Q&A: Is upsampling better?

Jan 8, 2021 16:30 · 937 words · 5 minute read

Today many network players and music player software on computers offer upsampling as do many DACs.

00:07 - That raises some questions: should you use upsampling and if so, should you upsample in the player or in the DAC - provided you have that choice? Let’s start with the first question: should you upsample? CD’s - and thus CD rips - use a sampling frequency of 44. 1 kHz, meaning that the music signal is sampled - measured - 44,100 times per second.

00:43 - According to Nyquist this gives a perfect registration of music signals up to 22 kHz and this is -at least in theory - correct.

00:53 - I’ll come back to that. As a consequence of the anti aliasing filters used during recording all audio information up to 22 kHz is registered but all information above 22 kHz is lost.

01:08 - Buy the way, in practice the filtering is set to 20 kHz.

01:13 - This is of no consequence since only the hearing of young people is capable of hearing 20 kHz.

01:20 - See my video “Hearing los at age”. I know there are people that think differently but I stick to the general scientific viewpoint.

01:29 - My ears are limited to around 14 kHz and I am still able to judge audio equipment better and faster than some of my young colleagues.

01:38 - So supersonic sound can not have any play in this.

01:47 - But despite the fact Nyquist promised us perfect registration and playback of music, we often hear differences when compared to acoustic sound, higher sampling frequencies and differences between audio equipment.

02:00 - The reason for this - apart from general quality differences between equipment - is the filtering needed for digital registration.

02:08 - Both during recording and playback, filters at 20 kHz are needed and it’s these filters that have clear impact on the sound.

02:17 - Watch my video “The truth about Nyquist and why 192 kHz does make sense” for more information.

02:28 - Now why would we want to upsample? All information above 20 kHz has been filtered out and won’t be regained by upsampling.

02:37 - In fact, the upsampled file will contain exactly the same information the 44. 1 kHz file has.

02:45 - But upsampling sounds better, you might say.

02:48 - And you are right. This again has to do with the filtering, in this case the reconstruction filter used in the DAC.

02:55 - That filter has an easier job with the higher the sampling frequency.

03:00 - Again, full explanation in my video “The truth about Nyquist and why 192 kHz does make sense”.

03:08 - These filters in true high-end DACs - as in expensive - are of very high quality and there upsampling often has limited to no effect on the sound quality.

03:20 - But in more affordable DACs these filters do have influence on the sound quality: the lower the sampling rate the lower the sound quality.

03:29 - In that case upsampling to the highest sampling frequency will make the job for the reconstruction filter in the DAC a lot easier, leading to a better sound.

03:39 - But only when the upsampling is done with high precision.

03:42 - The advantage of doing it in a computer or network player is that these have far more computational power than DAC chips have.

03:52 - Therefore the upsampling can be done better and thus lead to better sound quality.

03:59 - But what if your DAC can also upsample? Should you still use upsampling in the computer or network player? In general I would say yes but there are also DAC’s that offer high quality upsampling, usually the ones that do not use standard DAC chips or that use standard DAC chips but have the upsampling done by a separate processor in the DAC.

04:23 - The only way to find out is to try. Just select the best sounding option and if you don’t hear any difference, choose the one that’s the most convenient to you.

04:33 - One tip: always set the upsampling to do only integer upsampling, so 2 times, 4 times and so on.

04:42 - So 44. 1 kHz should be upsampled to 88. 2, 176. 4 and so on.

04:50 - 48 kHz should only be upsampled to 96, 192 and so on.

04:57 - If you, for instance, upsample 44. 1 kHz to 196 kHz, this requires more computational power while rounding will can occur.

05:07 - Which brings us to the end of this video. This is one of the Q&A videos I have made over time.

05:14 - Goto the Q&A playlist for answers on more questions.

05:18 - I also post a new video every Friday at 5 pm central European time.

05:22 - If you don’t want to miss that, subscribe to this channel or follow me on the social media so you’ll be informed when new videos are out.

05:30 - You can help me reach even more people by clicking the thumbs-up icon and mentioning my work on the social media.

05:38 - Many thanks to those viewers that support this channel financially, it keeps me independent and thus trustworthy.

05:43 - If that makes you feel like supporting my work too, the links are in the comments below this video on Youtube.

05:50 - I am Hans Beekhuyzen, thank you for watching and see you in the next show or on theHBproject. com.

05:56 - And whatever you do, enjoy the music. .