First thank you for doing this. As I am ATMOS mixing I and troubled by all the voodoo, and questions that can't be clearly answered. So what you have done is part of a test i wanted to do. I have heard "Don't use the bed use all objects" because it sounds better. Yet nobody can tell me how or why. So some of this you have answered with the first null test. ATMOS was designed for movie theaters and seems to make perfect sense for that use. But the high freq roll off suggested for ATMOS mixing seems to not make sense for music . So my question is if most of my clients will be listening on headphone, does the binaural render roll off high frequency? If not why roll of high freq for music mixing on speakers. I believe the high end roll off is only to make the experience more even in a large movie theater. Dolby seems unclear on many of these questions and I believe its because it matters more with binaural music mixing but for film post it matters less. ( if its used for helicopters flying over) Thanks again.
The binaural hrtf does have to play with high frequencies in order to replicate how humans determine where a sound is and where it is moving to. This is an issue for music I agree where continual shifting in tone doesn't translate well. There is a community of thought that all binaural processing would serve music better done by the engineer using the techniques of immersive mixing rather than leave it to the renderer. The problem though is the ADM for Dolby holds the audio source for both speaker and binaural so e.g. you cannot set binaural render mode off and apply your own because that will then end up in the speaker playback. Dolby Atmos as you say is a cinema technology so there will be limitations when it comes to having control of the actual sound because you have to give up quite a lot of sonic control to Dolby and Apple unfortunately. Of course there s nothing stopping you making binaural mixes with or without Dolby and not publishing via ADM simple 2.0.
@@balladeerstudio I've replied 3 times and it keeps disappearing. Try a search for Dolby ATMOS music studio best practices. You should find a document.
Great video! In my experience mixing atmos I've used the center channel to make a source sound more upfront, both in the speakers and binaural. This is incredibly useful with say an acoustic and lead vocal. The vocal can feel more forward than the acoustic, but they are both coming from the same place. I've found the center channel vs. phantom center to be way more apparent with Apple's Spatial Processing, with something in the phantom center feeling way further away than something in the center channel. I've noticed this especially with the atmos mixes of The 1975 albums that I believe were mixed by Andrew Scheps. The mixes that I feel translate the best to binaural (Dolby or Apple) all make great use of the center channel.
Binaural audio is supposed to be experienced externally from the head, so I am a little confused as to why it's heard in the centre of my head just as a stereo signal would be perceived. Interesting nonetheless.
Interesting! When you did the final comparison by subtracting the two channels, did you check to see if the difference we heard was mealy due to a flat amplitude difference? Thanks for your contributions!
OMG, I didn't! I assumed they were the same level. In fact, the max peak on the OBed recording is 1.259 dB louder than the object recording. I guess because it's two objects instead of one? When I compensated the gain correctly, it made the null test quieter, but still audible. The max peak on the corrected null test is exactly 4dB quieter than the max peak on the uncorrected null test. I wish I had caught that when I made the video. But still, the two versions are acoustically different. I also hear a slight difference. Thanks for pointing that out!
@@balladeerstudioGreat video, Did you check phase issues also in the null test? Maybe the Stereo Obed is adding some time, also it will be great to trying with the binaural settings in off position. Thanks!!!
@@genomamastering4777 Thanks for that comment. As I see it, the issue is that a stereo "bed" will be rendered binaurally as two channels, so inevitably there will be phase differences due to the algorithm. I don't see how to correct that away. I agree it would be useful to test with binaural settings "off", but I have heard that defeats the binaural algorithm which would, if true, not answer the question. I've thought of a better test.
@@balladeerstudio Thank you for your answer!!!. I been mixing music in Atmos for the last 2 years and this issue you mention in this video is very challenging when you use headphones and the binaural settings affects that, when I mix with speakers this issue is not so important. My main method now is to use OBED and I only use the bed for the LFE, tha´s allow me to make different Obeds with different Binaural settings and close or open the immersive field or only some parts of it. Thanks for your work.
Binaural rendering is done by convolving(filtering) HRTFs(Head Related Transfer Functions), but even in the academic field, right HRTFs haven't been got, especially front localization is really bad. Everyone knows it. It is hard to understand why Dolby Labs hastened to put Spatial Audio( for headphone) into commercial use.
Because it is a hybrid technology: ambisonic principles and stereo. In other words you hear the family stereo "in your head" sound AND ambisonic spherical harmonics.
Your question has a false premise and a misunderstanding: a phantom center only applies to sound over speakers and binaural only applies to headphones. So a phantom center (by definition this is a sound-source which appears in front of the listener) is not heard in headphones and binaural played over speakers is improper because you cannot introduce speaker cross-talk into a binaural signal. Also, you are actually only rendering 2.0 stereo in your video, not binaural. True binaural must have a spherical harmonics, HRTF applied or you simply have 2.0 stereo.
First thank you for doing this. As I am ATMOS mixing I and troubled by all the voodoo, and questions that can't be clearly answered. So what you have done is part of a test i wanted to do. I have heard "Don't use the bed use all objects" because it sounds better. Yet nobody can tell me how or why. So some of this you have answered with the first null test. ATMOS was designed for movie theaters and seems to make perfect sense for that use. But the high freq roll off suggested for ATMOS mixing seems to not make sense for music . So my question is if most of my clients will be listening on headphone, does the binaural render roll off high frequency? If not why roll of high freq for music mixing on speakers. I believe the high end roll off is only to make the experience more even in a large movie theater. Dolby seems unclear on many of these questions and I believe its because it matters more with binaural music mixing but for film post it matters less. ( if its used for helicopters flying over) Thanks again.
I'm glad I'm not the only one who has questions, ha! I don't remember hearing that about rolling off high frequencies. Where did you see that?
The binaural hrtf does have to play with high frequencies in order to replicate how humans determine where a sound is and where it is moving to. This is an issue for music I agree where continual shifting in tone doesn't translate well. There is a community of thought that all binaural processing would serve music better done by the engineer using the techniques of immersive mixing rather than leave it to the renderer. The problem though is the ADM for Dolby holds the audio source for both speaker and binaural so e.g. you cannot set binaural render mode off and apply your own because that will then end up in the speaker playback. Dolby Atmos as you say is a cinema technology so there will be limitations when it comes to having control of the actual sound because you have to give up quite a lot of sonic control to Dolby and Apple unfortunately. Of course there s nothing stopping you making binaural mixes with or without Dolby and not publishing via ADM simple 2.0.
@@balladeerstudio I've replied 3 times and it keeps disappearing. Try a search for Dolby ATMOS music studio best practices. You should find a document.
Great video! In my experience mixing atmos I've used the center channel to make a source sound more upfront, both in the speakers and binaural. This is incredibly useful with say an acoustic and lead vocal. The vocal can feel more forward than the acoustic, but they are both coming from the same place. I've found the center channel vs. phantom center to be way more apparent with Apple's Spatial Processing, with something in the phantom center feeling way further away than something in the center channel. I've noticed this especially with the atmos mixes of The 1975 albums that I believe were mixed by Andrew Scheps. The mixes that I feel translate the best to binaural (Dolby or Apple) all make great use of the center channel.
Binaural audio is supposed to be experienced externally from the head, so I am a little confused as to why it's heard in the centre of my head just as a stereo signal would be perceived.
Interesting nonetheless.
Interesting! When you did the final comparison by subtracting the two channels, did you check to see if the difference we heard was mealy due to a flat amplitude difference?
Thanks for your contributions!
OMG, I didn't! I assumed they were the same level. In fact, the max peak on the OBed recording is 1.259 dB louder than the object recording. I guess because it's two objects instead of one?
When I compensated the gain correctly, it made the null test quieter, but still audible. The max peak on the corrected null test is exactly 4dB quieter than the max peak on the uncorrected null test. I wish I had caught that when I made the video. But still, the two versions are acoustically different. I also hear a slight difference.
Thanks for pointing that out!
Correction video: It still shows a significant difference. ua-cam.com/video/F-iTia9xN94/v-deo.html
@@balladeerstudioGreat video, Did you check phase issues also in the null test? Maybe the Stereo Obed is adding some time, also it will be great to trying with the binaural settings in off position. Thanks!!!
@@genomamastering4777 Thanks for that comment. As I see it, the issue is that a stereo "bed" will be rendered binaurally as two channels, so inevitably there will be phase differences due to the algorithm. I don't see how to correct that away. I agree it would be useful to test with binaural settings "off", but I have heard that defeats the binaural algorithm which would, if true, not answer the question. I've thought of a better test.
@@balladeerstudio Thank you for your answer!!!. I been mixing music in Atmos for the last 2 years and this issue you mention in this video is very challenging when you use headphones and the binaural settings affects that, when I mix with speakers this issue is not so important. My main method now is to use OBED and I only use the bed for the LFE, tha´s allow me to make different Obeds with different Binaural settings and close or open the immersive field or only some parts of it. Thanks for your work.
Binaural rendering is done by convolving(filtering) HRTFs(Head Related Transfer Functions), but even in the academic field, right HRTFs haven't been got, especially front localization is really bad. Everyone knows it. It is hard to understand why Dolby Labs hastened to put Spatial Audio( for headphone) into commercial use.
Because it is a hybrid technology: ambisonic principles and stereo. In other words you hear the family stereo "in your head" sound AND ambisonic spherical harmonics.
Thanks for this wonderful Video :)
Your question has a false premise and a misunderstanding: a phantom center only applies to sound over speakers and binaural only applies to headphones. So a phantom center (by definition this is a sound-source which appears in front of the listener) is not heard in headphones and binaural played over speakers is improper because you cannot introduce speaker cross-talk into a binaural signal.
Also, you are actually only rendering 2.0 stereo in your video, not binaural. True binaural must have a spherical harmonics, HRTF applied or you simply have 2.0 stereo.
Binaural not should be used......is really terrible, HRTF not is ok.