Jim Alfredson

  • Content count

  • Joined

  • Last visited

Posts posted by Jim Alfredson

  1. At the least, pretty much every synth needs the electrolytic capacitors replaced, as they only have a lifespan of about 25 years or so. The power supply electrolytics can be especially dangerous as they might fail and cause voltage problems which can fry sensitive components downstream. This happened to a Yamaha CS50 I bought for cheap. The original owner shoved it in a closet for 30 years, then took it out one day and plugged it in and the dried out electrolytics in the power supply caused a voltage surge which blew out pretty much every proprietary Yamaha chip in the thing. Not fun.

    There's also calibration to get the synth back to factory spec. Some also contain foam that disintegrates over time that must be removed. And many suffer from problems with the keys / action. 

  2. Actually, when it comes to playing music, I do my best to enter a 'zone' where objectivity isn't present. So I take issue with the description that my music is somehow informed by a cold, calculated, scientific approach. I never studied music formally (I'm self-taught), I don't know much music theory, and can't even read music, so...

    But one of the reason I love recording is that it combines art and science. All I was trying to tell Allen is that the difference between 16bit and 24bit as a final delivery product is increased file size and that's it. As someone infinitely curious about the world, I find such claims interesting and they become something I obsess over for a little while, reading everything I can about the subject and performing experiments (like null tests and double-blind listening tests). I had hoped he would use that claim as a starting point for his own exploration of the science of sound. If one is really interested in how digital audio works, I don't see how learning something new is a threat.

    Allen says he can hear a difference. I never said he couldn't and would never presume to tell someone what they can and cannot hear. All I said is that our hearing is wholly subject to our brain and that it is very easy to trick ourselves. The only way to know for sure is to do a double-blind test. Otherwise there are too many biases to make an objective decision.

    But obviously he can do whatever he wants. I don't think it's worth leaving the community over.

  3. I'm just trying to spread some knowledge about how things work. Digital audio is based on mathematics and physics. There's no woo-woo there. Of course Allen and anyone else is free to do whatever they want, but if you're paying a lot of extra money for 24bit audio as the final product from places like HD Tracks and the like, you are paying for snake-oil. The only exception to that rule is if they mastered the 24bit tracks better, but that has nothing to do with 24bit as a format.

    As I've said repeatedly, recording at 24bit is smart because it gives you a lot of leeway with your gain staging and is better for internal processing (most DAWs process audio at 64bit these days). But for the final delivery product, it is useless.

    I did a blind test between the CD of David Bowie's last album, Blackstar, the 24bit files downloaded from HD Tracks, and the vinyl. The vinyl was easy to pick out due to the surface noise and the muted transients. The 16bit and 24bit, however, sounded identical. Neither me nor my friend were able to consistently pick one or the other. Of course, they could've made those 24bit files from the 16bit master, who knows? But I've done similar tests with true 24bit files. My engineer friend who I mentioned above claimed that 16bit does something to the low-end, tightening it in a way that's different than 24bit. Sorry, but it's just not there. If there was a difference, then a null test like the one from the video I posted earlier would reveal it. And it doesn't.

    Does any of this matter? Well, I personally like to be knowledgeable about things, for example knowing my own limitations when it comes to how my sensory systems work and how I can fool myself. I like to be wrong about things because it means I can learn something. I also like to be protected against someone taking advantage of me. 

    And we haven't even broached the high sample rate question yet (which can actually be making your audio worse). 96kHz is pointless. Anything above 48kHz is pretty much pointless. But that's another subject altogether.

  4. I've been recording since I was a young kid. I started on four-track cassette when I was 8 years old, moved to four-track reel-to-reel in my teens, then to ADAT digital audio tape, then to stand-alone harddisk recording and finally to computer-based DAWs by my mid 20s. I've recorded in iconic studios in LA, NYC, and Chicago. I've attended AES shows and hung with heavy-weights in the industry as well. I've studied under a world-class engineer that just happens to be local and as much as I love him as a friend and a teacher, I can honestly say that his biases have been revealed by double-blind tests as well. It's just human nature.

    The change from 16bit to 24bit only affects how much dynamic range there is. That's all. In fact, the dynamic range in a true 24bit system is so large that no current electronic device can adequately reproduce it and thus there really is no such thing as a true 24bit analog to digital converter. They are actually around 20bits in practice.

    It would be really easy to find out if your 24bit system is better than the same system at 16bits. Record something at 24bits, then down-convert it to 16bit, bring it back into your DAW, flip the phase on one of them, and do a null test. If you hear anything other than noise, I would be extremely surprised.

    An even better way to do it would be to have two of the same interfaces connected to two separate computers and record the same source in both 24bit and 16bit at the same time via a mic splitter. I could actually do that. It would be an interesting experiment. In fact, that would make a fascinating YouTube video.

    I don't think anyone here is calling you a schmuck, Allen. 

  5. No, I meant confirmation bias. From wiki: 

    " Confirmation bias, also called confirmatory bias or myside bias,[Note 1] is the tendency to search for, interpret, favor, and recall information in a way that confirms one's preexisting beliefs or hypotheses, while giving disproportionately less consideration to alternative possibilities.[1] It is a type of cognitive bias and a systematic error of inductive reasoning. People display this bias when they gather or remember information selectively, or when they interpret it in a biased way. The effect is stronger for emotionally charged issues and for deeply entrenched beliefs. People also tend to interpret ambiguous evidence as supporting their existing position. Biased search, interpretation and memory have been invoked to explain attitude polarization (when a disagreement becomes more extreme even though the different parties are exposed to the same evidence), belief perseverance (when beliefs persist after the evidence for them is shown to be false), the irrational primacy effect (a greater reliance on information encountered early in a series) and illusory correlation (when people falsely perceive an association between two events or situations)."

  6. 14 hours ago, AllenLowe said:

    though Jim, I will say I have A/B'd on 2 track digital, 16 vs 24 and heard the improvement. It may be for the reasons you say, but in a way that is the point; if 24 bit is not as fussy about levels and dynamic range, that tells me it is a better and more accurate medium. Am I missing something here?

    Unless it was a double-blind test, I would be wary of drawing any conclusions. Confirmation bias is very real and very powerful. If you know you're listening to 24bit, and you believe 24bit is better, then 24bit will sound better to you.

    The extra bits are just for low volume information. Essentially you are lowering the noise floor and pushing any quantization levels way way down there. For tracking, it's great (though with careful preparation and gain staging, 16bit sounds just as good). For the final product, there's no perceivable benefit. 

    Here's a great video outlining the difference between 24bit and 16bit. This guy does a null-test (phase reversing a 24bit track and a 16bit version of the same track made from the original 24bit file). The result is nothing but noise that is -83db (ie, not even audible...the natural noise of your room is far louder than that).


  7. It depends. For multi-tracking and using any kind of in-the-box processing, 24bit is better than 16bit for the reasons outlined above, mainly as a safeguard against bad gain staging. But if you were recording something say in stereo with no overdubs and very little post-processing and you set your levels correctly, 16bit would be fine. But most everything today is designed to record and sound best at 24bit (again, the analog side makes the biggest difference) and storage costs are low so why not use it?

  8. 5 hours ago, AllenLowe said:

    Jim, I'm a little bit shocked. Are you saying you don't think 24-bit as the original source recording is better than 16 bit? 


    24bit is only useful in the recording process, not for the final product. The only difference between 24bit and 16bit is the dynamic range. The dynamic range of 16bit is nominally 96db which is already far more than any final product needs or uses.

    The reason to use 24bit in the recording process is because it allows you to fudge your gain staging and not worry about it. So yes, I record at 24bit because it allows me to not necessarily worry about my gain staging when recording dynamic music like jazz or classical. I shoot for about -12dbfs (about halfway up most digital meters) as my peaks and even if the signal is much quieter, the noise floor of 24bit is so low that you can increase the gain after the fact and it won't suffer from quantization errors, aliasing, etc.

    But for the final mastered and delivered product, 16bit is more than enough. 

    Here's a great article on why and it even touches upon the subject of the usefulness of dither.


    From the article:

    So, 24bit does add more 'resolution' compared to 16bit but this added resolution doesn't mean higher quality, it just means we can encode a larger dynamic range. This is the misunderstanding made by many. There are no extra magical properties, nothing which the science does not understand or cannot measure. The only difference between 16bit and 24bit is 48dB of dynamic range (8bits x 6dB = 48dB) and nothing else. This is not a question for interpretation or opinion, it is the provable, undisputed logical mathematics which underpins the very existence of digital audio.

    So, can you actually hear any benefits of the larger (48dB) dynamic range offered by 24bit? Unfortunately, no you can't. The entire dynamic range of some types of music is sometimes less than 12dB. The recordings with the largest dynamic range tend to be symphony orchestra recordings but even these virtually never have a dynamic range greater than about 60dB. All of these are well inside the 96dB range of the humble CD. What is more, modern dithering techniques (see 3 below), perceptually enhance the dynamic range of CD by moving the quantisation noise out of the frequency band where our hearing is most sensitive. This gives a percievable dynamic range for CD up to 120dB (150dB in certain frequency bands).

    5 hours ago, AllenLowe said:

     I can hear the difference even in consumer recorders, geez it's not even subtle. 

    Again, I'm talking about the final product, not the recording process. By all means, record at 24bit. Many times it will sound better but that's because you don't have to worry about the gain. Case in point, I recently did an audio/video mix of a classical percussion duo from Michigan State. Unfortunately they recorded it themselves using a Zoom audio recorder / camera and some other GoPro style cameras. The Zoom was their main audio recorder and it was set to 16 bits. When they set the recording levels, they did so using the loudest part of the piece as their guide. This was fine for that particular movement, but unfortunately the first movement of the piece was orders of magnitude softer. And because they recorded it at 16bit, which does not have the extended dynamic range of 24bit (which is really just extended detail at very low volumes), when I gained up the audio there was all sorts of nasty quantization errors in it. I added some dither but the damage was done. If they had recorded it at 24bit, it would've been fine. Of if they had been more careful with their gain staging, it would've been fine.

    Also, consumer devices don't exactly have the best analog front ends. Most conversion chips are the same and don't have a sound. It's the analog electronics in front of the ADC and after the DACs that make the difference. 


  9. RIP Victor Bailey. I had the pleasure of meeting and hanging with him at the 2003 Park City Jazz Festival (where I played with Corey Christiansen and Danny Gottlieb). He was as nice as can be to this young and nervous musician and told me he loved Hammond organ. He also told me something I never knew; that his uncle was the great drummer Donald 'Duck' Bailey, who played and recorded with Jimmy Smith on a handful of classic Blue Note albums. He said that as a young boy, the trio of Jimmy, Donald, and Quentin Warren used to practice at his house in Philly! Can you imagine! No wonder he had the music in him.

    Serendipitously, at that same time I had been digesting the Joe Zawinul Syndicate World Tour CD for weeks. Victor played on a big chunk of it. He said that he got the gig literally days before and was reading everything for the first time on those recordings. That was when I realized I wasn't worthy. What an amazing musician and a beautiful human being.