Chord Hugo 2 Review (Portable DAC & HP Amp)

[Ken Tajalli]

For curiosity: do you know the audio latency of a modern mobo with integrated sound chip?

I would not, but I reckon it varies, from card to card.
The more processing there is the more delay there is.
And as I mentioned before, there are hardware and software buffers deliberately implemented for a smooth playback.
Sometimes you get lucky and there is a low-latency setting which one can trigger.
Sorry I can not be of much help.

 

[PeteL]

Cool, do we know how large that is roughly? (And I guess you mean that's independant of which filter is being used, ie they always have this large latency.) Have you tested latency between different DACS, what kind of ranges do we see, I think it would be an interesting inclusion if it's not a big time-suck.


Good points, I agree with your idea of relating it to screen refresh rate - I've got a 240Hz, which is a new frame every 4ms, I certainly wouldn't want DAC latency to be higher than that although I appreciate that it might be able to climb up a little from that before you start actually detecting that things like gunshots didn't correlate to the exact same time that you see the action happening on the screen........every ms of audio delay could count in some situations regarding reaction times, as we react to audio faster than vision and any extra ms added onto overall reaction time could mean the difference between a kill or a death. Actually, I would be interested in what the overall delays are in DACS. (I think there'd always be somekind of a latency & delay for audio that is an inherent part of the system on which the DAC is being used, so different DACS might have to be tested on the same system to work out comparative latencies rather than absolute latencies.....I'm not sure on this though. I'm also not sure how you would work out absolute DAC latency as you might not be able to seperate that from built-in system latency.....I'm also not sure how you would actually test latency......but if Amir were to include some latency testing I'm fairly confident he knows how he'd do it.)

OK, but even not bringing the video but just think "reaction time" the 10 ms I first stated was not pulled out of a hat, it is recognised in the recording industry for years as unnoticeable. If you strum your guitar and the sounds come 10 ms later, it's instant, it's no delay and believe me musicians will be even more picky than gamers about timing.
Think of it, in the air, sounds travel at 300 m/s. a 4 ms delay is the same thing as moving back 1.3 meters from your sound source. I know it's a bit abstract, but if a sound event happens 1.3 meter from you, would it feel like you hear it late?

 

[Robbo99999]

OK, but even not bringing the video but just think "reaction time" the 10 ms I first stated was not pulled out of a hat, it is recognised in the recording industry for years as unnoticeable. If you strum your guitar and the sounds come 10 ms later, it's instant, it's no delay and believe me musicians will be even more picky than gamers about timing.
Think of it, in the air, sounds travel at 300 m/s. a 4 ms delay is the same thing as moving back 1.3 meters from your sound source. I know it's a bit abstract, but if a sound event happens 1.3 meter from you, would it feel like you hear it late?

That's fine, but the longer the audio delay in ms in a gaming environment then the slower your reaction time will be, that can mean the difference between a kill or a death in an online multiplayer shooter. So I'm not saying it would sound unnatural and be noticeably desynced at the 34ms the other poster mentioned, just every ms makes an incremental difference to your reaction time, including the 10ms that you mentioned. I don't know what the overall and comparative latencies are of different DACS but I sure would like Amir to include that in his testing (including testing the latency of the different filters as long as it doesn't add an unwarranted length of time to the review procedure).

 

[dc655321]

Yep, that is definitely too much latency for gaming.


I also think that's gonna be too much latency for gaming (34ms) - well I wouldn't want to use it for gaming. At 34ms it might not mean you'd notice a desync but it would add 34ms onto your reaction time.

Out of curiosity, have you measured your reaction times? I’m sure I’ve seen apps for that somewhere…

I’ve seen a figure attributed to an nvidia study of 150ms for a talented (fast) gamer.

 

[nagster]

Compared the impulse response of Hugo2 with some DACs. (Apple A2049, D90SE, ADI-2 DAC FS v2, ASUS TUF Z270 mark1)
All of the PC output (USB and onboard) used ASUS TUF Z270 mark1 and ASIO4ALL v2.15 (Beta 3). The ASIO buffer size was unstable at 64, so I set it to 128. All USB outputs used the same USB port.

 

[nagster]

Split the image.
I combined them for comparison, but the characters were crushed.

 

[PeteL]

That's fine, but the longer the audio delay in ms in a gaming environment then the slower your reaction time will be, that can mean the difference between a kill or a death in an online multiplayer shooter. So I'm not saying it would sound unnatural and be noticeably desynced at the 34ms the other poster mentioned, just every ms makes an incremental difference to your reaction time, including the 10ms that you mentioned. I don't know what the overall and comparative latencies are of different DACS but I sure would like Amir to include that in his testing (including testing the latency of the different filters as long as it doesn't add an unwarranted length of time to the review procedure).

I have zero experience in gaming, but I am genuinely interested in what you mean exactly. Can you give a concrete exemple where hearing a sound event a few milisecond too late would make you die? I mean, once you hear a gunshot aimed at you, you're already dead anyway right, there's nothing you can react to once the shot is triggered? You mean reacting to the sound of someone coming out of the bushes behind you or something? Is this really what trigger a reaction from you in gaming or is it what you see on the screen? Perceptively they will happen at the same exact time, I have a hard time understanding. On the other way, if it's your own shot, I would assume it doesn't matter when you hear it. The game registered your action, it will never be based on when the sound come out of your headphones. Maybe something I just don't get. Now, about your screen 240 Hz refresh rate. I am not familiar with that neither but does the game actually generate 240 distinct frames per seconds? I would doubt so but willing to be educated. I would assume it's just some algorithm to maximise fluidity but maybe it's a new standard specific to gaming.

 

[staticV3]

Can you give a concrete exemple where hearing a sound event a few milisecond too late would make you die?

An example from CS:GO
An enemy is sneaking up on you from behind (silent). The last two meters he starts running to stab you in the back (audible footsteps).
Either you hear his steps fast enough to turn around and shoot him in the head, or you don't and die.

 

[PeteL]

An example from CS:GO
An enemy is sneaking up on you from behind (silent). The last two meters he starts running to stab you in the back (audible footsteps).
Either you hear his steps fast enough to turn around and shoot him in the head, or you don't and die.

OK I still highly doubt the difference between 4 and 10 ms would be what would Make a difference. It is pulling hair, 2 sounds 10 ms apart are perceived as exactly synchronised.

 

[bennetng]

One can always use a Realtek or PCIE gaming soundcard for games. No one force you to use an external DAC (Chord or not) for gaming.

 

[Ken Tajalli]

Split the image.
I combined them for comparison, but the characters were crushed.

First, thank you for this. Latency is not something I had thought about, but these are revealing.
On optical input standard DACs (not Chord), tend to have a very low latency as expected because optical is synchronous. Various filters, may have slightly higher latency because of extra processing they go through.
On USB output, being Asynchronous and having hardware buffers implemented, possibly even software buffers, latency gets larger. this is by default and expected.
Chord, as usual, has gone another way! The huge amounts of taps and processing means a much larger latency. Even on optical, being 35ms.
Chord uses a DPLL on optical (& Coax) and reclocks the signal using a small buffer. On USB, again the same applies and USB latencies add up to even make it larger.
Interestingly, various filters on Hugo2 only affect in-band response plus the total number of oversampling. The oversampling being 16fs or 256fs only delays the impulse a small amount, otherwise the shape of the response is the same for all filters (WTA filter shape?), unlike standard DACs such as the Topping (differing pre ring, post ring).
Topping and RME (I guess) use the same chipset, hence the similarity of filter responses.

 

[Robbo99999]

Out of curiosity, have you measured your reaction times? I’m sure I’ve seen apps for that somewhere…

I’ve seen a figure attributed to an nvidia study of 150ms for a talented (fast) gamer.

Yes, I did measure my reaction times to visual stimulus a number of years ago, I think it was around 200ms. I haven't measured my reaction time to auditory stimulus but it should be lower than visual, as our brains are wired to react faster to auditory. That's not really the point though, the point is that every ms of latency that is added onto the audio stimulus by the DAC then that is added onto your overall reaction time. Examples of the difference this makes in any multiplayer shooter - if someone starts shooting at you then you hear it, you then spin round immediately to where the sound is coming from and shoot back - if you're too slow and he's too accurate then you're dead, so each additional ms of latency can mean the difference between the game server deciding that you're dead or that you're alive based on both of the players actions. Other examples, in BF1 (which I used to play a lot and still a fair bit), if I hear someone shooting at me and I'm in the middle of reloading then evasive action is the only way to survive so I might immediately change direction and slide & dive behind any immediate cover, or just some plain old unpredictable random movements to evade the aim of the incoming bullets - but that is all stuff that I'd do instinctively and pretty much immediately and without conscious thinking the moment I get such an auditory cue.

I have zero experience in gaming, but I am genuinely interested in what you mean exactly. Can you give a concrete exemple where hearing a sound event a few milisecond too late would make you die? I mean, once you hear a gunshot aimed at you, you're already dead anyway right, there's nothing you can react to once the shot is triggered? You mean reacting to the sound of someone coming out of the bushes behind you or something? Is this really what trigger a reaction from you in gaming or is it what you see on the screen? Perceptively they will happen at the same exact time, I have a hard time understanding. On the other way, if it's your own shot, I would assume it doesn't matter when you hear it. The game registered your action, it will never be based on when the sound come out of your headphones. Maybe something I just don't get. Now, about your screen 240 Hz refresh rate. I am not familiar with that neither but does the game actually generate 240 distinct frames per seconds? I would doubt so but willing to be educated. I would assume it's just some algorithm to maximise fluidity but maybe it's a new standard specific to gaming.

Hi, I answered your questions in my reply to dc655321 above, but to add some further stuff that hasn't been covered. These auditory cues can come from behind, so you'd be reacting to stuff that is not in your field of vision. Well also you'd react in the same way if it was also in your field of vision, quite often you won't see an enemy player if he's under cover, but you would hear his gunshots at you, which emphasises where to look for the muzzle flash & tracers. You also get strong auditory cues for when you're actually being struck by bullets which again gives you more information that they're actually aimed at you and you've got to do something very quickly.

Re 240Hz, if you've got a powerful enough PC or the game is easy enough to run, then you can hit 240Hz or 200Hz or maybe lower, depending on the game, etc. But basically, the higher the refresh rate and higher the actual frames per second being generated by the game, then the more fluid & detailed the movement/information on the screen brought about through reduction of motion blur and just plain old more snapshots being generated per second of information being given to the gamer on which to base accurate & quick decisions. Visual reaction time is aided too as you'll see a player come round a corner a number of ms earlier, and because the rendition is clearer & smoother of the player coming round the corner you'll also be able to rationalise better & quicker who it is (enemy or other).....rather than with just a 60Hz screen and having two players move at each other fast around a blind corner the player might just appear in front of you in full form because the update rate is not fast enough to show his body moving smoothly around the corner to face you, as with 60Hz the previous snapshot would not show a player but then the next snapshot would show him full around the corner.......whereas with 240Hz you'd see maybe 20% of his body come round the corner, then in the next snapshot it would be 40% of his body, next snapshot 60%, next snapshot 80%, next snapshot 100% of his body is now around the corner & facing you - so it's not like he just magically appeared in front of you, which could be the case when using a 60Hz monitor, so a high Hz monitor & high fps will give you more information and prep you sooner as to what's going on which gives you time to make better & quicker decisions (as well as reducing motion blur). Have a look at following website for everything high refresh rate: https://blurbusters.com/

OK I still highly doubt the difference between 4 and 10 ms would be what would Make a difference. It is pulling hair, 2 sounds 10 ms apart are perceived as exactly synchronised.

The above post of yours Pete wasn't to me, but you might understand now why it's not about synchronisation, but instead the overall latency that is added onto reaction time, which can mean the difference between kill or death when the game server is deciding the outcome of 2 players reacting to each other in a confrontation.
EDIT: but to be clear, like I've mentioned in one of my previous posts, I don't know what the actual or comparative latencies are of different DACS out there, whether they be gaming or otherwise......I just know that theoretically lower can be better, and that I'd like to see DACS measured and compared like this if Amir would add it to his DAC reviews. To be honest, audio latency in DACS is not something I've spent time thinking about until it was brought up in this thread, as it hadn't really crossed my mind (beyond ensuring any EQ's I use in EqualiserAPO don't show a latency in the Analysis Panel of that software), but I realise that it could make some difference when it comes to gaming (DAC latency), especially when it was talked about some Chord DACs having around 1/8th second audio latency.

 

[Ken Tajalli]

OK I still highly doubt the difference between 4 and 10 ms would be what would Make a difference. It is pulling hair, 2 sounds 10 ms apart are perceived as exactly synchronised.

Sounds 10ms apart, are the peaks of a bass note at 100Hz!
No man or beast can tell them apart. Even at 50ms apart, it is like being able to clearly hear the peaks of a 20Hz note.
We are not splitting hair here, we are splitting atoms!

 

[Robbo99999]

Sounds 10ms apart, are the peaks of a bass note at 100Hz!
No man or beast can tell them apart. Even at 50ms apart, it is like being able to clearly hear the peaks of a 20Hz note.
We are not splitting hair here, we are splitting atoms!

(You missed the point (again - (another thread)), it's not about "telling apart" or synchronisation, it's about addition to overall reaction time in a game world - when viewed through the usage of a gamer, that's what we're talking about. My previous post in this thread expands on that idea. We're only talking about gaming use in relation to this point. If you don't use your DAC for "competetive" gaming, then it's not relevant to you, doesn't matter. But, when talking about the larger latencies of some Chord DACS the 1/8th second latency that Amir mentioned, then that would disqualify it for use for even casual non-competetive gaming even, due to synchronisation issues - when latency is that high.)

 

[Ken Tajalli]

(You missed the point (again - (another thread)), it's not about "telling apart" or synchronisation, it's about addition to overall reaction time in a game world - when viewed through the usage of a gamer, that's what we're talking about. My previous post in this thread expands on that idea. We're only talking about gaming use in relation to this point. If you don't use your DAC for "competetive" gaming, then it's not relevant to you, doesn't matter. But, when talking about the larger latencies of some Chord DACS the 1/8th second latency that Amir mentioned, then that would disqualify it for use for even casual non-competetive gaming even, due to synchronisation issues - when latency is that high.)

I am sorry, I was not pointing a finger at you.
I was just putting these time delays into perspective .
They are very small.

 

[BHazard]

You have to have some serious overconfidence as a company if you charge $3k and still use Micro USB.

 

[amirm]

On latency, if it is bad enough for watching video, then it is bad enough for playing games. My son stuck to an analog CRT monitor for the longest time due to latency of flat panels. And spends money on everything from network card to mouse and keyboard for fast response. When I watch him play the games, he is hitting keys 1000 times a second! Clearly timing matters to him.

 

[Ken Tajalli]

On latency, if it is bad enough for watching video, then it is bad enough for playing games. My son stuck to an analog CRT monitor for the longest time due to latency of flat panels. And spends money on everything from network card to mouse and keyboard for fast response. When I watch him play the games, he is hitting keys 1000 times a second! Clearly timing matters to him.

His finger moving at 1kHz?
Not healthy you know!
Joking aside, you are correct, in a game, where there is video, audio should not lag and go out of synch.
So what's a safe figure? is 100ms detectable?
Now we know not to use a Chord DAC in a gaming environment , even though it may look like a toy at a certain light!
But all this talk and measurements, has shone a light on the inner workings of Hugo2. The fact it uses a small buffer even on optical (and I know on coax too).
And that all four filters, look alike, have the same pre and post ringing, have the same time delays (almost) save for 16fs filters having slightly less latency.
It is all good information, to learn from.
So thanks everone.

 

[Jomungur]

I have zero experience in gaming, but I am genuinely interested in what you mean exactly. Can you give a concrete exemple where hearing a sound event a few milisecond too late would make you die? I mean, once you hear a gunshot aimed at you, you're already dead anyway right, there's nothing you can react to once the shot is triggered? You mean reacting to the sound of someone coming out of the bushes behind you or something? Is this really what trigger a reaction from you in gaming or is it what you see on the screen? Perceptively they will happen at the same exact time, I have a hard time understanding. On the other way, if it's your own shot, I would assume it doesn't matter when you hear it. The game registered your action, it will never be based on when the sound come out of your headphones. Maybe something I just don't get. Now, about your screen 240 Hz refresh rate. I am not familiar with that neither but does the game actually generate 240 distinct frames per seconds? I would doubt so but willing to be educated. I would assume it's just some algorithm to maximise fluidity but maybe it's a new standard specific to gaming.

Keep in mind in most competitive gaming is online. So the monitor latencies are additive with other latencies (e.g. mouse/keyboard input latency, which is why pro gamers usually use wired mouses, Internet lag, etc.) which can add up. Depending on how the game is engineered, it will usually prioritize synching graphics and sound.

If you are in an online fight in a 5v5 MOBA, there are hundreds of actions taking place in a few seconds. The game is usually hosted on a third party company server which is the master "clock"- it keeps track of the input/keystroke order among the 10 players. Having a slightly better latency is a big advantage for this reason alone. E.g. you make a move that is valid on your screen but your screen happens to be a couple of frames delayed from the game state of the server which received your opponent's input first, then the move is impossible from the server's perspective and you waste a move and possibly blow a cooldown.

I don't know why you would use an external DAC for a game. Many competitive players will scale down graphics and sound quality down to the bare minimum to minimize any processing lag, the exception being if there's a high graphics setting that will give you a particular advantage like detecting silhouettes or concealment.

 

[PeteL]

Keep in mind in most competitive gaming is online. So the monitor latencies are additive with other latencies (e.g. mouse/keyboard input latency, which is why pro gamers usually use wired mouses, Internet lag, etc.) which can add up. Depending on how the game is engineered, it will usually prioritize synching graphics and sound.

If you are in an online fight in a 5v5 MOBA, there are hundreds of actions taking place in a few seconds. The game is usually hosted on a third party company server which is the master "clock"- it keeps track of the input/keystroke order among the 10 players. Having a slightly better latency is a big advantage for this reason alone. E.g. you make a move that is valid on your screen but your screen happens to be a couple of frames delayed from the game state of the server which received your opponent's input first, then the move is impossible from the server's perspective and you waste a move and possibly blow a cooldown.

I don't know why you would use an external DAC for a game. Many competitive players will scale down graphics and sound quality down to the bare minimum to minimize any processing lag, the exception being if there's a high graphics setting that will give you a particular advantage like detecting silhouettes or concealment.

OK, but you are talking about key strokes input, graphic lags, and "moves". I was talking about sound. But anyway, yes, it makes sence to have the lowest latency possible for sound too but it's a choice I guess, just plug your headphones in your computer jack and not worry about those perfect multi taps filters... I find it also odd to chose a DAC based on that use case.