with a specific solid silver wire and then a different short cable made up of 5 teflon insulated wires each with much finer ga strands and this hooks to the main ATV board.. What is fun,, is,, when i make up that assembly of the dongle, cap network and wires to the board,, I take a special rig I have setup and pass 30A of 16V thru the wires of the whole assembly.. This 500 watts actually gets the wires hot. I have a crazy belief that this high current flow thru the wires establishes paths thru the crystal structure in the wires and kinda aligns it all. I also do some insanity high current variable frequency pulses thru it to work in the WIMA and Electrolytic caps.

So I think my analog engineering depth and my kinda odd approach to my engineering with materials science and physics alchemy is why my CRT projector, ATVX, Switch X and even the modded plasma TV and 7904 were all pretty easy to exceed what others had done. I have a different approach.

While I am talking about the Switch X here, the same concept goes on for the ATVX obviously.

 
Today its all just computer modeling..

That's a blessing and a curse.

Computer modeling is really valuable and lets anybody examine the currents at a particular point in a circuit, which is often really hard to do in an actual circuit using a current probe and a scope, of any kind. It shocks some people (bad pun) to learn just where the currents pass in any circuit. They always make a lot of assumptions, which aren't really right. These details make a big difference in the performance of any kind of circuit. That's true for any circuit much more complicated than a flashlight. Being able to probe a circuit with the click of a mouse is really valuable and a blessing.

Of course, this really is only accurate when you add the little things into the simulation that are often just ignored to make analysis simpler. Factors like parasitic R's, L's, and C's for example. The engineering assumptions and simplifications that are usually ignored in the first place often continue to get ignored even with computer simulation, where adding a part is a couple mouse clicks away. That's the curse. Simulation of an over-simplified circuit isn't really much help.

I guess I'd say that computer simulation can make an already good engineer better and give her or him more insight. My strong belief is that it's better to understand what's going on for the first, second, and third order effects of a circuit and getting those right first. Computer simulation speeds that process up and lets you do analysis that you'd otherwise skip because it's too time consuming and painful.

I've driven the bus for this thread kind of into the ditch, so I'll stop now.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu