Quick question about a couple of audio boards that I frequently use.
One is an ESI Waveterminal 192X card and the other is the E-MU 1820m box (plus its PCI card).
Both are ancient, in computer years, but I haven't noticed any change (but I'm am aging so that could be why). And their specs are excellent.
I did use them some 10 years ago to remaster tape to digital and they performed awesome. Used a much higher sampling rate and bit depth than was necessary but there was no harm in doing that as I had the disk space (uncompressed files, of course, in wav/PCM format).
I now need to remaster some more reel tapes to digital and the bottom line is: should I purchase a brand new pro audio card (or card/box combination as in a multiple component system, like the 1820m) since the two that I use are quite old? Also, is it true that the chips on the board should be allowed to get up to normal operating range...as in use them for a while before doing any actual work (yes, I may have been fooled on that one).
If so, on a different card setup for remastering, please do provide any recommendations.
BTW, the 1820m is my ultimate favorite device for playback due to the fine tuning that can be done to a digital signal before it converts to analog and sent to an amp and speakers. I have yet to find another device that can be a similar DSP processor for all needs, especially for equalization purposes.
Thanks!
Joe
One is an ESI Waveterminal 192X card and the other is the E-MU 1820m box (plus its PCI card).
Both are ancient, in computer years, but I haven't noticed any change (but I'm am aging so that could be why). And their specs are excellent.
I did use them some 10 years ago to remaster tape to digital and they performed awesome. Used a much higher sampling rate and bit depth than was necessary but there was no harm in doing that as I had the disk space (uncompressed files, of course, in wav/PCM format).
I now need to remaster some more reel tapes to digital and the bottom line is: should I purchase a brand new pro audio card (or card/box combination as in a multiple component system, like the 1820m) since the two that I use are quite old? Also, is it true that the chips on the board should be allowed to get up to normal operating range...as in use them for a while before doing any actual work (yes, I may have been fooled on that one).
If so, on a different card setup for remastering, please do provide any recommendations.
BTW, the 1820m is my ultimate favorite device for playback due to the fine tuning that can be done to a digital signal before it converts to analog and sent to an amp and speakers. I have yet to find another device that can be a similar DSP processor for all needs, especially for equalization purposes.
Thanks!
Joe
Hi
It depends on the technology. Despite being old, your interfaces probably use sigma delta DACs and ADCs, and these are insensitive to ageing afaik.
I think it could be a factor with multibit dacs using resistors. Values might drift and the dac might lose accuracy over a very long period? Never read anything about it though.
Maybe you should swap all electrolytics for new ones, of good reliable brand. PITA?
-Alex
It depends on the technology. Despite being old, your interfaces probably use sigma delta DACs and ADCs, and these are insensitive to ageing afaik.
I think it could be a factor with multibit dacs using resistors. Values might drift and the dac might lose accuracy over a very long period? Never read anything about it though.
Maybe you should swap all electrolytics for new ones, of good reliable brand. PITA?
-Alex
Last edited:
I think it is a good idea to do so. Let it all warm up, everything!
Jocko Homo said some crystal oscillators sound allright after power up, while others take a long period (many hours) to settle and sound best.
-Alex
Yep, I had to replace basically every single electrolytic capacitor in the box. One that I bought off eBay had a bunch of caps explode and magic smoke released when I turned the phantom power switch on the first time.
The EMU 1820m box gets really hot inside, which does not help the lifespan. The caps on the output of the switcher do not have a high enough ripple current rating I think as well. I replaced that first one with a solid polymer cap, it used to get really hot even with an opened box.
The drivers even work well in Windows 10 despite not being updated for years. The main reason I don't use the 1820m anymore is that I don't use a PC with PCI slots anymore and I don't want to waste a PCIe slot on an adapter.
Last edited:
I would recommend you not leave the 1820m box on 24/7 if you can. The box has very little ventilation and is packed tightly. Those AKM ADCs get really hot and the box turns into a toaster inside. You can unplug the RJ-45 cable at night if you must leave the PC on. I noticed you can hot-plug it back in without any ill effects.
I would use it cold without a second thought. I have never noticed an audible difference and I doubt there would be a significant measurable difference even. Either way, the thing gets so hot inside it will warm up within a few minutes regardless.
Last edited:
The exact type of DAC/ADC's is something I could probably look up. I have heard the term "sigma delta" but I am no electrical engineer. Computer science, yes, and I envy the EE's who know CS also. I am a very curious person and like to read the tops of chips and then find and read any technical documentation about them. You could say that I have a basic understanding about how they receive a signal and then measure it (quantization, it is called?).
As for "electrolytics", I assume you mean the capacitors and their quality? If so, one horrible story of mine is when two capacitors "blew their 'K' top imprint right off" and it did damage to the workstation. So, when servicing machines on a regular basis (like take a look inside every month to blow out the dust) and I will examine the tops of the caps. If any are bulging or popped, I guess I got lucky if the system was still working plus I can get it repaired. That video board's two caps blowing really bummed me out as it was a very nice and quality card plus it was not overclocked by the manufacturer or me. It ran nice and cool.
Joe
I think I heard about letting equipment warm up when it came to CRT professional monitors used by photographers. They let the CRT's get up to their stable operating temperature so the reds, greens, and blues are "dialed in" to what all of their previous work used. Consistency and the same color gamut are achieved.
Thank you for that information. I will inspect the PCI card and the external box's capacitors on a regular basis.
I truly do think I "chose luckily" when I bought the 1820m because, first, the sound and DSP is great and, second, when people sell them on Ebay, the price asked is still very close to the original retail price. I learned that one because I thought of getting a "backup one" but, wow, the prices asked were too much for a backup.
The PCI card caps don't fail to my knowledge, or if they do, it takes many more years to happen than the box. I've had the same PCI card since 2006 without any issues.
I am aware that it does get quite warm so it gets placed in a "standalone area" where its passive air ventilation can happen.
Also, when I bought it, I did read in the manual that the box can be disconnected at any time (like you mentioned) by unplugging its special network type cabling. If I'm using the workstation that has the 1820m installed and I'm not needing it for a project, it is disconnected to help extend its lifespan. I appreciate you passing along the information too.
I don't know why E-MU doesn't make an 1820m "version 2" that swaps the card's PCI slot type for a PCIe slot type. Plus, make any further improvements to the box's chips, the main DSP processing chip, cooling, more inputs/outputs, etc. Hearing how many people like the 1820m "system", I cannot see how E-MU couldn't easily sell them.
I haven't found anything that comes close to it. Maybe others know of a product similar to the 1820m system?
Thanks!
Joe
I am glad to hear your optimism and opinion.
I cannot believe I've had the thing for so many years now...as in, time flies pretty quick.
A toast to many more years from our 1820m's!
The problem with audio converters is the accuracy you need. For example, imagine you make a 16-bit DAC by just making binary-scaled current sources and switching them on and off. When the code goes from binary 0111111111111111 to 1000000000000000 (or from 1111111111111111 to 0000000000000000 when it is two's complement coded), 15 current sources switch off and the 16th switches on. The 15 that turn off ideally have a total weight of 32767 unit currents and the one that turns on 32768 unit currents, but when the one that turns on is 0.01 % too low, the signal actually goes in the wrong direction: from 32767 unit currents to less than 32765 unit currents.
One approach the chip manufacturers can use to solve this issue is to laser trim the current sources (or resistors or capacitors, depending on the exact kind of DAC) until they meet the tough accuracy requirements. In that case, mismatch between the long-term drift of the current sources could cause the DAC to become less accurate again. A similar story applies to ADCs that have such a DAC in a feedback loop.
Another approach is thermometer coding: using 65535 unit current sources and adding one when the input code increases. This quickly becomes expensive in terms of chip area and difficult in terms of wiring, especially when you go to 20 or 24 bits. Compromises are possible: thermometer coding the upper few bits and doing the rest binary scaled.
An approach that has become very popular is sigma-delta or delta-sigma modulation. For an ADC, this boils down to making a very course ADC that runs at a much higher rate than needed and applying feedback using a feedback DAC and a loop filter. These are designed such that they suppress the quantization noise of the course ADC within the band of interest. It only gets worse outside the band of interest, but that gets filtered off with a digital filter. When the course ADC and feedback DAC have only 1 bit, there is nothing that needs to match and no problem with long-term drift. Modern sigma-deltas often use a few bits rather than only one, but have some additional tricks (data weighted averaging, dynamic element matching) included to ensure that slight mismatch in the course ADC or the feedback DAC doesn't impact the performance much.
(You can also make a sigma-delta DAC: the course ADC then becomes a digital rounding function (or just taking the sign bit), the feedback DAC disappears, the loop filter becomes a digital filter and the output signal of the loop goes to a DAC with only one or a few bits and an analog filter.)
All in all, like Alexandre already wrote, a trimmed binary scaled converter might have an aging problem, but for a sigma-delta based converter that's very unlikely.
By the way, since you have a computer background, you might find this post interesting:
https://www.diyaudio.com/forums/dig...-sigma-delta-dac-operation-3.html#post5189375
It contains a Pascal software implementation of a sigma-delta modulator: it converts from a normal digital signal to a digital single-bit signal and back into a normal digital signal. The point of it all was to test a hypothesis of another forum member.
https://www.diyaudio.com/forums/dig...-sigma-delta-dac-operation-3.html#post5189375
It contains a Pascal software implementation of a sigma-delta modulator: it converts from a normal digital signal to a digital single-bit signal and back into a normal digital signal. The point of it all was to test a hypothesis of another forum member.
RME and Lynx make some similar products, but they cost more.
The 1820m uses Cirrus CS4398 DACs and AK5394A ADCs. They are both delta-sigma types. The AKM ADC is probably still the best on the market. The DAC is still pretty decent.
Wow! Now the above was awesome to read and gave me many avenues to explore.
(You can laugh at this) You easily blew away my first CD audio engineering book by Ken C. Pohlmann, "The Compact Disk Handbook: 2nd Edition" from 1992. One of my most favorite books as it started me on the path of wanting to extract the numerals from CD's and store them on HD units for math processing but back then, HD units didn't store all that much in GB, ya know? It was like, do one song at a time from a CD and with a Pentium 90MHz if I was lucky. I think the basic desktop machines back in 1994 still came with 6 GB drives. Until I was able to RAID together 80 GB drives...that was when the extraction began in earnest and PC storage of nearly all of my 300+ CD's was possible (the array physically - the disk units - looked like a leaning tower of Pisa as I just used a bent piece of sheet metal to attach the disk units to it for a vertical tower of horizontal HD units and wrote a note "Do NOT TOUCH this!!"). That Pohlmann book...I still revisit it for the great memories.
I do, of course, have many other books on the subject/c-lang-coding but they are quite old now (from the 90's and 00's) as I had to "put down" the DSP programming interest for a while. Until now. Now, I have all the freetime I need to read everything such as what you provided.
And the home "remastering audio (and video) room" is coming along nicely. The custom building of the equipment racks plus speaker/monitor units is nearly complete and the extra computer systems that I have laying around will be re-purposed for the actual work. Should be very fun.
Thanks for all of that information!!!
BTW, if you know of any awesome ADC/DAC chip, DSP, and similar books, please do let me know.
Joe
Thank you! I will visit that link and get the Pascal code. Yes, I do know Pascal. Remember old BBS's that people would use handset cradle modems aka acoustic coupler and then the Hayes Smartmodems (like in the movie "WarGames" with Matthew Broderick)? Pascal is what I wrote my very own BBS system in so that people could "call into" it...Turbo Pascal by Borland I believe. People laugh about it nowadays (younger people, heh) but I long ago moved to C and assembly for my programming projects. I also wrote one of those "automatic dialers" like was seen in the movie, searching for computer systems to answer, but I used somewhat randomized numbers unlike in the movie! And, I did find some...
Anyway, what you said sounds a lot like a proof that I needed to do with an FFT coding. Convert and store an entire audio track using a chosen windowing method and the FFT at any granularity. Then run that through the iFFT (inverse) and write that out. Compare the written out to the original. It was: If matching, everything going perfectly numerical precision-wise. I was quite happy many years ago when I could do that on my home PC! While, back then, it couldn't run in real time (due to processor speed) it was still great to have my very own infinitely adjustable equalizer and all I had to do was "wait" for processors to become faster so the code could run realtime. Also back then, I was offered, by an interested CSCI professor, some time to run my code on my university's computer science department mainframe (not a big deal because I'm sure many other students had done the same thing too) but something came up and I wasn't able to try that (on the university's mainframe). Oh well.
Thanks again!
Joe
I will investigate those brands. It is kind of OK if the equipment costs a lot but then it has to deliver. The 1820m system, to me, was quite expensive when I bought it but it has delivered many years of enjoyment so that is why I'm willing to pay a price for great stuff. It is like automobile audio: if you buy the expensive stuff, don't worry about loudness as you will get perfect sounding audio plus the loudness meaning lots of SPL headroom (I have said to friends that my Jeep's "PC-computer-touchscreen, JL Audio, and Polk system" will "hurt you before you hurt it" hearing-wise so don't crank it without knowing the dangers). The same with my home audio - lots of headroom.
While I do know the "specs" of each ADC/DAC that I come across (including motherboard built-in ones), I don't fully believe in them until someone with an electrical engineering background can really confirm them. Sometimes the S/N ratio doesn't specify at which bitrate and depth you get it, as an example, and I have to assume it must be at the highest the card can perform (ie 192kHz, 24-bit with the 192X card) and not at the usual "CD rate" of 44.1 and 16bit.
I've also had problems with motherboard based chips. Even when they are very high in specs, I often find that noise is introduced somehow and usually from a harddisk's reading and writing of data (like that stream of data flowing through copper on the motherboard is coupled to the onboard audio chip somehow). The manufacturer offered suggestions but none really worked so that is why I like the separate cards and boxes. It was probably a grounding design fault.
Thanks. Also, do you happen to know if there is an 1820m "users forum" on this site? I have some ideas that I would like to try with the 1820m but cannot figure out how to do. And maybe they cannot be done.
The conundrum is: I don't know which forum location to ask my questions in and if there are any people out there with the 1820m also.
Another thing about the 1820m is that I'm pretty sure I've found, in a subdirectory, all of the files which "create" the applet of the 1820m. Therefore, a savvy user could create their own applet (different colors, form/shape, etc) and I've always wanted it to be larger since I use hi-res monitors.
Thanks,
Joe
I don't know of many other 1820m users here, and I don't use any of the advanced features really so I might not be of too much help. There are probably some pro audio forums where you may find other owners.
Even if you found UI assets for the skins, I would say it's probably unlikely you can totally restructure the Patchmix applet without access to the source code.
RME makes very good stuff, I'd check them out. Creative / EMU killed this line a long time ago and the last driver was released in 2011. Their products are comparable in analog performance to the 1820m depending on the model.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.