Skip to content
HearingTracker
Log in

Phonak Spheric mode has long processing delay

This topic is closed.

34 replies1.6k viewsLast reply

Jump
Link to this topic

While Sphere have received glowing reviews from many on this forum, a few people didn’t like them at all.

It seems that a consensus emerged that with open fittings, Sphere mode can only do limited noise suppression, and therefore some wearers might be disappointed.

I’m sure this is true, but there is more to it. I’ve always wondered whether the chip adds to the processing delay. Since I got Spheres yesterday, I was able to measure:

And boy, does AI ever add to the delay!

The screenshots show what happens when I press a (dog-training-) clicker to generate a sharp sound. The orange signal is the original signal (clicker sound picked up by the measuring microphone) and blue is the output of the hearing aid.

Without AI: about 6ms delay. (Similar to other brands I measured)

Screenshot from 2025-08-20 16-21-51

In Spheric mode: 15ms delay!

Screenshot from 2025-08-20 16-12-20

Most studies regarding processing delay seem to find that once the delay in an open fitting exceeds 8-10ms, it will be detrimental to speech comprehension.

So there we have it. Don’t use Spheres with open fittings. Maybe try ActiveVent receivers.

16 people in this conversation9 min readJump to latest

It would be very interesting to see the same test on the other ’s like the vivia and others.

Indeed! I don’t own these though.

The equipment I used for this test only costs about $10, so maybe somebody else can step in. I will post more when I have time.

When the EP30s came out, my Audiologist told me that she was going to get training from ReSound. I asked her to question the ReSound rep about delay on Vivia / EP30. The answer was ‘it’s negligible’.

I’m sure Phonak would give the same answer :wink:

Edited

There must be a misunderstanding on your part - nobody claims that it does.

The difference in processing delay comes from turning the feature on or off.

[EDIT]: @hearingaidobserver1 , now I understand your question. You are referring to me mentioning ActiveVent receivers. These receivers will provide a closed fitting in noise and an open fitting in quiet situation.

With a closed fitting latency doesn’t matter so much because the delayed sound from the doesn’t mix with natural sound from the outside world. That’s why they are probably a good solution for people who would normally use an open fitting, but still want to have the benefits of AI noise suppression.

In reply to @hearingaidobserver1

If you want processing delay not to be detrimental for speech, you must avoid interference between sound processed by the hearing aid and direct sound going through the ear canal entrance.

That avoidance could be done by more closed couplings, the closer, the better, although you have to find balance with pinna effect, low frequency hearing, feedback threshold, etc.

Edited

Thank you for posting this. Super cool! I like analysis like this!
One reason the delay can sound unpleasant with open fittings, especially for those with good low-frequency hearing, is because of the comb effect. This video, at 7:20min, explains the comb effect. I’ve experienced it with the Rextons I tested (with vented domes): it sounded metallic, like being in a tunnel, very unnatural.

Widex Moment Hearing Aid Review | ZeroDelay Signal Processing23

In my experience, a delay often causes a robotic-like or weird sound to both speech and background. However, given all the sound is mushed up an processed with Spheric mode anyway, it never really bothered me.

In reply to @OcramSagev

I add about comb effect here:

High Frequency Sound Issues
I suffer from high frequency hearing loss, as well as certain indications of cognitive impairment. To help with both issues I’ve chosen Widex SmartRIC 440 hearing aids. I continue to struggle with my “audi” to eliminate high pitched sound from going over into sibilance, however. High pitched/powerful vocals & relatively high pitched broadcast speech are primary culprits. I’ve demo’d Oticon Intents & experienced very similar results, so I can believe that the problem might be something generic li…

.
.

High Frequency Sound Issues
Wait a moment: have you ever tried streaming before, with your phone? If so, was sibilance a problem then? Or was that issue a problem only when listening with hearing aid microphones? If the latter, the problem is probably mainly with feedback or oscillation, NOT THE PHONE. Found another useful link, someone on the forum provided it: [Screenshot_20250405_122924_Chrome] Source: To be sure, try it on another phone with the same audio samples that previously caused sibilance. Try to GEN…

.
.

High Frequency Sound Issues
It’s quite possible, because when the difference of 3k Hz and up is less than 10 dB between amplified and “through vent/leaks” sound, the risk of comb filtering is increased. There was also a risk of imbalance between medium- and high-frequency sounds, causing sibilance. As you can see in the last slide, “flutter” occurs when the vent and amplified path interact if there is a small decibel difference. Oh there, there he was “chastising” there laughing It was just an insight. innocent …

Interesting analysis. I’ve been using the Sphere’s for almost a year and I honestly don’t notice a delay when in the Spheric program. Bluetooth latency is actually higher than your findings for the Spheric AI processing and I do sometimes notice a very slight disconnect between the sound and mouth movements on videos when streaming.

I’m actually happy to trade a bit of latency for better speech recognition at noisy restaurants.

Jordan

That’s a really useful comparison.

I haven’t seen this anywhere in the reviews: given that the majority of newer users are likely to have milder losses, this is clearly a big red flag.

In reply to @JordanK

It works in your case given that you hear the large majority of the sound directly through the aids. The problem we’ve seen elsewhere occurs with more open vents and sound arriving from the outside vs the processed sound arriving 16uSec later.

But if the same problem is with Roger devices, regarding they have a 19 ms delay?

@JordanK, if the processed sound is over 10 louder than direct sound, it is highly probable that you don’t perceive distortions (as a rule of thumb).

@Bimodal_user said:

But if the same problem is with Roger devices, regarding they have a 19 ms delay?

There will be a big intensity difference, especially if the sound source is further away. If the sound is from the immediate sound field, you’d likely not be using the Roger device or if it wasn’t providing a benefit you could turn it off.

Edited

I’m new to Sphere , and am having an issue that my provider can’t really change. Using iPhone 12 with 18.6.1, I when texting, I often hear a hissing sound, as well as a strong increase in the volume of my voice, for the duration of my outbound, or the inbound transmission. My SonicCare toothbrush causes the same sounds. My hearing loss is high frequency, profound, and I have custom molds. The provider has been online with Phonak several times and made the suggested changes in my presence. No real improvement. Any advice? Thanks.
EDIT: I didn’t mention that a real PITA for me occurs when playing my acoustic guitar. I’m beginning to think that most of my symptoms are actually the units going into Spheric mode. Regardless, when that change happens, I find it unacceptable performance for units with this price point. How about you guys?

Does anyone know what the delay with LE audio and aurocast is? I@m assuming it’s not great but don’t have any figures on it. I do notice a significant difference when using iPhone low energy blu tooth compared to standard. For me this is very noticeable because each time I touch my phone, it reads back to me with the screen reader If I use Phonak, it takes ages to respond. If I use something with low energy blu tooth, it’s like using airpods, it feels almost instant, a bit like using the iPhone without an accessory.

Edited

latency depends on several variables: codec used , distance , Bluetooth version , interference, etc.

can go as low as 50ms in ideal conditions.

BT classic can go as low as 100ms (SBC,AAC) or may be as low as 50 with Qualcomms aptX. Apple uses AAC for example. These numbers are to give you an idea of the order of magnitude.

The delay you describe “Phonak , take ages to respond “ also includes phone performance , how the software was written, how the OS prioritizes tasks, etc — not just the BT latency

Hopefully I explained this in an easy way to understand .

This is very interesting and I am not terribly impressed with Phonak Sphere, I didn’t realise that there was a processing delay and 15 minutes is a long time and I thought the chip was supposed to help things along?

Related discussions