2. digital
  3. 166765-161490

NO! Jitter does not always matter.

I see a lot of discussion of jitter on this forum. A lot of it makes sense. However, I think a huge advancement in DAC chips about 13-14 years old was missed. Some people noticed that a big change in lower cost DACs occurred about aa decade ago, but probably do not know why.

Almost all modern DACs are multi-bit sigma-delta based implementation. They are not 1-bit. They have a multi-bit DAC with a sigma-delta based modulator.

Prior to the mid-late 2000's, the multi-bit DACs were based on effectively resistor ladders. So effectively a 3-4 bit resistor ladder DAC fed by a sigma-delta modulator. As it was resistor based, the amount of energy transferred was a function of both the DAC value, and also the timing, i.e how long each resistor was turned on.

Starting in the mid-late 2000's, audio DACs started to be made with switched capacitor DACs. A bank of capacitors replaced the resistors in the above DACs.  The capacitors would be charged, and depending on the digital value from the sigma delta modulator, the charge would be transferred to the output.  BIG CHANGE!  -- With resistors, the amount of energy transferred was a function of the DAC value, and time.  With switched capacitors, the energy transferred was now only a factor of the DAC value. Time was effectively removed from the equation.

Where a resistor based mutli-bit sigma-delta DAC or worse single bit sigma-delta DAC would drop to 70-80db dynamic range with 1nsec jitter, switched capacitor DACs could maintain 100, 105, even 120db of dynamic range.

I know it is inconceivable to many audiophiles that a relatively low cost DAC could compete with a megabuck DAC, especially when a lot of that money goes into jitter control, but that is what technology can do.

https://www.electronicproducts.com/understanding-clock-jitter/#
sugabooger
10 responses Add your response
sugabooger OP
89 posts
 
Tomcy,


This forced other manufacturers to compete with other technologies. Async sample rate conversion was put on chip at low cost (ESS), cypress came up with another switched cap implementation. Pulse timed as opposed to purely input timed resistor implementations also came out to do the same thing. The newest chips treated well, which does not cost much money and just takes competent design (or iterative with testing), are pretty amazing. Discrete designs are expensive as they have to compete with modern high density semiconductor mfg with discrete parts. Just the large size makes it harder.
sugabooger OP
89 posts
 
If you don’t know who Barco is, then claiming best ever for any projector is an embarrassing statement. Those CRT projectors you love. I was leading the team making what many believe we’re the best CRT projectors ever made. I was also leading the development as we moved into DLP. I think Barco has still outsold Christie (formerly Electrohome) in DLP theater projectors though they make a fine and successful product. They always kept us on our toes. My former colleagues are moving towards laser based systems now.
teo_audio
2,197 posts
 
In terms of your back patting yourself about video, I will say one thing before you embarrass yourself. I spent several years managing advanced engineering at Barco. I am sure you have heard of them. I am quite certain that what is being put together even in low cost models for DLP today, far exceeds what you were doing in 2006. The advances in the base engine far exceed what was possible in 2006, not to mention the electronics were for the most part fully integrated in the TI engine, and all the magic was buried in custom ASICs for volume products, or massive FPGAs for lower volume vendors such as us. You may want to take pause before typing more.

Actually NO.
tomcy6's avatar
tomcy6
2,597 posts
 
Starting in the mid-late 2000's, audio DACs started to be made with switched capacitor DACs.

So, is any  switched capacitor DAC better than any switched resistor or ladder DAC?  Can you tell us what some of your favorite  switched capacitor DACs are?
mapman's avatar
mapman
22,308 posts
 
Interesting.  I did notice inexpensive  DACs suddenly  got much better several years back due to technical advances relating to jitter. 
fuzztone's avatar
fuzztone
5,629 posts
 
More more 
teo_video. 
Livin’ in the past
sugabooger OP
89 posts
 
teo_audio,


I suggest you read my post again. Ladder DACs have the same high sensitivity to jitter as any DAC that is resistor based. It is less than a single bit resistor DAC, but it is still substantial.  Only the most esoteric ladder DACs, and only very recently, can achieve the performance of almost 10+ year old switched capacitor DACs. Ad-hoc comments without anything to support it carry little value.

In terms of your back patting yourself about video, I will say one thing before you embarrass yourself. I spent several years managing advanced engineering at Barco. I am sure you have heard of them. I am quite certain that what is being put together even in low cost models for DLP today, far exceeds what you were doing in 2006. The advances in the base engine far exceed what was possible in 2006, not to mention the electronics were for the most part fully integrated in the TI engine, and all the magic was buried in custom ASICs for volume products, or massive FPGAs for lower volume vendors such as us. You may want to take pause before typing more.
teo_audio
2,197 posts
 
Like everything, they are getting better. Almost back to what we had at the peak of ladder dacs that are implemented properly. I was aware of the quasi (arf ’n’ arf) dacs, and do appreciate them.

However, I’m not letting them off on the act of me and everyone else having my/our face(s) shoved into crap ---and being told it was better - for nigh on 30 years.

Expect it again. ie, class D.

CRT projection VS digital projection. same same. It took decades for digital projection to equal the best in CRT projection. This I know as I had one of the best CRT projectors made in my living room. Re-built by me. And my work of DLP engines and electronics/etc, which I stopped doing in 2006 or so, is still not equaled by DLP projection, to this day. I can take apart any DLP projector today and improve it for any eye to see... (I did this recently with a modern DLP engine, just to check on that)

You would have seen that work in DLP projectors today and be enjoying for at least a decade now... it but the manufacturers approached think of the intellectual property as the smallest part of the equation and is meaningless. so they get nothing as they treat the invention value as nothing. They get back what they put out, so they get nothing from me. They sell projectors regardless. They really don’t care.

I’ve got the finest image fidelity single chip 4k DLP projector in existence, sitting in a box, one of a kind... peak DLP...and it hasn’t been touched in a over a year.


oldhvymec
5,101 posts
 
I second that. I'm quite ignorant as to the actual workings of a DAC.

My old Krell is just dandy for it's time, but it can just have fits with some of the sources. A newer 200.00 DAC, CAN just trounces it.. Not always but streaming, a lot of the time.. CD or SDAC not at all. That old Krell can hold it's own.. Great two channel two..

I have a 100.00 tube preamp with a 10.00 DAC chip.. It will give that Krell a run for it's money.. Actually quieter NOW, too..

The Krell is class A driven If I remember, just not discrete. Time for some new parts and a reclock? I'm pretty analog.. :-)
erik_squires
16,130 posts
 
Thanks for posting this.

I am one of the few who did notice this and have posted about it regularly that new DAC's sound so much better, especially with Redbook sources. This would explain why.

Also, I'm a big fan of AKM DACs.

Best,

Erik
Log in to respond
More to discover
80x80
AIRES CERAT KASSANDRA SIG
80x80
T+A DAC-200
| 45-day Trial and Free Shipping! at JaguarAudioDesign.com
80x80
EMM Labs DA2