Time for some Stax SR-Sigma love......

[John Buchanan]

The Stax Sigma series panoramic earspeakers.

1. Models and nomenclature

The Stax SR-Sigma panoramic earspeaker was introduced in 1977. There have been 3 versions officially released and 1 after-market version commissioned.

1. The original low bias, grey grilled 1977 Sigma model (bias voltage of this model being 230V or “Normal”). The driver used was later recycled in the very successful Stax Lambda Semi-Panoramic earspeaker. The first release Sigma had a fabric coated, round bodied cord that connected the earspeaker to its amplifier/transformer.
   
2. Later versions of the normal bias Sigma (1987) used a lower capacitance, flat black cable that was recycled back from the (then) current Stax SR-Lambda earspeaker.
   
3. The later Sigma Professional (1987) version introduced the 580V Professional driver then used in the Stax SR-Signature earspeaker. This earspeaker had black grilles, lower capacitance chocolate coloured cables, and a chocolate coloured headband, compared with the original black cables and headband.
   
4. A later version, the Sigma/404, was a Sigma rebuilt with high bias SR-404 drivers and their corresponding very low-capacitance cables.
   

2. History and conception.

The Sigma earspeaker design was the result of a complete reassessment of how headphone sound is perceived. Up until that point, listening to headphones or speakers were considered completely different experiences. Headphones were designed to inject sound directly into the ears with as much sound isolation as possible between channels and also from the external environment, reflecting their communications genealogy. The drivers were parallel to the pinnae (= ear flaps) and either intra-aural, circum-aural or supra-aural.

Speaker listening has the drivers at a great distance from the ear canals and sound produced has to traverse a great number of direct, reflected and partially absorptive pathways before arriving at the ear canal, allowing much more modification of the sound as well as left and right channel blending. The drivers are also in front of the listener and perpendicular to the plane of the pinnae.

Naotake Hayashi, the genius behind the original Stax company, decided that one of the differences between speaker and headphone listening was a result of that very isolation and direct aural injection inherent in the design of all prior headphones. He decided to make a headphone that would actually sound like listening to speakers in a partially reflective/absorptive room. The genius lay in his actual recreation of a room around each ear – a revolutionary concept that has never before or since been replicated. Each earcup was meant to approximate a partially absorptive and partially reflective series of surfaces for the headphone driver (now in front of each pinna and perpendicular to them, as per speaker listening) to bounce sound off and then into the ear canal. In other words, the direct injection principle was thrown out the window and now the drivers were only heard after firing sound into the ear canals via a reflection - and some absorption – from an internal lining of mineral wool. I am guessing that the ear speaker cages had to be constructed pervious to air, rather than designed with a solid body, for weight considerations (viz. a solid body construction would have been too heavy for comfortable wearing). Possibly there were also enclosed cavity effects to deal with if the headphones were sealed. Indeed, weight has been one of the main complaints levelled at the only enclosed Stax design, the Stax SR-4070 Monitor. The mineral wool lining of the cages, apart from reflecting and absorbing sound, much like a normal listening room, also provided some hermetic sealing of the cages allowing reduced front to back sound cancellation around the periphery of the drivers. In other words, the drivers could have bass (the lack of which has been a criticism of the relatively similar design AKG K1000), but not as much as if a solid body had been used. I am guessing that a solid body Sigma had been tried and discarded due to comfort and sound considerations, so a compromise between weight of the headphones and sound quality as well as bass extension was reached.

3. Comfort.

The Sigmas, although bulky and laughably unfashionable, have been engineered to be exceptionally comfortable on your head. It is literally easy to listen for hours without your pinnae contacting the metal inner grille of the drivers (Lambda series) or the earpads themselves (SR-007). The small listening cavities of those 2 phones also leads to sweaty ears, which is far less apparent with the Sigma series.

4. Sound.

a. Low bass and bass.

The sound of the Sigma always has slightly reduced very low bass because of some residual front-to-back driver cancellation through the mineral wool earcup lining, but beyond that point, the earspeaker’s sound is very hard to fault compared to what one is used to. The bass that is present, until the very low bass roll-off, is of excellent quality. Certainly there are no bass instruments that move back and forward in the soundfield, nor do they completely disappear as they descend the scale, as I heard with the Jecklin Float Electrostatics playing, for example, Tony Levin’s descending Stick run during Projekct One’s “Live At The Jazz Café” Track 3. On the Jecklins, Tony appears to walk out the studio door as the run descends to subterranean levels. On the Sigmas, he’s in the studio and hasn’t moved a muscle.

b. Midrange.

The reproduction of vocals comes as close to free of sibilant emphasis as possible. This is truly what you hear in live, unamplified music. Particularly realistic are piano and voice – the smoothness of the sound is just as relaxing on replay as it is live. Indeed, after attending a piano concert in a relatively reverberant wooden hall, the Sigma/404s got the extreme dynamics of the piano without any of the brightness at higher volumes that the Lambda Nova Signature exhibited. This exquisite piano reproduction is unique to this headphone, in my opinion

c. Treble

The later substitution of the 580V “Professional” bias drivers (either the Lambda Signature or the Lambda 404 headphone driver) to replace the original “Normal” bias driver, along with an upgrade of the original headphone cable, allows a reduction in the marked high frequency roll-off and a flatter extension in the very bottom-end reproduction compared with the original low bias Sigma. In my opinion, there are no drawbacks from this modification whatsoever and the top end sounds both smooth and evident, rather than smooth and rolled off as in the original Sigma

d. Correct volume level.

The Sigma/404 really shines a light on just how successful Mr Hayashi’s earspeaker design concept really is. Another bonus I have noted with these earspeakers is that it seems to be relatively easy to dial up the “correct” volume of sound – they just sound “right” at that point. Although this is not unique for these phones, I haven’t heard a headphone with such a pronounced “correct” volume level for a track. The bass and treble just seem to be in perfect balance with the midrange only at that volume. Peter Walker of Quad was a great proponent of the “correct” volume theory apparently.

e. 3-D sound.

Then there is the seemingly increased 3D space that these headphones portray – the sound stage seems to be actually in front of the head, with some front to back space, compared with the usual line-between-the-two-ears imaging. This is something I’m not as good at hearing, so I will leave it to others to give their impressions.

These differences allow greater appreciation of albums that were mixed for speakers in the standard control room, because that is exactly what the Sigmas replicate. I would guess that apart from very low frequency roll-off, these earspeakers could be the greatest and most accurate magnifying glasses for mixing evaluation ever made.

5. Associated equipment needed.

I have found the SRD-7 Pro or SRD-7 Mk 2 will do an admirable job of driving these very power hungry monsters with a good power amp pushing them. If you wish to drive them with a direct drive electrostatic amp, I would suggest, at the very least, using any of Stax SRM-717/SRM-727/SRM-T2, Kevin Gilmore/Spritzer’s revamped T2 or the Blue Hawaii SE/Solid State Electrostatic amps. They are all powerful enough to drive these and the SR-007 Mk1/Mk2 series. The Lambda series are far less power hungry, despite using the same drivers as the Sigmas. Distance from the ears and absorption by the damping material may both account for these efficiency differences. The volume control levels for the Sigma/404 and SR-007 Mk1 I once owned were identical for the same reproduced volume – i.e. they seem to be equally inefficient - compared with a Lambda Nova Signature.

6. Credits.

Finally, one has to admire a designer who actually truly said nay to any marketing considerations. These phones are laughably big and ugly, but if they were anything else, could they sound as good? Thank you, Mr Hayashi for not listening to the form-over-function naysayers, and I bet there were plenty of those in Stax board meetings in 1976/1977

Lastly, I would like to thank Edstrelow for the inspiration to upgrade my Sigmas to Sigma/404s – something I did on faith and have never regretted it for a minute, and Webbie64 for making me realise the error of my ways when I briefly thought about selling them.

7. Postscript - Stax Sigmas high bias earspeakers compared – Pro vs Sigma/404.

The 2 Sigma high bias phones essentially differ in 2 main physical ways.

1. The headphone cable used for the original Sigma Pro was the same as the one used on the Lambda Signature – not as wide as the one used for the Sigma/404 hybrid, which were first seen on the Lambda Nova Signature.
   
2. The drivers are different also – the Sigma Pro uses the reported 1 um Lambda Signature driver, whilst the Sigma/404 uses the (2 generations) later 1.35 um drivers first seen in the Lambda 404.
   

Otherwise, the shells of the 2 earspeakers, apart from minor colour variations, are identical. The Sigma Pro driver appears to be very slightly more efficient than that of the Sigma/404.

In my set up so far, the two have been compared through the Studer D730 -> Apogee Mini-DAC -> Audio Research LS5 Mk2 -> Studer A68 -> Spritzer Pro SRD-7 bias + transformer box.

As seems to be the order of the day, the results are not what I expected. The Sigma Pro is far better than its reputation suggested and even bests the Sigma/404 in a couple of areas.

Firstly, the deep bass is slightly more evident and the mid bass is much tighter and slightly less resonant in the Sigma Pro. Bass drum has slightly more punch rather than smeared thud. This has been one of my only criticisms of the Sigma/404.

The midrange is about the same with both, but the treble is slightly more evident with the Sigma Pro. It’s close to a line call there, however.

As for dynamics – the Sigma Pro does dynamics somewhat better than the Sigma/404. The Lambda Signature driver/cable just sounds a little faster than the 404 driver/cable. This tends to alleviate some of the complaints about a mushy low end of the Sigma series.

The original Sigma low bias was far too rolled off at the top end in particular, despite the magical Sigma midrange being present there. It appears that either the Signature or 404 driver implants are a successful remedy to this, and give a phone that had huge promise a push into reference territory.

In summary, despite every single report to the contrary, I’m loving it! Flame suit on, hearing aid batteries fully charged LOL. Stax SR-009? Who cares?

Edited May 9, 2012 by John Buchanan

[faust3d]

I love my Sigmas, they are my favorite stat phones after O2, closely followed by LNS. Soundstage is superb. Here is even more info.

[Kabeer]

I love my Sigmas, they are my favorite stat phones after O2, closely followed by LNS. Soundstage is superb. Here is even more info.

Yo Faust, your wiki is coming up with your "dont steal" nurse when your on your site. Seems like it thinks your stealing from yourself .

Or have you just been a naughty boy ahaha.

[faust3d]

Yo Faust, your wiki is coming up with your "dont steal" nurse when your on your site. Seems like it thinks your stealing from yourself .

Or have you just been a naughty boy ahaha.

I don't see any problems. I think it's just the way you access it.

[Kabeer]

I don't see any problems. I think it's just the way you access it.

Curious...its working now. All yesterday it was giving me the nursey. Ah well no biggie.

[spritzer]

Woot a Sigma thread!!! To start off, I figured this was worth a repost with pics:

I posted this over on HF but it should also be here. What I did for those not in the "mafia" is to rip apart a SR-Sigma and a SR-404, swap drivers and cable and then put them together again as a SR-Lambda and SR-Sigma 404. The Sigma line used Lambda drivers or rather the SR-Lambda used the SR-Sigma driver design so swapping them is relatively easy. Stax has been doing this to broken Sigmas for the last few years but why pay them to do something I can do in a few hours.

Before anything like this could be done I put a lot of hours into ripping the SR-Sigma apart, cleaning it thoroughly and removing any glue residue that was left over. That is a nasty, time consuming job and leaves you with very sore fingers and no fingernails. I didn't do any mods to the original design so the drives are out of phase like they are supposed to be and I even replaced the glue that holds the damping material in place and provides a better seal.

SR-Sigma and SR-404 drivers side by side. It's startling how much things have changed in 30 years...

Here are most of the parts that make up the headphones. The damping material was still drying out after the cleaning treatment and the arc was still in pieces after a thorough cleaning. Also missing is all the glue needed to mount the drivers and the damping material.

The cable has been fitted and glued in place, drivers mounted to the metal baffles and all the glue to hold the damping material is in place.

Drivers soldered to the cable and damping material fitted. The only issue is that the soldering connections for the bias are now facing the baffle and thus useless so I had to be creative. It would be possible to solder the cable before the drivers are mounted to the baffle but it would be much tougher to work with.

Here they are ready with the "left over" SR-Lambda. I used Sigma earpads but Yamasinc has used Lambda pads for some reason in the past. I know that the Stax factory uses Sigma pads for this job so I went with that. I haven't had much time to play with them but first impressions are positive and they outclass the SR-Sigma Pro, that's for sure.

Now I've done a few since I posted that but the only thing that has changed is how I handle the drivers. Now I remove the protective grill and mount it to the other side of the drivers so that I have easy access to solder the connections. It's more work but eliminates any potential issues. The last two sets have also had the drivers permanently glued to the baffle.

[faust3d]

The last two sets have also had the drivers permanently glued to the baffle.

How does that change the sound?

[spritzer]

Doesn't change a whole lot. It's just me being OCD and going into overkill mode after drivers were knocked free during postage.

[Dusty Chalk]

Eliminates rattle?

[HeadphoneAddict]

I borrowed a pair of Sigma Pro from Plaidplatypus for a few weeks last year, and I thought the Sigma Pro sounded too soft and very rolled off in the treble. I liked the soundstage and the mids, but that was the only thing about them that I liked. The soundstage of the stock Pros even beat my K1000 for cohesiveness and placement, but the amount of missing detail was significant. I tried them with the GES and SRD-7 Mk2 Pro with a 12 watt speaker amp, and they worked better off the SRD-7 Mk2 Pro. Then I heard them last July after being upgraded to a Sigma 404, and the Sigma 404 sounded much more detailed and extended than the original (even on the GES), but still nothing like the Lambda, O2 or K1000.

[mypasswordis]

Is it me or is HPA improving with impressions? Still a bit too much mention of extraneous gear, but otherwise helpful. Next thing you know, he will start saying cables make little to no difference.

[jinp6301]

Random question, why does this woman wearing the sigmas have a black eye?

[Dusty Chalk]

Big Venetian Snares fan?

[HeadphoneAddict]

Is it me or is HPA improving with impressions? Still a bit too much mention of extraneous gear, but otherwise helpful. Next thing you know, he will start saying cables make little to no difference.

I did just tell the guy who bought a portable amp from me that my TWag LOD sounds as good as those costing 2x more, and it's not worth paying more than the TWag. But, that doesn't qualify because I told him the TWag was more detailed and transparent than the Oyaide LOD that I threw in with the amp to seal the deal. But I do appreciate the positive feedback and I'll keep trying.

EDIT - See, I didn't even post what amp I sold or all the LOD that I compared with each other...

[Iron_Dreamer]

I still need to listen to Sigmas one of these days. I can't think of a more desirable headphone I haven't had the chance to audition.

[faust3d]

Random question, why does this woman wearing the sigmas have a black eye?

Don't know. I am looking for an answer but I am yet to find it.

[padam]

I've just fired up the eXStatA today with the Sigma Pro (finally!!). The usual roll-off problems remain like I heard before but what is there sounds very nice - vocals are especially great and of course there is that huge headstage.

The amp does not seem to have quite enough 'oomph' to drive it via RCA, but XLR seems to be just fine.

[faust3d]

Roll-off in Sigmas is funny in a way that it is, for me at least, sounds very natural and speaker like. Overall presentation remind me of very HQ planar ribbon system. Awesome headphones that sound very lifelike and big.

[John Buchanan]

The amp does not seem to have quite enough 'oomph' to drive it via RCA, but XLR seems to be just fine.

I've found this on the Stax SRM Monitor and the SRM-717. Not sure why - I suspect the balanced Stax inputs leads to a slightly less bright result?

[spritzer]

The only difference using XLR's is more input voltage and thus more ultimate voltage swing. That's it.

[John Buchanan]

Been listening to the Sigma/404s all afternoon (Studer D730 -> Apogee Mini-DAC -> Audio Research LS5 Mk II -> Studer A68 -> SRD-7 (Modified to Pro standard with a Spritzer board)) Sounds absolutely fabulous! The smooth smooth top end allows playing of Keys To Ascension without any pain. Great way to feel better. These really are great phones.

[TommyTunes]

Going to have to start looking for a pair of Sigma's to upgrade. They certainly look interesting. Just ordered Omega SR-007's from Elusive Disc today so it'll be a while before I have the extra cash.

[John Buchanan]

Stax Sigma - one of the few headphone systems that seems to suggest a correct volume level - much like Peter Walker of Quad suggested. They seem to open up at a certain volume. Anyone else experienced this? I really like them through the SRD-7. One day, a T2 or BHSE or KGSSHV!

[spritzer]

Most of the stat's are like that for me. This is the main reason why I despise steppers since they never have the correct setting...

[John Buchanan]

Birgir, for some reason the Sigma is the easiest to dial up the correct volume - the LNS less so. I agree regarding steppers. What volume control did you put in your balanced SRM1 Mk 2? A Quad ALPS RK27? Let's see pictures!!!!!