Delta 1010 and Mytek

[johnR]

I was thinking about using my Delta 1010's S/PDIFs for connecting to a high end converter. I considered buying a Rosetta 800 to connect to my 4 x Delta cards' S/PDIFs but after I read about clock recovery problems on S/PDIF inputs I decided against it. Basically hardly any sound card S/PDIF inputs have decent low-jitter clock recovery, so even using an excellent master clock the recovered clock signal can still contain too much jitter to use with the card's own converters without signal degradation.

Brandon, I was interested to read that your Mytek only gave a marginal improvement over your 1010s. Were you using its internal clock?

John

[brandondrury]

To be honest, I never got THAT deep with it. I never really considered the fact that the S/PDIF inputs on a given sound card would be crappy. I should probably read up on this subject because it is worth looking into.

I was always under the impression that if you used high end converters even with a cheap interface, you would get maximum results. It appears that there is another link in the chain that I never thought about.

Were you using its internal clock?

Yeah, I'm using the clock on the Mytek's as the master and the Delta 1010s are feeding off of that.

Brandon, I was interested to read that your Mytek only gave a marginal improvement over your 1010s.

I'm not sure who these humans are that have life changing experiences with gear. It just doesn't happen to me. There is such a wide range of acceptable tones in recording and the difference from preamp to preamp is so subtle 99.9% of the time that I seldom care about that either.

A person can get robo-anal about the engineering end and stress out about a .25dB difference in high end or on this or that, but these monstrously subtle differences don't seam to have anything to do with monster pro mixes.

Generally speaking, if you take a good home recording and compare it with a mega budget major label recording in a similar genre, the differences are extreme and these differences tend to go WAY past the differences in gear.

I do have to say that there is some synergy to gaining 0.25% here and 0.25% there. In fact, the only way get to that magical 100% in engineering is to focus how all these tiny things add up. It's kind of like bowling in that the only way to get a 300 is to have all strikes. If you throw all strikes but 2, you end up with 240 or whatever.

I also think that having the mentality of robo-analness with engineering that leads to a person getting high end pres and converters is required to be a great engineer. Of course, these types of people will be anal about everything....how they tie their shoes....baking a pizza to absolute perfection....and taking 3 hours to wash a car even if it is going to rain the next day.

Personally, I don't see the benefit of 100% great engineering. There are a lot of new recordings that sound enormous. So what! The music sucks.

So, for me, I've decided that I don't give a damn about mic pres. I'll just use what I was suckered into buying years ago. I really don't give a damn about converters. I'm certainly not going to upgrade past the Myteks and I'm not going to worry much about the click jitter stuff either.

Instead, I'm going to focus on finding real talent and focus on my own songwriting. My goal is produce great records....not great examples of audio engineering.

These gear snobs typically shun on $100 dynamic mics even though though a guy like Bono refuses to use anything else these days. There is a direct and obvious contradiction between the gear snobs and most of the people actually cranking out socially significant music.

Brandon

[johnR]

Good points. When you take into account that most of the finished recordings will end up in lower resolution formats like CD or MP3, that makes ultimate engineering perfection even more pointless.

[brandondrury]

That's part of it, but I don't consider CDs to be "low resolution". I am aware that there are methods that can more accurately portray detail in music by using great bit depth, higher sample rates, etc. My goal is not really to "portray detail". My goal is to create some thing that effects people on an emotional level.

I think that CDs are more than capable of impacting an individual on an emotional level. I have vivid and pleasing memories of my teen years spent driving around listening to cds. Depending on the situation and the MP3, I think that mp3s can also make people feel something. (If a person couldn't feel anything listening to an mp3, they wouldn't use mp3s!)

I REALLY don't understand the people who are really big on 96Khz audio but don't have a damn clue about room acoustics. You see these retards in the home theater magazines. There are people who'll dump $150k on a pair of speakers for music but aren't even aware of the enormous havok the acoustics in their room are offering. I guess they are just deaf.

For the true audiophile to really get what he wants, he needs one of those harnesses (like something they would strap you in on the The Matrix) to keep your head from moving. If we want to get scientific with this whole "fidelity" issue, we can't move our heads off axis. It's possible that the frequency response of the room could change and that 6 inches of movement may reduce fidelity.

I don't know about you, but I don't consider the notion of being strapped into a chair like some sort of prisoner as being all that enjoyable. I listen to music because I do enjoy it.

I know of no person who sits glued to their chair (which is in the optimal listening position) while they really enjoy music. I'd rather turn on the radio while I'm washing dishes or driving or something.

So, the way I see it, we should strive to maximize the impact of our recordings on cds and not give a damn about pushing it any further. For some, the point is "fidelity". For me, the point is "fun".

Brandon

[johnR]

Cheap but effective upgrade for Delta 1010: replace the long host cables with short ones.

(Boring technical description follows).
One of the weaknesses of the 1010 is that the clock oscillator is on the PCI card but the converters are at the other end of a 10 foot cable in the breakout box. There's plenty of opportunity for jitter to be added to the clock signal on its way to the converters, and being budget priced gear it doesn't have an extra jitter removal stage in the breakout box (like you would find in high end gear).

I tested this today. First I played back a 24 bit recording of an acoustic guitar through one of the 1010's analog outputs, patched straight to one of its inputs to record a copy. Then I copied the copy and so on until the signal had been through the DA & AD converters six times.

I then changed the standard host cable to a short one (about 3 feet) and did the same test. Comparing the sixth generation copies showed a surprising difference. The copy with the short cable sounded clearer and cleaner than the copy done with the long cable, which sounded like it had a veil over it by comparison.

OK you probably wouldn't notice the difference on a single pass through the AD converters, but if you mix on an analog console like me there's an extra pass through the DA & AD converters, so I think it's a worthwhile upgrade.

I used 1 meter IEEE1284 25-pin D male-male computer cables with a gender changer on one end, total cost around $30 per unit. Make sure they are IEEE1284 and not the cheaper RS232 cables, which definitely won't work.

[brandondrury]

I then changed the standard host cable to a short one (about 3 feet) and did the same test. Comparing the sixth generation copies showed a surprising difference. The copy with the short cable sounded clearer and cleaner than the copy done with the long cable, which sounded like it had a veil over it by comparison.

Would you mind posting these audio files? I'd love to hear the difference. I've always wanted to do the "repeater" test.

Brandon

[johnR]

Would you mind posting these audio files? I'd love to hear the difference. I've always wanted to do the "repeater" test.

Brandon

OK I'll see what I can do. They're in 32-bit floating point BWAV format, which is what my Ardour DAW uses (let me know if you can't use these and I'll convert them). It'll be a day or two before I get a chance to upload them.

BTW the tests were done at a 44.1kHz. I'd expect the difference to be bigger at higher sample rates.

[conroy]

Interesting observation. I'll bet that shorter cable improved the sound. There's a lot going on in that Delta 1010 interconnect cable. You'd get even better quality sound without the "gender changer". I wonder if there are any higher quality 25 pin cables out there ie; fancy copper, teflon dielectrics, gold pins etc.

[johnR]

Here are the files. They're pretty short. I don't have much web space so I had to dither them to 16 bit, but you can still hear the difference when you A/B them.

http://www.sound-man.co.uk/delta1010_longcable.wav
http://www.sound-man.co.uk/delta1010_shortcable.wav

Conroy, I did worry about the gender changer causing a nasty impedance mismatch, but even with that there was a clear improvement with the short cable. I think much of the problem is caused by crosstalk with all the other stuff in that cable, so the shorter the cable the better.

After changing the cable I also tested the internal clock vs. external clock from another Delta's S/PDIF. The internal clock sounded slightly better, but there was much less difference than with the change of host cable. In fact with the long host cable you probably wouldn't notice any difference between clocking methods (assuming you use short, high quality S/PDIF cable).

[johnR]

I wonder if there are any higher quality 25 pin cables out there ie; fancy copper, teflon dielectrics, gold pins etc.

I did consider making my own cables with individually shielded clock conductors, but soldering 25 twisted pairs is a lot of work.

[brandondrury]

I've got a question for you guys. (I'm away from home and don't have audio on this computer).

Assuming that studio logistics weren't ideal and you wanted to get your computer out of the control room (without dumping a bunch of cash on an isolation rack) where would you use the longer cable? Would you a longer cable in between the breakout box and the PCI card (this thread seams to say no)? Would you use longer cables in between preamps and the Delta 1010 (before the breakout box)?

Just curious.

Brandon

[conroy]

Use the longer cables between the pre-amp and the A/D converter. Use a good quality cable and hit the Delta 1010 hard with lots of signal from your high end mic-pre. I have had success with DH Labs Pro Studio cable. Check out;

http://www.silversonic.com/docs/products/ProStudio.html

Keep the cable between the A/D and the PCI card/computer short. Again, use the highest quality digital cable possible. This would be the ideal scenario for the Delta 1010. If you want to take it another step, try giving the Delta 1010's power supply box first crack at the wall outlet instead of plugging it into a power bar. Better yet, give the Delta 1010 it's own spur to the mains panel in your building. Digital and analogue apparatus should generally be kept on separate spurs.

There are many steps that can be taken to improve the sound/signal path in your system. You can get a better sound with the gear you are already using.

[johnR]

All good advice. I use long cables between mic pres and the Deltas, which are right next to the computer (now with short host cables). Any decent mic pre will have a low enough output impedance to drive fairly long cables. If you put your pres close to the sound source you can use short mic cables which also helps to maintain quality. Standalone mic pres with fully balanced outputs will be better than the pres in a budget-level console, which are unlikely to have fully balanced direct outs.

I tend to avoid the higher sample rates on the Deltas as they don't sound as good as they should (probably due to increased jitter), but they are capable of good results at 44.1 or 48kHz. Another thing you can do is insulate the 1010s from each other and from the rack by using nylon washers and sleeved bolts. I got around 3dB reduction in background hum by doing that (and yes you can get hum loops with balanced gear).

[johnR]

Another thing. The first few channels on the Delta 1010 are close enough to the incoming 9Vac supply to pick up some hum. Not a lot, but it's there. For critical signals with lots of low level detail you might want to stick to the higher channels.

[conroy]

Excellent observations. Yes channel 1 and 2 do sound a bit noisier for being close to the power supply. I always use these channels last or for the least important elements in a mix. I did most of my modifications on channels 7 and 8. They are the most "souped up" channels. Upgrading the filter caps on the PCI card can also bring substantial improvements in bass and dynamics. It smooths out the jitter.

I didn't realize that the 1010 doesn't sound as good at higher sampling rates. Is this compared to other converters or compared to itself?

Those nylon washers sound like a good idea. I'd love to see your studio!

[brandondrury]

Out of curiosity, I tossed both sound files (long and short cables with the Delta 1010) into my DAW and flipped the phase on one and boosted the level by 10dB.

There is definitely some sound being cancelled, but it is definitely subtle. While A/Bing the tracks, I think the difference between the two is not as great as the placebo effect.

-------

Another question. I use a Mytek AD96 going into the Delta 1010 via S/PDIF. Since I VERY seldom use the breakout box these days (I don't drums in my house anymore) do the rules of S/PDIF cable apply apply the same way as the Delta 1010's breakout box cable does?

In other words, am I better off with a long cable coming out of my preamps or a long S/PDIF cable.

Thanks,
Brandon

[johnR]

Upgrading the filter caps on the PCI card can also bring substantial improvements in bass and dynamics. It smooths out the jitter.

That was going to be my next mod. I suspected it might improve things.

I didn't realize that the 1010 doesn't sound as good at higher sampling rates. Is this compared to other converters or compared to itself?

Before I modified my 1010s I couldn't hear enough of an improvement at 88.1 or 96 kHz to warrant filling the disks twice as fast. Admittedly I haven't A/B'd them against other converters, but a PAR review in 2002 said the same thing. I should try the repeat test at higher rates though; the shorter host cable might make a difference.

Those nylon washers sound like a good idea. I'd love to see your studio!

At the moment I have a portable setup and have to hire a room to record loud stuff. I hope to be fixing that situation soon.

If you're not averse to soldering, there are a couple of other upgrades that can be done. The +/- 15V power supply for the op amps in the breakout box is a bit of a joke. If you rectify 9V ac you usually only get around 12-13Vdc so there's a capacitor-based voltage doubler to boost the voltage high enough to feed to a pair of 15V regulators. Unfortunately the boost caps are undersized so the voltage isn't boosted high enough for the regulators to do anything. This means the op amps are being fed a ripply, noisy, unregulated power supply. To make things worse, higher sample rates draw a heavier supply current and pull down the voltage rails even further, so ripple and headroom are both worse at higher sample rates.

My solution was to put in larger, industrial grade boost capacitors, change the rectifiers for Schottky diodes and replace the regs with low dropout types. The +/- 15V regs dissipate more heat when this is done, so I made some big heatsinks (they need cutouts to clear the other components on the PCB). Now the 15V rails are clean and don't droop when the input voltage drops slightly. To keep it all cool I put a 40mm fan next to the left side of each 1010 in the rack, lined up with the ventilation slots. I also replaced the plastic external power supplies (yuk) with a hefty homemade PSU, with separate floating secondaries for each unit, which is bolted into the back of the rack.

[johnR]

Out of curiosity, I tossed both sound files (long and short cables with the Delta 1010) into my DAW and flipped the phase on one and boosted the level by 10dB.

There is definitely some sound being cancelled, but it is definitely subtle. While A/Bing the tracks, I think the difference between the two is not as great as the placebo effect.

-------

Another question. I use a Mytek AD96 going into the Delta 1010 via S/PDIF. Since I VERY seldom use the breakout box these days (I don't drums in my house anymore) do the rules of S/PDIF cable apply apply the same way as the Delta 1010's breakout box cable does?

In other words, am I better off with a long cable coming out of my preamps or a long S/PDIF cable.

Thanks,
Brandon

Some of the difference got lost in the conversion to 16 bit. Listening to the original 32 bit float files through my Beyer DT 250s, the difference was more obvious. Another engineer I work with said he was surprised it was so obvious when he listened to them. (I got a second opinion to eliminate placebo effect ;-)). Some of these mods could be described as nitpicking, but put enough nitpicks together and you can get a real improvement for little cost.

The jitter problems being discussed only really apply to the clock signal used by the A-D and D-A converters. In a purely digital transfer jitter isn't so important, as long it isn't so severe that data is lost. Your Mytek will have decent jitter reduction, so it will be quite tolerant of a long S/PDIF cable. (The 1010 host cable is only a problem because it doesn't use jitter reduction).

To summarise:
If you still want the option of using the 1010 converters when you run out of channels on the Mytek, use short S/PDIF and long preamp cables. If you really never need the 1010 converters, it doesn't matter.

[conroy]

The +/- 15V power supply for the op amps in the breakout box is a bit of a joke. If you rectify 9V ac you usually only get around 12-13Vdc so there's a capacitor-based voltage doubler to boost the voltage high enough to feed to a pair of 15V regulators. Unfortunately the boost caps are undersized so the voltage isn't boosted high enough for the regulators to do anything. This means the op amps are being fed a ripply, noisy, unregulated power supply. To make things worse, higher sample rates draw a heavier supply current and pull down the voltage rails even further, so ripple and headroom are both worse at higher sample rates.

My solution was to put in larger, industrial grade boost capacitors, change the rectifiers for Schottky diodes and replace the regs with low dropout types. The +/- 15V regs dissipate more heat when this is done, so I made some big heatsinks (they need cutouts to clear the other components on the PCB). Now the 15V rails are clean and don't droop when the input voltage drops slightly. To keep it all cool I put a 40mm fan next to the left side of each 1010 in the rack, lined up with the ventilation slots. I also replaced the plastic external power supplies (yuk) with a hefty homemade PSU, with separate floating secondaries for each unit, which is bolted into the back of the rack.

This makes sense. I upgraded many of the small caps on the inputs and outputs to Black Gates. I skipped the large power supply caps at the time. I thought the expense might not be worth the sonic gains. Perhaps I should go for some larger booster caps but not bother with buying the audiophile ones. I seem to remember those big Black Gates going for at least $40 a pop. I can see how beefing up those values would help out the op amps.

You mentioned using fans. I've gone to a lot of effort to keep my computer quiet. I'm a little concerned about the fan noise. The heatsinks are a great idea though.

It sounds like you know what you're doing. It's nice to know that there is someone else out there who feels that the Delta 1010 is worth modifying.

[brandondrury]

Hey dudes,

I'm glad to see that you guys are going balls to the wall to make sure that your Delta 1010 is tip top. (I'm also glad to see it does'nt have to cost $3k per channel either).

I don't mean to change the focus of other threads, but I'm curious if you are paying this much attention to detail to every piece of the chain. In other words, are your room acoustics so good that the length of the cable from the Delta 1010 is the weakest link?

Don't get the wrong idea here. I'm glad we have people that do what to make the best sounding records possible. I was just curious if you guys were focusing on the Delta 1010 as the project of the month or if you really are putting this much effort into improving the Delta 1010 constantly.

Brandon

[conroy]

Hey dudes,

are your room acoustics so good that the length of the cable from the Delta 1010 is the weakest link?

Don't get the wrong idea here. I'm glad we have people that do what to make the best sounding records possible. I was just curious if you guys were focusing on the Delta 1010 as the project of the month or if you really are putting this much effort into improving the Delta 1010 constantly.

Brandon

It's true, room acoustics are important. I'd rather hear a crappy recording rig in a great room than vise versa. We're talking about working with what you've got. I used to record with analogue tape. I loved the sound. A/D D/A conversion became a sore spot for me. It's the moment where a lot of magic and life can get sucked out of a recording. Anything that can be done to improve this step is worthwhile. My Delta 1010 sees constant use. Any small improvement I can do to this converter exponentially affects all the work I do.

Concerning Apogee products. I'm curious about the Big Ben. Would it's improvement upon the Delta's clock justify it's price? Has anyone tried it?

[conroy]

JohnR,

I dig the cable length sound sample test. The short cable definitely sounds better. A little smoother with more pleasant highs. It breathes. I'm hearing slightly better dynamics. I'll have to try getting my Delta 1010 closer to the computer. I'm also thinking about those power supply mods. Thanks for all the great ideas.

[johnR]

I've gradually improved my Delta 1010s over the last year or so. I pay the same kind of attention to the rest of my setup, when I have time. With a bit of electronics knowledge it's usually possible to make worthwhile improvements to modestly priced gear. Of course it's a waste of time if you have bad acoustics, as you'll be able to hear just how bad your room is with your improved gear.

Word clocks like the Big Ben might not deliver the improvement you hope for. S/PDIF is inherently jittery, and the 1010 uses a consumer grade S/PDIF receiver chip which will do very little to remove jitter introduced by the cable, isolating transformers, etc. There are good explanations of this that can be found with Google. I don't trust the word clock input on the 1010. The clock signal travels along the host cable (picking up crosstalk from data signals along the way) to the PCI card, where it hits a cheap phase-locked loop chip that will do very little to remove jitter. It won't lock to 96kHz either, unless your particular PLL chip is performing above spec.

One advantage of a separate clock though is that all the cards get their clock signals in phase, so you don't have to worry about splitting a stereo pair across two breakout boxes for example. I clock 3 Delta 1010s from a Delta Audiophile card in the same box, via a 3-way S/PDIF splitter (with very short cables).

This isn't ideal, but my next improvement will probably be to clean up the supply rails to the clock oscillator. Crystal oscillators vary slightly in frequency with supply voltage fluctuations, so if the supply isn't clean the clock will be jittery.

I wouldn't bother with audiophile caps in the power supply. The characteristics that make them good in signal paths can cause problems in other situations. The low ESR allows them to resonate very strongly at high frequencies, whereas the higher resistance in cheaper caps damps out resonance for example. The most important ratings for those booster caps are ripple current and temperature. They used to explode regularly in early 1010s, so use the highest rated caps you can get.

[johnR]

I dig the cable length sound sample test. The short cable definitely sounds better. A little smoother with more pleasant highs. It breathes. I'm hearing slightly better dynamics. I'll have to try getting my Delta 1010 closer to the computer. I'm also thinking about those power supply mods. Thanks for all the great ideas.

Glad you found this useful. If you want to discuss any of this in more detail feel free to PM me.

[abechap024@hotmail.com]

Hi guys! Finally I've been looking everywhere for some mod tips on the delta 1010s. I am taking your advice to heart, and at first when I read your post, I was discouraged because , I felt like my delta wasn't good enough and all of the sudden all these recordigns I did on it suddely sounded like shiet :-P You know what gear lust i'm talking about! Anyway just wanted to ad my 1 1/2 cents worth. I did some digging and the stock opamps on the delta are JRC and they cost about as much as a cheap lollypop! what I did recently was took a gamble - see the opamps in the thing are DMP sizing (surface mount stuff - a little wider than standard soic) well there is nothing that I could find on mouser or digikey besides other jrc rubbish to replace the deltas opamps. So I orderd some soic Burr-Browns - OPA2134UA. 16 for the input side of the card. Now I wasn't sure if the soic would be wide enought to be able to replace the jrc's but i was determind to replace them. They arrived today and after taking out one of the jrcs (just get your fine tipped adjustble Iron on the hot side, and slip a .9 or less guitar string under the little chips "feet" and heat and pull) the Soic fit in there great with room to spare! I was very pleased!! Now i recommend some tweezers if u have em cause its tiny! but after soldering in the 2 new BB opamps in channel 8 (because you guys were sayign 8 is farthest away from the power supply therfore least noisy) I plugged the thing in and loaded up my computer and it waas working! no smoke (thank god!) and loaded up my DAW and first recorded a stock channel of me talking and then the BB channel, and listened back with my Assoiciate (roomate) and We were both floored! the JRC sounded fizzy and unfocused compared with BBs' tight focused sound! Its hard to put into words but on the Black Lion Audio page they were saying the delta uses the same convertors as the digi002, and after doing this opamp switch out the "fuzzyness" of the delta is gone! I can't wait to change out the caps and do the power supply mods you all have been talking about~! Bu I can't beleive what a change in tone this delta has now. I can't wait to do some recording on it to REALLy test it. But I've noced from A/B stuff that its harder to tell with just one track how the preamp/convertors/etc will sound in a context of a song -. And with the opamp switch out the tone shift was obvious.Well Thanks guys! I do have some quesitons on the power mods and cap mods - this stuff Is all still a little mystifying! If you feel motivated could you point me in a direciton of what caps to replace?? And the green cap ?? any info would be most helpful, and if I coudl be of anyhelp just let me know!
AC