Jabra Elite 8 Active Gen 2 vs Elite 10 Gen 2

Nov 09, 2024 12:39 PM By WhizzWr

19

Thank you Jasper for the extensive explanation. 😊 I get your point, on listening the pink noise recording I can hear less hiss on the Elite 8 compared to Elite 10. I suppose this corresponds to the 3dB differences you mentioned in term of SpNR. Regarding the inclusion of noise gate kicks time, maybe rather than timing the noise gate, you can average the Amplitude differences over a time series. Or is it already done like that? I can only see frequency-domain plot on the linked page, so I’m not sure. If it is already like that, then the score should be (mostly) representative of the recording. I believe what most people would like to see, is not only the SnR, but also is the intelligibility of overall speech, and this seems to be the goal of ITU-T P.OLQA (referenced also in the linked page, hope in the future we can have that 😉). I would keep the score, because it is what it is based on the current methodology, but I imagine it would br helpful to the reader to mention that Elite 10 has slightly faster noise gate, so on this speech test, and possibly in real life, there is less chance that your voice got swallowed by noise when it kicks in. Ignoring the intelligibility on 0:22, I think the differences of 0.2 is sensible. I hear less hissing with the Elite 8, but the voice are reasonably intelligible with both Elite 8 and Elite 10. This is just my opinion, I’m not an audio/RF engineer, just I have a bit basic in signal processing. Cheers.

Hey again,

I’ve shared your suggestions with the test developers. I think it’s a good idea to measure the amplitude over a period of time, but we don’t at the moment. We’re definitely looking into upgrading the mic and noise handling testing down the line, and P.OLQA is definitely a standard for good reason.

Back to the two reviews: I’ll add a line about the noise gate shortly.

Thanks again for your input!

Thank you Jasper for the extensive explanation. 😊

I get your point, on listening the pink noise recording I can hear less hiss on the Elite 8 compared to Elite 10. I suppose this corresponds to the 3dB differences you mentioned in term of SpNR.

Regarding the inclusion of noise gate kicks time, maybe rather than timing the noise gate, you can average the Amplitude differences over a time series. Or is it already done like that? I can only see frequency-domain plot on the linked page, so I’m not sure.

If it is already like that, then the score should be (mostly) representative of the recording. I believe what most people would like to see, is not only the SnR, but also is the intelligibility of overall speech, and this seems to be the goal of ITU-T P.OLQA (referenced also in the linked page, hope in the future we can have that 😉).

I would keep the score, because it is what it is based on the current methodology, but I imagine it would br helpful to the reader to mention that Elite 10 has slightly faster noise gate, so on this speech test, and possibly in real life, there is less chance that your voice got swallowed by noise when it kicks in.

Ignoring the intelligibility on 0:22, I think the differences of 0.2 is sensible. I hear less hissing with the Elite 8, but the voice are reasonably intelligible with both Elite 8 and Elite 10.

This is just my opinion, I’m not an audio/RF engineer, just I have a bit basic in signal processing.

Cheers.

Nov 05, 2024 08:19 PM By RtingsUser8027768

1

I agree with your observation. The noise gate on the Elite 10 Gen 2 kicked in earlier hence providing a clearer speech at that 0:20 mark. I am also confused as to why the Elite 8 score higher in this category.

Hey, I’ve also replied to WhizzWr’s comment about this, but I’ll just quickly mention that the score is based on which pair cuts out more noise. In this case, the Elite 8 Active Gen 2 cuts out around 3 dB more noise. One of the limitations of the Noise Handling test currently is that it doesn’t account for the speed of noise gates. Part of this is because some mics use their microphone capsule pattern to reject offset noise, which can’t be timed, for example, while with others that use a software-based noise gate, it can be hard to detect when the gate has kicked in (either because it’s subtle or it’s extremely fast).

Ultimately, a 0.2 difference in the overall score for Noise Handling is small enough that, ideally, people will listen and make a choice based on what they hear. As reflected in the scores, we don’t consider either of these mic systems and their noise handling to be significantly better or worse than each other. I’m not certain how we would approach doing this, but do you think we should also score for the speed of noise gates? Thanks!

Nov 02, 2024 05:30 AM By WhizzWr

19

I don’t fully get the scoring result of “Speech + Subway Noise Audio Sample”. To me they are very similar, in which aspect and timestamp does the Elite 8 gen2 beat the Elite 10? If anything around 0:20 (the part where the recording said “THIS is……”) the noise gate on the Elite 10 kicked sooner, and therfore you get to hear more intelligible speech.

Hey there, So, the Noise Handling test is scored using objective measurements that account for the speech to noise ratio (spNR), but it doesn’t score the noise gate’s delay. Generally, some of the challenges of scoring based on the delay of a noise gate are with some mics it’s rather difficult to notice a noise gate kicking in because it’s not obvious, or they don’t have a software-based method of rejecting noise. As for how the scores work out in this case, the Elite 8 Active Gen 2’s spNR is 12.54 dB and the Elite 10 Gen 2 spNR’s is 9.29 dB, and the higher number scores higher because cutting out an extra 3 dB of noise is significant.

We hope that the demos serve to encourage people to also make decisions with their ears, too. I agree with you that both microphone systems sound similar and attenuate noise well, and an argument can be made that the Elite 10 Gen 2 has a slightly faster noise gate, even if it technically attenuates roughly 3 dB less noise in our tests. Do you think this distinction should be clearer in the text for the Elite 8 Active Gen 2? Thanks!

Nov 02, 2024 05:30 AM By WhizzWr

19

I don’t fully get the scoring result of “Speech + Subway Noise Audio Sample”. To me they are very similar, in which aspect and timestamp does the Elite 8 gen2 beat the Elite 10? If anything around 0:20 (the part where the recording said “THIS is……”) the noise gate on the Elite 10 kicked sooner, and therfore you get to hear more intelligible speech.

I agree with your observation. The noise gate on the Elite 10 Gen 2 kicked in earlier hence providing a clearer speech at that 0:20 mark. I am also confused as to why the Elite 8 score higher in this category.

I don’t fully get the scoring result of “Speech + Subway Noise Audio Sample”.

To me they are very similar, in which aspect and timestamp does the Elite 8 gen2 beat the Elite 10?

If anything around 0:20 (the part where the recording said “THIS is……”) the noise gate on the Elite 10 kicked sooner, and therfore you get to hear more intelligible speech.