Graham Audio LS8/1 Signature Edition & a New System

[jonstine]

Back to Benchmark AHB2 Amps

For the first time in the history of my audio hobby, I have purchased an amplifier for the second time. A few years ago, in my current listening room, I used a pair of Benchmark AHB2 solid state amps, bridged for mono operation with first my Harbeth M40.2 speakers and then my Gradient 1.4 speakers.

I sold that pair of Benchmark AHB2 amplifiers when I moved to active speakers, the Dutch & Dutch 8c, and later did not need separate amps with my Sanders 10e speakers since the speakers came with one Sanders Magtech amp and I logically purchased another to fill out the biamplified amplification needs of the 10e speakers. I sold both the Sanders amps as part of the sale of the Sanders speakers.

My recent experience with the 750-watt-per-monoblock Audio by Van Alstine AVM M750 amps showed me that I'm a power snob. While the smaller Van Alstine AVM M225 (225 watts per monoblock) sounded great in isolation with my Graham LS8/1 Signature Edition speakers, having heard the more powerful M750 made audible for the first time with these speakers the sense of unlimited power reserves which comes with yet more power on tap. However, while these amps sounded yet better than their smaller siblings, in my system they made way too much noise through the speakers--mostly hum--to be tolerated long term. Thus, I returned the M750s for a refund. As jonstine's comments above and elsewhere on WBF suggest, this noise problem with the M750s may be just a peculiarity of my set up.

Very few amps are truly silent in terms of quiescent hum and noise, in my experience. Usually, with no music playing, if I put my ear an inch or two from the drivers I can hear some noise. I forgive such noise as long as it is inaudible from the listening position, which, in my case is some 55 inches from the drivers.

Thus, my goal in replacing the Van Alstine DVA 750 amps was to acquire amps which provided a similar "endless power" sensation, while still being silent, both in terms of audible chassis hum and audible hum, hiss, or other noise through the speakers.

I know that the powerful Sanders Magtech amps usually are not silent enough. When I once used them with my Harbeth M40.1 speakers in another room of my house, I could hear hiss from several feet in front of the speakers. However, they work in a noise-free manner as part of the Sanders 10e system where they are combined with the dbx VENU360 Loudspeaker Management System. That unit can very effectively act as an input gain control for the power amps, providing, as it does, analog domain controls for both input gain and output voltage. Thus, the dbx can easily lower the effective gain of the Sanders amps so that hiss normally audible through the speakers is silenced. Part of proper gain structuring of a professional audio system is using high gain sources, together with a gain control/volume pot or pad at the amplifier input to allow the gain of the amps to be turned down to just a few dB. This lowers the noise and increases the signal to distortion and noise audio throughput of the system. The Sanders system is unusual among consumer audio systems in allowing for this sort of gain structuring.

The Benchmark Media company, having a foot in both the pro audio and consumer audio camps, provides a gain control switch on its AHB2 amps. In its lowest gain position the AHB2 amps have the truly low gain which ensures compatibility with the high source voltages used in pro-audio system. At this setting the amps require 9.8 volts input for rated output power. While my Lumin X1's rated output is only 6 volts, I know from my prior experience with this combination that the low gain setting goes loud enough with all source material in my room, as well as sounding best and providing the absolute least quiescent noise from the speakers. See my comments on this arrangement here. The Benchmark amps are extraordinarily quiet with any setting of the input gain switch. I hear no transformer hum from the amp chasses with my ear closer than inch from each chassis. And in their low gain setting, I hear absolutely no hum, hiss, or other noise from any of the speaker drivers even with my ear on axis with any of the speaker drivers and within an inch of the speaker grill cloths. This applies for any setting of the Lumin X1's Leedh-processed digital volume control.

My sonic evaluation of the Benchmark AHB2 amps is unchanged since I last used them. If anything, system improvements in the interim have only further revealed their level of sonic refinement in terms of perceived lack of background noise, clarity, lack of distortion, immense sense of power reserves, lack of coloration, three-dimensionality, envelopment, etc. See my 2018 discussion of the AHB2 sonics here.

In the years since its introduction, some, while acknowledging its matchless specs in terms of low noise and distortion combined with remarkable power output given its small and light form factor, have criticized the Benchmark AHB2's sonics by saying it is lacking in macro dynamic swings. I hear this, but I hear it differently and hear it as a plus for the Benchmark. The Benchmarks lack the overblown-sounding midbass of some other amps which can provide a sense of greater dynamic power. It also lacks the ultimate degree of "splash" on large transients which some interpret as a dynamic limitation. I think this "splash" is actually caused by a blast of slightly distorted sound in the higher frequencies. The Benchmark is ever precise, controlled, and clean, clean, clean, which amounts to "beautiful" sound in my book. Likewise, the Benchmark midbass is fully ample and certainly not lean in any way, just more controlled and defined/less distorted. This "uncovers" the Benchmark's lower bass, revealing more detail, definition, and power in the bottom two octaves.

In terms of size and weight, as well as cool running, the Benchmark AHB2 is roughly equivalent to the Audio by Van Alstine DVA M225. The form factor is a bit different, that's all. The AVA is narrow and deep. The Benchmark is wider and less deep. They are both easy to carry and fit well within my "downsized" system form factor goals. However, the Benchmark AHB2s, used in bridged mono, give that "limitless" power impression, whereas the AVA does not. And while the AVA is quiet enough running, the Benchmarks have noticeably more silent backgrounds. The AHB2 are, for all practical purpose, absolutely silent runners, totally concealing the vast power on tap.

Continued…

While I understand the need for attenuation is system specific, and not necessarily a requirement for everyone, they have made a huge impact with what’s being sent from the Lampi to the amps.

The SR Purple fuses are specific to the M750s however, and as long as you orient them the correct way (in my case both fuses have the “S” in SR closest to the fuse knob or exterior of the chassis) they make a difference. I’ve got the SR Master fuse in the Lampi, but not sure that I want to buy 2 of those for the M750s. I suspect they would be better, but that’s a huge price jump for 2 of them.

Regardless, with the addition of the SR Purples and the Rockwell attenuators, at 200 hrs the M750s are singing better than ever, and I am incredibly pleased with them. What a hobby!

 

[tmallin]

Bass Extension & Power: Can You Have Too Much of a Good Thing?

The potential bass extension of the LS8/1f floor-standing version of the Graham LS 8/1 model (see https://onahighernote.com/shop/spea...ing-model-of-the-derek-hughes-ls8-1-monitors/) is a potential plus, at least for some rooms. The floor stander currently lists for $11,500 in the USA.

For my small room, I'm no longer so certain. I have had several speakers in this room which measured flat-in-room to 20 Hz or below (Dutch & Dutch 8c, Harbeth M40.2, Sanders 10e, Gradient 1.4, Stirling LS3/6 + Swarm Sub quartet, KLH Model 12, AR-3a). The Graham LS8/1 may not measure flat in the bottom octave, but subjectively the bass from these sub-less speakers is every bit or more as satisfying as any of those others in my small room. In my room, the Graham bass strikes an exemplary balance between actual extension and cleanliness and detail, not audiby exciting structural resonances from the room and having no audible speaker-box or driver resonances to muddy up the low end.

Also, I note from the specs that the LS8/1f model is just about one meter tall. That's at least two inches shorter than the stand-mount model on the Graham stands. That means the tweeters and best listening position are another two inches lower, making the lower tweeter only about 32 inches above the floor. This means that either tall footers or tilt-back of the speakers would be mandatory for near-field listening.

Someone noted that the low-height problem of the floor-standing version of the floor-standing version could easily be dealt with by adding a platform of substantial thickness, such as a cinder block, between the speaker and the floor. That's a good point. Standard cinder blocks are 8" x 8" x 16". That only gives you two choices of additional height. With a 32" tweeter height for the speaker sitting on the floor, that would give you options of 40" and 48". The first is quite reasonable for a normal domestic chair. The second is a bit high for something other than a high stool.

There are obviously other options such as lumber of various sizes. For more money, there are also handsome purpose-made audio platforms made from maple and walnut (see, for example https://butcherblockacoustics.com/collections/audio-platforms or https://www.mapleshadestore.com/airdriedmapleplatforms.php. In addition, granite or marble platforms are a possible attractive alternative. The Tontrager website for the reference stands intended for Harbeths says that the stands optimally should be set atop slate plates; see https://www.tontraeger-audio.com/lang/en/produktdetails-stands.html

Any "rocking" problems at the interface of speakers and supports or supports and floor should be addressed. I know from experience that allowing your speakers or their supports to "float" atop carpet, rather than spiking them to the floor beneath the carpet, takes care of any rocking/resonance problems at that interface. I also know that putting a layer or two of terrycloth toweling between the speaker and the wood or cinder block support should prevent any rocking/resonance at that interface.

I do wonder, however, how much the sound of speakers is affected by changes in the speaker height. Perhaps electronic EQ could address any adverse effects.

I noticed when I moved the Graham LS8/1s from my initial stands--Skylan stands which had a solid top plate, 20 inches off the floor--to the Graham stands which are 4+ inches lower and have no top plate, and lowered my listening height by the corresponding amount, that the sound from my new lower listening position warmed up in the midbass and lower mids. In my installation this was all to the good.

I have to wonder how much of this sonic change was caused by (1) the lower Graham stand height, or (2) the lack of a top plate on the stand, allowing the speaker cabinet to resonate more into "free air", or (3) my lower seated height. I did not attempt to separate these effects. But I suspect that at least for some speakers in some rooms, changing these factors alone or in some combination can be expected to produce sonic changes, sometimes for the better and sometimes not.

But another reason for my switch to a smaller speaker from the Sanders is the difficulty of moving heavy and bulky speakers in and out of my listening room and navigating the right-angle turn in my stairway getting speaker boxes up to my second floor listening room. Since I change speakers every year or two, this is a practical consideration for me.

The floor-standing LS8/1f does not occupy any more floor space than the stand-mount version, but it is heavier and the box is much larger. I no longer want to wrestle large heavy boxes up and down the stairs. The two stands of the Signature version of the LS8/1 are nested in a single box about the same size and weight of one of the speaker boxes, so those boxes are all manageable.

I also believe that stand-mount speakers have an advantage over floorstanding versions because of the way the stand allows the drivers and enclosure to be surrounded by vacant air, improving the staging and imaging. Further, I'm pretty sure that the longer larger wood panels involved in making a stand mount into a floor stander create bigger challenges for eliminating box resonances. Boxes of dimensions like the two-cubic foot Graham and its ancestors (basically two feet high and one foot in the other two dimensions) seem to have an advantage in terms of dealing with box resonances. Measurable resonances certainly are there, but they are not obnoxious to the ear.

The lossy screwed-together construction of the stand mounts would seem to be a better approach than the glued-together construction of the LS8/1f, but it could be that once the shape and size of the box is changed, the glued construction/MDF may actually be equivalent. We'd have to see accurate box resonance measurements to be sure.

I'm not sure if what I hear in my small room from speakers with greater bass power/extension than the LS8/1 is what others call "room roar" or "overloading" the room with bass. But if I disconnect the high frequency drivers (easy with the Sanders given its electronic controls and many speakers with bi- or tri-wire connections) or just EQ the mids and highs way down in level, it is easy to hear room contents and parts of the building structure within the walls and floors, and even my hot-water heating pipes and registers, buzz or resonate with bass.

With some speakers with 20-Hz-flat bass extension I've used in this room (the Sanders 10c and 10e, and the Gradient 1.4 in particular) it was also clear that the bass drivers and cabinets buzzed or resonated when reproducing bass tones even when the SPL was down around a measured 80 dB. The Dutch & Dutch 8c cabinets and drivers were exemplary for not producing any spurious mechanical noises regardless of SPL or frequency, but the structural buzzing room resonances were still there in full force.

With the LS8/1, I hear nothing but clean sound from the drivers, cabinets and room at any SPL I care to listen. And given the frequency balance of the speakers in my room, I'm not always trying to crank them louder to bring up the bass or hear more upper range detail. I'm sure that helps keep the in-room sound clean.

 

[Ron Resnick]

Equalization: None vs. the Schiit Audio Loki Max
But then about three days into my audition, I began to have my doubts. Something was wrong. I still heard no real difference between the processing turned on but with the bands all set to flat and the electronic bypass of all the filters. But the tremendous sense of spatial envelopment I remembered having before adding the Loki Max to the system seemed reduced. In addition, recordings seemed much less dynamically unrestrained than before.

I did an experiment to enable a relatively quick A/B comparison.

This effectively took the Loki Max physically and electrically totally out of the signal path, bypassing the relay and resistor which Schiit says are the only electronic components still in the signal path when the bypass mode is enabled.

The sonic change was unmistakable. Now, even with the extra 1/2 meter balanced cable and its extra XLR connection in the signal path, the space and macro dynamics I remembered from my unequalized system were back. The difference was instantly audible and clearly "better" rather than just a difference.

For now I will eschew all electronic equalization in my new system.

Your opinion may well differ on that judgement. Your system may well not reveal the level of space/envelopment and dynamic contrasts my new system does. Certainly my prior Sanders system did not.

Hello!

Do you have a theory as to why the Loki Max obscured or eliminated ambient spacial cues and restrained macrodynamics?

 

[Ron Resnick]

Have you ever experimented with the Manley Mid-Frequency EQ device?

 

[tmallin]

The sonic effect I heard must have been due to the relay and resistor left in the signal path with the Loki Max physically connected. As I noted at the time in post #3 of this thread, the sonic problem went away when I physically removed the Max from the signal path by connecting the balanced input cable directly to the balanced output cable. I did not investigate further since by that time I was also quite disgusted with the, to me, totally inadequate user interface which I detailed in post #4 of this thread.

And, no, I've never experimented with the Manley Mid-Frequency EQ device. As I've mentioned before, I do not buy tube equipment. The sharp Q equalization description of what the unit does from the Manley website does not sound like something I would want to insert in my signal path.

 

[Ron Resnick]

The sonic effect I heard must have been due to the relay and resistor left in the signal path with the Loki Max physically connected. As I noted at the time in post #3 of this thread, the sonic problem went away when I physically removed the Max from the signal path by connecting the balanced input cable directly to the balanced output cable.

Thank you!

I wonder if the RC or RL switched in by the Loki Max positions introduces some phase anomaly which obscured or undermined ambient spacial information with you previously were hearing.

A sense of dynamics no longer being unrestrained perhaps could come from the inability of the Loki Max to swing voltage to the extent of your line stage's ability to swing voltage.

 

[tmallin]

Another Tweak

Back in Post #45 I listed a number of sonic tweaks which improved my streaming sound. Here is another tweak I recently discovered which seems to improve the sonics I hear when streaming via the Lumin App as opposed to Roon. In my system, streaming via the Lumin App sounds at least a bit better than streaming via Roon. This tweak might also help when streaming via other proprietary streaming software where the Roon Ready function of your streamer can be turned off as is possible in the Lumin App.

1. In the Lumin App Options menu, turn off "Roon Ready" if Roon is not being used.

2. In the Roon App on the computer which acts as your Roon Core, under Settings>Audio, disable all device choices.

3. Close and exit the Roon App on the computer which acts as your Roon Core. Then, turn off that computer.

4. Open the Roon App on your system controller (in my case, the iPad which controls my Lumin X1 streaming DAC). In this Roon App, under Settings>Audio, disable all device choices. Then close this Roon App and exit from it (swipe up on an iPad).

In my case, I had long been practicing steps 1. through 3. I only recently discovered that step 4., disabling all the device choices in the Roon App on the system controller, was a key to yet-better sound from the Lumin App. I have no idea why this matters, but it seems to make a positive sonic difference, in my system, at least.

 

[tmallin]

Enter the Lumin L2 Music Library & Network Switch

Over the last couple of years, I have mostly listened to my own music files ripped from my CDs via Roon. Those music files were hosted on my Mac (or before that Windows) computer and played through my stereo room rig via Roon. This is because I determined that attaching USB or even ethernet connected devices to my Lumin X1 streamer had detrimental effects on the sonics I heard from internet streaming from Qobuz, Tidal, or even internet radio stations. This was true even with the earlier Lumin L1 music library.

Over that time I gradually came to the conclusion that optical filtering via my GigaFOIL v4 INLINE Ethernet Filter inserted in the signal path just before my Lumin X1 was the best way to "clean up" the noise and other nasties otherwise audible from streaming sources via the X1 and the Lumin App. This also provided the best sound I've heard in my stereo room from Roon (Tidal, Qobuz, internet radio, and my local music files). In my installation, which involves a 90+ foot run of ethernet cable from my router, the GigaFOIL can make a subjectively HUGE improvement in sound quality as I've discussed in earlier posts of this thread, such as #5.

I heard the Lumin L2 at AXPONA 2023, however, and was quite impressed with its streaming presentation even under difficult hotel show conditions. When the L2 was actually released to consumers I became even more interested in its potential to isolate both itself and the Lumin X1 from conducted internet noise via fiber optical links. Recently I decided to pull the trigger on purchase of the L2.

Yes, it's expensive, at 4,500 USD for the version with 4 TB of built-in solid-state storage which I purchased. It's 3,500 USD with no storage and 5,500 USD with 8 TB of built-in solid-state storage. But for the first time the Lumin App now delivers both the best sound I've ever heard from my owned local music files and the best sound I've ever heard from all internet sources--Qobuz, Tidal, and internet radio stations of every ilk.

For a diagram of the streaming signal path I'm using, see the second diagram at this link. I'm using all of Lumin's recommended fiber optic pieces. The details are: Xfinity Gigabit internet service, feeding Xfinity xFi Advanced Gateway, feeding 90 feet of CAT6 shielded ethernet cable, feeding 5 feet of Blue Jeans Cable Cat 6 Patch Cords ethernet cable, feeding TRENTNet TEG-S51SFP 4-port Gigabit Switch with SFP Slot powered by 5V output of Keces P3 LPS, feeding Startech SFP Module feeding one meter of LC-LC Duplex 9/125µm Corning ClearCurve Single Mode Bend Insensitive Fiber Optic Patch Cable, feeding Startech SFP Module, feeding Lumin L2 Music Library & Network Switch, feeding Startech SFP Module, feeding one meter of LC-LC Duplex 9/125µm Corning ClearCurve Single Mode Bend Insensitive Fiber Optic Patch Cable, feeding Startech SFP Module, feeding Lumin X1 streaming DAC. The Leedh-processed volume control of the X1's balanced analog outputs directly drive my Benchmark AHB2 amps which directly drive the Graham LS8/1 speakers. No preamp, no electronic EQ.

Note that the additional cost of creating all this fiber optic isolation around the Lumin L2 and X1 is about $300. That's the total for the TRENDNet switch, four SFP modules, and two fiber optic cables.

The GigaFOIL is gone since it's optical isolation is now unneeded given the fiber optical connections I'm now using.

The Lumin L2 makes my system sound better yet than the GigaFOIL did. Not a huge difference, perhaps, but it's quite a substantial difference. And it's all in the right direction of making things sound yet more three dimensional, natural, and relaxed, while yet adding more audible musical detail without adding any high frequency brightness or nastiness. Musical lines are easier to separate/hear out and in fact often I hear musical lines which seemed totally buried before. With classical music, the result is "more like analog." What I mean by that is that the sound is more like unamplified live music heard from a great seat in a favorable venue in terms of perceived lack of distortion, clarity, subjective frequency balance, three dimensionality, and dynamics.

There's more. I'm still discovering the wonders of the Lumin L2. There's more improvement than I first thought. Music flows better and makes more sense rhythmically. By comparison, the timing of the music was a bit "off" before and now it's spot on and much more engaging. I hate to bring up the old PRaT concept, but that's another aspect that seems to have been greatly improved.

With the addition of the L2, I'm yet more convinced that many who insist on using electronic equalizers to tame the excessive highs of the majority of commercial recordings are in fact reacting at least partially, if not primarily, to the high frequency nasties present in most digital signal paths, not so much to the absolute level of the highs. Sure, many recordings still sound brightly balanced with the L2, but tonal brightness with the nasties so much reduced is entirely listenable and not worth electronically equalizing when such equalization adds sonic problems of its own. At least that's the way I hear it.

There were some flies in the ointment in terms of getting my music files loaded onto the L2. See this post of mine over on the Lumin forum and the responses of others to it. I would not have had any problems had I been aware of the need to add third-party software to my Mac before attempting to drag and drop files from my Mac hard drive onto the NFTS-formatted L2. And now it seems that Peter Lie, Lumin's Firmware Lead, says in a response he posted, that the SMB method of adding files to the L2 is the way to go anyway, being superior for the L2 than the drag-and-drop method that gave me such problems. You don't need any extra software to add files to the L2 via the SMB method. If you decide to take the L2 plunge, you will be forewarned and forearmed.

But now that the music is loaded, those teething problems are rapidly receding in the rear-view mirror. I'm basking in the glow of the best sound I've yet had in my stereo room.

 

[DaveH]

I have the Graham 8/1s as well. I have gone a different route with amplification -- I am using a Luxman mq88. My previous experiences with British monitor speakers taught me that the least complicated route to great sound was quality, moderately high powered SS amps. I was surprised by the apparent synchronicity between the 25wpc Luxman and the Grahams. Then again, the Grahams are more transparent and less warm than the older Harbeths I have owned, and have a better, more natural bass response than the Ls3/5as. Maybe it's the apparently stable impedance curve and the relative small room they're in. I am currently loving the combo, more than I enjoyed the Grahams with the LFD NCSE or a McIntosh 2105.

 

[godofwealth]

Great write up! I’ve listened to BBC designs since the mid 80s when as a grad student I bought a pair of Spendor SP1s. I now own the SP 1/2e and the Harbeth Monitor 40.1s. But my first love has and always will be with Quads (I own three pairs, the 57s, the 2805s and the 2905s). I find the crossover induced effects of the BBC designs too distracting. The change in directivity is also problematic. Ultimately I don’t think the BBC designs can overcome the theoretical limitations of the basic design, however clever the skullduggery in the thin walled cabinet construction and the polymer cones.